Believe It: Atheism Less and Less a Political Taboo

The US Constitution prohibits religious tests for public office, however, being an atheist in politics has been a powerful political taboo in our nation. New research finds that this bias against candidates who don’t believe in God has notably weakened. A poll conducted by Lake Research Partners for the American Humanist Association and the Center for Freethought Equality and funded by the Stiefel Freethought Foundation shows that being nonreligious, agnostic, or atheist need not be considered an impediment to a candidate’s electoral success.

Talking with candidates running in the 2018 midterm election about the political cost of identifying as an atheist was the impetus for the poll. A candidate in a very red district, where the last Democratic opponent received less than 20 percent of the vote against the Republican incumbent, said he couldn’t possibly identify as an atheist because he couldn’t afford to lose any more voters. He said he automatically lost voters by identifying as a Democrat, more since he is pro-choice, and even more with his support of LGBTQ equality. This begged the question: Would the supporters of a pro-choice, LGBTQ equality Democrat care if their candidate was also an atheist? Reliable data was needed to answer this question.

The survey results show that a candidate’s atheism is a non-issue—or even a plus—for pro-choice, pro-marriage-equality Democratic voters, fully 74 percent of whom saying a candidate being “not religious” or “agnostic” would make no difference in their vote and 72 percent saying a candidate’s “atheism” would make no difference in their vote. Among those with a preference for a nonreligious/agnostic candidate, 14 percent are more likely to vote for him/her (including 10 percent much more likely) and 7 percent are less likely (including 4 percent much less likely); the difference is modest among those voters with an inclination toward an atheist candidate: 10 percent are more likely to vote for him/her (8 percent much more likely) and 13 percent are less likely (7 percent much less likely). Nontheistic, progressive Democrats in non-swing districts should no longer feel hesitant to be public about their religious identity. And while it still could be a challenging factor in swing districts, it’s no longer the taboo it once was.   

These results are supported by the 2016 and 2018 underdog campaigns of Gayle Jordan, who ran for the Tennessee State Senate. In 2016 Jordan earned 25.6 percent of the vote against the Republican incumbent Jim Tracy, and her religion was never mentioned during the campaign. After Tracy took a position as Tennessee director for rural development in the Trump administration, Jordan decided to run again in the special election for the seat in early 2018. Her second race was notably different. The Lt. Governor, the chairman of the Republican Party of Tennessee, and her Republican opponent’s campaign publicly attacked Jordan for being an atheist and made this a focus of the race. Although Jordan again lost the election, she increased her vote total to 28.3 percent more than a two-point increase over 2016. The attacks appeared to have had no effect on the race, except to expose the leaders of the Tennessee Republican Party as appalling religious bigots.

In a related example, Democratic Congressman Jared Huffman (CA-2) announced in November 2017 that he is a humanist and agnostic—the only public nontheist in Congress. The percentage of his vote in the 2018 midterm election was 76.4 percent, which is just slightly less than 2016 (76.9 percent) and more than his 2014 midterm (75.0 percent) results. Being agnostic was not a factor in his re-election.

Fortunately, the bias against atheist and agnostics is diminishing and the reason may simply be demographics—simply put, the number of religiously unaffiliated Americans is growing rapidly. The Pew Research Center uses the term “nones” for the religiously unaffiliated, which includes people who identify as either atheist or agnostic and those who say their religion is “nothing in particular.” According to Pew, the “nones” have grown from 16 percent in 2007 to 23 percent in 2014, and comprise the largest “religious group” in the Democratic Party. With a third of millennials in the “nones” category, the religiously unaffiliated community will no doubt continue to grow.

Since 1958, Gallup has asked Americans if they would vote for a well-qualified presidential candidate who was an atheist. In the first poll only 18 percent answered in the affirmative. Four decades later (1999), for the first time a slim majority said they would vote for an atheist. In Gallup’s 2015 poll, 58 percent said they would vote for an atheist candidate. The willingness to vote for an atheist presidential candidate varies greatly by generation: 75 percent of those age eighteen to twenty-nine, 63 percent of thirty- to forty-nine-year-olds, 50 percent of those fifty to sixty-four, and 48 percent of those sixty-five and over. Considering political party, 64 percent of Democrats, 61 percent of independents, and 45 percent of Republicans would be willing.

