Number 26: Communities, social capital and public policy: literature review
This report was published by the former Department of Families, Community Services and Indigenous Affairs (FaCSIA).
- Executive summary
- 1. Introduction
- 2. Community and neighbourhood
- 3. Social capital
- 4. Strengthening communities
- 5. Concluding comments
Strengthening families and communities is an important goal of government policy. In recent times, interest in community affairs has been highlighted by the development of measures of social capital, which aim to summarise and measure the benefits of communities. A wide range of disciplines, including sociology, demography, political science, psychology and economics, have provided the framework for studies of community and social capital. In this review we explore the meaning and relevance of community and social capital drawing on all these fields, although the main basis for the review is our own specialist area of economics.
Economic studies of community
Economists view community both as something that can be explained by economic phenomena but also as an explanation for economic outcomes; that is, the community itself can be the object of interest, or the nature of community can help to explain other objects of interest. To enable unambiguous interpretation, it is necessary that key variables in any economic analysis be rigorously defined. Too often the concept of community is not given precise definition. Nevertheless, the community as a focus of attention has been important in the sub-discipline of regional economics.
Perhaps the most pervading view of community is that of a spatially defined group of interest—a small town, a suburb or neighbourhood. While it is certainly possible to speak of communities of interest and virtual communities, the spatial paradigm is the most compelling and the one employed in this review. Regional economics assumes the spatial definition of community as given. A number of important implications arise from the application of economic perspectives to community development.
First, a mainstream economic perspective suggests that economic phenomena can largely explain the long-term growth and development of towns, cities and regions, and therefore of communities. These economic phenomena include increasing returns to scale, external economies or spillovers1, and transport costs. While governance structures, the whims of important individuals, social, cultural, language and religious imperatives can all influence the pattern of growth and development of towns and cities, the enduring drivers are economic forces. Theories about the development of communities have evolved from largely descriptive central place theory, through export base modelling, to more recent views that encompass agglomeration economies and endogenous growth.
Additionally, an economic paradigm can be used to characterise and classify regions (and communities) and to explain phenomena that affect communities. In Australia issues of rural depopulation and the drivers of regional growth have become of key importance.
Finally, there are a number of economic tools that are useful for describing development within regions and communities. Regional input-output models seek to capture all of the important measured economic linkages between economic agents within the region and between it and the outside world. Private sector development within a region is amenable to analysis by all manner of business management tools and there are public sector evaluation methods such as cost- benefit analysis and social indicator analysis.
A second line of research embodying concepts of community is the neighbourhood literature. This literature began with sociological studies that aimed to explain differences among neighbourhoods and consequential differences in the life chances of residents. A number of themes have been stressed in this work. First, there is the theme of mobility. If populations are fully mobile they will adjust to neighbourhood disadvantage by exiting. However, physical and social capital fixity may constrain movement.
As in the regional economic literature, a second key theme associated with neighbourhood performance is that of the human capital externalities that arise from the characteristics of the neighbourhood. Numerous studies indicate that the educational achievements, employment prospects and career success of residents are affected by the nature of the neighbourhood.
Social scientists from a number of disciplines have found it useful to characterise the strengths and weaknesses of communities using the concept of social capital. Social capital refers to the networks and norms that enable collective action (World Bank 2002). Characterising the structure of social capital involves describing the size and density of networks, while the content of social capital includes the degree of trust and the prevalence of reciprocity within networks.
Social capital is often assumed to be positive at least from the perspective of the community to which it belongs. However, great care is needed when aggregating social capital since the social capital of one community may be created at the expense of another, or the strength of one community may be to the detriment of others.
Nevertheless, researchers have shown statistical associations between high levels of social capital and a range of benefits, including the development of pluralist democracy, improved physical health, self-rated happiness, public safety and enhanced economic performance and efficiency. In this review we have suggested that some of these associations are merely correlational, while others are almost certainly causal.
Australia appears to have high levels of social capital in comparison with other Western countries. Among Australian communities there is variation depending on urbanity, levels of income and ethnicity. A much-debated issue is the interpretation of high intra-group levels of social capital among immigrant groups. While there are some economic advantages arising from dense networks within groups, these can sometimes be at the expense of more valuable contacts with 'mainstream' groups, especially once the initial phase of immigrant adjustment is complete.
Public policy aims to enhance social capital and community outcomes. However, it is not clear whether policies that target communities rather than individuals make the most effective use of public money. In some circumstances, targeting will unnecessarily reward those not in need and divert resources from those who are in need. On the other hand, community-based policies may promote social capital and generate externalities thereby increasing effectiveness.
Evaluation of Australian and UK community-based programs suggests some useful guidelines that may be applied to community policies. These include: (a) multi-agency formulation and delivery of programs including representation of the private, public and the voluntary sectors; (b) an emphasis on capacity building and sustainability; (c) leveraging of funding from a variety of sources; (d) strategies that aim to develop human capital as well as physical capital; (e) emphasis on building linkages between communities; and (f) planned and longitudinal (not just post hoc) evaluation of programs.
Some concluding issues
In spite of the difficulties of defining communities, there does appear to be increasing interest in and public support for community-based service delivery. The potentially increased responsibility of community organisations raises issues in relation to their governance and accountability. We have reviewed concerns that many do not adequately represent all key stakeholders and may not be well equipped to handle increased government funding.
Finally, some issues of measurement. A number of studies have been undertaken in Australia measuring social capital and indicators of community progress more generally. A problem with compendia of social indicators is that it is not obvious what weight should be placed on individual indicators. With this in mind, we have reviewed several approaches to aggregating measures of social capital, including the future possibility of including both stock and flow measures in the National Income Accounts.
For the present, this review has suggested four key principles for measuring community strength and social capital:
- distinguish between the structure (size, density) of community networks and their content (norms of trust, reciprocity and participation)
- specify precisely the geographical area and/or type of organisation to which measures apply
- develop measures of community/networks/social capital that can be validly presented at both individual and aggregate levels
- always assess the costs and benefits of social capital empirically and never assume that all social capital is benign.
The purpose of this literature review is primarily to clarify issues and assess evidence in relation to:
- concepts and competing definitions of community and social capital
- measurement—how best to measure community integration, social networks and social capital
- empirical claims about the sources, benefits and potential costs of stronger communities and social capital.
It is not the aim of the review to conduct a detailed evaluation of policy programs and other interventions intended to enhance community and social capital. We include a section on policy programs at the local, regional and national levels, but our intention is mainly to classify the types of programs adopted and to canvass the arguments for and against them. We do not attempt detailed evaluation of programs, let alone make recommendations to improve them. Those tasks are important but need to be the subject of a separate study.
In the English-speaking world, three institutions are usually described as being mainly responsible for human welfare: families, markets and governments. In many European countries, especially those with a Catholic or corporatist social tradition, a fourth institution, the community, is generally included (Esping-Andersen 1990; Goodin, Headey, Muffels & Dirven 1999; Tam 1998). The principle of 'subsidiarity', enshrined in European Union treaties and laws, is supposed to ensure that all decisions that can be taken at local community levels are taken there.
In English-speaking countries, concern about 'community' seems greatest during periods of profound social change. This happened during and after the Industrial Revolution (Laslett 1971; Polanyi 1944) and it is happening again now in this period of globalisation. At such times, it is claimed that the main institutions supposed to promote human welfare cannot cope, or are not doing a good job. Presently, there are claims that families, markets and states are letting many people down. Nearly everyone lives in nuclear families and if the adults cannot find paid work in the labour market, the family has to rely on state benefits for its livelihood. This may lead to 'dependency' or 'marginalisation' or 'social exclusion'—a situation of long- term non-participation or low level participation in the social and economic life of the community. There are, in fact, increasing numbers of people in Australia, Britain and many other Western countries who are primarily dependent on state income support for their livelihood (McClure report 2000).
One possible remedy for social exclusion is a revival of 'community'. A practical step, discussed in the McClure report but not included as a formal recommendation, would be increased participation in voluntary or community work, that might be unpaid, or as the McClure report suggested, be paid at a higher level than basic state benefits (a participation supplement). Voluntary participation by significant numbers of people could perhaps replace economic participation (paid work) as a major daily activity, although the intention was, in many cases, that it should be an intermediate step leading to paid work in the mainstream workforce. Voluntary work would help to fulfil the participants' side of the 'mutual obligation' between themselves and society, and it might make a valuable contribution to their own and other people's welfare.
In Australia this particular policy thrust appears to have won some level of tripartite support—from the Coalition, Labor and the Democrats. However, for social participation to occur on a large scale, there would arguably need to be significant stimulation and perhaps increased financial support for the community sector (or, as it is variously called, the voluntary sector or the 'third sector').
Social capital and the community sector
The concept of community has recently been taken more seriously in political and social discourse in English-speaking countries. Arguably coincidentally, the analogous but different concept of 'social capital' (Putnam 1993, 2000; Fukuyama 1995; Cox 1995; Latham 1998; Adams & Hess 2000, 2001; SEU 2001) has become important in public discourse. Earlier attempts, partly through the OECD, to advance the somewhat similar concept of 'the active society' (Etzioni 1968; OECD 2001) were apparently less successful.
Use of the term social capital is presumably intended to imply that it is an important stock, which it is useful to think of, alongside physical capital (plant, computers etc.) and human capital (education and relevant work experience), as stocks which might be included in national income or other economic accounts (see section 4 for discussion of this possibility). In principle, and perhaps in practice, the stock of social capital in the country could be measured by the value of the assets of community groups, which include money received in government subsidies and contracts as well as voluntary donations, and also the size and density of their memberships and networks. Of course, stocks are primarily valuable in so far as they generate flows, which provide benefits/welfare to some sets of clients or consumers. The flows generated by community groups can be measured by the active participation of their members in group activities, including activities to benefit others, and by income and expenditures. Client benefits/welfare would presumably have to be assessed by evaluation research, including data on client perceptions and satisfaction levels. As implied above, implementation of McClure's tentative recommendations on community work would both require and promote a considerable increase in community sector stocks, flows and, presumably, client welfare.
The current political importance of the concept of mutual obligation as a basis for welfare reform needs to be stressed. Public policy initiatives stand a much greater chance of success if they are based on ideas and language that command a degree of multi-party support. It appears that in Australia, and also in Britain, the language of community and mutual obligation appeals to most sections of the political spectrum.
The Coalition and its supporters respond particularly favourably to the concept of mutual obligation between welfare recipients and society. This mutual obligation, it is held, requires some form of work from welfare recipients—if not paid work, then voluntary work. The Coalition and its supporters tend to look favourably on an increased role for voluntary groups (sometimes identified with charities), rather than supporting maintenance or continued expansion of state activities. The community sector and voluntary activities are also associated with giving by business and other charitable giving—activities which the Prime Minister, Mr Howard, has personally sought to encourage through the Community Business Partnership Board.
The ALP and the Australian Democrats, like Prime Minister Blair's New Labour in Britain, prefer the term 'reciprocal obligation' to 'mutual obligation', and their objection to 'welfare dependency' places somewhat less emphasis on the point that recipients do not work than that they suffer 'marginalisation' or 'social exclusion' (see e.g. Latham 1998; Tanner 1999). However, like the Coalition and its supporters, they respond favourably to proposals for voluntary work and increased participation in an expanded community sector. Social capital is a concept particularly favoured by the Centre-Left. Promoting social capital and social entrepreneurship is a key plank, perhaps the key plank, of the 'Third Way' (Giddens 1998, 2000). This is not to suggest, of course, that 'Third Way' ideas in general, or a focus on social entrepreneurship in particular (discussed in section 5), command multi-party support. Several government ministers have developed sharp critiques of the 'Third Way' (e.g. Vanstone 2001).
The cross-party support for stimulating the community sector and increasing voluntary participation as an element of welfare reform is mirrored in the mission statement of the Department of Family and Community Services (FaCS), the main department in the Australian Government with responsibility for communities. FaCS identifies three social policy outcomes in its current strategic plan—stronger families, stronger communities, and economic and social participation. FaCS' Research and evaluation framework (2001) indicates that stronger communities are to be promoted by 'government partnerships with the community and business sectors in the social safety net', and says that key policy questions are, 'How do government programs impact on communities and how can FaCS provide better assistance to rural, regional and urban communities that are experiencing economic or social problems?'
The key themes of the stronger communities research program are:
- understanding the factors which give rise to strong communities and those that can weaken communities
- identifying opportunities to promote and assist in the development of strong communities
- exploring volunteering in the community
- examining housing options and reflecting on the interactions between housing and non-housing related outcomes for FaCS customers
- examining factors that give rise to homelessness and opportunities to intervene
- indigenous focused research.
The significant level of cross-party support for stronger communities and a stronger community sector, given further impetus by the McClure report, can be viewed as creating enhanced opportunities for FaCS to propose and win support for initiatives in this area.
Differing frameworks—from public choice theory to community
A final introductory point: some writers (e.g. Edwards & Foley 1998; Adams & Hess 2001; Giddens 1998, 2000) draw a sharp contrast between the theory and language of public choice theory, which has greatly influenced public policy and management in the last twenty years, and the 'new' language of community, social capital, trust and partnership. Public choice theory is explicitly based on an economic model of motivation. The Nobel Laureate, James M. Buchanan, and his followers in the Virginia School, assumed that politicians, public servants and other political actors were rational egoists concerned to maximise their own career goals, not the welfare of the general public (Buchanan & Tullock 1962; Niskanen 1971; Hood 1998). The best way to achieve public policy goals was seen as being to redesign institutions and develop binding contracts which included incentives to motivate service suppliers to do in their own interests what 'the principal' or 'purchaser' viewed as being also in the interests of citizens or consumers. Purchaser-provider contracts based on competitive tendering were a key practical application of public choice theory and principal-agent theory in Australia and overseas.
However, in recent years public choice theory, and especially public management, has moved in directions that are arguably not incompatible with ideas about building community, social capital and trust. Many management theorists and practitioners regard purchaser-provider contracts and service agreements as primarily ways of clarifying goals, gaining efficiencies and reducing transaction costs (Williamson 1996). Beyond that, the American public management theorists, Osborne and Gaebler (1993), whose book Reinventing government has been very influential in Australia, view contracting as providing, among other benefits, prospects for mobilising the altruistic motives and resources of voluntary agencies and community groups in the service of 'community owned government'. Clearly, in this particular formulation, the new public management and proponents of policies to stimulate community and social capital are allies not opponents.
In this context it is interesting that, in conversation with the authors of this review, Robert Putnam said that he does not regard social capital theory and public choice theory as irreconcilable. Public choice theory could be viewed as a set of prescriptions for institutional design, which could, among other things, provide an appropriate framework for a flourishing community sector. The framework could provide the right incentive structure for habits of reciprocity to develop, even without initially high levels of trust. A final point: it can be argued that detailed, 'black letter', highly legalistic government contracting is inimical to the development of reciprocal partnership-like behaviour. However, more open-ended contracts, which are now becoming common, and which leave some risks and potential gains open to later negotiation, are compatible with the development of effective partnerships (Williamson 1996).
The following dot points indicate the main directions of the review:
- The concepts of community: We define communities and the community sector focusing on location-based concepts.
- Community and economics: While community is not an entity recognised in formal economic frameworks we explore the ways in which communities enter economic measurement. Communities may be explained by, or help explain matters of interest to, economics. In using the notion of community, economics draws on understandings developed in other disciplines such as sociology, geography and psychology. The sub-discipline of regional economics (including economic geography and urban economics) provides the principle framework for economic studies of community.
- Neighbourhoods: A particular area of social science literature known as neighbourhood studies is focused on the idea that the locality in which people live will influence their behaviour and their life chances. We explore the literature to outline the ways in which this aspect of community is useful in explaining behaviour.
- The concepts of social capital: Drawing heavily on the work of Raymond Bourdieu (1986), James Coleman (1988), Robert D. Putnam (1993, 2000) and Francis Fukuyama (1995)—and also their critics—we review concepts and proposed definitions of community and social capital. In general, and with the intention of being uncontroversial, many writers assume 'the more social capital the better'. However, the main academic contributors to the literature agree that there is such a thing as 'bad' community or negative social capital (e.g. criminal networks and perhaps 'old boy' networks) as well as beneficial social capital. So what is the 'downside' of social capital, and how can it be recognised, measured and countered?
- The sources of social capital: This section reviews issues to do with the sources of social capital in the family and in ethnic groups, especially immigrants. Better educated and financially better off people are known to be more likely than others to participate in political and other community activities. But what if a society is deficient in social capital? Can it be created in a deliberate way by public policy or other interventions, or can it only develop slowly in a 'natural' way over long periods?
- Consequences—the alleged benefits of enhanced community and social capital: We review the empirical evidence—and debates and gaps in the evidence—on the benefits of community and social capital to pluralist democracy, health, immigrant upward mobility, happiness (self-rated) and, particularly, economic performance and efficiency.
- Australian social capital in international perspective: This section presents data on Australian social capital—particularly levels of trust and confidence in institutions and fellow citizens, and levels of participation in voluntary organisations—together with some international comparisons. The comparative data appear to show that Australia has relatively high levels of social capital.
- Public policy initiatives in relation to community and social capital: Australia and Britain have a fairly extensive history of trying to stimulate and support community groups through public policy initiatives. These have come under such headings as 'community development', 'regional planning' and 'area based planning'. The community sector has also been heavily involved with government in recent years as a contractor under purchaser-provider contracts, delivering a vast range of government services, especially in the human services field. The increased experience of the sector in government-related work may provide an excellent basis for further expansion in the welfare field. After reviewing historical developments, this section focuses on recent initiatives in community development, many of which are tagged as 'place management' or 'targeted area' programs. Most examples of such programs in Britain and Australia are neighbourhood level initiatives—usually in run-down areas—to reduce crime and vandalism by stimulating previously passive residents to support each other and reject anti-social behaviours (Latham 1998; Smith 1999; Adams & Hess 2001; Adamson, Dearden & Castle 2001; Rhodes et al. 2002). We also review more limited regional level initiatives in community development, and national proposals as contained in the McClure report.
- The way forward? Urgent issues of measurement and third sector governance: In this section we make suggestions in relation to two sets of issues on which progress appears urgently needed. One is the development of principles for measuring the integration of communities and the development of social capital. The other is community sector, or third sector governance. The reason for focusing on the first set of issues is that in modern government significant policy developments are most unlikely to be undertaken unless there is prior agreement on how to evaluate and measure progress. The reason for concern with the second set of issues is the observation that, if major policy responsibilities are to be transferred to the community sector, effective governance and accountability procedures need to be in place.
