By Ryan P. Burge, Eastern Illinois University

About a month ago I saw a little exchange between Paul Djupe and Michael Wear on Twitter. It all started with this tweet:

I like Michael for a lot of reasons, I read his book Reclaiming Hope, and found his perspective refreshing. I think that his voice is one of the loudest and most effective for trying to get Democrats to not give up on the religious vote, and more specifically in this case, white evangelical Protestants. We all know the story: liberals/Democrats have a religion problem. While the reasons for this vary, did you know that, did you know that nearly half of white people who identify as a liberal in the United States claim no religious affiliation? That’s a staggering number. The tweet above references that larger argument: Democrats are continuing to pigeonhole themselves by ignoring the faith community. Paul, Michael, and I went back and forth on this point a little bit on Twitter, but it stuck in my brain. Did Clinton really do that much worse than Obama with religious folks?

Let’s start by looking at this question fairly broadly: where did Obama do well that Clinton did not? The map above displays the difference in vote share from 2012 to 2016 for the Democratic candidates. If one were looking for places were Clinton heldpace with Obama in 2012 it would be all through the south and even in the west. However, the flipside of that is she did much worse than Obama in the upper midwest in states like Michigan, Iowa, and Ohio. She also struggled in rural parts of the Rust Belt like Pennsylvania and upstate New York. I am not going to dwell on this too much, but if you want to read more about the margin of victory I recommend a piece from John McCormack at the Weekly Standard entitled, “The Election Came Down to 77,744 Votes in Pennsylvania, Wisconsin, and Michigan.”

How do the places that Clinton did poorly overlay with where evangelicals live in the United States? The above map visualizes high concentrations of evangelicals in yellow and low concentrations of evangelicals in purple (the data comes from the 2010 Religious Census). I think an appropriate alternative title for this map would be: Where is the Bible Belt? Remember, Clinton did poorly in the upper Midwest and the Rust Belt. Where do a lot of evangelicals live in high concentrations? Nowhere near those two places. Pennsylvania and Wisconsin are not particular hotbeds of evangelical activity. If Clinton would have boosted her support among white evangelicals, it doesn’t appear that it would have any substantive impact on her electoral college vote total from this view.

Let’s get even more granular, though. The above scatterplot compares the number of evangelicals in a county in 2010 on the x-axis with the vote share that the Democrat candidates received in 2012 and 2016. The line is a LOESS line denoting the relationship between the two variables under study. You know what is remarkable about this graph? There are basically no parts of the evangelical concentration spectrum where Hillary Clinton did better than Barack Obama. If she didn’t do well with evangelicals then she should have done better than Obama among more non-evangelical counties, right? Nope. It seems that are nearly every point along the scale that Obama received about 5% more of the vote share than Hillary Clinton did. The trend lines stay remarkably consistent from the lowest levels of evangelicals to the highest. Clinton just did worse than Obama, period.

While above scatterplot gives us a lot of insight into the relationship between Clinton and religious counties, it only does it through the lense of evangelicalism. I wanted to approach this from a more tradition neutral perspective. The scatterplot above has the same vote share on the y-axis but now the x-axis is the percentage of people in each state that said they attend religious services at least once a week. I will note that there are smaller sample sizes here because things are aggregated at the state level but the CCES is over 50,000 respondents per wave, so the margins are reasonable (however, don’t put a lot of stock in this, Utah didn’t really move 15 points in two years). For Clinton and Obama the relationship between these two variables is statistically indistinguishable. As a state had a higher share of people who attended frequently, they were less likely to vote for the Democrat both in 2012 and 2016. In fact, it’s somewhat staggering how consistent these lines are for both Clinton and Obama. Again, when measured against the evidence, Hillary Clinton’s lack of success with those who are highly religious was no different than the trend that emerged in 2012. Democrats struggle with people of faith at the national level.

Finally, I wanted to see if Hillary Clinton had targeted religious states in a way that was fundamentally different than Obama did in 2012. The above scatterplot has the number of field offices opened by each candidate on the x-axis (this data comes from Joshua Darr). Again the LOESS lines look fairly similar. It seems that Obama deployed more resources in states that were slightly less religious than Clinton did in 2016, but that may be a function of small sample sizes to calculate weekly attenders. The other result from this graph is how Clinton had fewer field offices than Obama overall. Obama had over 100 offices in both Florida and Ohio, while Clinton only placed 75 in Ohio and 72 in Florida. So, from this view it doesn’t appear that Clinton tried to concentrate offices in less religious states and relative to her total campaign effort, it may be more.

How do we make sense of all of this from a religion and politics perspective? Did Hillary Clinton have a problem drawing support from people of faith? Yes, she did, but so did Barack Obama. Neither of them had any real success in courting white evangelicals. So, if I were to advise Democrats running for office in 2020 on how to be engage with faith communities, what would my advice be? First, I would waste no time trying to sway white evangelicals to vote blue in 2020. Four in five of them will pull the lever for Trump again. If I wanted to reach out to evangelicals, I would target people of color. I have written about how Hispanic evangelicals are much more moderate, and how African Americans are a steady source of support for Democrats. Second, I would tell Democrats that the rapidly changing religious landscape of the United States should make it easier to win elections in the future. Two facts support that: religious nones are over 30% of the population and rising, while white evangelicals constitute about 18% of the population and are aging rapidly. As long as Democrats can continue to find support in the “nones” then the path for victory for Republicans will narrow. Unfortunately, that process did not occur quickly enough for Hillary Clinton.

Ryan P. Burge teaches at Eastern Illinois University in Charleston, Illinois. He can be contacted via Twitter or his personal website. The syntax for the post can be found in this gist.

Featured Image Credit: The Denver Post