Civic Engagement and Social Cohesion: Measuring Dimensions of Social Capital to Inform Policy (2014)

1

Introduction

1.1. WHY MEASURE CIVIC ENGAGEMENT AND SOCIAL COHESION?

People’s bonds, associations, and networks—as well as the civil, political, and institutional characteristics of the society in which they live—can be powerful drivers affecting the quality of life among a community’s, a city’s, or a nation’s inhabitants and their ability to achieve both individual and societal goals. Civic engagement, social cohesion, and other dimensions of social capital affect social, economic and health outcomes and, therefore, measurement of these phenomena is in the public interest.

The development in 2000 of the Social Capital Community Benchmark Survey by the Saguaro Seminar at Harvard University advanced the idea that distinct dimensions of social capital could be identified and measured. The survey built on the work of Coleman (1988), Putnam (2000), and many others who have argued that attributes commonly discussed under the rubric “social capital”—political participation; engagement in community groups and associations; connectedness with friends and family and neighbors; attitudes toward and relationships with neighbors, government, groups unlike one’s own, and the like—are often positively associated with a range of desirable outcomes in such areas as health, altruism, compliance with the law, child welfare, and even self-reported well-being. However, those attributes may in some instances contribute to negative outcomes as well, depending on how community and group

resources are used.¹ Portes (1998) provided a balanced assessment of both the positive functions of social capital—for example, as a source of network-mediated benefits that are important for occupational mobility—and of the potentially negative consequences of the same processes, such as when privileged access to jobs by a specific ethnic group (or graduates from certain colleges) restricts the opportunities of outsiders.² While this capacity for negative effects is generally accepted, it does not nullify the widespread view that steps to increase social capital under conditions that lead to social benefits should be pursued (see, e.g., Halpern, 2004; OECD, 2001).

Data on civic engagement, social cohesion, and other aspects of social capital—terms we define below—have been collected for many years and for many purposes. To varying degrees, such data have been used to document conditions of policy importance, inform and enlighten the public more generally, and underpin social science research. Studies of these phenomena have raised critical questions, about casual relationships for example, but have also introduced new ways of thinking about the workings of civic society.

For half a century, the U.S. government has collected data and produced statistics on political participation and more general aspects of civic engagement; comparatively less has been done to measure social cohesion. Voting and registration data were first collected in the November 1964 supplement to the Current Population Survey (CPS) of the Census Bureau; data collection has been biennial since then. Data on volunteering were first collected in an April supplement to the CPS and again in a May supplement to the 1989 CPS.

With funding from the Corporation for National and Community Service (CNCS), an independent federal agency, annual collection of data on volunteering began with the September 2002 CPS supplement. Beginning in 2003, the American Time Use Survey (ATUS), administered monthly to outgoing rotation groups of the CPS, has collected time-use diaries on relevant activities, including volunteering, political participation, and other aspects of civic engagement. The ATUS also featured a module on

__________________

¹At the extreme, Satyanath et al. (2013) trace, town by town, how the rise of Nazism was facilitated by unusually high levels of social capital—specifically a dense network of clubs and associations—in Weimar, Germany.

²Putnam has described this side of social capital as well. His public view evolved shortly after writing Making Democracy Work, in which he defined social capital as something that had to be positive for society, to explicitly acknowledge that social networks can lead to negative consequences. See http://www.the-american-interest.com/articles/2008/1/1/bowling-with-robert-putnam/ [May 2014].

subjective well-being in 2010, 2012, and 2013; it included questions on both experienced well-being and life evaluation.³

In reviewing these efforts and offering guidance for their continuation and improvement, the panel synthesizes and adds to the foundations developed by many others. Beginning in 2006, the National Conference on Citizenship (NCoC; a congressionally chartered independent organization), CNCS, and the Center for Information and Research on Civic Learning and Engagement at Tufts University (CIRCLE) undertook to develop indicators of “civic health,” drawing from several ongoing surveys and specially commissioned small-scale surveys. The goal of the partnership was to insert relevant questions into federal surveys and, in particular, to establish a regular supplement to the CPS. In 2008, funded by CNCS, the November CPS supplement became the Voting and Civic Engagement Supplement; it included questions related to “civic health” in addition to those previously asked about voter and nonvoter characteristics and trends. In 2009 and 2010, a shorter, temporary list of questions was fielded in the supplement. The 2012 module was suspended for budgetary reasons, but both the civic engagement and volunteer supplements were restored with something close to the original battery of questions for 2013, albeit with half samples.

