Religion, by my estimation, was a more prominent issue in the runup to the 2020 election than in other recent presidential campaigns. This reflects two factors. First, President Donald Trump over the past four years has made a concerted effort to promote policies and programs designed specifically to increase support among White evangelical Christians. Second, Joe Biden is only the fourth Catholic major-party nominee in U.S. history (along with Al Smith in 1928, John F. Kennedy in 1960 and John Kerry in 2004), and Biden made his faith a significant part of his presidential campaign.
Did either of these factors make a difference in the election outcome? The answer to that question is complex. But -- at this stage in our understanding of the dynamics of the 2020 election -- I would say that if Trump's or Biden's campaign shifted the vote of these two core religious groups, the impact was fairly minimal and/or difficult to document.
White Evangelical Voters
There are two primary sources of information about voting by religious groups this year. One is a continuation of the traditional exit polls conducted by Edison Research. These involve stopping voters at actual voting places both on Election Day and during early voting periods, supplemented by phone surveys of absentee and early voters. The other is AP VoteCast, a massive effort to understand how people voted, carried out by The Associated Press and NORC at the University of Chicago. Their surveying uses a number of different sample sources, with interviews conducted by phone, by mail and online in the days and weeks before the election.
The AP VoteCast survey shows that 81% of White evangelical Protestant voters went for Trump this year, compared with 18% who voted for Biden. The Edison exit polls estimate that 76% of White evangelicals voted for Trump, 24% for Biden.
Like any survey (and perhaps more than most surveys because of their complicated methodologies designed to capture both in-person and early/mail-in voting), the numbers produced by both of these research endeavors are estimates with margins of error around them. But even with these caveats understood, it is clear that Trump received the significant majority of the White evangelical vote.
This is not new, of course. Republican presidential candidates have been receiving the significant majority of White evangelical Protestants' votes for many election cycles. The White evangelical vote in 2020 is not only close to the vote of the same group in 2016 (to which I will return below), but also close to what it was when Mitt Romney challenged Barack Obama in 2012, in the John McCain/Obama contest in 2008, and when George W. Bush ran against John Kerry in 2004. We thus have five straight elections in which the White evangelical vote (as defined by exit pollsters) has come in strongly in favor of the GOP candidate.
Trump and his campaign recognized this structural pattern prior to his bid for the presidency in 2016 and targeted evangelicals in that campaign. Subsequently, once in office, Trump has done a great deal to maintain this connection with evangelical-friendly policy decisions and public pronouncements. At the same time, Biden's campaign this year made its own concerted effort to reach White evangelical voters (along with other religious voters). Biden appointed evangelical Josh Dickson as his campaign's national faith engagement director, and Dickson in turn worked diligently to reach out in many ways to the evangelical community. Plus, the Biden campaign hoped that Biden's personal Christian faith and very public discussion of the role of religion in his personal life might increase his share of the evangelical vote.
Did any of this make a difference? In answering that question, we have to rely on the Edison exit polls, since there was no 2016 AP VoteCast survey to compare to the VoteCast numbers this year. Edison in 2016 showed an 80% vote for Trump among White evangelical Protestants and 16% for Hillary Clinton. That compares to Edison's 76% Trump and 24% Biden estimate this year.
If we assume these figures are accurate indicators of the vote of the underlying population in both years, we would conclude that Trump's share of the White evangelical vote fell back slightly (by about four percentage points), while the Democratic share of the vote (Clinton in 2016 and Biden this year) increased by about seven points. Of course, both year's figures are estimates, as noted, with associated margins of error, and there were changes in Edison's methods in 2020 (compared with 2016) to account for the increase in early voting. So I'm cautious about reading too much into the relatively small shifts between 2016 and 2020, although the probability is higher than 50-50 that Biden may have gained on a relative basis. (The AP VoteCast surveys use a different methodology than Edison, but it's worth keeping in mind that its estimate of the White evangelical vote is almost exactly what Edison estimated in 2016.)
Could the small shift in the vote of White evangelicals have made a particular difference in certain key states? This argument has been advanced in reference to Georgia, a state where about a third of the vote is classified as White evangelical Protestant.
A comparison of 2016 to 2020 Edison exit polls in Georgia does show a shift. In 2016, the estimate was that Trump received 92% of the White evangelical vote and Clinton received 5%. This year's preliminary estimate is that Trump received 85% and Biden 14%. Michael Wear, a faith adviser during the Obama administration, is of the opinion that these Democratic gains in Georgia could have made a difference in the state's outcome. (Biden won the state by over 14,000 votes -- a flip from 2016, although the vote is currently undergoing a full hand recount.) "It's just an unbelievable swing," Wear said. "[Biden] basically tripled Clinton's numbers with White evangelicals. To be clear, if Biden would have performed as poorly as Clinton did four years ago among White evangelicals, he would have lost Georgia and this election."
