Summer courses. 2007

SOCIAL CAPITAL MEASUREMENT AND APPLICATIONS.

Laura Iisakka Statistics Finland. Helsinki. Finlandia

Summarie

There is still no single, universal definition of social capital. The concept of social capital is most com-monly used to refer to social networks, norms of reciprocity and trust. These characteristics of the social structure promote interaction between individuals, facilitate the coordination of activities and support the attainment of collective and individual goals. Furthermore, social interaction generates a greater sense of community spirit.

In spite of the absence of a universal definition, the measurement of social capital has attracted much research interest in recent years. Still there is no generally accepted way to measure it. Distinctions are often made between key components of social capital, which are then used to measure the concept. There is general acceptance of some key indicators suitable for the measurement of social capital, e.g. networks and trust in other people. Measurement of networks often make a distinction between bonding and bridging networks (see Putnam 2000, p. 22–23.). Bonding networks describe closely-knit groups of people who are very similar to one another and who know one another. Bridging refers to networking among different kinds of people or groups. Measure of bridging networks consicts usually indicators of participation in different organisations and bonding networks deals about meeting friends and relatives.

Whose social capital should we measure? Social capital is mostly understood as a group rather than an individual attribute. In practice, this community level concept is usually measured by using individual based surveys and still, social capital is often measured indirectly by using proxies. Recently it seems to be generally accepted that social capital can be seen both as a property of an individual and of a society or community. David Halpern (2005) has distinguished social capital in to three different levels which are: 1) micro or individual level, 2) meso or community level and 3) macro or society level. Halpern points out that the interaction between these levels is extremenly significant.

The most well-known studies about social capital and its measurement are propably Robert D. Putnam’s Making Democracy Work (1993) and Bowling Alone (2000). In Making Democracy Work Putnam fo-cused on the role of social capital in explaining regional differences in economic and social develop-ment between southern and northern Italy. Economical development was better in regions where people had more social capital, such as they belonged to civic organizations, where interested about local mat-ters, read newspapers etc. In Bowling Alone, Putnam was interested in long term trends of social capital in the USA. He also measured social capital in the American States and found that social capital dif-fered from very low to very high level between different states. His study was based on an extensive database collected from different materials.

Social capital measurement is usually based on examining different indicators that represent social capi-tal’s key components. This has been done in Finland (Social Capital in Finland -statistical review 2006), as well as in Australia (Aspects of Social Capital, 2006) and in Canada (PRI’s reports in 2005). Some comparisons between different regions and countries has been done, also. The problem is that its not very easy to compare social capital between countries because the data used in the measurement is usu-ally different in different countries. Analyses that integrate different data materials usually combine different indicators as well as questions that are worded differently in different countries.

Social capital measurement has usually been done in the level of community, like comparisment be-tween different countries or regions or on a local or neigbourhood level. Reseachers are usually inter-ested in the level of social capital and the trends. Measures are often based on just the few single indica-tors. For example, in many cross-national studies trust is often measured on the basis of one single question from the World Values Survey. There are also few and hopefully growing amount of studies based on more wide-ranging and extencive measurement, such as on multivariate analysis (Sabatini 2005). Actually, social capital understood as a multidimensional concept needs to be analysed more wide-range. Biggest problems for this are the lack of comprehensive data and also the lack of generally agreed definition or as well as generally agreed way to measure this concept.

The sources used in social capital measurement usually consist of 1) existing interview and question-naire materials, 2) register data and statistics and 3) separate datasets on social capital. Measurements often make use of various parallel datasets (e.g. Putnam 1993; 2000). Most statistics and materials used in measuring social capital have originally been compiled for some other purpose, which certainly ups the challenge and detracts from the comparability of the results. A further source of difficulty is that community-level characteristics have to be examined against data describing the actions and opinions of individuals. (Simpura 2002.) A single dataset rarely provides a strong enough foundation for a compre-hensive and diverse examination of social capital, but it is necessary to use side by side a variety of different materials at different levels. Measurement of social capital has suffered from the lack of suit-able data. Indeed researchers in English-speaking countries have in recent years been working to collect separate datasets on social capital. Questions on social capital were included in Australian General So-cial Survey (GSS) collected in 2006 as well as in Canada already in the year 2003, also part in GSS-survey. The bank of questions was used also in the 2004 General Household Survey (GHS) in Great Britain.

Discussions and debates on the measurement of social capital have largely taken a theoretical and con-ceptual perspective. Theoretically reseachers are interested about whether social capital is really a form of capital or about whether social capital is a property of a community or an individual. Internationally, there has been a growing interest in the early 2000s to incorporate social capital in official statistics and to develop a harmonised system of measurement. Most of the work to compile statistics on social capi-tal is concentrated in the English-speaking countries of the world: national statistical offices in the UK, Canada, New Zealand and Australia have worked on various projects to develop tools and methods for the measurement of social capital (Harper & Kelly 2003; PRI 2005a; Spellerberg 2001; ABS 2004).

These statistical frameworks are for technical reasons based on questionnaire and interview materials collected at the individual level even though national statistical offices mainly consider social capital as a character of community.

The New Zealand Statistical Office introduced its first framework for the measurement of social capital as early as 1997. The aim was to determine whether social capital had potential for policy development at both the local and central government levels. A revised and modified version of the framework was published in 2001 (Spellerberg 2001). Its four main components are:

1. Behaviour: Helping and supporting others (giving time, money, blood, etc.), partici-pation in formal and informal networks, compliance with rules and norms, wider interest in society.
2. Attitudes and values: Trust and reciprocity, attitudes to government and social institutions, at-titudes towards self and others and confidence in the future.
3. Population groups: Demographic factors, family, culture, employment and communication.
4. Organisations: Numbers, type, size, structure and cooperation between organisations.

