Exploring civic virtue and behaviors across the Christian tradition
Emerging findings from the National Survey on American Health
Author
David Weerts & Alberto F Cabrera
Published
September 22, 2023
Purpose
Using the 2021 National Survey on American Civic Health, this project addresses three questions:
What is the political orientation of Christian and non-Christian communities?
How do civic engagement vary across Christian and non-Christian communities?
How do Christian and non-Christian communities differ across indicators of national civic health?
Organization
This document reports the analyses followed to answer these three research questions. The results are organized into four sections: 1) Descriptive statistics; 2) Research Question # 1; 3) Research Question # 2; and 3) Research Question # 3.
Descriptive statistics
Our approach to civic virtue is grounded on the civic health model (Weerts, Cabrera & Van Dorn, 2021). The model postulates that civic health is composed of civic knowledge, predispositions towards constructive political deliberation and civic engagement. In turn civic health is the product of several factors pertaining to personal beliefs and values, faith-based organizations, family upbringing, high school and college experiences as well as work experiences. Positive manifestations of civic health are confidence in pluralism, trust in government and civic engagement. Table 1 reports the descriptive statistics of the variables used to answer our two research questions.
Faith groups. Based on 63 religious self-reported preferences, we identified four faith communities (see table 1) in a 2021 national sample of 4990 americans. Sixty percent of our sample identified themselves as either protestants or Christians. The third largest group is made of Americans who reported no religion affiliation at all (31%). Almost 9% self described as members of non-christian communities (e.g., Judaism, Mormon, Islam) (see table 1). Our estimates seem to consistent with national trends. According to Statista, 35% of the US population self-described as Protestants (35%), Catholics (22%), or members of other religious communities (9%) in 2021. Our sample, however, overestimates the proportion of non-faith affiliated individuals by about ten percentage points (21% vs 31%).
Faith-community ideology. We relied on two items to identify the extent one’s religion community supported either political oriented progressive causes (e.g. LGBTQ, racial justice, environmental conservative causes ), or conservative political causes (e.g., pro-life, traditional family values). We identified three communities: 1) those that supported neither political orientation, 2) those who were progressive, and 3) those who stressed conservative values. Nearly 60% of respondents report that their religious communities support neither ideologies. Respondents from conservative communities outnumber those from progressive religious communities (26% vs. 15%) (see table 1).
Civic class. The National Survey on American Health study reveals a high degree of diversity of Americans’ civic identities. The intensity of civism varies across theses five classes (Classes of civic engagement in the US population). The lower levels of civism are confined among submarines followed by tender boats. The higher levels of community involvement are found among lightships followed by tugboats and cruise ships. Accordingly, fifty-five percent of our sample is made of active citizens, while forty-four percent display low predispositions for civic engagement (see table 1).
Confident pluralism. Our scales of tolerance for political discourse and humility in political discourse are consistent with Inazu’s concept of confident pluralism (Inazu, 2015, 2018). Our confirmatory analysis indicate that each of these two scales is highly reliable. The reliability of Humility in Political Differences is 0.828 and the one for Tolerance for Political Discourse is 0.893 (Confident pluralism & trust in government). The average tolerance score was 17 in a scale ranging from 5 to 25, while the average humility score was 16 in a scale ranging from 3 to 15 (see table 1).
Trust in government. Public trust in government is key for governmental legitimacy, and political stability (e.g., Useem & Useem, 1979). Our scale of public trust taps on perceptions the government can be trusted to do its best for its citizens, and that the public institutions are helpful and fair in their treatment of all citizens. The reliability of this scale is high (CR = 0.871) according to our confirmatory factor analysis (Confident pluralism & trust in government). The average trust in government score was 8.9 in a scale ranging from 3 to 15.
How do political orientations vary across Christian and non-Christian communities
Table 2 and Figure 1 report religious ideology within each of the four religious communities under consideration. Faith communities and political ideology are associated ( χ2 (6) = 189.7 ), but the degree of association is small (Cramér’s V = 0.17). However, there are important trends. Across protestants, Christians and members of other faiths, the common denominator is attending religious communities that neither supported progressive or conservative politics. Support for progressive values is the highest among members of other faiths communities (26%). On the other hand, almost a third of protestants and Catholics belong to religious communities whose political leaning is conservative.
