On one hand, I am concerned about the phenomenon of white Christian nationalism in the United States. In some churches and religious networks, some people believe that the nation ought to be fundamentally white and Christian and that Christianity somehow undergirds this idea. White Christian nationalism competes with more emancipatory and inclusive theological views, but it’s influential.

On the other hand, I am still drawn to the old, Tocquevillian argument that participating in the governance of an organization that contains any diversity can teach skills and values of deliberation, tolerance, and inclusion. And churches can be such organizations.

Therefore, I am worried about people who identify as white and Christian without actually participating in maintaining communities or wrestling with the complexities required by self-government. At the same time, I am cautiously optimistic that asking people to take responsibility for real religious communities will strengthen democracy, because members can learn how to hold a group together, manage disagreements, select good leaders, interact with outsiders, and hold each other accountable.

Thus I would hypothesize that:

• H1 Identifying as a white Christian will be associated with racial anxiety and intolerance, because white Christian dominance is threatened today by growing racial diversity and greater equality. But …
• H2 Participating in groups that offer an element of self-governance will promote democratic values and therefore should correlate negatively with racial anxiety. (Especially if the group is diverse).

Churches are manifestations of Christian identity and places where some people receive racialized messages, but they are also groups that can offer various degrees of diversity and self-governance. Therefore:

• H3 Among whites, church membership (a proxy for Christian identity) will be associated with racial anxiety and intolerance. However,
• H4 Controlling for religious identity as well as race, church participation will be associated with higher tolerance. (This hypothesis is similar to the findings of Ekins 2018.)

Nationally representative survey data collected by Data for Progress from August 4-30, 2022 offer a chance to test these hypotheses. The sample is 1,899 likely voters (who differ from non-voters in the US resident population). The survey asks several questions about religious belief, identity, attendance, and participation.

The survey also asks several attitudinal questions about diversity. I chose: “It would bother me if, one day, white people were an ethnic minority in the United States.” I will label disagreement with that prompt “inclusiveness,” for the sake of using one word.*

Two limitations of the survey: It doesn’t ask about the diversity of respondents’ religious congregations or which religion people belong to. However, 94% of white Americans who identify as religious are Christian (2% are Jewish), so a relatively modest error results from assuming that white religious respondents are Christian and their congregations are churches.

I have run Ordinary Least Squares (OLS) regressions to predict inclusiveness based on various other questions in the survey and to test the four hypotheses stated above.

H1 is correct: The importance of religion to a person is negatively correlated with their inclusiveness (when controlling for demographics). Being white, over 45, and not being college educated are also negatively associated with inclusiveness in this model.

Another way to look at the same pattern is to restrict the sample to whites (most of whom are either Christian or non-religious) and omit race from the regression. For whites, being religious, over 45, and not college educated are negatively associated with inclusiveness.

H2 is not correct: Contrary to what de Tocqueville or Dewey might predict, a person’s total number of active group memberships is negatively associated with inclusiveness. The more groups you actively participate in, the less likely you are to accept whites becoming a minority. This relationship is statistically significant but small. When looking only at whites, the relationship is no longer significant, but it isn’t positive, as I would have hoped.

Different group memberships evidently have different relationships with racial anxiety. The bivariate correlation between religious-congregation membership and inclusiveness is -.058** (Pearson), but no other forms of membership are correlated to a significant extent with this outcome. In a regression with group memberships as the only independent variables (i.e., without demographics), religious groups and sports are negatively correlated with inclusiveness; art groups are positively related.

H3 is correct: Being a member of a religious congregation is associated with less inclusiveness (controlling for demographics).

H4 is correct: Religious participation helps.

In a model that includes prayer, attendance, religiosity, congregational membership, and religious participation, membership has a significant positive correlation with inclusiveness, whereas religiosity is a negative correlate. When looking at whites only, congregational membership is a positive; religiosity and attendance are negative.

The previous sentence is my takeaway. For white Christians, being actively involved in a church builds the values that we need for a pluralistic democracy. At the same time, simply identifying as Christian and hearing Christian messages is associated with intolerant values. We need the first effect to outweigh the second one, and that is a struggle in an era when most associations (including churches) have weakened while mass and social media have expanded.

*Data for Progress actually ran a survey experiment and asked some people whether they would be bothered and others whether they would not be bothered. This difference in question-wording yielded somewhat different results. I combined the two items into one measure.

See also: churchgoing and Trump; the relevance of American civil religion to K-12 education; the Dutch secret; the prospects for an evangelical turn against Trump.