The Australian statistical framework is one of the most comprehensive systems developed for the measurement of social capital. The Australian concept of social capital is built around networks, which are divided according to their type, quality, structure and transactions taking place within the networks.

The ABS framework for the measurement of social capital:

1. Network qualities: norms (trust, reciprocity, cooperation),common purpose (participation, civic participation, helping others, friends).
2. Network structure: size, number, intensity, density and openness, etc.
3. Network transactions: sharing support, knowledge and information, negotiation, sanctions.
4. Network types: bonding, bridging, linking, isolation.

Much work has also been done in the UK to develop a framework for the measurement of social capital. The UK measurement framework has five main dimensions. In addition, it includes examples of indica-tors with which each dimension can be measured. The main dimensions of the UK statistical frame-work are as follows:

1. Social participation
2. Civic participation
3. Social networks and social support
4. Reciprocity and trust
5. Views of the local area

The OECD and the World Bank have also been active in developing statistics on social capital (see e.g. Grootaert et al. 2004). Their main interest has been to explore social capital alongside human capital. The OECD has hosted to international conferences on the measurement of social capital. The main fo-cus at the conference in London in 2002 was on the experiences gained in different countries with the systematic measurement and analysis of social capital. The second conference in Budapest in 2003 was convened to develop an internationally harmonised measurement system and questionnaire for social capital. The statistical framework outlined in Budapest comprises four main dimensions of social capi-tal: 1) social participation, 2) social networks and support, 3) reciprocity and trust and 4) civic participa-tion (OECD 2003).

The World Bank launched in 1996 the Social Capital Initiative (SCI), which has involved the develop-ment of a model for the measurement of social capital. It has been particularly interested in developing methods of measuring social capital as part of an action programme aimed at preventing poverty and at boosting economic growth in the developing countries. The World Bank has used various different sta-tistical frameworks. One example is the Integrated Questionnaire (SC-IQ) that was developed at the World Bank in the early 2000s for the measurement of social capital. SC-IQ covers the following themes (Grootaert et al. 2004):

1. Groups and networks
2. Trust and solidarity
3. Collective action and collaboration
4. Information and communication
5. Social cohesion and inclusion
6. Empowerment and political action

As can be seen above, work in developing international harmonised statistical standards is still in the blueprint stage. The key element of the concept of social capital in its measurement frameworks is that of networks. There can be found some common dimensions from many of these statistical frameworks which are trust, social participation, civic participation as well as formal and informal networks.

In addition to measure the level and trend of social capital, there has been a growing need to measure also the total volume of social capital. Especially economists consider that it is quite necessary to bring social capital concept and it’s measurement to the side by national accounts or to satellite accounts. Essential question then is that to what extent social capital can be quantified in money values and are there ways in which it can be measured in a meaningful way. To be fully usable for such purposes, so-cial capital needs to be assessed at national and regional levels in an internationally or inter-regionally comparable way. Attempts to proceed in this direction have been many but achievements scarce. The problem is that it has not been possible to measure the total volume in Finland as well as in other coun-tries either. Some detailed empirical works where some kind of total index of social capital can be found. For instance, Fabio Sabatini’s has done a multivariate analysis in Italy (Sabatini 2005).

Another relevant strain in the economic debate of the last few years is the discussion of the World Bank on extending the notion of capital into aspects of wealth and sustainable development (Where is the Wealth of Nations, 2006; see also Hamilton and Ruta 2006) in which social capital has been mentioned. Hamilton and Ruta (2006), find out that what they call intangible capital has a majority share in all capital, and that this share is the highest in the economically most advanced countries. Intangible capi-tal includes human capital, governance, institutional effectiveness and all other capitals not mentioned elsewhere. The last group, all other capitals, includes social capital. As for the method of assessment, Hamilton and Ruta mention the method of “difference” for all forms of intangible capital. This method does not include direct measurement of the volume of capital and changes therein for any forms of in-tangible capital.

There’s also a common need to find out more about the multidimensional nature of social capital. It is not satisfactory to examine just a single dimension of the concept and consider it as representative of the concept as a whole, and analyze its effects on the economic performance. This is a reason why there’s a general need to create a way to measure total volume of social capital in different countries which is still missing in the measure of social capital. These measures could then be used to make esti-mates for national accounts, comparisons over time, comparisons between both at international and territorial level inside a country as well as to analyze the impact of social capital on competitiveness and economic growth.

OUTCOMES OF SOCIAL CAPITAL

One of the key aspects of the concept and related research is that social capital is thought to have some outcome, that social capital has an impact and influence on something. The reason why social capital is so popular is partly because it is thought to have societal meaning and implications, it is thought to bring benefits to individuals and more broadly to communities and societies. It is often used to explain the success and well-being of societies, communities or individuals.

Examples of the outcomes of social capital include improved public administration and democracy and greater well-being in society (Putnam 1993; 2000), improved health (Hyyppä 2002, Nieminen 2006), or economic growth and efficiency (Hjerppe 2003; Knack & Keefer 1997). Social capital seems to affect also on companys’ productive capabilities and overall well-being at work (Pärnänen 2006). Social capi-tal is one alternative means for explaining the unexplained, or residual, variation in economic develop-ment between countries or regions (e.g. Hjerppe 1998, 2003; Woolcock, 2000). David Halpern (2005) has explored the impacts of social capital on job opportunities, incomes, the safety of the neighbour-hood, health and well-being. There is also evidence of a link between crime and social capital (PRI 2005c). Although it is often difficult to establish causal relations, a selective focus on specific factors does produce quite clear results in the case of health, for example.