In 2009, the effort to make civic health and related indicators a staple of the federal government’s statistical programs obtained statutory support in the Serve America Act (H.R. 1388). This act reauthorized and expanded national service programs administered by the Corporation for National and Community Service, and called for “sponsored data collection” for assessment of civic health indicators related to “(A) volunteering and community service; (B) voting and other forms of political and civic engagement; (C) charitable giving; (D) connecting to civic groups and faith-based organizations; (E) interest in employment, and careers, in public service in the nonprofit sector or government; (F) understanding and obtaining knowledge of United States history and government; and (G) social enterprise and innovation.” The Act directed the Census Bureau and Bureau of Labor Statistics to collect this information—along with data that would allow analysis “by age group, race and ethnicity, education level, and other demographic characteristics of the individuals involved” (H.R. 1388, p. 75)—annually if possible, to inform the civic health assessment volumes published by NCoC.

Much of the intellectual content underlying the first (November 2008) CPS Civic Engagement Supplement was compiled by or originated with

__________________

³Historical time use data are also available from surveys fielded by the University of Michigan in 1965, 1975-1976, 1981, and 1985 and by the University of Maryland in 1992-1994, 1998, and 2001.

the Harvard Saguaro Seminar. As noted above, the seminar began the Social Capital Community Benchmark Survey (SCCBS), the first large-scale measurement of social capital variables.⁴ From this foundation, the Saguaro Seminar convened an informal steering group of social scientists to advise on what questions should be included in the CPS supplement module.⁵ Since its beginning, the seminar’s mission has been to improve social capital data and measurement and to investigate ways to build social capital at community and other levels.

Countries other than the United States have recognized the public importance of civic engagement and social cohesion and have initiated data collection programs for their measurement. In some cases, national statistical offices have been the leaders: one example is Statistics Canada, with such efforts as the 1996 General Social Survey on Social and Community Support.⁶ Work on broad social well-being concepts is also underway in international agencies. For example, the World Bank, in an effort to understand causes, manifestations, and consequences of poverty, has engaged in a number of efforts to measure community engagement in developing countries through its Global Social Capital Survey.⁷

Statistical programs to measure population well-being were given additional impetus by the influential Report by the Commission on the Measurement of Economic Performance and Social Progress (Stiglitz et al., 2009). The key argument in that report was that gross domestic product (GDP) alone is not a satisfactory measure of the welfare of a population. The report recommended a shift in the focus of measurement from market production toward “people’s well-being,” and cited the relevance of social capital and its association with self-reported well-being.⁸

__________________

⁴For details, see http://www.hks.harvard.edu/saguaro/communitysurvey/ [February 2014]. The SCCBS (N = 30,000) was also fielded in 1992-1994, 1998, 2001 and 2006. An abridged 5-10 minute version of the 25-minute Benchmark Survey has also been developed.

⁵For a detailed account about the process whereby the 100+ questions from the SCCBS were streamlined into the much abbreviated CPS Supplement instrument, see Hudson and Chapman (2002).

⁶Franke (2005) comprehensively documented the many efforts internationally to define and measure social capital and related concepts.

⁷Among the “ground up” initiatives by the World Bank was the work by Narayan and Pritchett (1999) to construct a measure of social capital based on individuals’ associational activities, and trust in people and institutions, using the Tanzania Social Capital and Poverty Survey.

⁸The idea that societal well-being and progress should be measured more broadly than GDP long predated this report; it was most conspicuous during the social-indicators movement among social scientists and public policy analysts during the 1960s. The international standard for compiling national accounts—the UN System of National Accounts—has long recognized this to be the case.

Policy Relevance

Informing policy decisions is the primary rationale for government statistics.⁹ Expanding statistical coverage of topics previously unmeasured frequently follows from research findings that identify factors influencing social conditions and behaviors that have obvious program and policy importance. For example, in early childhood development, research documented how the early treatment of children bears on subsequent educational and employment outcomes.