Chi-square results
Code
load("~/Documents/2021 Civic Aptitude Survey/2023 Exploring Civic Virtue/american_civic_health_survey.RData")faith_ideology<-chisq.test(x =american_civic_health_survey$faith_cmty, y =american_civic_health_survey$ideology_comb)faith_ideology
Pearson's Chi-squared test
data: american_civic_health_survey$faith_cmty and american_civic_health_survey$ideology_comb
X-squared = 189.71, df = 6, p-value < 2.2e-16
Table 2. Political orientation within faith communities
Variable
Faith community
Other Faiths, N = 4211
None, N = 1,5221
p-value2
Protestant, N = 2,0111
Catholic, N = 9871
Neither
824 (54%)
445 (58%)
178 (57%)
442 (74%)
Progressive
203 (13%)
93 (12%)
81 (26%)
115 (19%)
Conservative
497 (33%)
234 (30%)
54 (17%)
43 (7.2%)
1 n (%)
2 Pearson’s Chi-squared test
Code
faith_ideology<-american_civic_health_survey%>%select(faith_cmty, ideology_comb)%>%na.omit()faith_ideology<-faith_ideology%>%group_by(faith_cmty,ideology_comb)%>%summarise(n =n())%>%mutate(p =n/sum(n))%>%mutate(prop =round(p, digits =3))color_ideology<-c("Neither"="grey41","Conservative"="red","Progressive"='#064a81')faith_ideology%>%mutate(faith_cmty =recode(faith_cmty,"Other"="Other Faiths"))%>%ggplot(aes(x =faith_cmty, y =prop, fill =ideology_comb))+geom_col(colour ="black", position ="fill", alpha =0.8)+geom_text(aes(label =scales::percent(prop)), position ="stack",vjust =+2.1, color ="white", size =3)+scale_y_continuous(labels =scales::percent)+scale_fill_manual(values =color_ideology)+scale_size(guide =FALSE)+labs(y ='', x =" ", fill ="Political ideology")+theme_minimal(base_size =14)+theme( plot.title.position ="plot", axis.text.x =element_text(face ="italic", size =10), panel.grid.major =element_line(linetype =3, color ="white"))+labs(title ="Political ideologies within faith communities")
Figure 1: faith-comunity-ideology
Research question # 2:
How do civic orientations vary across Christian and non-Christian communities?
Table 3 and Figure 2 display the distribution of civic classes within each of the 4 faith communities under consideration. Overall, there is a significant association between one’s religion affiliation and one’s civic orientation ( χ2 (12) = 227.53). Although the magnitude of the association is small (Cramer’s V = 0.12), there are important civic trends across faith groups. Non catholic denominations are more prone to belong to the two highest civically oriented groups. Nearly 60% of them are tugboats or lightships compared to Protestants (35%) and Catholics (33%). The least civically committed individuals are those who don’t belong to faith communities. Nearly 60 percent of them are either submarines or tender boats.
Chi-square results
Code
load("~/Documents/2021 Civic Aptitude Survey/2023 Exploring Civic Virtue/american_civic_health_survey.RData")faith_civism<-chisq.test(x =american_civic_health_survey$civic_class, y =american_civic_health_survey$faith_cmty)faith_civism
Pearson's Chi-squared test
data: american_civic_health_survey$civic_class and american_civic_health_survey$faith_cmty
X-squared = 227.53, df = 12, p-value < 2.2e-16
Table 3. Civic orientation within faith communities
Variable
Faith commmunity
Other Faiths, N = 4211
None, N = 1,5221
p-value2
Protestant, N = 2,0111
Catholic, N = 9871
Civic class
Disengaged
233 (12%)
86 (8.7%)
51 (12%)
336 (22%)
Voter-only
591 (29%)
311 (32%)
80 (19%)
526 (35%)
Charitable-Passive
484 (24%)
265 (27%)
90 (21%)
330 (22%)
Active-Benevolent
559 (28%)
250 (25%)
157 (37%)
279 (18%)
Fully Engaged
144 (7.2%)
75 (7.6%)
43 (10%)
51 (3.4%)
1 n (%)
2 Pearson’s Chi-squared test
Code
faith_class<-american_civic_health_survey%>%select(faith_cmty, civic_class)%>%mutate(faith_cmty =recode(faith_cmty,"Other"="Other Faiths"))%>%mutate(civic_class =recode(civic_class,"Submarine"="Disengaged","Tender boat"="Voter-only","Cruise ship"="Charitable-Passive","Tugboat"="Active-Benevolent","Lightship"="Fully Engaged"))%>%na.omit()faith_class<-faith_class%>%group_by(faith_cmty,civic_class)%>%summarise(n =n())%>%mutate(p =n/sum(n))%>%mutate(prop =round(p, digits =3))color_fill<-c("Fully Engaged"='#929d37',"Active-Benevolent"="red","Charitable-Passive"="orange","Voter-only"='#064a81',"Disengaged"='grey41')faith_class%>%ggplot(aes(x =faith_cmty, y =prop, fill =civic_class))+geom_col(color ="black", position ="fill", alpha =0.8)+geom_text(aes(label =scales::percent(prop)), position ="stack",vjust =+1.4, color ="white", face ="bold", size =3)+scale_fill_manual(values =color_fill)+scale_y_continuous(labels =scales::percent)+scale_size(guide =FALSE)+labs(y ='', x =" ", fill ="Civic class")+theme_minimal(base_size =14)+theme( plot.title.position ="plot", axis.text.x =element_text(face ="italic", size =10), panel.grid.major =element_line(linetype =3, color ="white"))+labs(title ="Figure 2 Civic classes within faith communities")
Figure 2: civic-faith-comunity
Research question # 3:
How do Christian communities differ from other faith communities as it relates to tolerance for political difference, humility in political discourse and trust in government
All anova tests for tolerance (F-test = 10.3, df = 3, p < .01) humility (F-test = 9.19, df = 3, p < .01) and trust ((F-test = 20.07, df = 3, p < .01) were significant, indicating that indicators of civic health vary across faith groups (see table 4). Bonferroni tests revealed differences among the four faith communities across the three national health indicators.