- 2.1 Community and community sectors
- 2.2 Regional economics
- 2.3 Some Australian policy implications
- 2.4 Neighbourhood studies
Historically a community has always meant a group of people who, at a minimum, shared a location, a geographical base. More recently, we have started to refer to communities of interest, as in the phrases 'the business community', 'the sporting community' and 'the criminal community'. A special type of community of interest is a virtual community, a community of people who interact via the Internet or some other software.
In this paper we are primarily concerned with geographical communities, which we take to be FaCS' responsibility. We also follow Stone (Stone & Hughes 2000; Stone 2001) and make a consistent distinction between the community at large and the community sector.
It is hard to give the community sector a neat definition. A pointer to the difficulty is the multiplicity of names by which it is known: the voluntary sector, the third sector (the private and public sectors being the first and second), the not-for-profit sector, and the non-government (NGO) sector. Several of these definitions say what the sector is not, rather than what it is, but all of them convey something of what is meant.
Rather than attempt a single neat definition, perhaps it is simplest and best to say that the segment of the community sector which government is most concerned to relate to and stimulate is the segment which engages in a great deal of pro-social, altruistic, caring activity. The activists and carers in this sector include many more-or-less unpaid volunteers, as well as (usually, and in all big organisations) paid staff. Funding comes in part from voluntary donations, although government contracts and subsidies are also key parts of the revenue base. Some of these organisations were founded as charities, and were so-called, although that term is now frowned on as denoting a top-down, noblesse oblige orientation. More recently, numerous other community groups have developed, pursuing a range of environmental and social goals. Major community sector organisations that clearly belong to the pro- social grouping include the Salvation Army, the Yooralla Society, the Brotherhood of St Laurence, Rotary Clubs, and country fire brigades.
Having indicated that part of the community sector to which government and government policy most closely relate, it perhaps matters little how we choose to classify borderline organisations. Many sporting organisations, clubs and hobby groups, for example, are clearly not either private, for profit organisations, nor do they belong to the government or public sector. They may also engage in pro- social, caring activity, though usually of a restricted kind. Somewhat more relevant, but borderline in terms of how the community sector is usually thought of, are organisations like universities and regional health networks. These receive most of their funding from government and are, in general, close to the public sector, although their employees do not fall under public service acts.
Communities defined by location
As we have noted, the most common definition of community is a group of people who have a common interest and in the simplest situation the common interest is defined in a spatial way—communities are groups of people who live close together. A reasonable question is: what forces explain people living together?
Economics provides one paradigm to explain the spatial forces that lead to the development of communities.2 While other explanations can influence development, the most enduring urban forms will be the result of primarily economic forces. By examining the spatial structure of populations, economic geography develops models that can explain existing geographic patterns by external economies, by increasing returns to scale, by transport costs and by the natural advantage of proximity to materials of value.3 In this framework, community has a derivative role—it is a result of economic forces rather than a shaper of them. Factors such as cultural, language, religious, social and governance groupings may modify the development of the economic geography and define the characteristics of a community, but they do not explain the existence of the community.4
The underlying notion of economic geography is that population clustering and the associated location-based communities essentially result from an economic imperative. While in history we may tell stories about the importance of religion, governance and topography for the development of communities, in contemporary societies the fundamental drivers are economic forces. Communities form because they are the most efficient way of organising activity and they grow because they obtain some economic advantage over alternative communities.
Community growth and development reflects the characteristics of spatial geography, the nature of the towns and cities and their interaction with one another. Regional economics provides tools to explore evaluate and explain communities.
Traditionally, economic analyses view individual actions as maximising personal utility subject to some economic constraint (Blaug 1992). It is not surprising then that the effects of location and an individual's environmental context share a relatively minor position in economic literature.5 However, spatial factors have received greater attention in particular economic sub-disciplines, especially regional economics. Regional economists have developed a set of analytic tools to explain the existence of towns and the structure of regions. The first role is useful in providing an historical context while the second is important for the development of contemporary policy.
Development of regions
Intrinsic to the development of regions is the growth of towns, and initial attention to regional economics was focused on urban growth and development. Von Thunen (1826) was one of the first to explain urban development. He postulated that there is a gradient of land rents surrounding a town and in making production decisions farmers trade-off land rent and transportation costs. This principle leads to the most intensive uses of highest price land closest to town. The same model has been applied to a city, substituting commuters for farmers and the central business district for the town centre. Householders trade-off house prices against transport costs.
Marshall (1920, p. 271) suggested that external economies (externalities) provide a rationale for producers locating together. He nominated three examples: knowledge spillovers, the deepening and broadening of markets allowing for specialisation, and the backward and forward linkages brought about by a large(r) pool of local labour.
Note there is a tension between external economies associated with geographic concentrations of industry (for instance, commuting costs and knowledge spillovers) and diseconomies (such as congestion). While diseconomies of size are fairly constant, external economies vary according to the natural advantage of the city, leading to the possibility of variations in optimal city sizes.
A more atheoretic approach derives from the central-place notion of Christaller (1966) who noted that rational consumers on a uniform plain will minimise their travel time. In such situations regional growth will develop a lattice like structure with urban centres of different size in a hierarchy—a large number of market towns will be arranged around a larger administrative centre, which is in turn focused around a regional capital, and so on. An extension of the Christaller notion proposes urban centres defined by the purchasing power of consumers rather than distance.6 The central place idea and its variants provides a description of the (observed) spatial form of a population but it does not show how or why this form emerged.
North (1955) was among the first to tackle an explanation. He suggested the 'export base' approach, which is concerned with the induced effects on a region deriving from the so-called basic industries, which export their output. Demand by workers in basic industries provides the motivation for the development of local industries. However, the approach does not explain what induces the growth of the basic industries in the first place. Nevertheless, the economic base-multiplier model gives four insights: (a) the interaction between economies of scale and endogenous market size can lead to a cumulative process of agglomeration; (b) it is essential to study not only static equilibria but also dynamics; (c) dynamics involve the possibility of discontinuous change; and (d) change may not be symmetric as between growth and decline.
Externalities and increasing returns
Tiebout (1956) designed a public finance model of local government expenditures that emphasised the importance of both mobility between local government areas and the 'external economies' prevalent in the provision of public goods. His contribution to the literature comes from his acknowledgement of the unique nature of local government expenditure due to its narrow spatial sphere. Hence, externalities play a greater role in the distribution of local government expenditures than other public funding.
In the 1980s and 1990s, location played an increasingly important role in other distinct areas of research. The new growth and new trade literature of, among others, Paul Romer and Paul Krugman emphasised the role of location and its relation to increasing returns (Romer 1986; Krugman 1991). That the advantages of location can lead to increasing returns has important implications for economic growth, trade and industry analysts alike. Krugman's (1991) core-periphery model is a prime example of the role of location in the development of industries that, in turn, attempts to explain much of the distribution of production in the United States (US).
New growth economists, Krugman's core-periphery model, and even Jacobs' theoretical explanation of the growth of cities all focus upon the importance of location (Jacobs 1961). Each details the importance of shared resources, infrastructure development (and the consequent drop in transportation costs), location-specific demand and the role of increasing returns. Increasing returns derive from many of the above factors but also from externalities—knowledge or technological spillovers. The notion of spillovers between individuals is also central to research on neighbourhood externalities, which is discussed shortly.
Government and regional policy
The basic rationale for state and federal government interest in regional economics is so public policy can take account of problems peculiar to the regional level. If regional information is incorporated into the policy process, it is more likely that resulting policies will both achieve their goals and be consistent with regional variations. By contrast, local governments are unlikely to be concerned with achieving consistency within regional economies, but will generally focus solely on benefits and costs to their own locality.
Applying these notions, we might postulate that the role of economics in relation to community development lies in three separate but closely associated discipline areas. These are: community economic analysis that is primarily concerned with aggregate measures of community performance; business management measures that are concerned with evaluating particular development proposals for communities; and public sector evaluation methods.
The three strands are differentiated on the basis of purpose. Community economic analysis has the primary role of describing a community as a whole, its make-up, linkages and the connections with other communities. It may also be concerned with the pattern of communities within some larger entity like a region or nation. The business management component is concerned with monitoring and guiding the performance of particular industries. It assumes a conventional profit-motivated private sector with limited altruistic concern.
Public sector evaluation methods are useful where the government is the provider of economic activity, a regulator of activity, or a potential investor in such activity. Government activity is warranted where there are external benefits resulting from activities that cannot be captured by private firms and typically involves the provision of infrastructure. Thus, this latter area covers proposals for development that are likely to require or receive public funding.
Regional economists have worked in Australia and elsewhere to analyse the comparative advantage of regions. In the US, economists have contributed to and analysed the development of research triangles and other developments that make use of synergies between business needs, university research and government policy in particular regions (Porter 1990; Saxenian 1994).
A key issue of policy concern is the viability of small rural towns. There are several factors driving rural depopulation. These include: (a) technological improvements in agricultural production and transport, (b) economies of scale and scope in agriculture, (c) decreasing returns to agriculture, and (d) the fact that most agricultural products are inferior goods. A major driver of rural depopulation is technological innovation that leads to economies of scale so that fewer people are required to make a given amount of product. The inferiority of agricultural based goods means that as incomes rise demand for them (or the agricultural component of products) rises more slowly leading to a fall in the relative importance of agricultural production and a fall in relative returns.
The issue of rural depopulation is sometimes confused because, while small country towns may decline, regional centres may grow since the first port of call for departing townspeople is larger regional centres. A focus on regional centres misses the decline in the country towns.
Coping with rural and regional growth and decline is among the most contentious of policy issues, as the approach has to involve choice among places. Not all places will be able to participate in growth in the long term. Rural regions need public policies that will allow rural residents to take advantage of the improved mobility and opportunities offered by innovation in transport and telecommunications, so that their social allegiance and group identity can be transferred to larger places. This would allow the refashioning of the established social organisation in a way, and at a pace, that does not destroy the essence of most local communities.
Beer, Bolam and Maude (1994) examined growth and development of non-capital urban centres. These were defined as cities of more than 10 000 people that were not state or federal capitals. They noted that the pattern of urban development in Australia was the product of its geography, economic structure and the historical period in which it was colonised by Europeans. When compared with other nations in the developed world there are fewer non-metropolitan centres in Australia and a smaller proportion of the population lives in them. Urban development in Australia took place independently on six separate sites, with economic activity geared to production for export markets. The combination of a mercantile economy, strong central administration by the colonial authorities, a nodal, centralised transport system and the fact that the capitals were also the major ports gave rise to a pattern in which the metropolis dominated.
Economic restructuring and population change have expanded the role and significance of non-capital cities within the national economy. They are contributing to the erosion of the existing state-based urban systems and the development of a national urban system. In the past, non-capital cities were principally centres for agricultural and pastoral regions; large and sometimes prosperous mining towns; manufacturing centres for highly protected industries; and key nodes within the transport system. Important changes in recent times include a strengthening of the role in the provision of community and other services; rapid growth in tourism, recreation and retirement industries; increased national prominence in manufacturing; and further development of centres for wholesaling and retailing.
A second major policy concern is the regulation of housing and town planning. Community is determined in part by urban form. The location, nature, proximity and environment created by the housing stock may enhance or discourage communal relationships. Living on a highway is not congenial to talking to one's neighbour. Modern housing estates are designed as cul-de-sacs to provide communal open space that encourages interaction, limits traffic flow and provides a high degree of safety. Modern apartments have clearly defined communal living spaces for interaction. Australian housing form has been unique among developed countries in its penchant for detached single family/household dwellings, though it is not clear whether this facilitates or hinders development of community (Winter & Stone 1999; Watling 1999).
Characteristics of communities
Maher (1999) noted a trend towards decentralisation, which he attributed to a changing role of cities from agglomerations involved primarily in industrial production to those concerned with information. He argued that this is having a substantial impact on both the nature and location of work, and on growing social polarisation. The informational economy takes place on a much greater spatial scale than the industrial economy, and a growing number of functions no longer need to be centrally located.
Baum et al. (1999) investigated the characteristics of communities across metropolitan and urban Australia and how they reflect the forces of change. Their research involved identifying the dimensions of expansion and decline in communities, and describing and accounting for the diversity of community outcomes. Communities were differentiated and classified according to a battery of dynamic and static measures, including labour force engagement and disengagement, the structure of industry and occupations, levels of human capital and incomes, and the concentration of advantage and disadvantage.
Communities of opportunity were characterised by high employment growth, above average reduction in unemployment, positive growth in incomes, greater concentrations of workers in expanding occupations and industries, commensurate levels of human capital, greater concentrations of high-income families, greater labour force attachment, low concentrations of disadvantaged families, low concentrations of public housing and fewer families facing housing financial stress. Communities of vulnerability had the opposite features to those listed above.
Baum et al. (1999) noted that it was the levels of human capital, household income, engagement in and disengagement from work, and involvement of labour in industry sectors and occupations linked to the new services and information economy that explained most of the variation across SLAs.7
In interpreting the results for policy making they distinguished between policies that focus on the characteristics of people, and those that focus on the characteristics of places (or communities). They claimed that, although some of the findings strengthened the emphasis on people prosperity (as opposed to place prosperity), the distinctive geography of the nation's vulnerable communities showed that place was still a powerful determinant of social and economic outcomes. Some places had developed people prosperity better than others. This was shown by the distinct spatial concentration of some of the clusters of community opportunity, including the transformation of inner-city suburbs, the location of specific remote area extractive-based towns, and the new regional-tourism-based communities of opportunity.
They stated that a people prosperity policy alone would be unlikely to be effective unless there was some hope for people to reach out to, and become engaged in, those emerging parts of the economy that have global and national linkages; and a place-only strategy would not be very effective in locations with little chance of engagement with major growth forces operating in the economy. They concluded that it might be worthwhile scrutinising communities of opportunity to see if their successes are related just to chance and good luck or to strategic development planning and action, including intangible factors such as leadership and institutional capacity and flexibility, as well as competitive factors such as their resource base and infrastructure.
The findings of Baum et al. are supported by Bray (2000). Bray uses ten classifications of location type based on population and the relationship to the state capital to explore the variation in performance between types and across states. He reviews and ranks performance in relation to demography, human resources, labour force characteristics, income, source of income, housing and services.
Bray concludes that there are many diverse contributing factors to the location of development. Consequently, generic 'special area assistance' is unlikely to be successful. Rather, assistance should be focused on the particular circumstance of a location and of the needs of those who live there.
The central idea of neighbourhood studies is that the characteristics and nature of the localities in which people live generate external effects or externalities on their life chances. Most work on neighbourhood effects has been done in the US, particularly in ethnic ghettoes where one might expect the effects to be particularly strong. Australian neighbourhoods are almost certainly more heterogeneous, with lower concentrations of extreme deprivation and low social capital (but see Vinson 1999). However, it seems reasonable to conclude that sufficient Australian research has been done to infer that neighbourhood effects here are significant both for educational aspirations and employment.
Jacobs combined multiple theoretical aspects into her analyses of the economies of cities. She used sociological, educational, geographical, evolutionary and economic theory (Jacobs 1961, 1984, 2000). Her 1961 publication of The Death and Life of Great American Cities was a response to what she saw as a decline in the socio-economic climate of many American cities. At the core of this demise were mistakes made by urban planners who failed to understand many of the basics of location theory. She also made the link between urban planning and the importance of information transfers and, while not putting it in these terms, she also alluded to the importance of positive demonstration effects in declining neighbourhoods.
A number of researchers have taken elements of the new growth literature (especially the focus on increasing returns) and combined it with analyses of cities that grew from Jacobs' research. Macrosocial and representative agent models are used to describe the distribution of production and human capital formation within cities and the resultant ramifications for inequality (Benabou 1993, 1994; Durlauf 1994, 1995). These models start with the realistic assumption that residential income segregation exists due to differences in individuals' and families' budget constraints. This residential segregation leads to distinct differences in human capital externalities between neighbourhoods, as individuals' and families' budget constraints are a function of their human capital. Lucas (1988, p. 19) emphasised the role of human capital externalities in human capital formation: 'human capital accumulation is a social activity, involving groups of people in a way that has no counterpart in the accumulation of physical capital' (Lucas 1988, p. 19).
Human capital externalities operate within neighbourhoods and affect the cost of individuals' human capital accumulation. The cost of education for individuals in neighbourhoods with large positive human capital externalities is lower than in neighbourhoods with low human capital levels. Therefore, the long-term interaction between residential segregation, neighbourhood human capital externalities and human capital formation creates increasing stratification over time. In the absence of other factors this would lead to ever-rising inequality. This strand of research provides the broader framework for analyses of how neighbourhood externalities affect decision-making, emphasising the long-term costs to both efficiency and equality of differences in neighbourhood externalities.
Kain (1968) suggested the spatial mismatch hypothesis to describe the inability of the supply of labour to adjust to shifts in labour demand. The hypothesis focused on the role of location in economic outcomes, yet did not touch on the role of spillovers or transfers between individuals or groups of individuals.8 It identified geography as the explanatory variable in the disequilibrium in the supply and demand of black employment.
However, if labour was perfectly mobile, geographic pockets of inequality would be transitory and the notion of ghettos would be unsupportable. That the spatial mismatch hypothesis is still a subject of analysis indicates that either labour or particular types of labour are relatively immobile or that there is a steady inflow of economically disadvantaged people and an outflow of economically advantaged people from particular areas that gives the impression of an area remaining constantly economically disadvantaged. The pertinent issue in the spatial mismatch hypothesis lies in identifying patterns of labour mobility in response to changes in the demand for labour that cause pockets of inequality. A wealth of literature in a variety of disciplines analyses population mobility patterns. Economic analyses vary but a large number measure mobility in response to labour demand shocks.
While differing in their estimated rates of adjustment, Bartik (1991) and Blanchard and Katz (1992) found high levels of mobility in response to demand shocks in the US labour market. However, both these studies use aggregate data and cannot therefore be interpreted as refuting Kain's hypothesis that focuses on black workers. Other investigations have identified the immobility of the most disadvantaged segments of the population following labour demand shocks.