The rationale for expanding government data collection draws on sociological theory about why phenomena now summarized under the label “social capital” are broadly consequential for the functioning of societies. This theory dates most notably to Alexis de Tocqueville (Democracy in America) and to Emile Durkheim (1964, p. 28), who wrote that “A nation can be maintained only if, between the state and the individual, there is interposed a whole series of secondary groups near enough to the individuals to attract them strongly in their sphere of action and drag them, in this way, into the general torrent of social life.” Drawing on this theory, scholars have comparatively recently begun systematically studying dimensions of social capital and outcomes relevant to policy. Below are examples (some of this literature is reviewed in more detail in Chapter 2)

• Measures of isolation or lack of social connection (such as the Social Network Index, which takes into account marital status, frequency of contact with other people, participation in religious activities, and participation in other club or organization activities) has under some conditions been as predictive of premature death as such clinical risk factors as smoking and hypertension (Berkman and Glass, 2000; Pantell et al., 2013; Steptoe et al., 2013).
• Neighborhood networks and characteristics have a significant impact on crime and safety (Sampson, 2006).
• The condition and development of social infrastructure help explain a community’s resilience to natural disasters, such as hurricanes (Adger et al., 2005).¹⁰
• The effect of immigration and ethnic diversity on the social cohe-

_______________________

⁹“Policy” extends to beyond government actions; corporations, universities, churches, charities, and other organizations also have policies that can be informed by data on civic health and elements of social capital. For example, many institutions have “diversity policies” that can be better informed through an understanding of society provided by government statistics.

¹⁰Klinenberg (2013) discussed how cities adapt and may best survive.

These are suggestive rather than definitive research findings, but they are sufficient to warrant greater investments in data gathering for policy purposes. In the area of public health, the need for evidence linking social capital to risk factors such as smoking or obesity is an obvious example.

Public Information and Research Needs

A related rationale for improved data is the need for descriptive statistics that inform general public awareness about the state of society, where it has been, and where it is going. Data produced by government agencies that enter the official statistical system have common attributes, including high-quality standards, transparency, accessibility, and related professional norms. These norms guide practice in national statistical offices around the world and have been codified in principles promoted by the U.N. Economic and Social Council (see Box 1-1).¹³

In the United States, because of their importance to decision mak-

__________________

¹¹See, among others, Farley and Alba (2002), Hirschman (2001), Portes and Rumbaut (2001), Rong and Brown (2001) and Waldinger and Feliciano (2004). Massey et al. (1993) is a seminal work depicting the role of networks in migration.

¹²Van Kemenade et al. (2006) found that “having access to close networks of people from the same cultural origin—as well as to programs that support these networks—is associated with the social and economic integration of immigrants in the host county and with their well-being.”

¹³See, also, Principles and Practices for a Federal Statistical Agency, a report periodically updated by the U.S. National Academy of Sciences’ Committee on National Statistics (National Research Council, 2013b), which identified “four basic principles that statistical agencies must embody in order to carry out their mission fully:

(1) They must produce objective data that are relevant to policy issues,

(2) They must achieve and maintain credibility among data users,

(3) They must achieve and maintain trust among data providers, and

(4) They must achieve and maintain a strong position of independence from the appearance and reality of political control.”

The book also described 11 important practices to uphold these principles.

ers, some series—including gross domestic product (GDP), the consumer price index (CPI), and unemployment statistics—have been designated key economic indicators and special rules have been devised to ensure their unbiased and orderly dissemination. Because these statistical series are closely tied to economic policy and in some instances are used to adjust key government programs such as the level of social security payments, they appear on a publicly scheduled cycle. Many surveys are conducted less frequently or less regularly but, nevertheless, generate information that is useful to researchers and for descriptive monitoring purposes; the CPS supplements on civic engagement and volunteerism are examples. Over time, as knowledge deepens, these data may become essential to informing policy (or markets, or other kinds of decision makers), and the timing and process by which they are collected and disseminated may change accordingly. Indeed, the potentially critical importance

of social capital variables in explaining and perhaps predicting change in society is a strong argument for data collection.

Data that are initially primarily descriptive, when accumulated over time, may allow researchers to test hypothesized relationships among variables. For example, correlational analysis has demonstrated an association between neighborhood characteristics and school performance. When these are strong and consistent correlations, taking action can be justified—even in the absence of fully developed causal tests.