Although the bias against nonbelievers is diminishing, the Lake Research poll demonstrated that the bias still exists. When asked to choose between a candidate who believes in God or a candidate who does not believe in God, respondents strongly chose the believer—with 60 percent supporting the candidate who believes in God and only 6 percent supporting the candidate who doesn’t, while for 31 percent of the voters it made no difference (3 percent were undecided or did not know). These results were markedly different by political party affiliation: 46 percent of Democrats voted for the believer (10 percent supported the nonbeliever and 41 percent were indifferent) compared to 81 percent of Republicans voting for the believer (3 percent for nonbeliever, 15 percent no difference). For those independent respondents who did not identify with either major party, the results were more in line with Democratic voters (48 percent voting for the believer, 5 percent voting for the nonbeliever, and 42 percent having no preference).

However, this preference for candidates who believe in God nearly disappears when policy stances are included in the question. Surprisingly this holds true for Republicans as well.

When asked if they would vote for a candidate who stood for most of the things they believed in but who didn’t believe in God or who was nonreligious, versus a candidate who was religious but didn’t stand for most of the things they did, 60 percent of respondents chose the nonbeliever and 24 percent chose the religious candidate. Party identification narrows with support for the nonbeliever: 67 percent for Democrats (19 percent voting for the religious candidate and 15 percent did not know); 54 percent for Republicans (31 percent voting for the religious candidate and 15 percent didn’t know); and 57 percent for independents (20 percent voting for the religious candidate and 23 percent didn’t know).

Removing the comparison to a religious candidate from the polling question, the bias against candidates who don’t believe in God further erodes when respondents were simply asked if they would vote for a nonbeliever or nonreligious candidate who stood for most of the things they did. 72 percent would vote for the nonbeliever/nonreligious candidate and 16 percent would not (11 percent didn’t know). The variance between party identification again narrows with support for the nonbeliever/nonreligious: 79 percent for Democrats (12 percent wouldn’t and 9 percent didn’t know), 68 percent for Republicans (22 percent wouldn’t and 11 percent didn’t know) and 69 percent for independents (13 percent wouldn’t and 17 percent didn’t know).

These results demonstrate that religious identification is a less important factor than policy stances when voters make their decisions about which candidates to support.

These preferences about political candidates hold even when using several different identifiers for those who don’t believe in God (not religious, agnostic, atheist):

• “Not religious” – 71 percent of Democrats reported this made no difference in determining their vote, 15 percent were more likely to vote for the person, and 9 percent were less likely to vote for them. For Republicans, 56 percent said there was no difference, 15 percent were more favorable to the candidate, and 24 percent were less so. For independents, 69 percent saw no difference, 11 percent were more likely to vote for the nonreligious person, and 16 percent were less likely to vote for them.
• “Agnostic” – 65 percent of Democrats reported this made no difference in determining their vote, 16 percent were more likely to vote for the person, and 13 percent were less likely to vote for them. For Republicans, 51 percent said there was no difference, 14 percent were more favorable to the candidate, and 30 percent were less so. For independents, 70 percent saw no difference, 9 percent were more likely to vote for the agnostic, and 11 percent less likely to vote for them.
• “Atheist” – 64 percent of Democrats said this made no difference, 10 percent said it was favorable, and 20 percent said it wasn’t. Among Republicans, 40 percent reported no difference, 9 percent were more favorable, and 48 percent were less. For independents 62 percent saw no difference, 8 percent were more favorable to an atheist, and 24 percent were less likely to vote for an atheist.

Although a plurality of Republicans (48 percent) say that they would be less likely to vote for an atheist candidate who shared their positions, most younger Republicans (including 68 percent of those under thirty-five and 54 percent of those under fifty) say that it makes no difference. Notably, support for nonreligious/nonbeliever candidates among black voters is lower than the overall voting population. However, a 58 percent majority of black voters would still support a nonreligious/nonbeliever candidate who otherwise shared their views on most major issues.

The findings from this survey and other research should encourage candidates and elected officials to be authentic about their religious beliefs. Discriminatory attacks on atheist candidates and elected officials will no doubt continue, but as revealed by this research, such attacks will probably not affect the outcome of an election and may, in fact, offend voters in both political parties. People who have stayed out of the electoral arena because of the bias against atheists, agnostics, and “nones” should be encouraged by these findings and consider running for political office. Our political system depends on the active participation of all our citizens. The humanist, atheist, and agnostic community needs to be fully and openly engaged in our democratic process.

You can see the poll questions, top line results, and methodology here.