Declining demand in distinct industries will particularly hit disadvantaged segments of the population if their employment is concentrated in these industries. Gregory and Hunter (1995) identify declining manufacturing employment as a determinant of increasing spatial inequality in the distribution of jobs in Australia in the 1980s. The adverse effects on black workers of declining demand have been well documented in the US. Johnson and Oliver (1992) and Acs and Danziger (1993) found blacks suffered disproportionately from a declining American manufacturing sector. Juhn (1994) did not distinguish by race but found significant adverse effects on less educated workers of changes in the American industrial structure. Industries particularly susceptible to cyclical movements in the macroeconomy may also leave segments of the labour market disadvantaged when an economic downturn is particularly harsh. Regardless of industry concentration, if there is less mobility among the less skilled, spatial inequality will rise following declining employment levels in a given area (Topel 1986).
Bound and Holzer (2000) found shifts in local labour demand explained much of the regional variation in employment levels in the US in the 1980s. The effects of these shifts have had the greatest impact on the wages and employment of the less educated, least experienced and on black workers. Therefore, in the US the immobility of less advantaged workers contributed to the rise in spatial inequality and 'appear[s] to have contributed importantly to the relatively greater deterioration of their employment and earnings in declining areas in the 1980's' (Bound & Holzer 2000, p. 23).
There is also the issue of ethnic groups remaining in particular neighbourhoods. This has been well documented in both the sociological and economic literature (Kasarda 1989; Borjas 1995). Both of these investigations identify the relative mobility of particular ethnic communities.9 Yet, this immobility is not presented as evidence of the spatial mismatch hypothesis but of the concentration effects that facilitate the transfer of neighbourhood externalities. The spatial mismatch hypothesis identifies the impact of changes in the demand for labour on geographic inequality. Research on neighbourhood externalities augments this work by analysing the influence of geographic inequality on individuals' socio-economic outcomes.
In the context of regional analysis, recognition of the importance of externalities dates back to Tiebout (1956). However, the first empirical analysis of the influence of neighbourhood spillovers in the mainstream economic literature appeared in 1982 and was concerned with youths' labour market and educational outcomes (Datcher 1982). Datcher introduced neighbourhood externalities as an explanatory variable of youth outcomes. Significantly, the paper extended economic interpretations of location by moving away from the more strict supply and demand framework of the spatial mismatch hypothesis. Datcher cited the intergenerational consequences of sustained inequality and the negative effects of black people's community of origin as her reasoning for moving beyond this supply and demand analysis. To capture these effects, community of origin characteristics were used as independent variables whereas, previously, only spatial factors that identified specific regions or size of place of origin were included in economic analyses.
In models predicting the years of education, the log of 1978 hourly earnings, and the log of earnings in 1977, Datcher found all of the family variables had the expected signs and most were significant. Family income had no impact on the education of sons but raised hourly earnings for white sons and annual earnings for both races. Parental education had a positive impact upon sons' education, while the number of siblings lowered human capital investments. As would be expected, strong correlations existed among the community variables. Average neighbourhood income had a significant positive impact on education and income, while the percentage of the neighbourhood that was white was only significant for blacks. Undoubtedly, these last two findings were the most important of Datcher's results. It was the first time in economic analysis that socio-economic characteristics of neighbourhoods were found to influence economic and educational outcomes, and this was an important advance on the spatial mismatch hypothesis that had dominated debate of spatial inequality.
Datcher's findings have been criticised for suspected omitted variable bias. A lack of variables to fully capture family spillovers may have biased upward the magnitude of the reported neighbourhood effect. Paternal (or head of the student's household) occupation was not included, nor were parents' verbal scores. However, the small effect of parental aspirations found by Datcher indicates this bias is probably relatively small (Jencks & Mayer 1990). Many economic analyses of neighbourhood externalities attract this criticism. Neighbourhood effects are biased upwards if externalities emanating within the family are not properly captured. Conversely, family effects are biased upward if neighbourhood effects are not fully captured. However, capturing both types of effects remains an issue, exacerbated because researchers within both the economics and sociology disciplines traditionally incorporate and place emphasis on different family and neighbourhood characteristics.
Corcoran et al. (1990) also used PSID data but included a more complete set of parental characteristics to identify family spillovers. He included five neighbourhood characteristics, but only welfare receipt and the unemployment rate were found to influence outcomes (Corcoran et al. 1990).
George Borjas' research on neighbourhood externalities led to significant growth in the analysis of neighbourhood effects. He identified the transfer of skills across generations of migrant families in the US, but it was his incorporation of neighbourhood factors into intergenerational transfer that was most notable (Borjas 1994, 1995, 2001). Borjas used census data to identify segmentation by both ethnicity and income and showed the effects of neighbourhood factors increased with persistent segmentation.
His model framework follows Becker's (1964) human capital model where utility- maximising parents invest in the human capital of their children. Ethnic capital was included as an independent variable due to its externality effect on an individual's education and was found to be a significant determinant of an individual's outcomes.
Most studies investigating neighbourhood externalities use models similar to that of Borjas. A handful of studies have also attempted to measure neighbourhood effects by creating their own dataset. This possesses the advantage of creating a dataset with variables relevant to investigations of neighbourhood externalities. This is especially important when considering the problems of omitted variable bias in research that fails to fully capture individual, family and neighbourhood effects.
Case and Katz (1991) used a survey of 1200 youths in inner city Boston designed to assess the determining factors in the outcomes of youths in a city experiencing strong labour market growth. Information was obtained regarding youths' labour market experience, current living arrangements, family background, social contacts and neighbourhood characteristics, childbearing and marital history, schooling experiences, personal history, illegal activities, drug use and personal opinions and outlooks (Case & Katz 1991). The actions of family members and youths' peers were found to significantly influence analogous behaviour in youths. Outcomes analysed included criminal activity, drug and alcohol use, childbearing out of marriage, education and church attendance. The analysis was deliberately designed as an epidemic model so that contagion theories (emphasising the notion of 'like begets like') could be tested in both the family and neighbourhood contexts. The level of peer and/or parental criminal activity was found to have a greater effect on the probability of youth performing a criminal act than other family and peer characteristics. Yet the analysis did not show that neighbourhood externalities operate solely in a 'like begets like' manner. They did not investigate the 'role model' influence of adults within a neighbourhood.10 However, Case and Katz claimed their 'findings suggest that (neighbourhood) families may influence inner city youths directly through the actions of their children, or perhaps indirectly, with influence working from adult neighbour to parent, and from parent to child' (Case & Katz 1991, p. 24).
A number of recent Australian studies have sought to identify the existence of neighbourhood effects, demonstrate their association with various outcomes, and trace the direction of causality. Kelley and Lewis (2002) provide an overview of the issues in relation to youth employment outcomes. Drawing on many of the sources discussed above, they conclude that youth labour force activity is likely to be limited by neighbourhood effects.
Overman (2000) examined the extent to which the teenage dropout rates from full-time education were effected by neighbourhood characteristics. He tested neighbourhood effects at two levels, large (postcodes) and small (collectors' districts) using data from the Longitudinal Survey of Australian Youth (LSAY). Using the large definition of neighbourhoods only, Overman found that labour dropout rates were associated with the proportion of adults with vocational qualifications. This suggested two possible mechanisms, through local labour market demand or through social networks. In the former situation, high vocational qualifications might occur when local labour markets sought workers with low educational qualifications. Alternatively, environments in which there were high proportions of vocational qualifications might encourage low educational attainment. When variables representing the small definition of neighbourhood were added to the model, he found the probability of dropout rates was reduced with high concentrations of vocationally qualified adults, counteracting the effect of the variable representing the large definition of neighbourhood. The socio-economic status of the small neighbourhood also was important with higher status neighbourhoods. Overman concluded that the second mechanism was more likely; that is, local labour market conditions influence educational drop-out rates.
Starting with a cross-section from the LSAY, Andrews, Green and Mangan (2002) showed that neighbourhood effects had a significant effect on unemployment in both high and low income areas. They then used the panel nature of the LSAY to show important unobserved heterogeneity.11 When this was controlled for, neighbourhood effects remained important in low income neighbourhoods.
Jensen and Seltzer (2000) estimated the effects of individual, family and neighbourhood effects on youths' post-secondary education aspirations. They used a survey of Melbourne secondary school students and found significant neighbourhood effects in youths' educational aspirations.12 To measure the influence of neighbourhood externalities they compared the probability of youths with median individual and family characteristics aspiring to post-secondary education in different neighbourhoods. They found substantial variation between the educational expectations of youths living in low and high-unemployment neighbourhoods.
Most of the literature reviewed in this section has focused on either ghetto communities or highly disadvantaged inner-city neighbourhoods in the US. That significant neighbourhood effects were found in Melbourne, a city with less heterogeneous characteristics than American cities, indicates neighbourhood externalities are not specific to particular communities. They do not operate solely in neighbourhoods at either end of the wealth spectrum. Criticisms of an insufficient sample size could be levelled at Jensen and Seltzer's research.13 However, their findings are robust and the inclusion of both paternal and maternal characteristics is rare in economic analyses.
They were unable to identify the transmission mechanisms through which neighbourhood externalities affect educational aspirations, yet they did find evidence of the epidemic models described above. Youths were asked to report their peers' education plans and these were found to influence youths' aspirations, even when coupled with a neighbourhood characteristic in the regression model. While this was not a perfect measure of the influence of peers' aspirations, it did suggest promising avenues for further research comparing the interaction of peer and neighbourhood effects.
As yet, no consistent findings on the impact of neighbourhood externalities on youth outcomes exist. While some research has found neighbourhoods play a significantly large role in determining outcomes, other research has found the influence to be insignificant. Variation in estimation methods, the types of neighbourhood proxies and diagnostic testing have hindered progress towards commonly accepted heuristics in neighbourhood research.
Substantial variation exists in findings on the influence of neighbourhood externalities. Neighbourhood externalities have been found to significantly influence youths' educational outcomes (Crane 1991; Borjas 1995). However, among those that have found significant neighbourhood effects, their magnitudes have often been small and/or significant only among particular demographic groups (Datcher 1982; O'Regan & Quigley 1998). Several studies have found no significant neighbourhood effect (Brooks-Gunn et al. 1993; Plotnick & Hoffman 1999). Other studies have found the influence of youths' peers to play a more important role than neighbourhood effects upon youths' outcomes (Case & Katz 1991; Evans, Oates & Scwab 1992).
Inconsistency in the findings of previous research stems from problems in correctly specifying neighbourhood externalities. Previous research has been unable to identify the most appropriate proxy of these externalities (Jencks & Mayer 1990). A number of studies have found only particular neighbourhood proxies to significantly influence socio-economic outcomes (Corcoran et al. 1990; Duncan, Connell & Klebanov 1997; Plotnick & Hoffman 1999). Difficulties in estimating neighbourhood externalities can occur if neighbourhood proxies are incomplete or endogenous to family characteristics, or if unmeasured family characteristics affect both youths' outcomes and their residential location or choice of school. Recent research suggests that neighbourhood proxies that measure characteristics analogous or proximate to the dependant variable may better capture neighbourhood externalities. For example, a neighbourhood proxy that measures levels of education within a neighbourhood may provide a more accurate estimation of the influence of neighbourhood externalities on the education of youths living in that neighbourhood (Ginther et al. 2000; Manski 1990). Manski (1993) explored difficulties in identifying neighbourhood externalities in his analysis of what he termed the 'reflection problem'. This occurs with analysis of 'whether the average behaviour in some group influences the behaviour of individuals that comprise the group' (Manski 1993, p. 532). Critical to overcoming this problem is the relationship between sets of independent variables and the outcome variable. Neighbourhood characteristics that are more related, or what Ginther et al. term 'proximate', to the outcome variable may overcome potential bias incurred with the reflection problem. However, this may vary depending on the extent to which family background characteristics are included in the model estimation (Ginther, Haveman & Wolfe 2000).
- 3.1 Social capital: competing definitions
- 3.2 What produces social capital?
- 3.3 The claimed benefits of social capital
- 3.4 Australian social capital
Social capital is a currently fashionable term—a fashionable metaphor really—and may well be a useful one. It is largely a rebadging of what sociologists previously called social networks. Measures of social capital (discussed below) are quite largely measures of linkage, of networks. The rebadging is intended, in part, to lend importance to networks/social capital by associating them with physical capital and human capital as 'stocks' which deserve public policy attention and in which society should 'invest'.
Definitions of social capital range from the relatively narrow to the broad and have been much debated. Some economists, including two Nobel Laureates, have objected to use of the term on the grounds that it suggests a misleading analogy with physical capital (Arrow 1999; Solow 1999). Investment in physical capital, Arrow points out, always involves foregoing current benefits for future gains, and physical capital can be sold or otherwise transferred to others. Neither is necessarily true of social capital. Contrary to Arrow's view it could reasonably be argued that the same points could be made about differences between physical capital and human capital. Human capital like social capital is not lost if transferred to someone else (e.g. by education) and may even be augmented by use (Ostrom 1999; Caveye 2000).
A very narrow definition of social capital is given in a recent review of the literature by the American economic theorist, Joel Sobel (2002), who based his definition on that of one of the earliest writers in the field, the French sociologist, Raymond Bourdieu (1986). Sobel wrote that 'social capital describes circumstances in which individuals can use membership in groups and networks to secure benefits'. Notice that this definition makes social capital an attribute of individuals (albeit in relation to others; that is, in a social context) and that it says nothing about the frequency or quality of interactions among group or network members. That is, it says nothing about rates of participation or degrees of trust and reciprocity. An advantage of this narrow definition is that it concentrates solely on structure (groups, networks) and leaves open for empirical investigation the causes and consequences, costs and benefits of social capital. A possible disadvantage is that the definition does not point towards a sensible way of determining whether an individual's social capital is going up or down, unless one assumes that if a person joins more groups and networks they will automatically gain more benefits.
Partly on these grounds, Stone (2001) prefers a somewhat broader definition. She distinguishes between the 'structure' and 'content' of social capital. By 'structure' she means networks measured on various dimensions including size, density (membership as a percentage of eligibles, e.g. standard measures of trade union density), and whether they are relatively open or closed to non-members and other groups. By 'content' she means mainly the prevalence of relations of trust and reciprocity within the structures. Using this definition, we infer that social capital can be thought of as 'structure multiplied by content'; that is, an individual has more social capital the more structures they belong to and the greater the degree of trust and reciprocity within those structures. Notice that this definition also leaves open questions about the causes and consequences of networks, and whether particular networks are beneficial or harmful to society as a whole (e.g. from an economic efficiency standpoint).
In some of his work Putnam seeks to deal with this last issue by equating social capital to what he calls 'civic community' or 'civic trust' (Putnam 1993, 2000; Helliwell & Putnam 1999). In this formulation, a distinction is made between 'bridging' organisations—organisations that cooperate with and build bridges to other organisations—and 'bonding' organisations whose members seek only to benefit each other. This formulation was explicitly a response to critics who pointed out 'the downside of social capital'; the near-certainty that some organisations and networks have negative social consequences (e.g. criminal networks and perhaps 'old boy' networks), and that most organisations compete with others in their field, so that gains to one are losses to others (Portes & Landolt 1996). Against Putnam's revised formulation, it can be argued that all organisations have bonding aspects; they cannot be classified straightforwardly as bonding or bridging. Further, defining out of social capital those networks and organisations which only look after their own members may have the effect of discouraging research on them, and so limit our ability to accurately assess the net benefits of all networks. This last task should, logically speaking, involve subtracting the harm done by some from the benefits produced by others.
Another definitional issue is whether it is preferable to view social capital as a property of individuals or collectivities, including states or society as a whole. Our own preference is to begin by defining and measuring social capital as a property of individuals (see section 5). This has the advantage of allowing conventional economic analysis in which individuals are assumed to join networks with a view to gaining market and non-market benefits (Glaeser, Laibson & Sacerdote 2000). Individual social capital may generate positive or negative externalities that must be taken into account in assessing overall social benefits and costs. Put another way, we think it best to begin by treating social capital as a stock possessed by individuals and then investigate the flows and outcomes resulting from use of this stock. In section 5, we will go on to suggest reasons why it can also be useful to aggregate the stocks of individuals to provide measures of the stocks of (say) a local community, region or nation. Only by doing so can we assess whether society's total social capital is increasing or decreasing. But it still remains an empirical question whether existing or additional stocks have beneficial or harmful social, economic and political consequences. We should never simply assume the more social capital, the better.
This section discusses the sources or antecedents of social capital. The argument is that trust and reciprocity develop quite 'naturally' in stable families and in ethnically homogeneous groups, especially immigrant groups. It is also clear that better educated people are more likely than the less educated to participate in political and other civic activities. Societies and localities that have high and relatively equal incomes are also characterised by high levels of social capital. But what if a particular society or locality lacks social capital? Some writers appear to believe that trust and reciprocity can only be built up over long periods—perhaps hundreds of years—while others have documented instances where working social capital appears to have been developed within a generation. The last part of the section deals with the time it takes to develop social capital.
Families are often said to be the foundation of social capital (Putnam 2000; Winter 2000). Winter (2000) suggests important linkages between family life and social capital. On the one hand, particular aspects of a secure family life, including the development of self-esteem and trust, appear to be associated with the generation of social capital in civil society. On the other hand, where family life is characterised by 'familism' (the elevation of family and kinship ties above other social types of obligation), then social capital within civil society is limited.
Putnam (2000) assessed the impact on social capital of the breakdown of extended families in the US—the so-called atomisation of society into nuclear families. He also took account of the increasing numbers of single person households and couples without children. His conclusion was that these changes account for no part of the decline of social capital in the US during the last thirty years. The main contributing factor was the difference between generations. People born in the 1920s were the last highly participatory generation. Since then there has been a decline in civic-mindedness. The post-war baby-boomers who came of age in the 1960s appear to have become disillusioned with public institutions and to have adopted a privatised lifestyle. Later generations are much the same. Television watching and, more generally, the substitution of passive entertainment for more active involvements is a secondary factor explaining the decline in civic activity.
Researchers at the Australian Institute of Family Studies (AIFS) have made a strong case that in this country a stable family life has been a key underpinning of civic activity (Winter 2000a, 2000b; Stone 2001). Stone (2001) has outlined a continuing AIFS project to investigate the role of families in the development and sustainability of strong communities, and to identify family circumstances associated with active economic, community and political life.