1.2. CHARGE TO THE PANEL

Statement of Task

The formal charge or statement of task to the Panel on Measuring Social and Civic Engagement and Social Cohesion in Surveys was as follows:

This charge recognizes a number of related concepts and terminologies that are introduced here and considered in greater detail in Chapter 2. There are few standardized definitions for these terms, and terminological confusion, inconsistency, and ambiguity characterizes much of the research literature on which this report draws and summarizes.

“Civic engagement” has been characterized as comprising the activities of individuals that are oriented toward making “a difference in the

civic life of…communities and developing the combination of knowledge, skills, values and motivation to make that difference. It means promoting the quality of life in a community, through both political and nonpolitical processes” (Ehrlich, 2000). Activities include but are not limited to participating in community organizational life through elections, attending public meetings, and joining in community projects. Civic engagement can occur at neighborhood and local levels, and also at national and international levels. Volunteerism is one defining characteristic of civic engagement in that most if not all such activities are discretionary.¹⁴ Although voting and direct political engagement are the most frequently measured indicators, they constitute a subset of what is treated as civic engagement.

Social cohesion refers to the extent to which groups—from communities to nations—are bound together by harmonious relations, work together, and feel obligated to act toward common purpose. Social cohesion is difficult to measure, given its many and complex dimensions: a shared sense of morality, values, and common purpose; levels of social order; extent of social solidarity created by income and wealth equalities; social interaction within and across communities or families; and sense of belonging to place. Inversely, as articulated by Forrest and Kearns (2001, pp. 2128-2129), “by implication, a society lacking cohesion would be one which displayed social disorder and conflict, disparate moral values, extreme social inequality, low levels of social interaction between and within communities and low levels of place attachment.”

The geographic unit of analysis (spatial scale) is an essential dimension of social cohesion. Neighborhoods, states, or other groups could be in conflict with one another while demonstrating strong social cohesion internally. This possibility puts a premium on being clear in specifying how social cohesion is formed and that it functions at levels from family to countries, and many levels between. Important research in this area includes work by Sampson et al. (e.g., 2012b) on “collective efficacy,” the willingness of a community’s residents to intervene on behalf of the common good,¹⁵ and by Putnam and others on bonding and bridging capital that manifests as social cohesion within and across group structures

__________________

¹⁴Fischer (2010) identified the historical roots of volunteerism in 18th century and discussed the persistence of both the attitudes and institutions that sustain and reproduce it.

¹⁵Sampson’s work focused on the social cohesion of Chicago residents in terms of their inclination to get involved in righting social disorders, like children skipping school and hanging out on a street corner, children spray-painting graffiti, children disrespecting an adult, or residents fighting in front of one’s house (Sampson et al., 1997).

respectively.¹⁶ In contrast to civic engagement, which can be measured at the individual level and then aggregated to describe groups, social cohesion is a group property to begin with, and study of it often requires more complicated research methods.

Social capital is used in our report as an umbrella term; civic engagement and social cohesion are often, but not always, treated as dimensions of social capital (in Chapter 2, we give greater attention to the multiple dimensions of social capital). These constructs, though malleable, are treated in this report with as much specificity as feasible—thus, for example, we refer to voting, neighborhood resilience, and connectedness with friends rather than civic engagement, social cohesion, and social capital, except when a label is needed to denote the full breadth of phenomena under consideration. In so doing, we accept the conclusion of Sobel (2002, p. 145) who argued that, even though “the strengths of the analogy [to other forms of capital] are not persuasive enough to justify the terminology,” the use of the term “social capital” is justifiable because existing literature has established “convincing evidence that the topics under the social capital umbrella are worthy of study, and application of economic principles can provide important insights. A vague keyword is not sufficient reason to condemn a promising line of research.”

We are further guided by practice in statistical agencies. As described by Ruston (2002, p. 14), the U.K. Office of National Statistics identified five dimensions of social capital (used as the umbrella term): Social Participation, Social Engagement, and Commitment; Level of Empowerment, Control, Self-efficacy; Perception of Community; Social Networks, Social Support, and Social Interaction; and Trust, Reciprocity, and Social Cohesion. In a World Bank publication, Grootaert et al. (2003) develop an “Integrated Questionnaire” for measuring social capital across six domains: Groups and Networks, Trust and Solidarity, Collective Action and Cooperation, Information and Communication, Social Cohesion and Inclusion, and Empowerment and Political Action.

