Following in the tradition of Alexis de Tocqueville, academics and social commentators have long been interested in the study of voluntary associations in the United States. Tocquevilles oft-quoted Democracy in America offers the proposition that the level of broader political participation can be traced to membership in voluntary associations. A variety of associations--from religious to secular, from fraternal to philanthropic--are viewed as requisite components of a healthy civic life in a democracy. The assumption is that association membership, in and of itself, fosters community engagement.

But can we ascribe to all voluntary associations the same social outcome? Are associations of every type equivalent in terms of their impact on participation in political and community life? These are not questions of only theoretical interest. In the late 20th century, a relatively new form of voluntary association referred to by many as the "small group movement" has emerged. Small groups, and support groups in particular, are appearing on the American social landscape in great numbers and varieties. This movement has been viewed as a retreat from participation in large-scale, rational, bureaucratic structures, and a return to more intimate, community-based, voluntaristic associations.

While the increasing size of the small group movement is not in doubt, academics are not in accord when it comes to the impact of this trend. From this scholarly disagreement arose our research questions: Does small group membership foster increased participation in the wider community, as suggested by Robert Wuthnow? Or does small group membership discourage civic involvement and encourage self-absorption, as suggested by Robert Putnam? Or does it, perhaps, depend on the type of the group? Using data collected by Wuthnow in his survey of small group membership among the American public, we endeavored to critically examine and test competing hypotheses about the impact of small group membership on community involvement and political participation.

We first examined whether differences in civic engagement are linked to the characteristics of individual group members. Perhaps self-selection occurs if people with specific characteristics (concerning age, race, sex, and class) are attracted to groups that engage in certain activities and not others. Second, we identified various types of small groups (religious, secular, and mixed) and how these variations might result in different levels of member participation in the civic sphere. Third, we sought to determine whether properties of a group, and member experiences in the group, have an effect on the broader civic engagement of members. To that end, we looked at the role played by the structure of a group (how formal is it?) and its interpersonal climate (do members feel close?) in contributing to the development of "social capital" that "spills over" into broader social participation.

Setting Up the Study

In collecting his data, Wuthnow and his associates conducted in-person interviews with 1,021 participants in small groups. On average, the groups were comprised of about 25 members. Respondents were asked to identify their group as one of the following: youth group, singles group, couples group, womens group, mens group, bible study group, prayer fellowship, house church, covenant group, self-help group, anonymous group, Sunday school class, special interest group, discussion group, support group, and therapy group. Based on that identification, groups were classified as secular ( "Parents Without Partners"), religious ("Wednesday Bible Study"), or secular and religious ("Alcoholics Anonymous"). This resulted in 412 respondents in secular groups, 113 in religious groups, and 443 in mixed groups.

To measure an individuals participation in the wider community--we call this civic engagement--we looked at each persons response to the following question: "As a result of being in this group, have you done any of the following?" The alternatives listed included: (1) "worked with the group to help other people in need outside the group"; (2) "participated in a political rally or worked for a political campaign"; (3) "donated money to a charitable organization, other than your church or synagogue"; (4) "became more interested in social or political issues"; (5) "became more interested in peace and social justice"; and (6) "changed your attitudes on some social or political issues." Seeking to get a broad picture, the first three activities are about behaviors representing involvement in community affairs while the last three involve changes in attitudes or thinking about public issues. (Respondents answers to the six active and passive statements revealed a statistically acceptable level of internal reliability, thereby providing some confidence in the measure of "civic engagement.")

Testing the Variables

Characteristics of the Individual. A fair amount of previous research has indicated that education and occupation are positively associated with civic participation. This is not surprising, as education provides communication skills and organizational abilities, and certain occupations provide opportunities to exercise leadership outside the workplace. We thus expected to find that a small group members education, occupation, sex, age, and income would be related to level of civic engagement. Instead we found that none of these individual characteristics affected small group members level of interest and activity in the civic sphere. How do we account for this unexpected finding? Two factors come to mind: First, our sample included only people who already belonged to a small group. Since they were already "joiners," and our sample did not include "non-joiners," it is difficult to detect the effect of individual characteristics on civic engagement. Second, it is possible that those who belong to groups that provide "support and caring"--a defining characteristic of many small groups--do not fit the prototype of the voluntary association member referred to in much of the literature. They may join small groups because of their small size and their focus on inward-looking, expressive goals, in contrast to the more instrumental orientation of most larger voluntary associations.

However, we did find that individual characteristics were related to the type of small group that one joins. Members of secular small groups have more education and income, and higher status occupations. In contrast, members of religious small groups have less education and income, and lower status occupations. Members of mixed secular/religious groups are in the middle on education and occupation, but they have the lowest incomes of all three group members. In sum, small group members status characteristics do not affect their level of civic engagement, but do influence the type of small group to which they belong.

Type of Group. Does the type of small group influence civic engagement? Some scholars have argued that, because of the distinct religious social capital created by church-built social networks, groups embedded within religious organizations are more likely than other groups to produce civic and religious volunteers. Our research, however, raises doubts about half of that proposition: members of religious small groups are actually the least likely to be interested and active in civic affairs. Members of mixed groups are the most civically engaged, followed by members of secular groups.