It is well established that well educated people are considerably more likely to participate in civic activities than others. Many of the early studies focused on political activity and showed that lawyers, teachers and other occupational groups with high educational levels, and also communication skills and work patterns which quite readily permitted some attention to political activity, were most likely to get ahead in politics (Lane 1959; Milbrath 1965; Verba & Nye 1972). Working class people, by contrast, only get into politics in small numbers and mainly then if sponsored by a trade union. More generally, volunteering is strongly associated with educational attainment (Verba & Nye 1972).
Ethnic homogeneity and immigrants
American research, initiated by Jane Jacobs (1961) and Herbert Gans (1967), showed that ethnically homogenous groups, especially in immigrant enclaves or ghettoes, often form tight social groups whose members assist each other when they are in business and politics, and when family and health crises occur. Gans referred to these areas as 'ghettoes of hope'. The fact that people are living together in close proximity, struggling in a new country but somewhat optimistic, generates the motivation for cooperation and reciprocity. By contrast, in ethnically diverse localities it has been found that networks are poor and civic participation is low (Knack & Keefer 1997; Alesina & La Ferrara 2000).
Laboratory experiments and also survey research have confirmed that, when individuals are socially similar, both trust and perceived trustworthiness tend to be high (Glaeser et al. 2000). However, trust and trustworthiness decline when experimental partners are of different race or nationality.
Research in areas where very low income blacks live in the US has indicated that when people are no longer reasonably optimistic about the future social networks are weak (Wilson 1987). Furthermore, because ethnic groups tend to form bonding not bridging social capital, there are limitations on how far their members tend to progress in their businesses and careers.
There is fairly strong evidence, both at national and local levels, that rates of civic participation are higher in places with higher and also more equal incomes (Knack & Keefer 1997; Alesina & La Ferrara 2000). Milner (1994) reports that participation— particularly political participation—is highest of all in the West's most egalitarian countries: Sweden, Norway, Denmark and the Netherlands. So it appears that the economic homogeneity of an area (like ethnic homogeneity) is conducive to development of trust and participation.
Can social capital be created quickly?
The political science literature on the relationship between political culture and democracy certainly implied that norms of trust and participation take many decades to develop, and that their existence or absence is a major explanation for stable democracy in some Western countries but not others (Almond & Verba 1963). Putnam, in his account of developments in Northern and Southern Italy, also seems to imply that a 'civic culture' can only develop slowly and 'naturally', and cannot be deliberately constructed (see also Fukuyama 1995).
There is, however, a quite different area of research—'new institutionalism' in political science and economics—that suggests that in the right institutional conditions trust and reciprocity can develop quickly. Elinor Ostrom (1990, 1999) has studied the governance or management of common property resources (e.g. irrigation projects) in developing countries and tried to discover how farmers in some areas have succeeded in devising and adhering to arrangements which enable them to cooperate in maintaining the common property in good working order and share its output (see also Sobel 2002). She uses the theory of repeated games to show how, in some instances, farmers were able to construct generally understood and easily enforced rules of behaviour, with minimal use of sanctions. Progress in making cooperative arrangements was rapid in some areas. However, it was not clear that external public policy intervention could have helped, in part because the collective-action rules the farmers devised appeared to depend heavily on their own assessment of local interests which might not have been readily observed by outsiders.
Another indication of how collective-action rules might be devised and enforced, even where there is little pre-existing trust, comes from experiments comparing the behaviour of Japanese and American subjects (research by Yamagishi and colleagues, reviewed by Sobel 2002). Prior to the experiments, which involved subjects being given monetary rewards for cooperative acts, the Americans reported higher levels of general trust—and that is the way they initially behaved. However, in later experiments subjects were encouraged to develop their own group rules that could involve monitoring and penalties for non-cooperative behaviour. The Japanese subjects devised effective rules and succeeded in increasing their collective rewards. The experiments may be taken as illustrating the possibility that well constructed institutional arrangements can promote cooperative behaviour even among people who do not share norms of trust.
Advocates of the importance and value of social capital point to a lengthy list of improvements in human welfare that allegedly flow from social capital. The work of the American political scientist, Robert D. Putnam is central (1993, 2000). Putnam's most famous claim in Making Democracy Work (1993) is that high levels of social capital are a prerequisite for pluralist democracy. This is a claim originally based on comparing Northern and Southern Italy. In the North, local government and democracy flourish. In the South they are stillborn. As noted above, Putnam defines social capital mainly in terms of trust and participation. He examines the history of the two parts of Italy and finds that in the North numerous flourishing community organisations, with large actively participating memberships, existed in mediaeval times and certainly in the nineteenth century before political democracy came to Italy. There was also evidence of trust among fellow citizens and trust in some public and community organisations. In the South, by contrast, people trusted only family members. Public organisations, in so far as they existed, were regarded as self-interested, often corrupt, certainly untrustworthy. So the North had the social capital basis for democracy and the South did not.
In a later book, Bowling Alone (2000), Putnam traces the vicissitudes—mainly decline in recent years—of community groups and social capital in the US. Bowling Alone provides a comprehensive summary of recent trends in all manner of social indicators of civil life in the contemporary US. The book offers a much-expanded list of the claimed benefits of social capital and suggests a wide range of remedies for curing what Putnam sees as the current social malaise. The claimed benefits arise in relation to:
- pluralist democracy
- physical health
- happiness (self-rated) and mental health
- public safety, vandalism and crime
- economic performance and efficiency.
Social capital and pluralist democracy
The first major studies that traced the linkages between social capital and pluralist democracy (although they did not use the term social capital) were conducted in France (Crozier 1964), Italy (Banfield 1958), the US (Lane 1959; Milbrath 1965; Verba & Nye 1972) and cross-nationally in the US, Germany, Italy, Britain and Mexico (Almond & Verba 1963). They all found positive relationships between interpersonal trust, participation in community groups, and participation in and support for democratic parties and politics. However, questions were raised about causal direction—did trust and community participation cause political participation, or was it the other way round? Putnam's Northern versus Southern Italy comparison was crucial in helping to resolve these questions by showing the historical sequence. More recent cross-national studies and detailed analyses linking community to political participation in the 50 American states have confirmed that social capital (defined as trust and voluntary participation in community activities) is causally related to democratic political participation (Barnes, Kaase et al. 1979; Putnam 2000).
People with rich social networks/capital appear to suffer less illness and live longer than people with poor networks. Further, it is clear that the relationship is not spurious. Several longitudinal studies in the US, Japan and Scandinavia have shown that the quality of people's social networks measured at an earlier point in time predicts morbidity and mortality in later years (Putnam 2000). People who have poor networks are between two and five times more likely to die, compared with matched individuals (same income, education etc.) who have high quality family and social networks (Berkman & Glass 2000). Joining a community group (if at present you belong to none) reduces your risk of dying in the next year by about the same amount as giving up smoking (and might be easier to do), but visiting friends or talking with them on the phone would be equally good (Berkman & Syme 1979; House, Robbins & Metzner 1982; Blazer 1982; Orth-Gomer & Johnson 1987; Welin et al. 1985). People with rich social ties are less likely even to catch colds (Cohen et al. 1997).
Good social networks are particularly important for the health of older people (Seeman 1996), but the benefits to children's health are also clear. In the US, poverty is, as one would expect, the best predictor of low birth weight and of childhood morbidity and mortality, but social capital appears to be the second most important factor in analyses in which other variables implicated in ill-health (including, of course, poverty) are controlled for (Putnam 2000).
Happiness and mental health
In the last 25 years or so there has been a great deal of research on self-rated happiness and life satisfaction (the ratings are typically given on longish scales—for example 0-10 scales with end points labelled 0 = very dissatisfied and 10 = very satisfied). There are fairly strong bivariate correlations between social network/ social capital measures and happiness; correlations in the range 0.20 to 0.35 (Diener 1984; Headey & Wearing 1992; Veenhoven 1996). However, the quality of people's networks depends quite strongly on the stable personality trait of extraversion, which is also (and more strongly) associated with happiness. So the picture seems to be that extraverts are happier anyway and just a bit more so if they have diverse social networks. More introverted people are less happy and less inclined to join groups anyway, but if they do so their wellbeing is enhanced (Headey & Wearing 1992).
There is a considerable body of research which claims that good social networks improve mental health and, more specifically, that they reduce anxiety and depression, and buffer or reduce the impact of adverse life events and experiences (experiences like divorce and unemployment) on mental health. Certainly, there are numerous correlational studies and studies using recall data which show positive relationships between social networks—especially the availability of a close, intimate partner or friend—and better mental health (House et al. 1982; Pearlin & Schooler 1978; Thoits 1983). However, the correlational studies plainly do not demonstrate causation, and studies based on recall are open to objections of 'retrospective bias'. That is, survey respondents who have in fact coped better than others with adverse life events and who report lower levels of anxiety and depression may also, at time of interview, report better social networks than respondents who are anxious or depressed and have coped less well with adversity. Reports on current or past social networks could well be biased by one's current mental state.
To counter these objections, or rather to investigate their force, a small number of studies have employed a prospective (rather than a retrospective or recall) design. Survey respondents are interviewed at least twice. When they are first interviewed, baseline levels of social support and mental health are established. Then at the second (and subsequent) interviews they are asked to record the life events and experiences that have happened to them in the meantime. Also, their current (new) levels of mental health are recorded. Researchers can then assess whether respondents with better baseline social networks show smaller declines in mental health (or perhaps no decline at all) in the face of adverse experiences than respondents with poorer baseline networks.
The few studies that have employed this stronger research design, including an excellent Australian study conducted at Australian National University (ANU) (Henderson et al. 1981), show that, net of other factors, good social networks either do not, or only slightly, reduce (buffer) the impact of adverse events and experiences on mental health. In the ANU study a statistical model was developed in which baseline social network measures were appropriately viewed as partly consequences of personality traits (including extraversion) and their impact in buffering the impact of adverse events (like divorce, death in the family and job loss) was assessed. The buffering effect was not substantial. (It is relevant to note that if one looked only at the cross-sectional correlations, which can of course be computed from the data, one would mistakenly infer that the quality of social networks was associated with better mental health). So our conclusion in reviewing this research—a conclusion that some would reject—is that the best available evidence does not support the view that good social networks are much help in coping with adversity.
Public safety, vandalism and crime
Rates of crime, gang fights and vandalism have long been known to be higher in neighbourhoods characterised by 'social disorganisation'—more or less the opposite of social capital (Sampson 1995). Further, criminologists have shown, and modern police work usually accepts, that catching criminals depends much less on sophisticated detective work than on neighbours being concerned for each other and informing on criminals whom they have seen or just suspect to be at work (Moore 1978). Concerned, vigilant neighbours are mainly found in areas with high overall levels of social capital. Several American police chiefs have become famous for trying to stimulate social capital (as well as good police-public relations) and then make use of it to improve both crime prevention and arrest rates (Sparrow, Moore & Kennedy 1990).
In trying to understand teenage vandalism and violence, sociologists from the 1920s onwards have pointed to the increased influence of macho, peer group influence in localities where adults are relatively anomic, do not exert effective influence and certainly do not join organisations. In these localities youths tend to drop out of school, or leave as soon as it is legal, and may hang around in gangs (Sampson 1995).
The same debate about correlation versus causation occurs in relation to social capital and public safety as occurs in the other fields reviewed in this section. It is pointed out by critics that the statistical association between low social capital and high vandalism and crime could be just due to the tendency of similar types of people to congregate together in particular neighbourhoods. So it is possible that unsociable or just plain anti-social people choose to congregate together in the same areas and have no inclination to join either formal or informal groups. They and their teenage children also tend to engage in a range of criminal, semi- criminal and aggressive behaviours. Perhaps if the same people were transferred to high social capital areas, they would behave just the same. Several studies have attempted to investigate the force of this critique by comparing the lives of youths who stay in low social capital neighbourhoods with those who move out to higher social capital neighbourhoods.14 The evidence is fairly strong that there are substantial 'neighbourhood effects' and that the movers' propensity to violence and crime is reduced. They are also more likely to get a tertiary education. A severe critic might still object that no controlled social experiment has been done, and high-risk adults and youths who move to an area with fairly high social capital have at minimum consented to move, while others might have flatly refused and virtually barricaded themselves in.
Overall, the evidence on the beneficial effects of social capital for public safety is quite convincing, and has in fact convinced policy makers, police and other influential people in a number of places in Australia, Britain and the US. In a later section of this report we review a range of specific policy programs in these countries to improve social capital mainly with the aim of reducing crime and vandalism in run-down neighbourhoods.
As noted above, there is no recognised field of community economics per se. There is, however, a burgeoning literature concerned with the economic consequences of social capital. Much of this literature is concerned with the inclusion of social capital as a factor of production in conventional economic models and applying tests to see how well it explains, is associated with, or is determined by economic phenomena (Schiff 1998; Alesina & La Ferrara 2000; Knack & Keefer 1997; Knack 2001; Helliwell & Putnam 1999).
There is really little doubt that individuals with high levels of social capital and rich networks benefit economically. Trust and good networks plainly reduce the transaction costs of doing business. One can get information more quickly, and less time needs to be spent drawing up watertight contracts and guarding one's back legally (for a detailed review, see Knack & Keefer 1997). Much more contentious is the issue of whether individuals with good networks are essentially benefiting at the expense of less well connected fellow citizens, or whether a society is in aggregate better off if it is characterised by rich social networks.
The folklore says that paid jobs and high status positions are often gained through 'old boy' and perhaps 'old girl' networks. Interestingly, research by the American sociologist, Mark Granovetter (1973), has shown that jobs are more often obtained via 'the strength of weak ties'. That is, A wants a job and B knows it, but does not have one to offer himself—but he knows C, who knows D, who knows E who does have a job available. If A can network effectively, then the linkages from B through to E can be made and the job will be landed; if not, it won't. As Corcoran, Datcher and Duncan (1980) wrote, 'Most workers find their jobs through word of mouth'. The teledex is often the key (Burt 1992).
It is not just at the top end of town that contacts count. Herbert Gans (1967) reported that one reason why many immigrants in the US are so successful economically—they have higher incomes in the second generation than native-born people—is that they sponsor each others' careers and prefer to do business with each other rather than with 'outsiders'. As noted above, Gans coined the phrase 'ghettoes of hope' to characterise the social and economic processes which he found to be at work in Greek, Irish, Italian, Polish and other enclaves (see also Portes & Sensenbrenner 1993). One consequence of this phenomenon is that it is often the case that a particular trade becomes the monopoly or near monopoly of a particular ethnic group; Chinese laundries and the Jewish rag trade are cases in point.
Australian researchers have found the same processes going on in this country, where it is also the case that most immigrant groups are better off in the second generation than the native born (Evans 1989). Evans also shows that immigrant entrepreneurs flourish particularly in large immigrant groups in which English language skills are poor. They have a reasonably large but not wealthy market, and a large pool of labour to tap into. Workers within their community, especially those with poor English, can do better than they otherwise would. A corollary of Evans' work, which she spells out, is that ethnic businesses have no competitive advantage in small ethnic groups or ones in which English skills are adequate.
Later research has added an important caveat to the work on ghettoes of hope. It is pointed out that contacts in one's own ethnic group can usually only take you a moderate distance up the career and business success ladders. To make it big, one needs contacts in the wider community; one needs 'bridging networks', or one might say 'bridging' social capital (Green, Tigges & Browne 1995).
In the US, as noted above, it has been found that the very poorest ethnic groups and ethnic ghettoes are not characterised by effective social networks. William Julius Wilson (1987), an expert on American ghettoes, documented the long-term structural disadvantages confronting ghetto dwellers, and in this context pointed out that where hope has died effective networks were rarely established. Wilson believed that poor networks were more responsible for the lack of jobs in poor neighbourhoods than lack of transport and other types of infrastructure. The spatial mismatch hypothesis, discussed in section 2.4 is an alternative view. This view holds that one reason for unemployment in some poorer suburbs is that the demand for unskilled labour is quite low, even though that is where much of the available labour is to be found (Kain 1968; Ellwood 1986).
Cross-country comparisons also support the contention that social capital has a positive association with economic growth. Knack and Keefer (1997) found a significant positive association between trust and civic norms and economic growth among 29 market economies. Temple and Johnson (1998) re-examined and reworked data reporting 'social capability' for mainly developing countries from 1960. They tested the predictive power of various measures of social capability on the subsequent per capita growth performance of the countries over the period 1960 to 1985.
More contentious is the value of social capital as a contributor to the performance of high technology industries and 'research triangles'. AnnaLee Saxenian (1994) compared the fortunes of Silicon Valley in California and Route 128 in Massachussetts in the early 1990s when their branch of high tech industry suffered a downturn. Silicon Valley coped much better than Route 128 due, Saxenian claims, to stronger social capital. Business people in Silicon Valley were used to sharing information and also socialising together. In the recession, they extended their cooperation to include community organisations (Saxenian 1994). On Route 128 standard business practices of secrecy and not sharing with competitors prevailed.
Saxenian's account has been challenged by the Berkeley Roundtable on the International Economy (BRIE), which does a great deal of consultancy work in Silicon Valley (Cohen & Fields 1998). BRIE's experience is that alliances in Silicon Valley form, dissolve and reform as a matter of pragmatics in the endless drive for innovation. In this view, social capital and norms of trust are irrelevant. The institutional goal of innovation, which all players share, helps to create temporary alliances among players who are self-interested (standard economic rational actors).
So this account gives an institutional and quasi-game theory explanation of the successes of Silicon Valley. In broad terms, BRIE's reading of events recalls Axelrod's (1984) well-known proof that cooperative behaviour (punctuated by occasional 'tit-for-tat' retaliation) is more effective for all parties than continuously competitive behaviour in what are termed 'iterated prisoner's dilemma' situations. These are situations where the players have roughly equal power and can see that they are going to interact with each other over long periods of time. Generally, one would think that major businesses in the same geographical area can reasonably be regarded as being in an iterated prisoner's dilemma. This may seem a rather abstract proposition, but Axelrod's theorems are of great potential practical importance, as is recognised in the burgeoning field of 'cooperation theory'.
In this section we consider levels of social capital in Australia. We compare Australian levels of trust and performance in voluntary activity with performance elsewhere. We consider differences within Australia and report on research on social capital in relation to immigrants, small communities and health.
Australia in an international perspective
International comparative evidence appears to show that Australia is quite well endowed with social capital, and it might perhaps be inferred that greater public policy efforts to make effective use of this capital would have reasonable prospects of success.