In the research literature, Putnam (1993, 2000) emphasized social values (especially trust) and social networks (especially voluntary associations) along with values and norms as pre-conditions for a well-functioning civil society and prosperous economy. Civic engagement—participation in public affairs—is part of Putnam’s conception of active

__________________

¹⁶“Bonding capital” stands in contrast to “bridging capital,” which refers to the type of social capital that links or cuts across different communities or groups. The extent and balance of bridging and bonding social networks help determine whether a community, even if civically active, is civically unhealthy, characterized by many sociometric islands that are not interconnected. Beyond the United States, Varshney (2001) studied the correlation between the presence of interethnic networks (bridging) versus intra-ethnic ones (bonding) on ethnic violence in India.

citizens producing cohesive societies. Individuals’ social relationships and connectedness is another frequently identified component of social capital. Bourdieu (1986, p. 249) emphasized these: “The volume of social capital possessed by a given agent…depends on the size of the network of connections that he can effectively mobilize.”

“Civic health” also appears in the charge to the panel and could be accepted as an appropriate construct for organizing a set of indicators and for setting measurement priorities. Judging from the many civic health indexes in use (e.g., the NCoC state and national indexes), it is clear that many working in the area believe that term to be useful. Civic health has the added benefit that it is a concept that can be applied at the national level as well as to smaller geographic designations. Rating the civic health of a city, state, or nation involves a normative assessment drawing on a full array of measures and indicators ranging across civic engagement, social cohesion, and other aspects of social capital. In this sense, it is analogous to an assessment of economic health, which may be based on a range of measures that can be given different weights. One analyst might weight the employment rate or the number of jobs created more heavily while another one might give greater weight to wage rates, price inflation, or income inequality. In a similar way, civic health involves a normative assessment of the state of social capital in some geographic unit. However, there is little theory on what elements or factors constitute civic health, and little support to date for treating it as a single index. For example, we would not expect to find general public agreement on the optimum rate of divorce, let alone how heavily to weigh that variable in a civic health index. Consequently, the panel decided to focus on the more measureable and agreed-upon dimensions of social capital, focusing on civic engagement and social cohesion.

Interpreting the Statement of Task

In order to be responsive to the statement of task, the panel was required, at a minimum, to assess steps to improve data collection on dimensions of social capital in a manner that effectively informs research and policy—and to assess the role of the U.S. statistical agencies in the enterprise. Sponsors of the report requested guidance on information to be collected in government surveys, particularly in the CPS supplements. This assessment involved (1) assessing the CPS Civic Engagement and Volunteer Supplements, currently the most visible federal statistical system efforts to measure social capital; (2) evaluating which dimensions of key constructs are most amenable to measurement in the supplements; and (3) providing guidance on question content.

While this panel was convened to offer guidance about the CPS

supplement—a task that shaped what was examined and concluded—it was clear that the broader context had to be examined to support the report’s conclusions and recommendations. Restricting this study to the CPS—or even to current federal data collections—would overlook the possibility of alternative data sources, such as administrative records and digital data, whether from the government or other sources.

Our primary focus, however, is the appropriate role of the federal statistical system in improving measurement of social capital through its population surveys. The recommendations and conclusions herein acknowledge the growing importance of building strategies capable of exploiting the potential of nonsurvey data to supplement and work in coordination with the more traditional (and, at this point, more scientifically established) survey approaches mastered by National Statistics Offices over many decades. Consequently, we review the importance of methods and opportunities to link data systems—whether survey or nonsurvey based, government or private—to maximize the policy, information, and research value that can be extracted from them, taking up such questions as

• Which social capital variables (dimensions) are most needed for policy, research, and general information needs, and which are measureable?
• Which aspects of social capital are currently measured best and which are measured less well?
• What are the most promising approaches—survey and nonsurvey, government and nongovernment—for collecting information on key variables or indicators?
• What should be the role of the federal statistical system, recognizing a rapidly changing data collection environment that includes declining survey viability (in terms of costs and response rates), declining budgets of statistics agencies, and the emergence of other data—organic, big data, Web-based—that can substitute for and complement traditional government surveys?
• How might big data—the vast range of digital information produced daily, mainly in the private sector and usually for purposes other than statistical and research—be linked or otherwise used?