Wear's conclusions may be overstated, given that changes in other groups' votes may also have affected the overall results in Georgia. But it does appear likely that Biden increased his share among White evangelicals, and with the closeness of the race, that change could have made a difference.
The Catholic Vote
The Edison exit polls estimate that 52% of all Catholic voters went for Biden this year, and 47% for Trump. The Edison exit polls in 2016 showed a 46% Catholic vote for Clinton, and 50% for Trump. These estimates thus reflect a three-point downturn for the Trump vote among Catholics and a five-point uptick for the Democratic candidate (Biden compared with Clinton).
The AP VoteCast estimates of the national Catholic vote this year show an almost even split: 49% of Catholics voted for Biden and 50% for Trump. This would represent a very small gain for Biden when compared with Edison's 2016 polling, with no change for Trump.
How different are these voting patterns among Catholics compared with previous elections? Available data show that Kennedy received roughly 80% of the Catholic vote in 1960 (estimates vary), but that was a different time with a different partisan structure. By 2004, when Kerry was the Catholic nominee for the Democratic Party, Catholics went for Bush (52%) over Kerry (47%). In 2008, when Obama defeated McCain by seven points overall, Obama won the Catholic vote 54% to 45%. And in 2012, Obama got 50% while Romney got 48%.
So, in a broad sense, the Catholic vote was split in each of the last three elections, with minor fluctuations from year to year. The Democratic candidate received 50% in 2012, 46% in 2016 and 52% in 2020, according to exit polls. Biden's 2020 percentage of the Catholic vote is also down slightly from the 54% that he and his running mate (Obama) received in 2008. So it's certainly safe to conclude that Biden's Catholicism did not result in a startling or substantial uptick in support for the Democratic ticket among Biden's fellow Catholics compared with Democratic performance among Catholics in previous elections.
It is important to keep in mind that the impact of Biden's Catholicism on Catholic voters is complicated and not necessarily positive. Biden has consistently run into Catholic opposition over the years because of his deviation from Catholic doctrine on moral issues such as abortion and same-sex marriage. There may have been countervailing forces this year among Catholic voters -- some who liked Biden because he is Catholic and others who were put off by Biden because he is not Catholic enough in terms of his adoption of Church positions.
The argument has been advanced that even a small shift in Catholic votes this year could have cost Trump in the Midwestern swing states with significant Catholic populations -- Wisconsin and Pennsylvania, in particular. These states, won by Trump by very small margins in 2016, flipped to Biden by very small margins this year.
Josh Dickson, the Biden campaign's faith engagement director and hardly a neutral observer, said: "In the Midwest, we saw gains that in a number of ways outpaced our margin of victory. The reason we won in these key states is because of the coalition we built. I think the work we did to engage evangelicals and Catholics undoubtedly helped us get there."
The aforementioned Michael Wear also has argued that Biden's increased strength among White Catholics in these Midwestern states was instrumental in his victory there. But we don't have reliable data to allow us to estimate the relative change in the Catholic vote in these states from 2016 to 2020, given that the Edison exit polls did not include the religion question in Wisconsin and Pennsylvania in 2016 nor in Wisconsin in 2020. As is the case relating to the impact of White evangelicals in Georgia, it is thus difficult to pinpoint the causes for small changes in these states' vote with certitude.
Americans' religious identities were clearly related to their vote in this presidential election, and in that sense, religion did factor into the outcome. About one-fourth of all voters were White evangelical Christians, and they voted overwhelmingly for Trump, providing a core segment of his base vote. About as many voters were "nones" -- those with no formal religious identity -- and 65% of them voted for Biden, providing him a key component of his winning coalition. Additionally, almost all non-Christian groups (those who identify with a religion that is not Christian) voted strongly for Biden. And the rough split in the Catholic vote -- over a fifth of all votes cast -- was an instrumental factor in the outcome as well.
But the impact of both campaigns' efforts to increase vote share among various religious groups in this year's election is a bit murkier at this point, in part because it is difficult to assess the significance of small changes in the samples used to estimate the vote and also because final data and subsequent analysis are still pending. This election -- as was the case in 2016 -- ended up with very close swing state results. It is thus tempting to say that if this group or that group had voted in marginally different ways, the outcome would have been different. It is certainly possible that Biden's pickup of votes from White evangelicals and Catholics in specific states could have made a difference in his wins in those states, but I think we need to wait a bit longer for more data and more analysis to verify those conclusions.