Australians undertake a large and increasing amount of volunteer work (Evans 2000). This is in contrast to the US and some other countries where Putnam (2000) reports that volunteer work, and social capital generally, are in decline. Evans (2000) reports that 33 percent of Australians undertook 'volunteer or charity work' in an average week, and these people typically put in 7.9 hours. Churchgoers volunteered at over twice the rate of others.
Table 1 reports the results of an international comparison of participation in voluntary work constructed from the International Social Science Survey (ISSS) for 1999.15 Respondents in 13 countries were simply asked whether they had done 'any volunteer or charity work in the past 12 months'.
Source: ISSS data 1998-1999, Australian social monitor, vol. 3, no. 1, July 2000.
The data show Australians rating very highly on commitment to voluntary work. They rate second just behind New Zealanders and just ahead of Americans (Evans 2000). Evans' results need to be set alongside data from Australian Bureau of Statistics (ABS) time use surveys reported by Winter (2000). In the 1992 time use survey, 22.4 per cent of Australians undertook an average of 89 minutes per week of volunteer work. Then, in 1997, 19.8 per cent of Australians reported an average of 109 minutes per week of volunteer work. In the ABS surveys, data on volunteer work were gathered in the context of tracking all 24 hours of the day (i.e. the sum of activities must add up to the total time available). By contrast, the question in the ISSS does not constrain respondents to account for all of their time. So in terms of establishing levels of time spent on volunteer work, the ABS data are likely to be more reliable. However, the ISSS surveys used the same question in all countries listed in Table 1, so while respondents may tend to overstate their level of volunteering, the ranking of countries seems likely to be accurate.
Trust in other people, in one's fellow citizens, is usually regarded as a key affective component underlying social capital (Paxton 1999). International evidence on levels of trust is available in the World Values Study (WVS) (see also Janssen 2000).16 Table 2 shows rankings for 1995-96 and 1998-99. Respondents were asked whether they generally trusted strangers.
Source: WVS data (1995 and 1999-2000)
There were only minor changes in rankings during this period, so we will focus on the rankings for 1998-99. Australia ranked fourth in level of trust among the countries listed, and first among the English-speaking countries. The results show that the Scandinavian countries (Denmark, Sweden) and the Netherlands scored very high (Janssen 2000). The majority of people (nearly two-thirds) in these countries gave a positive response when asked if they generally trust strangers. The international differences are huge. In most Western European countries only about a third of the population expressed trust in strangers, and in France the figure was 22 per cent.
There is a lively debate among social scientists about whether Putnam is right in believing that trust and participation are declining in other Western countries besides the US. The evidence in Table 2 shows no consistent pattern of change in levels of trust between 1995-96 and 1998-99; in some countries levels appeared to increase a bit, in others to decline. In relation to participation, the evidence seems even more problematic. In a recent review the OECD (2001) suggests that patterns of participation have changed, with formal groups declining, but with single issue and ad hoc groups perhaps attracting increased interest and support. In most countries, including Australia, there has clearly been a decline in membership and participation in long established voluntary groups, including trade unions and churches. However, participation in single issue groups and social movements may be on the increase (Evans & Kelley 2002).
Variations of social capital in Australian communities
We saw that the level of trust in the Australian society is quite high, compared with most other Western countries. The Australian Community Survey (N = 8500), carried out in 1997-1998, reported variations in trust in different sections of society (Hughes, Bellamy & Black 2000). Trust was measured by four questions and correlated with a range of personal and other characteristics. Important correlates were level of education, degree of altruism and the socio-economic status of the locality. Poorer areas were characterised by considerably lower levels of trust than richer areas. Of people living in the poorest 25 per cent of areas, 57 per cent scored positively on a scale of trust, compared to 71 per cent in the best off 25 per cent. Lack of trust was higher in urban areas, with only 35 per cent of people in the poorest urban areas reporting positive scores. It should be noted that this study appears to treat a poor socio-economic environment as a cause of low social capital, whereas earlier studies we reviewed claimed that low levels of social capital can contribute to the poor economic performance of an area.
A number of recent studies have made estimates of social capital in small communities. Onyx and Bullen (2000) surveyed 1200 adults in five communities in New South Wales. The five communities consisted of two rural, two outer metropolitan and one inner city community. Respondents were asked 68 questions about their attitudes, links with family and friends, and participation in community life. The authors identified the importance of economic wellbeing as a necessary but not sufficient condition for the generation of social capital. Using factor analysis, they identified a single underlying social capital factor that had eight components. The components comprised four 'building blocks' associated with capacity building, and four associated with arenas of operation. The four building blocks were proactivity (taking initiative) in social contexts, feelings of trust and safety, tolerance of diversity and an appreciation of the value of one's own life.17 The four arenas were participation in community, neighbourhood connections, family and friendship connections, and work connections.
The analysis showed large differences among communities. Rural communities had higher overall levels of social capital, and specifically stronger feelings of trust and safety, and greater social capital in the arenas of participation in community and neighbourhood connections. However, they had lower levels of proactivity in social contexts and lower tolerance of diversity. Economic wellbeing was related to the four capacity building blocks, but not the four arenas. The authors concluded that capacity- building social capital is a necessary but not sufficient condition for economic wellbeing. However, they conceded that, since their research was cross- sectional not longitudinal, it was not possible to be confident of causal direction.
Differences in social capital among immigrants
In the last few years there have been several Australian studies of social capital in ethnic communities. Giorgas (2000) studied six immigrant groups: Germans, Dutch, Hungarians, Poles, Italians and Greeks. He found that levels of social capital among the last two were relatively high. Greek and Italian immigrants had a collective sense of identity. Greeks, in particular, saw themselves not only as individuals and family members, but part of the Greek-Australian community. This, claimed Giorgas, made them better able to organise and unite for a common purpose, and gave them greater collective command over resources. German, and especially Dutch, immigrants were quite different; they tended not to join ethno-specific groups and usually had high levels of contact with Anglo-Australians.
In general, Giorgas (2000) claimed that ethnic community group formation has served as a positive strategy for Australian immigrants. This type of social capital has helped to overcome social isolation and economic difficulties by providing employment opportunities and a sense of a familiar social milieu. He argued that social capital was used more effectively by groups with stronger cultural boundaries and a collective sense of identity.
Like Giorgas, Portes and Zhou (1993) considered resources within an immigrant's community crucial to improving chances of upward mobility—people could obtain both economic and moral support to help them on their way. For example, values regarding the importance of educational attainment and economic success could be maintained and transferred to the second and later generations. According to these authors, ethnicity can be considered a distinct form of social capital constructed from one's cultural endowments and including obligations and expectations, information channels and social norms.
From this work and also Evans' (1989) work on Australian ethnic entrepreneurs (referred to earlier), we can deduce when sticking together helps and when it hurts. Large immigrant groups, in which many members have poor English language skills, probably benefit, especially in starting businesses and getting jobs, by sticking together. However, if they are to get further ahead, they need to forge links with the Anglo-Australian community.
Members of small ethnic groups may not benefit at all; the intra-ethnic market for their businesses and labour is likely to be too small. In the case of ethnic groups whose members typically have no problems with English, like the German and Dutch immigrants studied by Giorgas (2000), it is hard to see why they would gain substantial benefits from sticking together. Their behaviour, in not joining ethno- specific groups and having close contacts with Anglo-Australian organisations, seems entirely rational—not the loss implied by Giorgas.
Social capital and health in Australia
Baum et al. (2001) report results of the Adelaide Health Development and Social Capital Project that explored social capital, community participation and health in the western suburbs of Adelaide. Data were collected in four ways: from 2542 respondents to a random sample mail survey, from in-depth interviews with 40 people from the survey, from a survey of 239 community groups and organisations, and from 25 case studies of the community groups.
The authors were particularly interested in the relationship between health (the presumed outcome variable) and participation in both ordinary social activities and in civic and political activities. Their main finding was that, controlling for age and socio-economic status, social participation had a strong link with health, but civic participation did not. This finding should not be seen as contradicting the mainly American results reviewed in section 3.3. The American results showed that, from a health standpoint, social and civic participation are effective substitutes for each other. However, the Adelaide data show that if you have rich social networks no additional health benefits will flow from also engaging in civic activities.
In the Adelaide study, levels of participation varied markedly by family type. Informal social participation was high for all family types, but dual-parent families with children were involved, on average, in more activity than people who lived alone or lived with unrelated adults. Children's friendships and their activities and interests were the conduit for much social participation by parents.
- 4.1 Local area initiatives
- 4.2 Regional social and economic plans with community building aspects
- 4.3 National level policy
We now review actual and proposed public policy initiatives intended to enhance community outcomes. Many of these interventions are also intended to enhance local social capital and 'partnerships' (see below) as a means to an end—the end being outcomes related to employment, crime, housing, education and so forth. There are three sub-sections: one on local area interventions, the second on regional level policy and the third on the central or Australian Government level. However, the latter two sections are brief; most recent effort, especially in Britain from where the best evaluation data are available, has gone into local area interventions.
The primary focus of targeted local area initiatives has been on attacking 'multiple deprivation' and 'social exclusion'.18 There are powerful arguments for and against area-based initiatives. It seems best just to lay these arguments out; there is no convincing way to resolve them. We then describe recent initiatives mainly in Britain and Australia, and finally consider evaluations and attempts to elicit 'best practice'.
The principal arguments in favour are (Smith 1999):
- There are identifiable localities that suffer disproportionately from problems of joblessness, run-down housing, crime and vandalism; in short, 'multiple deprivation'.
- Many of these areas have been run down for decades and their inhabitants are suffering to some degree from 'social exclusion'; that is, they are not able to enjoy a normal, mainstream lifestyle (assuming they want one). Many depend on government benefits.
- The fact that some areas have been deprived for decades indicates that 'trickle- down' theory—the theory that benefits of economic growth trickle down to the poor and socially excluded—does not always, or perhaps even generally, work (Robson et al. 1998).
- Problems compound each other in deprived areas; the sheer scale of the problems means that special initiatives are needed.
- Because problems are concentrated, a greater number of deprived people can be assisted if resources are geographically targeted.
Major arguments against local area targeting are as follows:
- Most deprived people do not live in deprived areas and so will be missed by area targeted programs.
- Area targeting can displace problems (e.g. unemployment and crime) to other, and particularly adjacent, areas.
- Area targeting can have the effect of encouraging people to remain in run-down areas instead of seeking opportunities elsewhere. It may be better to let the market work and leave some areas to decline.
- Many problems (e.g. unemployment) may be regarded as being generated at the national level and therefore need national solutions.
Despite trenchant opposition to local area targeting from some of the country's experts on poverty and social exclusion, including Peter Townsend (1979), since 1994 successive British governments have taken three major initiatives in area targeting. The Single Regeneration Budget (SRB) program, begun by the Conservative Government and expanded by Labour, has funded 1028 deprived areas, mostly local government authorities (LGAs) or smaller (e.g. a ward or two). About a quarter are in London and most others are in large cities. New Deal for Communities has funded 39 very deprived neighbourhoods (in many cases, run- down public housing estates often referred to as 'problem estates'), and the Neighbourhood Renewal Fund has targeted 88 deprived neighbourhoods in England (Rhodes et al. 2002).
Bids to central government (or more recently to regional development authorities) are made by local partnerships between the public sector, the voluntary sector and private sector (the 'three-thirds model') in the locality. Often the LGA is the lead partner with the task of trying to coordinate all relevant public sector agencies ('joined-up government' is the Blair Government's ideal in these matters), as well as the local partners. Competition is claimed to improve the quality of bids, but might be expected to mean that some highly deprived areas miss out. In fact, evaluations have found that all the most deprived urban areas of the country are receiving funds (Smith 1999; Rhodes et al. 2002). The funding bodies have recently used an Index of Local Deprivation (1998) to assist targeting.
One of the tasks of the local partners is to use central funding to obtain additional local funding. In fact, every pound of SRB funds has attracted an additional £2.7 of private expenditure and has leveraged £3.96 altogether (Rhodes et al. 2002). The partners are also encouraged to 'bend' mainstream (i.e. national) program funding into their area, but the main evaluation concludes that in this respect their efforts have been 'modest and variable', and have enhanced SRB funding by only 7 per cent. It is not surprising that other jurisdictions resist reallocation of mainstream funds to deprived areas that are already getting targeted assistance.
Interim evaluations generally agreed that in the early days of the SRB some local partnerships were shotgun arrangements made to secure central funding. More recently, 'capacity building funding' has been increased in order to help build partnerships and enhance skills in areas like negotiation, monitoring and evaluation (Rhodes et al. 2002). Private sector involvement has been particularly hard to secure on a continuing basis, in part because only specialised neighbourhood renewal companies can directly make a profit, but evaluators report that many companies have been prepared to sponsor job training and other programs to improve labour market skills (Smith 1999).
We find it hard to summarise evaluations of the British programs. The evaluations are conducted by academics acting as consultants to the government departments that run targeted programs. The studies have a generally positive tone, but the evidence from the most detailed case studies could be regarded as showing that only young people were effectively assisted and that there were considerable social barriers and disincentives in the welfare and tax systems for unemployed people to take work (see Rhodes et al. 2002 and, especially, evaluation of results for Chalkhill, Hangleton Knoll and Rochdale). Unemployed men, in particular, tend to form networks with other unemployed men, and it is difficult to encourage job search (Smith 1999). Smith notes that 'it is fairly clear that the bulk of the most deprived areas of nearly 20 years ago are still among the most deprived areas in the country'.
Australian local area targeting
Recent urban initiatives in Australia aimed at community building have also mainly been in run-down areas beset by high unemployment, vandalism and, in some cases, drug addiction. Two NSW government 'place management' pilot studies funded by the Premier's Department in Cabramatta and Kings Cross are of particular interest because they have been carefully evaluated (Nexus Management Consulting 1999ab; Walsh 2001). The evaluations are somewhat ambivalent. They show fairly substantial benefits to clients directly assisted, but note that unemployment and even drug addiction problems cannot be cured solely by local interventions because the supply of jobs and drugs is primarily affected by national policy. Further, these two pilot projects were demonstration projects, or what Cronbach (1983) calls 'super-realisations', to which local politicians, bureaucrats and community activists made contributions above and beyond the call of duty. The difficulty is that it is doubtful if such efforts could be replicated on a larger scale or even in numerous neighbourhoods simultaneously.
The evaluations conclude that two-year (short, sharp) crisis intervention programs of this kind have much to recommend them, that they should only be directed at discrete local areas, and that they are particularly appropriate when no single government agency could hope to solve local problems, so that a whole-of- government (break down the silos) approach is needed. However, short-term interventions are hard to reconcile with community consultation and 'ownership' of programs; there was some evidence in these two cases that the local communities regarded the programs mainly as state government programs, not as locally owned.
The former Leader of the Opposition and Labor MP, Mark Latham (1998), reported similar place management programs in Campbelltown (Sydney) in his own electorate. There, community activists, rather than just appointed place managers, appear to have taken the lead, and local residents have joined in to resist and report on vandalism and crime, partly through Neighbourhood Watch and local patrols. In Claymore (a housing estate in Campbelltown), it appears that crime and substance abuse have been much reduced (Latham 1998).
A further important Australian initiative at the local level, primarily in remote rural areas, is the Community Development Employment Program (CDEP). This is a labour market program for Indigenous Australians. It provides funds for work and training activities determined at the local level; activities deemed to have a community building focus. The scheme is administered by the Australian Government Department of Employment and Workplace Relations (DEWR), but wages are paid by the local CDEP organisation.
In statements issued at the time of the 2001 Budget the Australian Government recommitted support to CDEP but judged that it was not entirely satisfactory in areas where good job opportunities already existed. In these areas, indigenous people would be encouraged to use mainstream services rather than the CDEP (media release 22 May 2001).19
The Australian Government in recent years has sponsored several other innovative community-building programs in rural and remote areas. These include major efforts to revive community spirit and participation in the mining town of Broken Hill (Onyx and Bullen 1997, 2001) and promotion of the Rural Communities Program, which has been thoroughly evaluated and the lessons of successful local ventures disseminated (Bureau of Rural Sciences 2002). The Australian Government also sought to showcase Australian best practice in community building in its Can Do Community Initiative issued after the launch of the Stronger Families and Communities Strategy (April 2002). One aspect of community building, much emphasised by the government and less prominent in overseas literature, has been the development of local leadership both to launch initiatives and for conflict resolution (Black and Hughes 2001). This is particularly important in local areas where the exit of banks, a range of health services and, more generally, educated people has meant the loss of sources of 'natural' leadership. The government appears to agree with the 1994 McKinsey report on the growth of Australia's regions, which said that 'Given the task of rejuvenating a region and the choice of $50 million, or $2 million and 20 committed local leaders, we would choose the smaller amount of money and the committed leaders' (quoted in Black and Hughes 2001).
Finally, it should be noted that in seeking to promote best practice in community strengthening, the Australian Government and state governments have commissioned several reviews (of overseas and Australian programs and social and economic measures) that could be used to evaluate local programs (Black & Hughes 2001; Nexus Management Consulting 1999ab; Onyx & Bullen 1997, 2001; Salvaris et al. 2000; Walsh 2001; Zubrick et al. 2000). At the national level, the Australian Bureau of Statistics, following review of work undertaken by the OECD and other international sources (see particularly OECD 1998, 2001 and World Bank 2002), issued the important compendium, Measuring social progress (2002), which includes papers clarifying conceptual and measurement issues, as well as numerous measures reported in the volume itself. So there is now a wealth of proposed, as well as implemented, measures of community strengthening to draw on—both process and outcome measures—but, as discussed below, we believe that further efforts are needed to integrate and perhaps link them to existing systems of national accounts.
Best practice in local area targeting
Several attempts have been made to codify best practice in local area targeting (OECD 1998; McGregor & McConnachie 1995; Adamson et al. 2001; Rhodes et al. 2002). None seeks to identify a single model—all offer fairly general principles and guidelines. We summarise generally agreed principles:
- Multi-agency partnerships are usually seen as the primary mechanism for formulating and delivering area regeneration strategies. There is near universal agreement that the public sector, the private sector and the community sector should all be involved (the 'three-thirds' model), and many consider that local government should be the lead partner (Adamson et al. 2001). Private sector companies rarely make an effective lead partner.
- Capacity building (skill enhancement) is likely to be needed to enable partners to work together effectively. A secretariat and people with negotiating and grant- winning skills are likely to be required, partly to help the partners work together effectively. The aim should be to build 'sustainable communities' with partners committed to work together after special funding ceases.