Regarding tolerance, protestants and Catholics are more tolerant for political differences next to agnostics. In terms of humility in political discourse, Bonferroni tests indicate that protestants and Catholics display higher levels of humility in relation to people from other faiths. Catholics are humbler next to agnostics. Both Catholics and Protestants display similar average levels of tolerance and humility.
Protestants are less trusting of the government next to Catholics and people of other faiths. However, both Protestants and Catholics trust more the government than do members of non-faith communities. Moreover, members of other-faith communities trust more than government that non-faith affiliated individuals.
ANOVA test for tolerance for political difference and faith communities
Code
load("~/Documents/2021 Civic Aptitude Survey/2023 Exploring Civic Virtue/american_civic_health_survey.RData")one.way.tolerance<-aov(tolerance_tot~faith_cmty, data =american_civic_health_survey)summary(one.way.tolerance)
Df Sum Sq Mean Sq F value Pr(>F)
faith_cmty 3 666 221.90 10.35 8.67e-07 ***
Residuals 4937 105835 21.44
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
49 observations deleted due to missingness
ANOVA test for humility in political discourse and faith communities
Code
load("~/Documents/2021 Civic Aptitude Survey/2023 Exploring Civic Virtue/american_civic_health_survey.RData")one.way.humility<-aov(humility_tot~faith_cmty, data =american_civic_health_survey)summary(one.way.humility)
Df Sum Sq Mean Sq F value Pr(>F)
faith_cmty 3 275 91.78 9.187 4.65e-06 ***
Residuals 4937 49321 9.99
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
49 observations deleted due to missingness
Code
load("~/Documents/2021 Civic Aptitude Survey/2023 Exploring Civic Virtue/american_civic_health_survey.RData")american_civic_health_survey%>%select(faith_cmty, tolerance_tot, humility_tot)%>%mutate(faith_cmty =recode(faith_cmty,"Other"="Other Faiths"))%>%tbl_summary( by =faith_cmty, missing ="no", label =list(tolerance_tot~"Tolerance",humility_tot~"Humility"), statistic =list(all_continuous()~c("{mean} ({sd})")))%>%add_p(all_continuous()~"aov")%>%# add a header (which also unhides a hidden column)modify_header(statistic~"***F-test***", label ="**Variable**")%>%modify_spanning_header(c("stat_1", "stat_2")~"**Faith group**")%>%# add a function to format the columnmodify_fmt_fun(statistic~style_sigfig)%>%modify_caption("**Table 4. Tolerance, humility & trust across faith communities**")%>%bold_labels()
Table 4. Tolerance, humility & trust across faith communities
Variable
Faith group
Other Faiths, N = 4211
None, N = 1,5221
F-test
p-value2
Protestant, N = 2,0111
Catholic, N = 9871
Tolerance
17.3 (4.6)
17.4 (4.8)
17.0 (4.7)
16.5 (4.5)
10
Humility
16.1 (3.2)
16.4 (2.9)
15.5 (3.6)
15.9 (3.1)
9.2
1 Mean (SD)
2 One-way ANOVA
References
Inazu, J. (2018). Confident Pluralism: Surviving and Thriving through Deep Difference. Chicago: University of Chicago Press.
Inazu, J. D. (2015). A confident pluralism. Southern California Law Review, 88(3), 587-618
Morton, M., Dolgon, C., Maher, T., & Pennell, J. (2012). Civic engagement and public sociology: Two “movements” in search of a mission. Journal of Applied Social Sciences 6 (1), 5-30. https://doi:10.1177/1936724411436170
OECD (2022). Building Trust to Reinforce Democracy: Main Findings from the 2021 OECD Survey on Drivers of Trust in Public Institutions. OECD Publishing, Paris, https://doi.org/10.1787/b407f99c-en
Reich, R. B. (2019). The common good. Penguin Random House.
Useem, B. & Useem, M. (1979). Government legitimacy and political stability. Social Forces, 57(3), https://doi.org/10.1093/sf/57.3.840
Schoenherr, N., (2015). What is confident pluralism https://source.wustl.edu/2015/11/what-is-confident-pluralism/
Weerts, D., Cabrera, A. & Van Dorn, K. (2023). Tolerance, humility, and educational attainment: Does college attendance build capacity for confident pluralism? Round table. Annual meeting of American Association of Educational Research. Chicago, Illinois.
Weerts, D. J., Cabrera, A.F. & Van Dorn, K., (2022). Antecedents of Individual Civic Health. American Center for Political Leadership, Southeastern University, Lakeland, FL.