A key factor underscoring the need for multiple data collection modes and strategies, including those that might complement or substitute for traditional government surveys, is that much of what is interesting and important about social capital takes place at the community and neighborhood levels. When the objective is to improve the understanding of associations among variables that require highly localized neighborhood

or community information (e.g., at the block group or tract, in census terms), it will typically not make sense to do so by adapting national level, general population surveys; rather, specialized targeted studies may be more appropriate. Similarly, policy issues embedded in social capital, such as those associated with bridging and bonding social capital building strategies, necessitate disaggregation of information by relevant social groups—defined by race and ethnicity; urban, suburban, and rural makeup; and socioeconomic status. This need creates additional data demands, such as the need for larger samples or oversampling of groups of interest.

During its deliberations, the panel also agreed that a number of “big questions” were too ambitious to address meaningfully. We do not attempt to advance the notion of a unified theory in which, for example, the many dimensions of social capital might somehow be organized in terms of inputs that aggregate to some overall measure (analogous to economic accounting systems). Although elements of social capital, social cohesion, and civic engagement can be sensibly grouped into broader domains, it does not follow that these elements add up to a meaningful, overall measure that could be used as a key national indicator or monitoring statistic. In addition to the lacking conceptual precision, as noted above there is no theoretical basis for weighting various components of social capital when combining them into an “index.” Most scholars in this field agree and downplay the idea of creating aggregate indexes of social capital. Putnam (2001, p. 2), for example, commented on the impracticality of a general measure of social capital:

Whether or not such an integrated theory (and in turn framework for data collection) can ever be developed or makes sense is unknown at this time.

Additionally, although the panel was not explicitly charged with exploring the links between social capital indicators and measures of societal progress or well-being, these relationships are important. The growing movement of interest in subjective well-being and quality-of-life measurements, which was given impetus by the Report by the Commission on the Measurement of Economic Performance and Social Progress noted above (Stiglitz et al., 2009), has already generated insights into the role of social capital (as well as many other factors—ranging from income and employment status, to age and relationships, to access to green space

and neighborhood amenities) and people’s experienced well-being and life satisfaction. The relationship between people’s social connectedness, or lack thereof, and self-reports of the quality of their day-to-day experiences (and even their health) have, for example, been shown to be quite robust, particularly for the elderly.¹⁷

Well before the 2009 report, however, measures of subjective wellbeing have taken into account aspects of social capital. For example—though its validity remains highly contentious (Springer et al., 2006)—the Ryff Psychological Well-being Inventory (Ryff, 1989) includes a subscale for positive social relations (six items) to reflect the effect of supportive social relationships on psychological wellbeing and health. And, recently, Su et al. (2013) have included measures of “participation in local community” in a psychological well-being scale.

Many national statistical offices are pursuing data collection in the area of subjective well-being, and connectedness, civic engagement, and governance are frequently identified “domains” (along with more traditional ones, such as income, environment and health) that figure prominently in this work.¹⁸ The domains of well-being identified in Stiglitz et al. (2009)—material conditions, economic insecurity, personal activities, health, education, social connections, political voice and governance, personal insecurity, environmental conditions—include a distinct “social capital” flavor. The European Union Sponsorship Group on Measuring Progress, Well-being, and Sustainable Development, the OECD How’s Life? Initiative, and the Italian National Institute of Statistics have all adopted variants of the Stiglitz et al. (2009) approach to frame data collections. This reorienting of priorities has recast agency agendas (perhaps most notably in the UK Office for National Statistics) in such a way that the idea of measuring social networks and contexts and other aspects of social capital now fits in.¹⁹ This trend toward measurement of well-being

__________________

¹⁷See, for example, Saito, Kai, and Takizawa (2012) and Chappell and Badger (1989) on the relationship between social isolation and subjective well-being among the elderly; Boehm and Kubzansky (2012) on associations between positive psychological well-being and cardiovascular health; and Thisted (2010) on evidence suggesting that changes in wellbeing may work through physiological channels taking place at the cellular level.

¹⁸The OECD publication, The Well-being of Nations: The Role of Human and Social Capital (2001) specifically asked the question “What is the impact of social capital on well-being?” It then laid out the sketchy and mixed evidence at the time, and suggested research for studying the links to answer the question.

