Gender Identity, Religion and Political Ideology in 2023

This post has been unlocked through a generous grant from the Lilly Endowment for the Association of Religion Data Archives (ARDA). The graphs you see here use data that is publicly available for download and analysis through link(s) provided in the text below.

---

In the 2024 Presidential Election campaign there were hundreds of millions of dollars spent on advertising to convince voters to back either Donald Trump or Kamala Harris. But there was one spot that ran on television in a seeming loop. It was paid for by the Trump campaign and it focused on the issue of gender identity. It featured the following comment from Harris during a town hall in 2019, “Every transgender inmate in the prison system would have access (to gender affirming care).” It’s important to note that Harris did not campaign on the issue at all in 2024, but Trump seized on those previous remarks to try and paint the Democratic party as out of step with mainstream America on what has become a key Culture War issue.

The tagline for that advertisement was simple, “Kamala Harris is for they/them. President Trump is for you.” While it’s impossible to know if any specific campaign message was effective in winning over voters, this ad was discussed at length in major media outlets. In the Washington Post, Megan McArdle wrote an op-ed with the headline, “Trump’s ads on trans issues are effective. Harris has herself to blame.” Former Republican Governor of New Jersey, Chris Christie called it the most effective ad of the 2024 presidential election cycle.

But how do voters feel generally about the issue of gender identity? It’s something that polling has only really started asking about in the last couple of years. The Public Religion Research Institute (PRRI) polled over 5,400 Americans about a variety of topics related to gender identity back in March of 2023 and the results offer a lot of nuance related to how the public is thinking through issues related to gender.

One of the first questions on the survey is a simple one, “When it comes to gender, do you…” then they are given four response options:

1. Feel strongly that there are only two genders, man or woman

2. Think there are only two genders, but do not feel strongly about it

3. Think there is a range of many gender identities, but do not feel strongly about it

4. Feel strongly that there is a range of many possible gender identities

This is the distribution of these responses across the sample.

The majority of the folks in the sample chose the first option - they feel strongly that there are only two genders, but it’s important to note that it’s just a bare majority - 51%. Moreover, another 15% think that there are two genders but don’t feel strongly about their position. That means that about two in three Americans feel (either strongly or less strongly) that there are only two genders.

In contrast, the share of the sample who strongly believed the opposite (there is a range of possible gender identities) was relatively small at just 14%. About 20% of folks believed that there is a range of many genders, but didn’t hold that view strongly. In other words, it does seem like a significant portion of the public reject the view that there are more than two genders.

However, what role does religion play in this equation? I am using PRRI’s measure of religious tradition, which breaks the sample down into various Christian, non-Christian, and non-religious groups.

The data here points to the fact that most Christian groups have a fairly clear position that there are only two genders: male and female. Among evangelicals, two-thirds strongly hold this viewpoint - that’s fifteen points higher than the general public. Additionally, a slim majority of Catholics are also in this camp (51%). Among Christians who are not evangelical nor are they Catholic, 62% are convinced that there are only two genders. However, when you combine the “only two genders” response (whether strongly or weakly held) you find huge majorities who reject the view that there are three or more genders.

What about non-Christian groups? Jews are fairly evenly divided on this topic. About 44% take the position that there are two genders, while 56% are on the other side of the debates. It’s noteworthy that 26% of Jews strongly affirm the belief that there are many genders - that’s the highest of any group in this analysis. Among the non-religious it’s really interesting to note that only 54% of them believe that there are many genders. In other words, there’s not a group in this analysis where a large majority take the view that there are more genders than male and female.

Let me throw religious attendance into the mix now. The response options ranged from never to more than once a week. This is the share who strongly believe there are only two genders.

I think it’s fair to say that as religious attendance increases, there’s also an uptick in the portion of folks who are sure that there are only two genders. That’s definitely true for evangelicals. Among those who attend less than once a year, about 56% are convinced that there are only two genders. That percentage rises as frequency of attendance moves up. Among evangelicals who attend multiple times a week it’s nearly 90% of the sample. You can also see that same trend line among Catholics, although it’s a bit more muted. Among Catholics who attend Mass once a week, 57% strongly believe there are only two genders. It leaps to 81% of Catholics who attend Mass multiple times per week.

For the ‘other Christian religion’ group it’s a bit more of a mixed-bag. I think it’s fair to say with church attendance that there’s a clear delineation among those who attend no more than once a year versus at least once a month. The Jewish sample is really interesting because of how much of an outlier the “weekly+” category is compared to the other levels of synagogue participation. However, this is obviously a HUGE sample size problem. In a sample of over 5,400 there were only 26 Jews who reported weekly attendance. So, we can safely ignore those results unless they are confirmed with a much larger sample.

But, to this point we have only examined the issue of religion. Of course, every viewpoint in life is deeply related to all kinds of other factors: age, gender, race, education, and political partisanship. To disentangle all these things, I specified a regression model. We are trying to predict whether someone strongly believes that there are only two genders, male and female. I specified a logit regression model. The results are below - any coefficient that does not intersect with zero means that it is statistically significant. If the coefficient is positive, it predicts a greater likelihood of strongly believing that there are only two genders.

There is a total of one variable that predicts a lower likelihood of being sure that there are only two genders - education. As education increases, the propensity to hold this belief goes down. There are a handful of factors that don’t matter at all. They are: race, income and age. None are predictive in either direction.

However, the bottom four predict a greater likelihood of strongly believing that there are only two genders. They are gender, church attendance, self-identifying as an evangelical Christian, and identifying as politically conservative. But the magnitude of these variables is important to note.

The impact of being male is the same as church attendance. However, self-identifying as evangelical has twice the magnitude of either of those factors. But, without a doubt the strongest predictor of holding to the view that there are only two genders is identifying as a political conservative. That one factor is twice as predictive as being an evangelical Christian. It looks like politics matters a whole lot more than religion when it comes to views related to gender identity.

But I want to make that point clearer by estimating an interaction model of political ideology and religious attendance. I also included the same control variables as the previous analysis, things like age, income, education, gender, and race. Here’s what I get.

We can see that as religious attendance rises, there’s a clear increase in the viewpoint that there are only two genders. That’s the case for liberals, moderates and conservatives. For instance, among liberals who report never attending religious services about 12% of them strongly hold the view that there are only two genders. Among liberals who attend religious services multiple times a week, that share doubles to just about 25%. For conservatives, those figures are 73% among never attenders to about 88% among those who attend multiple times a week. Note that the slope of the lines is essentially the same for the three categories of political ideology.

But here’s the big takeaway from this graph - there is a huge chasm in views of gender based on political ideology. Think about this for a second. Among the most religious active political liberals about 28% strongly believe that there are only two genders. Among moderates who never attend religious services, that share is 40%. A never attending political conservative is three times more likely to strongly believe in only two genders compared to a liberal who attends more than once a week.

In other words, it’s not only religion that is doing the work here. It’s clearly politics.

Code for this post can be found here.