its generation. Brehm and Rahn (1997) found that trust is a function of engagement. The
literature suggests that churches generate social capital through engagement and the development
of civic skills. Developing civic skills generally requires more than just attending. Three
additional measures of religious involvement are incorporated in the study. Respondents
reported whether they volunteered through the church, whether they attended a small group at
the church and whether they held a leadership position there.
Respondents were also asked about how many friends they had in the congregation.
People with many friends in the congregation may have higher levels of interaction. They may
also develop more bonding social capital because of the interpersonal relationships. Finally,
respondents revealed whether or not within the last year they had heard a sermon about a social
or political issue. This question may actually reflect some congregational variation because most
people are likely to hear a sermon at the church they attend. Still, the question did not explicitly
reference whether they heard the sermon at their own church and thus the variable is treated only
as a measure of the religio-political input a respondent receives. There were no measures of
internal religiosity included in this data set. While I would have preferred to control for such
factors, previous research indicates that internal religiosity is not a strong predictor of social
trust. As Coleman observed, “perhaps it should be obvious that, absent any embeddedness in the
social, the likelihood of any social capital investment of individual spirituality is small”
(Coleman 2003: 37).
Other Measures
Demographic variables representing age (a three-category variable that distinguishes
respondents 18 through 29, 30-49 and to and over), education (whether or not a respondent
graduated from college), gender (1=female), marital status (1=married) and region (1=South),
and employment status (unemployed=1) are contained in the analysis. In addition because