by Ryan P. Burge, Eastern Illinois University
Sometimes you make a graph that is worth more than a tweet. I’ve been trying to visualize as clearly as possible the changes in partisanship among white Christians over the last forty years. A lot was changing during this period, including new cultural issues, racial conflict, the end of the Cold War, all of which was driving the movement of Southern Democrats to the Republican Party, among other things. By including church attendance, I want to see if being a part of a religious institution was fueling this change.
I ran a statistical model of identifying as a Republican with controls for education, age, and gender. I broke the sample up into the three types of white Christians: evangelicals, mainline, and Catholic, then left everyone else in the graph to give a comparison group. The results for each of the last five decades are visualized below (this is a three way interaction between worship attendance, religious tradition, and decade).
It seems worthwhile to take each decade on its own and point out some things that jump out to me. Let’s start with the 1970’s. Clearly, the outliers are mainline Protestants. They are easily the most Republican group at all attendance levels. About a third of never attending mainliners were Republicans, but that jumped to about 55% among those who attended church multiple times per week. White evangelicals do display a positive relationship between attendance and Republican identification, but even at the top end just 37% were Republicans. The white Catholic line is unbelievably low – just 15-20% of all white Catholics were Republicans in the 1970’s, which is not much different than the rest of the population.
The 1980’s begin to bring about some interesting changes. Mainline Protestants still stand out, but their line has become flatter. While ~55% of the most active of them were Republicans in the 1970’s, that’s now declined to 48%. At the same time, the evangelical line has begun to pitch up significantly. At the top end of attendance, Republican identity jumped about five percentage points between the 1970’s and 1980’s. The Catholic line has become completely flat now. A white Catholic who went to Mass every day was just as likely to be a Republican as one who never attended. But, the “everyone else” camp has pitched up. Now, about 38% of those most active of the rest were Republicans, compared to just 23% in the 1970’s. An explanation that seems plausible is that Ronald Reagan was fairly popular during the 1980’s, he took all but one state in his reelection bid in 1984.
The 1990’s are when the convergence between the two Protestant camps finally happens – white evangelicals at all attendance levels are now significantly more Republican. Among the never attenders, Republican ID jumped about seven points, but at the top end it was much larger (around thirteen percentage points growth over the 80s). But, the mainline Protestant line also shifts more Republican on the low attending side of the graph a few points as well. So this is not a case of mainline Protestants becoming less Republican, the distance between Catholics and evangelicals was shrinking because of evangelicals shifting significantly to the right. At the same time, the white Catholic line now begins to pitch upward – the more religiously active, the greater likelihood of identifying as a Republican.
The graph displaying the results of the 2000’s now clearly shows that unmistakable relationship between white evangelicals and the Republican Party. However, that’s only really true among those who attend church more frequently. Among the rare attenders, Republican affiliation actually drops about eight percentage points – this may be some white evangelical Democrats who are becoming disaffected. What I mean is that Democrats may have purposely been attending church as a way to limit their exposure to conservative political discussion. The same thing also appears among less active mainline Protestants. The Republican share among low attenders drops about seven percentage points, as well. The white Catholic line also shows some of this same pattern – those of lower attendance are less Republican than the same group in the 1990’s.
The most recent decade shows two dramatic changes. First, the most religiously active mainline Protestants are much less Republican now than the same group in the 2000’s. While 52% of white mainline Protestants who attended church multiple times a week in the 2000’s were Republicans, that has dropped by ten points by the 2010’s. The line is much flatter for mainline Protestants now – just a seven percent bump from low to high attendance.
The top end of attendance for white evangelicals hasn’t shifted in the last two graphs, but things have moved on the low attendance side. While just 28% of the never attenders were Republicans in the 2000’s, it has jumped to 36% in the 2010’s. The patternfor White Catholics is unmoved in the last two graphs. About a quarter of those who never attend Mass are Republicans, compared to 40% of frequent attenders.
Taken together here’s what we have. The share of Republicans among the most active mainline Protestants has dropped fifteen points since the 1970’s. At the same time that same attendance group among white evangelicals has moved twenty five points to the Republican side. The most active white Catholics have jumped from twenty percent to forty percent Republican in that time.
White Christianity has become more Republican. I know that the counterargument is that white mainline Protestants have moved in the opposite direction, but here’s why that doesn’t move the needle:white Catholics and white evangelicals together make up about 45% of the population, mainline Protestants are about 10%. It’s also important to note that mainline Protestants and Catholics are less frequent attenders than evangelicals. The fact that the groups that used to serve as a moderating force has shrunk significantly should not be lost on us. Against this backdrop the religiously unaffiliated have risen from 5% of the population in the early 1970’s to 23% in 2018.
The causal relationship between religion and politics has become somewhat of the Gordian Knot, among political scientists. There seems to be emerging a preponderance of evidence that politics is driving a lot of the shifts in religious affiliation and church attendance over the last several decades. Djupe, Neiheisel, and Sokhey found that the people who switched their affiliation in 2016-2017 were largely those who disagreed with their pastor about Trump. Michele Margolis makes a strong case in her book From Politics to Pews, that partisanship is now the primary lens through which we view the world. Djupe and I did some analysis that concluded that people who voted for Hillary Clinton to 2016 were about 50% more likely to leave their current church than those that voted for Donald Trump. The studies are now piling up – politics has become religion. And it seems there’s no turning back now.