Why do religious small groups yield those members who are the least engaged in civic affairs? Although such groups may produce a distinct form of social capital--religious social capital--we must also consider the context of its procurement. It is likely that within a religious group, information that is exchanged about opportunities for participating is overwhelmingly about church-sponsored ministries and social services. And, as time for voluntary activity and civic engagement is often limited, members of religious groups may choose to invest themselves in church-related activities as opposed to broader community affairs.

Characteristics of the Group. As reported above, members of secular, religious, and mixed groups differ in their education, occupation, and income. It would follow that the groups themselves may also differ in their internal structures and modes of operation. Members of varying educational and occupational backgrounds bring with them different skills and competencies (human capital), which may in turn produce different experiences within the groups (social capital). Of particular relevance to the issue at hand is the question of which small group experiences provide opportunities for learning about how to organize and participate in community groups with collective goals. For example, prior research on social movement organizations indicates that when groups have formal structures they tend to be more conducive to member participation in group activities, and when they have clear internal committee structures they tend to attract more members and more volunteer labor. With these ideas in mind, we classified the groups in terms of their degree of formalization (elected leaders, committees, regular meetings) and the level of cohesion/integration among the members.

We found that secular groups have the greatest degree of formalization, while religious groups have the least, and mixed groups have moderate formalization. Perhaps surprisingly, mixed group members report feeling the greatest sense of unity and belonging among the membership, while members of religious groups report somewhat less, and secular group members report their groups to be least cohesive and integrated. These findings suggest that in order for group members to feel a sense of belonging to the group--an emotional connection to other members--some form of structure is necessary; but if groups are too structured, they tend to erode feelings of belonging.

Relative Importance

The final task in our research on civic engagement was to bring together and assess the relative importance of the different factors. We used a statistical technique--linear regression--that permits the simultaneous examination of what is more important for determining civic engagement: the type of group to which one belongs (secular, religious, or mixed), the individual characteristics of the members (education, occupation, income), or the internal structure of groups (formalization and cohesion).

Our findings indicated that, controlling for the effect of individual and group characteristics, the most significant factor is the type of small group in which a person is involved. Members of a group with a mix of secular and religious activities are significantly more likely to be involved in civic affairs (an average involvement in 2.9 of the 6 activities listed earlier), while members of religious groups are least likely to extend their participation beyond their small group (average involvement in 1.7 activities). Members of secular groups are involved in 2.5 activities. (The differences in civic engagement between the three groups are statistically significant.)

Group characteristics--or what goes on in a group--also make a difference. The degree of formalization of a small group is both positively and significantly related to civic engagement. This finding suggests that the opportunities for modeling and practicing group-process skills are transferred by group members to the world outside the small group, specifically to the civic arena. We further found that the cohesion and integration that members feel about their small groups also has a positive impact on the level of members political interest and activity.

As for the individual characteristics of group members, the only one that continues to influence civic engagement is having completed some college (but not necessarily having a college or graduate degree). A members occupation, age, sex, income, and race do not have an effect on civic involvement after the influence of group factors is taken into account. Bearing in mind that the sample used in this study only includes "joiners" (i.e., persons who all belong to at least one group), our analysis suggests that these individual characteristics have very little independent effect on civic engagement.

Two Cheers for Small Groups and Civic Engagement

Given the recent spate of concern about civic disengagement expressed by academics and the popular press, the project of identifying new sources of community involvement and evaluating the claims about growing or declining "social capital" becomes increasingly important. While most of the public debate concerns the question of whether or not Americans are joining voluntary associations with the same frequency as in years past, our research suggests that social scientists may need to hone the question. For example, despite the claims of some analysts, the growth of the "small group movement" has produced neither self-absorbed, inward-looking "retreatists," nor a new wave of small group members who are active in a wide variety of community affairs. The actual outcome is more nuanced: some small group members are socially involved and some are not; the difference can be attributed to both the type of small group in which people are involved and the opportunities they provide for learning social and organizational skills and for establishing close relationships with other group members.

This analysis has important implications for those concerned with declining social capital and political apathy. Most obvious are some of the explicit findings: the growth of religious small groups could, counter-intuitively, result in declining involvement in the public social sphere. But that is not the only concern. Our research should also serve as a general warning against assuming that all civic engagement has the same social impact. For example, while much of the literature seems to assume that civic engagement is inherently good because it inevitably leads to a stronger civil society, we prefer to approach this normative view with the same caution that we applied to small groups. The small group members in our study who reported participating in a political rally could have been attending a rally for David Duke or one for Paul Wellstone. Does this difference matter? Is it one that should make us think about what we mean by civil society? Until there are answers to such questions about civic engagement, and in light of the discovery that participation in some small groups is not as favorable towards broader public involvement as some might have hoped, we offer two cheers for small groups rather than three.

The Communitarian Network
2130 H Street, NW, Suite 703
Washington, DC 20052
202.994.6118
comnet@gwu.edu