by EVAN STEWART
Nearly 30% of Americans say they have no religious affiliation. Today the so-called “nones” represent about 30% of Democrats and 12% of Republicans – and they are making their voices heard. Organizations lobby on behalf of atheists, agnostics, secular humanists and other nonreligious people.
As more people leave religious institutions, or never join them in the first place, it’s easy to assume this demographic will command more influence. But as a sociologist who studies politics and religion, I wanted to know whether there was evidence that this religious change could actually make a strong political impact.
There are reasons to be skeptical of unaffiliated Americans’ power at the ballot box. Religious institutions have long been key for mobilizing voters, both on the left and the right. Religiously unaffiliated people tend to be younger, and younger people tend to vote less often. What’s more, exit polls from recent elections show the religiously unaffiliated may be a smaller percentage of voters than of the general population.
Most importantly, it’s hard to put the “unaffiliated” in a box. Only a third of them identify as atheists or agnostics. While there is a smaller core of secular activists, they tend to hold different views from the larger group of people who are religiously unaffiliated, such as being more concerned about the separation of church and state.
By combining all unaffiliated people as “the nones,” researchers and political analysts risk missing key details about this large and diverse constituency.
Crunching the numbers
In order to learn more about which parts of religious unaffiliated populations turn out to vote, I used data from the Cooperative Election Study, or CES, for presidential elections in 2008, 2012, 2016 and 2020. The CES collects large surveys and then matches individual respondents in those surveys to validated voter turnout records.
These surveys were different from exit polls in some key ways. For example, according to these survey samples, overall validated voter turnout looked higher in many groups, not just the unaffiliated, than exit polls suggested. But because each survey sample had over 100,000 respondents and detailed questions about religious affiliation, they allowed me to find some important differences between smaller groups within the unaffiliated.
My findings, published in June 2023 in the journal Sociology of Religion, were that the unaffiliated are divided in their voter turnout: Some unaffiliated groups are more likely to vote than religiously affiliated respondents, and some are less likely.
People who identified as atheists and agnostics were more likely to vote than religiously affiliated respondents, especially in more recent elections. For example, after controlling for key demographic predictors of voting – like age, education and income – I found that atheists and agnostics were each about 30% more likely to have a validated record of voting in the 2020 election than religiously affiliated respondents.
With those same controls, people who identified their religion as simply “nothing in particular,” who are about two-thirds of the unaffiliated, were actually less likely to turn out in all four elections. In the 2020 election sample, for example, I found that around 7 in 10 agnostics and atheists had a validated voter turnout record, versus only about half of the “nothing in particulars.”
Together, these groups’ voting behaviors tend to cancel each other out. Once I controlled for other predictors of voting like age and education, “the nones” as a whole were equally likely to have a turnout record as religiously affiliated respondents.
The Conversation for more