- Special funding for deprived areas should be used to leverage additional funding from other levels of government and from the private sector. 'Bending' mainstream funding into deprived areas via 'single pot' funding mechanisms (e.g. at a regional level) is desirable (Rhodes et al. 2002).
- Strategies focused primarily on improving physical capital (housing, open space, transport) are rarely successful in the absence of human capital investment.
- It is necessary to establish linkages between deprived areas and adjacent better off areas, especially to take advantage of employment opportunities.
- Evaluation studies and reasonable management information systems at the local level are required to check on progress and value for money.
Australia already has three levels of government, so it is no surprise that a fourth possible level, the regional level, has not flourished. Nevertheless, during the last thirty years most federal and state governments have launched some regional social and economic programs. These initiatives have invariably involved extensive community consultation and have had, as one of their stated objectives, 'community development' or 'community building'. The record of these initiatives is mixed but it would be fair to say that, in general, they have not lived up to their claims. While a comprehensive discussion of the failure of regional policy (but not necessarily the failure of regional economic policy) is beyond the scope of this paper, the following selected list of activity illustrates the nature of regional policy.
The Whitlam Government in 1972-75 launched the Australian Assistance Plan (AAP) from its new Department of Urban and Regional Development (DURD). Regional level consultation was undertaken and cooperation between local governments encouraged. A related set of initiatives accelerating the growth of designated towns, notably Albury-Wodonga on the NSW-Victorian border. DURD was abolished when the Coalition was re-elected in 1975, but some state governments continued to show interest in aspects of regional consultation and planning. In Victoria, for example, the Coalition's Family and Community Services Program could be regarded in part as a continuation of the AAP. Regional consultative committees were set up and there were discussions among state departments about bringing their very different boundaries into line. These discussions largely failed.
State governments have instituted regional economic plans from time to time. The Cain Labor Government in Victoria (1982-90) encouraged non-metropolitan regions to develop and implement economic plans based on identified regional comparative advantage. Some of these plans involved extensive consultation and were detailed documents.
As we have noted, tools of regional economics can help provide an intellectual framework for the development of community building (social capital) programs. In part, as implied above, this involves analysis of regional comparative advantage and analysis of mismatches between demand and supply of unskilled labour (the labour mismatch hypothesis; see Kain 1968).
Since 1996, regional and rural electoral backlash has stimulated a wide variety of efforts by the Australian Government and some state governments to consult more widely in non-metropolitan regions and to offer infrastructure programs, subsidies and community development grants. The Commonwealth has funded rural transaction centres in remote rural areas to encourage local people to get together to arrange for Internet services, banking and so forth. In 2000, a new Regional Solutions Program was announced to provide grants for community planning, local projects and infrastructure development.
In several states, cabinet meetings are sometimes now held in regional centres and regional consultative forums of various kinds have been held. Infrastructure projects (e.g. in Victoria, fast train services linking Ballarat, Bendigo and Geelong to Melbourne) have been launched and plans made to improve Internet services in rural areas. As part of this thrust, governments have commissioned studies to develop measures of community development and social capital.
A decentralised program which appears to have borne much fruit at local government level is the Better Cities/Urban Design Innovation program. A large number of environmental initiatives have been taken under the banner of this program, with local government managers exchanging a great deal of information with each other, and presumably learning from each others' programs, in part through the program website.
The McClure report (2000) on welfare reform put up for consideration, but stopped short of formally recommending, major innovations in the role of voluntary work and the community sector. The ethical principle underlying the report was that people in receipt of state benefits had a 'mutual obligation' to contribute as best they could to the community. If they could not contribute paid work, they should consider voluntary work in the community sector as a means of discharging their mutual obligation. The government should discharge its obligation by active one- on-one case management which would include, for some people, finding voluntary work, providing training and, in some cases, assisting them to move on to paid work. The McClure proposals, although deliberately not costed, would plainly have required a substantial investment in the community sector. It was also proposed that payments to people doing voluntary work would be in the form of 'participation supplements' and, thus, would be modestly above state benefit levels. This would provide an incentive to do voluntary work and would compensate for the costs of working, including travel.
The McClure report was broadly welcomed by the three main political parties. To date, only a few proposals have been acted on. In the May 2001 Budget the Australian Government accepted that voluntary work could be a method of fulfilling one's mutual obligation. Thirty-five thousand places were earmarked for community work over the next three years. The proposal for a participation supplement was not taken up, either in this area or in relation to Work for the Dole. However, community work participants would get a Training Credit of $500 for future job training if they did the required minimum of 240 hours of community work in a year (Minister for Employment media release, 22 May 2001). This would go up to $800 if they did 390 hours of work. There was no announcement of an increase in funding to the community sector to provide jobs or training. It is much too early to judge what the practical impact of the McClure report will eventually be.
There is no evaluation evidence available in Australia or, so far as we are aware, overseas on the cost-effectiveness of programs to get unemployed people and other state benefit recipients to do voluntary work. The possibility has been much discussed in Germany and proposed in several states there (Erlinghagen & Wagner 2002). Research using the longitudinal German Socio-Economic Panel Survey indicated that unemployed people usually do little volunteer work and that those who become newly unemployed in recessions are most unlikely to volunteer. In Germany, as in Australia, most voluntary work is undertaken by people with good interpersonal skills and above average education. People who lack the requisite skills are unlikely to stick at volunteer work. Training unemployed people for the work would be essential for any hope of success, and would need to be carefully evaluated to see if it proved effective (Erlinghagen & Wagner 2002).
A more abstract but not unimportant concern is that if voluntary work is subsidised and participants paid (at least above state benefit levels), then it is not voluntary work at all. The community sector, it could be argued, just becomes an arm of the public sector, competing for government contracts, and not enhancing social capital, which is based on (almost defined as) voluntarily given trust and participation (Cox 1995).
In previous sections, we have reviewed the state of play regarding the concepts of community and neighbourhood (section 2), concepts of social capital (section 3.1), their sources and antecedents (section 3.2), consequences and claimed benefits (section 3.3), and public policy interventions designed to enhance them (section 4). In this section, we focus on two sets of issues on which we believe progress is urgently needed if public policy momentum for strengthening community and social capital is to continue to develop. These are principles for measurement of social capital and issues to do with the governance of community sector organisations.
The first set of issues is urgent because in modern government a proposed major policy development never gets far unless there is prior agreement on how to measure and evaluate progress or regress. So developments in the area of community building and social capital urgently need agreement on at least principles of measurement. Perhaps less obvious is the need to consider issues of governance and accountability in the community sector. If substantial new responsibilities are to be entrusted to the sector it needs to be governed effectively and in response to the appropriate stakeholders.
Principles of measurement—community, networks, social capital
In dealing with measurement issues in this section we propose to treat 'community', 'social networks' and 'social capital' as interchangeable. In other words, we assume that it means exactly the same thing to say, 'this community is improving' as to say 'the quantity and quality of its social networks is improving' and 'the quantity and quality of its social capital is improving'.
Principle 1: Distinguish between structure and content
The proliferation of terms relating to social capital and social networks confuses discussion and creates the impression that there is more disagreement about what should be measured than is actually the case. Following Stone (2001), we prefer to distinguish between the structure of networks (size, internal and external organisational linkages) and their content (trust, reciprocity etc). This is really very similar to distinguishing, as other authors have, between quantity and quality, or availability and attachment (Henderson et al. 1981).
There are two reasons for distinguishing between structure and content. One is to suggest that each measure of structure should be accompanied by and matched to one or more appropriate measures of content (see below). The other is to insist that relationships between the two should always be investigated empirically and not assumed. As noted earlier, it is a mistake just to assume that community groups and networks are always benign, and are characterised internally or in dealings with other groups by trusting, cooperative or pro-social relationships.
Measures of structure which are easily available and therefore commonly used, and which also appear reliable and valid, are:
- measures of organisational joining/membership: social, occupational and political organisations to which people belong
- measures of voluntary work: e.g. numbers of people who engage in voluntary or charity work
- measures of informal networks: family and neighbourhood linkages.
Built on these direct measures, which are usually obtained from sample surveys, are second order structural measures which analyse the initial data to obtain, additionally, measures of organisational or network density (e.g. percentage of employees who belong to unions) and measures of openness or closure (whether the network is linked to others via shared members, activities and alliances).
Measures of content usually focus on three dimensions:
- measures of trust and reciprocity: e.g. trust in institutions (such as government, media, business, unions, police)
- measures of friendship and intimacy: e.g. measures of friendship in the neighbourhood and intimacy within the family
- measures of participation/activism: these measures seek to establish whether people are active or passive members of organisations and networks (e.g. attend meetings, serve on committees, put in many hours of voluntary work).
Two important hypotheses about relationships between structure and content, which appear to have considerable empirical support, are that:
- Networks with open membership structures cooperate more effectively in coalitions than networks with closed structures.
- Open networks are more effective for obtaining information, while closed networks are more effective for securing adherence to rules and norms without costly enforcement mechanisms.
The first hypothesis derives largely from Putnam's (1993, 2000) work on organisations and networks in Italy and the US. The second is suggested by Sobel (2002) as a link between the work of Granovetter (1973), who found that open networks were excellent for job search, and Coleman (1988), who found numerous closed and dense networks (e.g. the network of New York diamond merchants) within which rules and norms were strictly adhered to without any formal enforcement mechanisms.
Principle 2: Specify the arena or area of activity to which measures apply
Some measures of social capital are extremely general and non-specific about the arenas to which they are supposed to apply. It seems essential for valid measurement that each indicator of structure or content should clearly relate to one of the following arenas:
- political organisations
- economic and occupational organisations or networks
- voluntary work and organisations
- social organisations or informal neighbourhood and friendship networks
- family networks, including extended family.
The principle that arenas should be specified may seem an obvious one, but examples of very widely used measures that break the principle can easily be found. Examples cited in this review are the World Values Study measures of 'trust in people' and 'trust in strangers'. A moment's thought suggests that people might well have mixed responses if they were asked separately about trust in, say, politicians and political organisations, their own occupational organisation and their neighbours. Inferences made from general questions, or inferences that extrapolate from one arena to another, are not likely to be valid.
Principle 3: Measures of community/networks/social capital can validly be presented at any level of aggregation
There has been a strident debate among students of social capital about whether it is 'best' conceived as a property of individuals or of aggregates like regions or nations. As noted above, Bourdieu (1986) and Coleman (1988) treated it as a property of individuals, whereas Putnam usually treats it as a property of aggregates.
Because measures of social capital and social networks usually require survey research, they are often initially reported at the individual level. However, they can then quite sensibly and validly be aggregated to the neighbourhood, city, state, or national level so that different locations (or the same location at different times) may be compared. In principle, both measures of central tendency (means, medians) and measures of dispersion (standard deviations and measures of inequality) may be compared.
Principle 4: The costs and benefits of social capital should always be assessed empirically and should not be assumed to be benign
By aggregating, we can assess whether social capital is increasing or decreasing (provided that all specific measures are valid as indicators of underlying social capital; for further discussion, see next section). The potential danger and error is simply to assume or imply that 'more is better'. The social benefits and costs—the welfare gains and losses—of social capital must always be a matter of empirical investigation, not assumption. Exactly the same is true of financial capital and human capital. Like social capital, they too can be used (or invested) well or badly, for high or low rates of return, and with greater or lesser costs and benefits to third parties. This last conceptual point is made by Elinor Ostrom (1990, 1999) who, in her own case studies of the management of property held in common by multiple owners (referred to earlier), has found numerous examples of both effective and ineffective networks.
Applications: Putnam's Social Capital Index for the American States (2000)
We now illustrate points relating to these principles of measurement by commenting on some available and proposed indicators of social capital. First, Putnam derived an index for the 50 American states from readily available sources. The components of the index and their correlation with it are shown in Table 3. Most of these components have counterparts in Australia.20 They capture the main dimensions of social capital. The structural dimensions of organisational joining, volunteerism and membership of informal networks are covered, as are the content dimensions of trust, friendship and activism/participation. However, the headings within Table 3 were constructed for Putnam's own presentational purposes, and might be regarded as lacking conceptual clarity.
|Components of Comprehensive Social Capital Index||Correlation with index|
Measures of community organizational life
Served on committee of local organizations in last year (per cent)
Served as an officer of some club meetings in last year (per cent)
Civic and social organizations per 1000 population
Mean number of club meetings attended in last year
Mean number of group memberships
Measures of engagement in public affairs
Turnout in presidential elections, 1988 and 1992
Attend public meeting on town or school affairs in last year (per cent)
Measures of community volunteerism
Number of non-profit organizations per 1000 population
Mean numbers of times worked on community project in last year
Mean number of times did volunteer work in last year
Measures of informal sociability
Agree that 'I spend a lot of time visiting friends'
Mean number of times entertained at home in last year
Measures of social trust
Agree that 'most people can be trusted'
Agree that 'most people are honest'
Source: Putnam (2000, p. 291)
The first three headings relate to what we have termed arenas (the local community, public affairs, voluntary work). The specific measures under these headings can each be straightforwardly classified as either structural or content related, although there is no matching of a content measure (or measures) to accompany each structural measure. The last two headings confuse matters by referring not to arenas but solely to content. In the case of 'measures of informal sociability', the arena can be inferred, but, as we have noted, the arena(s) for 'measures of social trust' is invisible and cannot possibly be inferred.
Finally, note that Putnam reports aggregate level item-total correlation coefficients that show each indicator, which appears quite reliable as a measure of the underlying index of social capital. (The correlations could be regarded as providing limited, although not really adequate, justification for the measures of social trust). In other words, more of one is generally associated with more of the others, so if one indicator goes up over time, then it will usually be correct to infer that total social capital has increased. This is a desirable feature of the index; we want to know whether total social capital is going up or down. However, remembering principle 4, it does not follow that increases in any or all of the measures is going to correlate with, let alone cause, improvements in welfare. Whether improvements in economic performance or health and happiness followed would be a matter for empirical investigation.
Applications: proposed social capital indicators for Australia
In April 1999, FaCS held a workshop on Indicators of Social and Family Functioning at which issues and measures relating to social capital were canvassed. Subsequently, headings for measuring social capital were published (Zubrick et al. 2000). Social capital measures already available related to neighbourhood violence and crime, social or cultural discrimination, and availability of family and community support services. Measures deemed desirable but currently unavailable were those relating to trust, social support (including kin and community contacts), social participation and engagement, reciprocity and civic involvement (Zubrick et al. 2000).
This approach appears to combine measures directly related to the structure and content of social capital with measures of the potential consequences. The measures said to be already available were all (with the exception of the availability of support services) measures of potential consequences. Thus crime, violence and discrimination are all possible consequences of low levels of social capital, or perhaps of social capital with negative consequences for welfare (e.g. well organised but mutually antagonistic criminal groups). The second list of measures deemed desirable but not yet available are all measures of either the structure or content of social capital, but there is no clear matching of structural measures with appropriate content measures. There is also no indication of how the measures could be aggregated to see whether national social capital had grown or declined.
In Victoria, following the Growing Victoria Together summit in March 2000, a Swinburne University group was commissioned to report on social benchmarks and indicators worldwide and to make recommendations for the state (Salvaris et al. 2000). The Swinburne group usefully reviewed government-sponsored projects which it regarded as best practice in a number of areas. These included national, regional, state, citywide and local projects to develop social indicators, and best practice models of national strategic planning, national social auditing, and community participation in policy making.
As with the FaCS project, the Victorian team specifically considered concepts and measures relating to social capital. They recommended forty different sets of indicators of 'social and community well-being'—one set relating to 'social cohesion and social capital' (Salvaris et al. 2000). The only measure identified as currently available was 'rate of philanthropic donations'. Other desirable measures were said to be measures of community networks, subjective trust in strangers, and an index of social exclusion.
A problem with these measures is that they do not appear to be based on a clear conceptual framework. Some of the measures clearly relate to social networks/ social capital but, again, there is no systematic matching of the two. Other measures, including perhaps the rate of philanthropic donations and certainly the proposed index of social exclusion, could well be consequences of social capital and do not have any clear validity as measures of the concept itself. Again, an issue not addressed is how to combine the indicators to assess whether social capital is increasing or decreasing.
National Accounts—could social capital be included?
We now consider a more difficult and speculative measurement issue. Could and should measures of social capital be included in the national income accounts? As noted in the introduction to this review, an obvious reason for rebadging social networks as social capital is to imply that social capital is a stock of sufficient importance to be thought of alongside physical capital and human capital.
In fact, a number of economists and other social scientists have proposed revisions to the national income accounts to make them more like national wellbeing accounts or accounts measuring social progress and not just economic progress (ABS 2002; Becker 1965; Eckersley 1998; Juster & Stafford 1985; Lancaster 1966; Nordhaus & Tobin 1972). The current drive in governmental and business circles to develop 'triple bottom line' accounts (economic, social and environmental) reflects similar priorities.
Perhaps the most ambitious research on these lines was funded by the American National Science Foundation and undertaken by a team at the University of Michigan, led by F. Thomas Juster (Juster & Stafford 1985, especially chapter 6). The aim of Juster's research was to propose and implement a new system of national accounts based on a similar approach to that of the Nobel Laureate, Sir Richard Stone, who developed the UN's System of National Accounts (Stone 1973). Juster proposed measuring stocks (capital account), flows (current account) and outcomes (current account). The stocks were physical capital, human capital, socio-political capital and environmental capital.21 Socio-political capital included social networks, voluntary organisations and the institutions of government, so it was more or less what we have been calling social capital. However, Juster did not propose specific measures of social capital stocks or, indeed, of flows (we attempt to do this below).
The most distinctive aspect of the whole framework is the central role given in the current account to uses of time (activities). Time and money were seen as being combined in the production and consumption of goods and services. As a result, stocks, including social capital, could be augmented ('investment') or run down ('disinvestment'). The ultimate outcome was viewed as wellbeing—happiness or life satisfaction.
Juster and his colleagues made substantial efforts to implement the framework, principally by conducting national time use surveys and measuring satisfaction with activities (Juster & Stafford 1985). However, the project eventually foundered, in part because the National Science Foundation and its reviewers believed that some measurement issues could not be satisfactorily solved and that the new system could not be fitted into the double-entry bookkeeping format of national income accounts (Ruggles 1981).
We believe it is valuable for policy purposes to distinguish between stocks and flows. Such a framework, applied to social capital, could lead policy makers to ask such questions as: 'Are stocks being run down—do we need to invest more?' and 'Could flows be increased and better directed to target clients?'. Juster and his colleagues produced sets of accounts for the private sector, the public sector and the household sector, but did not outline anything for the community sector. Here is our attempt for the community sector and social capital.