¹⁹The UK Office for National Statistics (ONS) announced in November 2010 that it would start measuring subjective well-being to help guide national policy. Prime Minister David Cameron spoke about how well-being indicators would be used as a new measure of the country’s progress, arguing that the government has the power to improve well-being by creating a climate in the country more conducive to the good life. Cameron discussed the shift to “measuring our progress as a country not just by how our economy is growing, but

more broadly has not gained the same traction among U.S. federal statistical agencies as it has in many other national statistical offices. While currently of great interest, the panel has judged the task of linking various social capital, cohesion, and engagement indicators to subjective (self-reported) well-being beyond the scope of the study.

The panel’s approach for conceptualizing data collection is to assess and prioritize measurement of various social capital components on the basis of three factors:

1. evidence connecting them to specific, measurable outcomes in such domains such as health, crime, education, employment, effectiveness of governance;
2. their value in providing descriptive information to better understand society; and, relatedly,
3. their research value.

The spectrum of “indictors” emphasized in this report includes those that have, in the research, been defined and broadly identified with social capital and for which there is some agreement in terms of their status as either socially positive (high levels of trust in neighbors, volunteering, voter participation, charitable giving) or socially negative (social isolation, extreme polarization, corruption, incivility in the public sphere).

A number of social environment characteristics—which may affect or be affected by the social capital of a community or neighborhood and which are in principle measurable—fall in close proximity to the concepts identified in the panel charge and could arguably have been considered by the panel. Some of these, such as changing family structure, intergenerational (social and economic) mobility, political and social polarization, and fairness and discrimination have vast research literatures of their own that span multiple disciplines. The recent research on intergenerational income mobility by Chetty et al. (2013) was one example of the fascinating and important work in these areas. Their finding that upward mobility patterns for local areas (defined by census commuting zone) correlated significantly with extent of residential segregation by income, school quality, a social capital index, and other variables related to civic engagement and community cohesiveness is indicative of the salient connections between these phenomena and the topics central to this report. Although

__________________

by how our lives are improving…not just by our standard of living, but by our quality of life.” The ONS was given the task of choosing several subjective well-being questions to be included in the Integrated Household Survey, the biggest source of social data on the United Kingdom after the census. See http://www.nationalaccountsofwellbeing.org/news/archive/David-Cameron-announces-UK-well-being-measure [February 2014].

the exact mechanisms at work studied by Chetty et al. (which included school quality and catchment area and the strength of transportation systems) requires more research, it is clear that there is substance and policy relevance to these research questions. The panel often felt compelled to consider these closely related phenomena, because they are so important, alongside the traditionally identified dimensions of social capital, and to weigh in on how they may interrelate with civic engagement and social cohesion. However, as acknowledged explicitly in Chapter 2, the idea that this panel could add meaningfully to the research addressing these big questions was unrealistic.

Report Audience

The audience for this report includes statistical agencies (both domestic and foreign), which oversee government data collection; the Corporation for National and Community Service, the study’s sponsor, responsible for fielding the most useful CPS Civic Engagement and Volunteer supplements possible; academic researchers, who have advanced the broader understanding of social capital dimensions and established the need to measure them; national and local policy makers who, ideally, put research findings to good use; community-based organizations that often are best positioned to enhance or initiate programs related to civic engagement and community betterment; and the general public, which benefits from information about its society.

1.3. REPORT STRUCTURE

The remainder of the report is structured as follows: Chapter 2 identifies and defines the key measurement constructs that have been raised in this introduction. We present our views on what kinds of data should be collected and offer thoughts on how to measure various components of social capital. In Chapter 3, the strength of the evidence tying these components to social, economic, and other outcomes is assessed and criteria identified for prioritizing measures and driving a data collection strategy. Issues of causality (as they relate to policy relevance) are explored in the context of this assessment. A number of key measurement and technical survey issues—some unique to the social capital context and some not—are discussed.

Setting up the discussion of recommendations for action, the comparative advantages of competing data strategies are weighed in Chapter 4. The role of the federal statistical system in data collection on civic engagement and social cohesion is considered, as are specific, potentially exploitable government data sources. In Chapter 5, attention is given to alterna-

tive methods of measuring civic engagement, social cohesion, and other dimensions of social capital being created by the rapidly changing world of data collection and statistics generation. Both government (“official statistics”) and nongovernment data strategies are discussed, along with experimental approaches that may involve pilot studies, public/private collaborations and partnerships, and exploitation of emerging technologies. These final chapters lay out next steps and a number of recommendations for advancing concepts, methodology, data collection, and research. The report’s appendixes present background information on various topics.

This page intentionally left blank.