- Financial capital: endowments and other accumulated financial assets of community groups
- Human capital: numbers of active members (office holders and those who do voluntary work) and the skills they have in running the organisation and doing voluntary work; passive members (pay subscription, receive mailings, don't attend/work)
- Networks and public support: linkages (measured perhaps by joint committee representation) with the government, private and community sectors; support and trust of the community sector among the public (reputation for good works, fairness, openness, lack of corruption)
Flows/annual transactions include:
- Financial-income: annual donations and membership subscriptions
- Financial-expenditure: annual expenditures on (a) voluntary work and (b) running the organisation itself
- Time: time spent by members in (a) voluntary work and (b) running the organisation itself
- Outcomes: voluntary work that is more or less satisfying to providers and clients.
Stocks and outcomes would usually be measured annually at the end of the financial year. Flows would be recorded in the usual way in the current account. It is suggested that a framework of this kind could be valuable at state and national levels to assess the performance of the community sector and judge where interventions were needed.
What kind of democracy?
In an important sense, the central governance issue raised by efforts to strengthen community, develop targeted area programs, and expand the role of the community sector are issues to do with what kind of democracy we want (Warren 1999). Australia has a conventional representative democracy in which most decisions are made by elected representatives on behalf of the public. However, if local communities are to play a more active role, with public, private and voluntary sector agencies involved in partnerships that both propose initiatives and seek to implement them, then a new kind of participative democracy may emerge. This would particularly be the case if community forums directly participated in decisions. So far as we are aware, Australian governments have not directly addressed these potential issues of democratic reform. The Labour Government in Britain has and is at least tentatively committed to increased participation. This runs counter to long-established habits of mind in most of the country. It is noteworthy that some local governments are reportedly resisting ceding part of their role to wider partnerships designated to run targeted area programs (Smith 1999; Rhodes et al. 2002).
Governance of the community sector—who are the stakeholders?
A second set of issues that requires attention concerns the governance and accountability of the community sector or third sector (Salamon 1992; Adamson, Dearden & Castle 2001). Community organisations are playing a larger role than in the past in delivering government services, mainly human services, largely due to the contracting out of services under purchaser-provider arrangements. Proposals are now afoot in Australia and elsewhere to further expand the sector as a key plank of welfare reform. In view of these developments, it is reasonable to ask if third sector governance and accountability meet the standards expected in a pluralist democracy.
Principles of governance, ethics and accountability in the public sector, and even the business sector, are widely debated and reforms are regularly instituted. By contrast, third sector governance sometimes seems caught in a time warp (Salamon 1992). Many community organisations are, on paper, run in the same way as they were when founded in the nineteenth century. They are governed by management committees whose members are usually volunteers. The volunteers are in some cases wealthy donors and in some cases represent other sponsors (e.g. churches) linked to the organisation. Generally, being volunteers, management committee members find it difficult to exert effective control over the paid executives and key staff of the organisation; probably even less control than remunerated board members of private sector companies.
Two sets of issues arise. One is whether internal governance structures are adequate, especially in large organisations receiving substantial government funding. The second is the chain of accountability linking funded organisations back to the executive branch of government, to parliament and the taxpayer.
The questions that need to be considered in designing or redesigning internal governance structures are really the same as for all democratic or partly democratic organisations:
- Who are the organisation's stakeholders?
- How and in what proportions should they be represented on a governing body so that they are likely to promote (rather than frustrate) the mission of the organisation?
- What payments or other incentives are needed to get committee/board members to do an effective job in setting policy objectives and progress—checking on their implementation?
The stakeholders of a large community organisation receiving funding from government would typically include its clients, its staff, the government department(s) from which funding mostly came, parliament (representing the public/taxpayers), perhaps local government and neighbouring businesses and householders in the area(s) in which it operated, sponsors and donors who provided an auspice and funds, perhaps other community and private organisations that were in the same type of work and with whom it had formed partnership or supplier links, and perhaps professionals with relevant expertise with whom good relations are needed (e.g. medical professionals, social workers, police, bankers).
In order to effectively promote the mission of the organisation, including work done for government, it would probably not be a good idea for these stakeholders to be represented equally. Perhaps tradition should be followed, and sponsors and donors should receive disproportionate representation. Clearly, consideration should be given to representation of government, clients, staff and other stakeholders who in most existing organisations are scarcely represented at all.
Government representation might seem especially problematic to many community organisations, which in some cases jealously guard their independence and have traditionally resented government interference and wanted funding with few strings attached. However, it makes little sense for governments to supply large capital or operating grants without representation on relevant committees and boards.
This brings us directly to issues surrounding accountability to government. The community sector in Australia has gained a lot of experience of these issues in recent years through extensive use of purchaser-provider contracts. These usually involve negotiating and signing service agreements that specify detailed performance standards, client throughput requirements, different levels of payment for different sorts of client, and (often) financial penalties if contractual requirements are not met. Reporting requirements, including detailed financial accounts, are usually stringent in the public sector and impose a considerable burden, especially on smaller, less professionally run organisations.
Regular attempts are made at Commonwealth and state level to streamline these processes and introduce more uniform and less burdensome reporting requirements. In many areas, service agreement negotiations drag on so long that agreements have often barely been signed when they run out—and a new agreement becomes due.
A further important governance and accountability question concerns the role of parliament and appointees like ombudsmen. So far, we have only discussed accountability to the executive branch of government. It could be argued that contracts with the community sector, performance standards and client complaints should be open to investigation by the legislative branch and ombudsmen. To date, governments have tended to resist these demands, partly on the grounds that commercial confidentiality of contracts is required to protect the business edge of private and community sector contractors. In so far as commercial protection is a relevant factor, the resistance may be justified. However, in so far as the executive branch may just be protecting its own work from scrutiny, its resistance has no justification in a framework of democratic accountability.
It is clear that there are strong reasons for continued research and policy development relating to communities and social capital. There seems little doubt that strong communities and social capital are associated with all kinds of benign outcomes, including stronger democracy, better educational performance by children, improved physical and mental health, and better local economic performance. Our assessment has been that some of the linkages are probably causal in the sense that changes in social capital appear to have preceded changes in outcomes, and the relationship does not seem to be spurious.
Another reason for continued interest and policy momentum is that, by international standards, Australia appears to be richly endowed with social capital. Perhaps these 'stocks' could be further expanded and perhaps they could be used more effectively. At present, community work is being encouraged on a small scale as a way of discharging mutual obligation requirements. It is important that these initiatives are planned and phased in a way that facilitates pilot testing and systematic evaluation; and that such evaluation is undertaken.
Areas of future research suggested by this review include:
- a comprehensive assessment of the current status and recent trends in community integration and social capital in Australia along the lines of Putnam's Bowling alone (2000), and including measurement of social capital in regional areas
- further investigation of the relationship between social capital and various economic and social phenomena, using Australian data, and investigation of the empirical relationship between social capital and economic performance, educational achievement, health, mental health and crime
- continued review and further investigation of whether stronger communities and social capital can be deliberately created—or their development accelerated—by public policy (involving theoretical work along the lines of Axelrod  and empirical work of the kind pioneered by Ostrom [1990, 1999])
- further development of the concept, theory and measurement of social capital—efforts to develop an economic theory of social capital and to ground measurement in theory seem particularly valuable (Glaeser, Laibson & Sacerdote 2000; Stone 2001)
- work on the development of Australian national and state social capital indices along the lines suggested by Putnam (2000)
- consideration of the possibility of developing aggregate measures of stocks and flows of social capital that might be used to provide a set of accounts for the community sector.
In most of these areas, research has already begun, both in Australia and overseas. The idea of promoting more integrated communities and deliberately seeking to build up social capital has captured the imagination of politicians and public servants, as well as academics and the media, so there is considerable momentum for continued research and evidence-based policy development.
1. External economies, or externalities for short, are external or 'spillover' effects; outcomes of market transactions on individuals and groups other than those party to the transactions. They are 'free' to these recipients, and, because they are not 'captured' by the parties themselves, may be ignored by them in their decision-making. Externalities may be positive (beneficial) or negative (costly).
2. However, the word 'community' is absent from the literature on economics; it does not even appear in the Palgrave dictionary of economics. There is no designated field of community economics, unlike, for example, housing economics, transport economics and family economics.
4. It follows that Government policy is circumscribed in its influence on community formation and development. The role of policy is to ameliorate and smooth change, to allow communities to adjust to their comparative advantage, to foster adjustment, to provide information about opportunities, and to regulate change.
5. Perhaps the best indicator of the importance placed upon the role of location in economic analysis is that the sub-discipline of economic geography is rarely included in mainstream economic thought. Relatively few economic geography subjects are taught at economic schools and researchers in the field can normally be found in schools of geography. Krugman (1991) laments this fact and his is the first attempt to marry economic geography with international economics, and in particular with the field of international trade.
6. The relationship may be expressed algebraically. Sites for markets are defined according to the spatial density of the consumers and their purchasing power. There will be a measure of spatial density M, that equates the sum of the ratios of the purchasing power and distance between consumers, P. Hence, where D is the distance between consumers and the potential site Mr and there are s consumers.
7. SLAs are statistical local areas designed by the Australian Bureau of Statistics to represent local communities. In most cases they are the same as LGAs (local government areas) representing the population and area within a given local government jurisdiction..
8. This is similar to a component of Wilson's (1987) explanation for the increasing social isolation of inner city communities. He emphasised labour market changes and the moves of middle and working class employees (rather than employers) away from the inner city to the suburbs.
9. In fact, Borjas finds quite severe immobility in inner-city ethnic communities. Analysing US Census data he found the average Hispanic migrant lived in a neighbourhood that was 35 percent Hispanic, the average second-generation Hispanic lived in a neighbourhood where the population was 33 percent Hispanic, and the average third-generation Hispanic immigrant lived in a neighbourhood that was 29 percent Hispanic. Hispanic immobility, although large, is surpassed by African-Americans who, on average, live in a neighbourhood that is 55 per cent African-American, while only 11 percent of the US population as a whole is African-American.
12. They included separate neighbourhood characteristics in separate models. The characteristics included were: the unemployment rate, the percentage of professionals, an ABS index measuring both education and occupation, and mean income. They were measured at the post-code level except for unemployment, which was measured by local government area.
16. Earlier waves of the WVS were known as the European Values Study (EVS). The question asked was, 'Generally speaking, would you say that most people can be trusted, or that you can't be too careful in dealing with people?' Data are available at wvs.isr.umich.edu/index.html
17. 'Proactivity' in a social context means doing unsolicited community-minded things like picking up someone else's rubbish in a public place; 'value of life' means people's feelings about their own value or worth to society.
18. Social exclusion has become the catchall preferred term, in the European Union, to cover what was previously termed poverty or disadvantage. However, it is by no means a precise term. Some authors use it to mean multiple deprivation, others quite clearly regard a person as at least potentially socially excluded if they have one problem (e.g. they are unemployed, or a single parent, or disabled).
19. A recent issue of the Australian Journal of Labour Economics (Volume 4, No 2, June 2000-01) is devoted to discussion of the CDEP by five sets of authors. The papers discuss the impact of the CDEP on indigenous jobs growth and unemployment, as a stepping stone to employment, on prospects for employment generation in remote Australia, on welfare-to-work in two community case studies, and on discouraged workers.
20. The International Social Science Survey Australia, conducted annually at the Australian National University, has asked a range of questions on social capital since 1994. In addition the development of the Household and Income Dynamics of Australia (HILDA) survey at the Melbourne Institute will provide information for the same individuals over time, enabling us to trace both age effects and generational effects.
21. Two other 'contexts' are also listed under the heading of capital: knowledge not embodied in individuals (and hence not human capital) and socio-political context (Juster and Stafford 1985, chapter 6).
Acs, G & Dazinger, S 1993, 'Educational attainment, industrial structure, and earnings: 1973-87', Journal of Human Resources, vol. 28, pp. 618-48.
Adamson, D, Dearden, H & Castle, B 2001, Community regeneration: review of best practice, National Assembly for Wales, Cardiff.
Adams, D & Hess, M 2000, 'Competitive tendering and contracting out: a case study from the Australian health industry', Australian Journal of Public Administration, vol. 19, pp. 49-59.
Adams, D & Hess, M 2001, 'Community in public policy: fad or foundation?', Australian Journal of Public Administration, vol. 21, pp. 13-23.
Alesina, A & La Ferrara, E 2000, 'Participation in heterogeneous communities', The Quarterly Journal of Economics, August, pp. 847-904.
Almond, GA & Verba, S 1963, The civic culture, Princeton University Press, Princeton.
Andrews, D, Green, C & Mangan J 2002, Neighbourhood effects and community spillovers in the Australian youth labour market, research report no. 24, Australian Council for Educational Research, Melbourne.
Arrow, K 1999, 'Observations on social capital', in P Dasgupta & I Serageldin (eds), Social capital: a multifaceted perspective, Washington DC.
Australian Bureau of Statistics 2002, Measuring Australia's progress, ABS, Canberra.
Australian Journal of Labour Economics 2000-01, 'Indigenous social policy and the labour markets', vol. 4, no. 2, June.
Axelrod, R 1984, The evolution of cooperation, Basic Books, New York.
Banfield, E 1958, The moral basis of a backward society, Free Press, Glencoe IL.
Barnes, S, Kaase, M et al. 1979, Political action: mass participation in five western democracies, Sage, Beverley Hills, CA.
Bartik, T 1993, 'Who benefits from local job growth: migrants or the original residents', Regional Studies, vol. 27, pp. 297-311.
Baum, S, Stimson, R, O'Connor, K, Mullins, P & Davis, R 1999, Community opportunity and vulnerability in Australia's cities and towns, University of Queensland Press, Brisbane.
Baum, F, Palmer, C, Modra, C, Murray, C & Bush, R 2001, 'Families, social capital and health', in I Winter (ed.), Social capital and public policy in Australia, Australian Institute of Family Studies, Melbourne, ch. 10, pp. 250-75.
Becker, GS 1964, Human capital: a theoretical and empirical analysis with special reference to education, Columbia University Press, New York.
Becker, GS 1965, 'A theory of the allocation of time', Economic Journal, vol. 75, pp. 493-517.
Beer, A, Bolam, A & Maude, A 1994, Beyond the capitals: urban growth in regional Australia, AGPS, Canberra.
Benabou, R 1993, 'Workings of a city: location, education, and production', Quarterly Journal of Economics, vol. 108, pp. 619-52.
Benabou, R 1994, 'Human capital, inequality, and growth: a local perspective', European Economic Review, vol. 38, pp. 817-26.
Berkman, L & Glass, T 2000, 'Social integration, social networks, social support and health' in L Berkman & I Kawachi (eds), Social Epidemiology, Oxford University Press, New York.
Berkman, L & Syme, S 1979, 'Social networks, host resistance and mortality: a nine year follow-up of Alameda County residents', American Journal of Epidemiology, vol. 109, pp. 186-204.
Black, A & Hughes P 2001, The identification and analysis of indicators of community strength and outcomes, report to the Department of Family and Community Services (February).
Blanchard, O & Katz, LF 1992, Regional evolutions, Brookings Institute, Washington DC.
Blaug, M 1992, The methodology of economics, Cambridge University Press, Cambridge.
Blazer, D 1982, 'Social support and mortality in an elderly community population', American Journal of Epidemiology, vol. 115, pp. 684-94.
Bound, J & Holzer, HJ 2000, 'Demand shifts, population adjustments, and labor market outcomes during the 1980s', Journal of Labor Economics, vol. 18, no. 1, pp. 20-54.
Borjas, G 1994, Long-run convergence of ethnic skill differentials, working paper no. 4641, National Bureau of Economic Research, Cambridge MA.
Borjas, G 1995, 'Ethnicity, neighbourhoods, and human capital externalities', American Economic Review, vol. 85, pp. 365-90.
Borjas, G 2001, 'Long-run convergence of ethnic skill differentials, revisited', Demography, vol. 38, no. 3, pp. 357-61.
Bourdieu, P 1985, 'The forms of capital' in J Richardson (ed.), Handbook of theory and research for the sociology of education, Greenwood, New York, pp. 241-58.
Bray, J Rob 2000, Social indicators for regional Australia, Policy Research Paper no. 8, Department of Family and Community Services, Canberra.
Brooks-Gunn, J, Duncan, GJ, Klebanov, P & Sealand, N 1993, 'Do neighbourhoods influence child and adolescent development', American Journal of Sociology, vol. 99, pp. 353-93.
Buchanan, JM & Tullock, G 1962, The calculus of consent, University of Michigan Press, Ann Arbor.
Bureau of Rural Sciences 2001, Evaluation of the Rural Communities Program, September.
Burt, R 1992, Structural holes: the social structure of competition, Harvard University Press, Cambridge.
Case, AC & Katz, LF 1991, The company you keep: the effects of family and neighbourhood on disadvantaged youths, working paper no. 3705, National Bureau of Economic Research, Cambridge MA.
Caveye, J 2000, 'The role of government in community capacity building', Department of Primary Industries, Queensland, information series, Q199804.
Christaller, W 1966, Central places in Southern Germany, translated by CW Baskin, Prentice-Hall, cited in HW Richardson 1969, Elements of regional economics, Penguin, Harmondsworth.
Cohen, SS & Fields, G 1998, Social capital and capital gains, or virtual bowling in Silicon Valley, working paper no. 132, Berkeley Roundtable on the International Economy.
Cohen, S et al. 1997, 'Social ties and susceptibility to the common cold', Journal of the American Medical Association, vol. 277, pp. 1940-44.
Coleman, J 1988, 'Social capital in the creation of human capital', American Journal of Sociology, vol. 94, pp. 95-120.
Corcoran, M, Datcher L & Duncan G 1980, 'Most workers find jobs through word of mouth', Monthly Labor Review, August, pp. 33-5.
Corcoran, M et al. 1990, 'Effects of family and community background on economic status', American Economic Review, vol. 80, no. 2, pp. 362-66.
Cox, E 1995, A truly civil society: 1995 Boyer lectures, Australian Broadcasting Commission, Sydney.
Crane, J 1991, 'Effects of neighbourhoods on dropping out of school and teenage childbearing', in C Jencks & PE Peterson (eds), The urban underclass, Brookings Institute, Washington DC.
Cronbach, L 1983, Designing evaluations of educational and social programs, Jossey-Bass, San Francisco.
Crozier, M 1964, The bureaucratic phenomenon, University of Chicago Press, Chicago.
Datcher, L 1982, 'Effects of community and family background on achievement', Review of Economics and Statistics, vol. 64, no. 1, pp. 32-41.
Department of Family and Community Services 2001, Research and evaluation framework, Canberra.
Diener, E 1984, 'Subjective well-being', Psychological Bulletin, vol. 95, pp. 542-75.
Duncan, GJ, Connell, JR & Klebanov, P 1997, 'Conceptual and methodological issues in estimating causal effects of neighbourhood and family conditions on individual development', in J Brooks-Gunn, GJ Duncan & L Aber, Neighbourhood poverty: context and consequences for children, Russell Sage Foundation, New York.
Durlauf, SN 1994, 'Spillovers, stratification, and inequality', European Economic Review, vol. 38, pp. 836-45.
Durlauf, SN 1995, A theory of persistent income inequality, Social Systems Research Institute working paper no. 9443, University of Wisconsin-Madison.
Eckersley, R (ed.) 1998, Measuring progress: is life getting better?, CSIRO Publishing, Melbourne.
Edwards, B & Foley, MW 1998, 'Social capital and civil society beyond Putnam', American Behavioral Scientist, vol. 42, no. 2, pp. 1-15.
Ellwood, D 1986, 'The spatial mismatch hypothesis: are there teenage jobs missing in the ghetto?', in R Freeman & H Holzer (eds), The black youth employment crisis, University of Chicago Press, Chicago, pp. 148-85.
Erlinghagen, M & Wagner, GG 2002, 'Can voluntary work be a substitute for gainful employment? Should welfare recipients earn their benefits by doing voluntary work?', Australian Social Monitor, vol. 5, no.1, pp. 1-6.
Esping-Anderson, G 1990, The three worlds of welfare capitalism, Polity Press, Cambridge.
Etzioni, A 1968, The active society, Free Press, New York.
Evans, M 1989, 'Immigrant entrepreneurship: aspects of ethnic market size and linguistically isolated labor pools', American Sociological Review, vol. 54, pp. 950-62.
Evans, M 2000, 'Charity work: international differences and Australian trends', Australian Social Monitor, 3, July, pp. 15-23.
Evans, M & Kelley, J 2002, Australian economy and society 2001: education, work and welfare, Federation Press, Sydney.
Evans, W, Oates, W & Scwab, R 1992, 'Measuring peer group effects: a study of teenage behaviour', Journal of Political Economy, vol. 100, no. 5, pp. 966-91.
Fujita, M, Krugman, P & Venables, A 1999, The spatial economy: cities, regions and international trade, The MIT Press, London.
Fukuyama, F 1995, Trust the social virtues and the creation of prosperity, Free Press, New York.
Gans, H 1967, The Levittowners: ways of life and politics in a new suburban community, Pantheon Books, New York.
Giddens, A 1998, The third way: the renewal of social democracy, Polity Press, Oxford.
Giddens, A 2000, The third way and its critics, Polity Press, Oxford.
Ginther, D, Haveman, R & Wolfe, B 2000, 'Neighbourhood attributes as determinants of children's outcomes: how robust are the relationships?', Journal of Human Resources, vol. 35, no. 4, pp. 603-42.
Giorgas, D 2000, Community formation and social capital in Australia, paper presented at the 7th Australian Institute of Family Studies Conference, Sydney, 24-26 July 2000.
Glaeser, EL, Laibson, D & Sacerdote, B 2000, The economic approach to social capital, working paper no. 7728, National Bureau of Economic Research, Cambridge MA.
Glaeser, EL, Laibson, D, Schienkman, JA & Soutter, CL 2000, 'Measuring trust', The Quarterly Journal of Economics, August, pp. 811-46.
Goodin, R, Headey, B, Muffels R & Dirven, HJ 1999, The real worlds of welfare capitalism, Cambridge University Press, Cambridge.
Granovetter, M 1973, 'The strength of weak ties', American Journal of Sociology, vol. 78, pp. 1360-80.
Green, G, Tigges, L & Browne, I 1995, 'Social resources, job search and poverty in Atlanta', Research in Community Sociology, vol. 5, pp. 161-82.
Gregory, R & Hunter, B 1995, The macro economy and the growth of ghettos and urban poverty in Australia, discussion paper no. 325, Centre for Economic Policy Research, Canberra.
Headey, B & Wearing, AJ 1992, Understanding happiness, Longman Cheshire, Melbourne.
Helliwell, J & Putnam, R 1999, Education and social capital, working paper no. 7121, National Bureau of Economic Research, Cambridge MA.
Henderson, S, Byrne, DG & Duncan-Jones, P 1981, Neurosis and the social environment, Academic Press, New York.
Hood, CC 1998, The art of the state: culture, rhetoric and public management, Clarendon, Oxford.
House, J, Robbins, C & Metzner H 1982, 'The association of social relationships and activities with mortality: prospective evidence from the Tecumseh Community Health Study', American Journal of Epidemiology, vol. 116, pp. 123-40.
Hughes, P, Bellamy, J & Black, A 2000, 'Building social trust through education', in I Winter (ed.), Social capital and public policy in Australia, Australian Institute of Family Studies, Melbourne, ch. 9, pp. 225-49.
Jacobs, J 1961, The death and life of great American cities, Random House, New York.
Jacobs, J 1984, Cities and the wealth of nations: principles of economic life, Random House, New York.
Jacobs, J 2000, The nature of economies, Modern Library, New York.
Janssen, C 2000, De smeerolie van de maatschappy; een cross nationale en historische studie naar sociaal kapitaal in Europa, Masters thesis, Tilburg University.
Jencks, C & Mayer, S 1990, 'The social consequences of growing up in a poor neighbourhood', in LJ Lynn & M McGeary, Inner-city poverty in the United States, National Academy Press, Washington DC, pp. 111-86.
Jensen, B & Seltzer, A 2000, 'Neighbourhood and family effects in educational progress', Australian Economic Review, vol. 33, pp. 17-31.
Johnson, J & Oliver, M 1992, 'Structural changes in the U.S. economy and black male joblessness: a reassessment', in G Peterson & W Vroman (eds), Urban labor markets and job opportunities, Urban Institute, Washington DC.
Juster, FT & Stafford, F (eds) 1985, Time, goods and well-being, Michigan Institute for Social Research, University of Michigan, Ann Arbor.
Kain, J 1968, 'Housing segregation, negro employment and metropolitan decentralisation', Quarterly Journal of Economics, vol. LXXXII, no. 2, pp. 175-97.
Kasarda, J 1989, 'Urban transition and the underclass', The Annals of American Academy of Political Science, vol. 501, pp. 26-47.
Kelley, J & Evans, MDR 1999, 'Data surveys: Australian and international survey data for multivariate analysis: the ISSSA', Australian Economic Review, vol. 32, no. 3, pp. 298-303.
Kelly, R & Lewis, PET 2002, 'Neighbourhoods and youth employment outcomes in Melbourne', Australian Journal of Labour Economics, vol. 5, no. 1, pp. 61-76.
Knack, S & Keefer, P 1997, 'Does social capital have an economic payoff: a cross country investigation', The Quarterly Journal of Economics, pp. 1251-85.
Knack, S 2000, Social capital and the quality of government: evidence from the United States, World Bank Policy Research Working Paper no. 2504, December.
Krugman, P 1991, Geography and trade, Leuven University Press, Leuven Belgium, and The MIT Press, Cambridge.
Lancaster, K 1966, 'A new approach to consumer theory', Journal of Political Economy, vol. 74, pp. 132-57.
Lane, RE 1959, Political life: why and how people get involved in politics, Free Press, New York.
Laslett, P 1971, The world we have lost, Methuen, London.
Latham, M 1998, Civilising global capital: new thinking for Australian Labor, Allen and Unwin, Sydney.
Lucas, RE 1988, 'On the mechanics of economic development', Journal of Monetary Economics, vol. 22, no. 1, pp. 3-42.
McClure, P 2000, Participation support for a more equitable society: final report of the Reference Group on Welfare Reform, Department of Family and Community Services, Canberra.
McGregor, A & McCannachie A 1995, 'Social exclusion, urban regeneration and economic reintegration', Urban Studies, vol. 32, pp. 1587-1600.
Manski, CF 1990, 'The use of intentions data to predict behavior: a best case analysis', Journal of American Statistical Association, vol. 85, no. 412, pp. 934-40.
Manski, CF 1993, 'Identification of endogenous social effects: the reflection problem', Review of Economic Studies, vol. 60, pp. 531-42.
Maher, C 1999, 'Locational disadvantage and concentrations: a review and evaluation of the findings', in K O'Connor, Houses and jobs in cities and regions: research in honour of Chris Maher, University of Queensland Press, Brisbane, ch. 1, pp. 1-57.
Marshall, A 1920, Principles of economics, eighth edition (1947 print), MacMillan, London.
Milbrath, LW 1965, Political participation, Rand McNally, Chicago.
Milner, H 1994, Social democracy and rational choice: the Scandinavian experiment and beyond, Routledge, London.
Moore, JW 1978, Homeboys, Temple University Press, Philadelphia.
Nexus Management Consulting 1999a, Kings Cross Pilot Place Management Project preliminary evaluation: final report, NSW Premier's Department.
Nexus Management Consulting 1999b, Final report—review of the Cabramatta Project, NSW Premier's Department.
Niskanen, WS 1971, Bureaucracy and representative government, Rand McNally, Chicago.
Nordhaus, W & Tobin J 1972, 'Is growth obsolete?', in Economic growth (Fiftieth Anniversary Colloquium V), National Bureau of Economic Research, New York.
North, DC 1955, 'Location theory and regional economic growth', Journal of Political Economy, vol. 63, June, pp. 243-58.
O'Connor, K, Stimson, R & Daly, M 2001, Australia's changing economic geography: a society dividing, Oxford University Press, Melbourne.
O'Regan, KM & Quigley, JM 1998, Spatial effects upon employment outcomes: the case of New Jersey teenagers, working paper, University of California, Berkeley.
OECD 1998, Integrating distressed urban areas, OECD, Paris.
OECD 2001, The well-being of nations: the role of human and social capital, OECD, Paris.
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 (CACOM), University of Technology, Sydney.
Onyx, J & Bullen, P 2001, 'Sources of social capital', in I Winter (ed.), Social capital and public policy in Australia, Australian Institute of Family Studies, Melbourne, ch. 5, pp. 136-64.
Ostrom, E 1990, Governing the commons, Cambridge University Press, Cambridge.
Ostrom, E 1999, 'Social capital: a fad or a fundamental concept?', in P Dasgupta & I Serageldin (eds), Social capital: a multifaceted perspective, World Bank, Washington DC.
Orth-Gomer, K & Johnson, V 1987, 'Social network interaction and mortality', Journal of Chronic Diseases, vol. 40, pp. 949-57.
Osborne, D & Gaebler, T 1993, Reinventing government: how the entrepreneurial spirit is transforming the public sector, Penguin, New York.
Overman, HG 2000, Neighbourhood effects in small neighbourhoods, Centre for Economic Performance discussion paper 481, London School of Economics and Political Science, London.
Paxton, P 1999, 'Is social capital declining in the United States: a multiple indicator assessment', American Journal of Sociology, vol. 105, pp. 88-127.
Pearlin, LI & Schooler, C 1978, 'The structure of coping', Journal of Health and Social Behavior, vol. 19, pp. 2-21.
Plotnick, R & Hoffman, S 1999, 'The effect of neighbourhood characteristics on young adult outcomes: alternative estimates', Social Science Quarterly, vol. 80, no. 1, pp. 1-18.
Polanyi, K 1944, The great transformation, Beacon Press, Boston.
Porter, ME 1990, The comparative advantage of nations, Free Press, New York.
Portes, A & Landolt, P 1996, 'The downside of social capital', The American Prospect, vol. 26, pp. 18-22.
Portes, A & Sensenbrenner, J 1993, 'Embeddedness and immigration: notes on social determinants of economic action, American Journal of Sociology, vol. 98, pp. 1320-50.
Portes, A & Zhou, M 1993, 'The new second generation: segmented assimilation and its variants', The Annals of the Academy of Political and Social Science, vol. 530, pp. 74-96.
Putnam, R 1993, Making democracy work: civic traditions in modern Italy, Princeton University Press, Princeton.
Putnam, R 1995, 'Bowling alone: America's declining social capital', Journal of Democracy, vol. 1, pp. 65-78.
Putnam, R 2000, Bowling alone: the collapse and revival of American community, Simon & Schuster, New York.
Reich, RB 1992, The work of nations: preparing ourselves for 21st century capitalism, Vintage Books, New York.
Rhodes, J, Tyler, P, Brennan, A, Stevens, S, Warnock, C & Otera-Garcia, M 2002, Lessons and evaluations from ten single budget regeneration case studies, Department of Transport, Local Government and the Regions, London.
Robson, B, Bradford, M & Tomlinson, R 1998, Updating and revising the index of local deprivation, DETR Regeneration Research series, London.
Romer, P 1986, 'Increasing returns and long-run growth', Journal of Political Economy, vol. 94, no. 5, pp. 1002-37.
Ruggles, R 1981, 'The conceptual and empirical strengths and limitations of demographic and time based accounts', in FT Juster & KC Land (eds), Social Accounting Systems, Academic Press, New York.
Salamon, LM 1992, America's nonprofit sector, The Foundation Centre, New York.
Salvaris, M, Burke T, Pidgeon J & Kelman, S 2000, Social benchmarks and indicators for Victoria, Consultants' Report for the Department of Premier and Cabinet, Institute for Social Research, Swinburne University of Technology, Melbourne.
Sampson, RJ 1995, 'The community', in JQ Wilson & J Petersilia (eds), Crime, Institute for Contemporary Studies Press, San Francisco.
Saxenian, A 1994, Regional advantage: culture and competition in Silicon Valley and Route 128, Harvard University Press, Cambridge MA.
Schiff, M 1998, Trade, migration and welfare: the impact of social capital, November, World Bank.
Seeman, TE 1996, 'Social ties and health: the benefits of social integration', Annual of Epidemiology, vol. 6, pp. 442-51.
SEU (Social Exclusion Unit) 2001, A new commitment to neighbourhood renewal, Cabinet Office, London.
Smith, GR 1999, Area-based initiatives: the rationale and options for area targeting, Centre for the Analysis of Social Exclusion (CASE) Paper No. 25, London School of Economics.
Sobel, J 2002, 'Can we trust social capital?', Journal of Economic Literature, vol. 40, pp. 139-54.
Solow 1999, 'Notes on social capital and economic performance', in P Dasgupta & I Serageldin (eds), Social capital: a multifaceted perspective, Washington DC.
Sparrow, MK, Moore, MH & Kennedy, DM 1990, Beyond 91: a new era for policing, Basic Books, New York.
Stone, Sir R 1973, 'A system of social matrices', The Review of Income and Wealth, vol. 19, pp. 143-66.
Stone, W 2001, Measuring social capital, Australian Institute of Family Studies Research Paper No. 24, February, Melbourne.
Stone, W & Hughes, J 2001, 'What role for social capital in family policy—and how does it measure up', paper presented to 'Family futures: issues in research and policy', Australian Institute of Family Studies Conference, Sydney, July 24-26.
Tam, H 1998, Communitarianism: a new agenda for politics and citizenship, Macmillan, Basingstoke.
Tanner, L 1999, Open Australia, Pluto Press, Sydney.
Temple, J & Johnson, PA 1998, 'Social capability and economic growth', Quarterly Journal of Economics, vol. CXIII, pp. 965-90.
Thoits, PA 1983, 'Conceptual, theoretical and methodological problems in studying social support as a buffer against life stress', Journal of Health and Social Behavior, vol. 23, pp. 145-59.
Tiebout, CM 1956, 'A pure theory of local expenditures', Journal of Political Economy, vol. 64, pp. 416-24.
Topel, R 1986, 'Local labour markets', Journal of Political Economy, vol. 94, pp. S111-S143.
Townsend, P 1979, Poverty in the United Kingdom, Penguin, Harmondsworth.
Vanstone, A 2001, 'The third way', speech July 13.
Veenhoven, R 1996, 'Developments in satisfaction research', Social Indicators Research, vol. 37, pp. 1-46.
Verba, S & Nye, N 1972, Participation in America, Harper and Row, New York.
Vinson, T 1999, Unequal in life: the distribution of social disadvantage in Victoria and New South Wales, Jesuit Social Service, Melbourne.
Von Thunen, JH 1826, The isolated state, trans. C.M. Wartenburg, Pergamon Press, London.
Walsh, P 2001, 'Improving government's response to Local Government—is place management an answer?', Australian Journal of Public Administration, vol. 2, pp. 3-12.
Warren, ME (ed.) 1999, Democracy and trust, Cambridge University Press, Cambridge.
Watling, R 1999, 'The high life: the apartment boom in Melbourne', in K O'Connor, Houses and jobs in cities and regions: research in honour of Chris Maher, University of Queensland Press, Brisbane, ch. 4, pp. 97-105.
Welin, L, Tibblin, G et al. 1985, 'Prospective study of social influences on mortality', The Lancet, April 20, pp. 1915-18.
Williamson, OE 1996, The mechanisms of governance, Oxford University Press, New York.
Wilson, WJ 1987, The truly disadvantaged, University of Chicago Press, Chicago.
Winter, I 2000a, 'Major themes and debates in the social capital literature: the Australian connection', in I Winter (ed.), Social capital and public policy in Australia, Australian Institute of Family Studies, Melbourne, ch 1.
Winter, I 2000b, Towards a theorised understanding of family life and social capital, Australian Institute of Family Studies Working Paper No. 21, April, Melbourne.
Winter, I & Stone, W 1999, 'Social polarisation and housing careers: exploring the interrelationship of labour and housing markets in Australia', in K O'Connor, Houses and jobs in cities and regions: research in honour of Chris Maher, University of Qld Press, Brisbane, ch. 1, pp. 59-79.
World Bank 2002.
Local Government Managers Australia website.
Zubrick, SR, Williams, AA, Silburn, SR & Vimpani, G (2000) Indicators of Family Functioning, TVW Telethon Institute for Child Health Research, Perth, Western Australia and Child and Youth Health Network, University of Newcastle.