2024 Election Post-Mortem: Muslims

A big shift in 2024 among a very hard group to poll

I’m gonna be honest here - the 2024 election wasn’t chock full of storylines about religion and politics. A lot of the same battle lines that were drawn in 2020 (and 2016) were still there in the most recent contest for the White House. Trump wanted to restrict abortion, the Democrats didn’t. Trump talked about a lot of gender issues, the Democrats tried not to talk about it so much. Trump did a pretty hamfisted job acting like he was a practicing Christian and making the point that he was going to defend the Bible from the secular left. But he’s been doing that kind of stuff since he burst onto the political scene years ago. The same old stuff - just in a slightly more updated package.

But I think that there was a single religious group that really could have shifted their votes significantly in 2024 - Muslims. With the conflict between Israel and Hamas raging, the Biden administration did a pretty terrible job of listening to the concerns of Muslim Americans. The New York Times ran this headline, “Inside Biden’s Broken Relationship With Muslim and Arab American Leaders” and the Washington Post wrote a similar story, “Muslim leaders decline White House Ramadan invitation.” So, there was ample fodder for news stories about how Biden may lose his edge with Muslims, especially in Michigan, and that could end up shifting the entire election. (Which is funny in hindsight because Michigan didn’t matter at all on Election Day.)

So, did that actually come to pass? Did Harris see a lot less support from the Muslim community compared to Biden four years earlier? Let me show you the data from the Cooperative Election Study in the last five presidential election years.

The Muslim vote has always favored the Democrats. That was especially true in 2008, when Obama basically did not quite win all of their support, but pretty close - 94%. He also did incredibly well in 2012 at 85%. Trump did manage to make some in-roads in 2016, driving his share up to 22% but then he really struggled in 2020 when Biden won 91% of the Muslim vote.

So what happened in 2024? Well, this data does tell a pretty interesting story - the Muslim vote was 61% for Harris and 33% for Trump. That’s easily the worst showing for a Democrat in the last five cycles. So maybe that whole narrative about handling the conflict did bear out at the ballot box.

However, I need to take a methodological aside for a minute here. After sniffing around these numbers for a couple of minutes, I’ve come to a conclusion - Muslims are a really hard population to poll. There are two reasons for that. The first is pretty simple: a small sample size. As I wrote in my most recent book, The American Religious Landscape, only 1% of the U.S. population identifies as Muslim. Even if you have a huge sample (and the CES does), it’s still going to be really hard to get your hands around a group where the N size is small.

So we’re already starting with a small sample - in many waves of the CES, it’s 500-600 respondents. But then we have to subset that to just Muslims who reported that they cast a ballot in the presidential election. I calculated that share and the numbers are making me scratch my head quite a bit.

For instance, in the 2020 data there were a total of 525 Muslims in the sample (the entire survey was 61,000). But guess how many indicated that they cast a vote on election day? Just 101. That’s a turnout rate of about 20%. That was also the case in the 2016 data, too. Now what’s weird is that in 2008, 2012, and 2024, the turnout rate was just slightly above 50%. How do we go from 20% turnout in 2020 to nearly 60% turnout in 2024? I have absolutely no clue. The small sample size is playing havoc, but here’s another possibility - immigration.

Immigrants who haven’t become naturalized citizens cannot vote in presidential elections. So maybe what’s happening is that a huge portion of Muslims are legally ineligible to vote. I can check that out pretty easily.

Welp, this wasn’t any kind of “Eureka” moment. Over the last four election cycles, the share of Muslims who were not eligible to vote because they weren’t citizens was just about 15%. So that explains just a very small portion of the lower turnout we saw in 2016 and 2020. What’s also striking to me is how the overall immigrant composition of Muslims hasn’t changed that much over time. In fact, the results from 2012 look almost exactly like the results from 2024.

About 40% of Muslims in the United States have been here for at least three generations, while about the same share are immigrants themselves. So why this huge vacillation in turnout? Your guess is as good as mine. But let this be a lesson - even with an incredibly large dataset that was collected under the watchful care of some of the best survey administrators and public polling experts, we still can’t really nail down what was happening with the Muslim vote.

So, given that caveat, let me just show you one more piece of analysis that I can have just a bit of confidence in - Muslim votes by mosque attendance. I broke this into just two categories: weekly attenders vs everyone else because of the small sample size that was mentioned above.

I think there are two big takeaways from me on this graph. The first is that among those Muslims who are less religiously active, there’s been no clear trajectory of their voting patterns since 2012. In 2012 and 2020 about 20% of these less active Muslims were Republican voters. It was ten points higher in both 2016 and 2024. Why this see-saw pattern? I have absolutely no idea.

I think you can clearly see the huge deviation among the weekly attending Muslims, right? In 2016 and 2020, Trump did very poorly with this group. He earned just 10% of their votes. But in 2024, the data from the CES indicates that 42% of weekly attending Muslims cast a ballot for the Republican candidate. That was four times higher than the previous election.

Is that really the case? Let me show you a bit of math: There were 60,000 people in the sample. Of that, 464 were Muslims. Of that the number who were weekly mosque attenders who recorded a vote for the president was 149. And this is the absolute best we can do with social science data. I know of no other data source (certainly not a publicly available one) that gets us any closer to the answer.

But if I get away from voting metrics a little bit and focus on questions that almost all the Muslims in the sample answer, that may help to round out the picture of Muslims and politics in 2024. For instance, looking at the partisan composition of Muslims over time can offer some nice insights into what is actually happening.

Before I had a whole lot of reservations in saying that the Muslim vote is shifting toward the Republicans, but this is some really strong secondary evidence that there is some movement afoot. At least three quarters of Muslims aligned with the Democrats in 2008, 2012, and 2016. But that shifted in a real and meaningful way in both 2020 and 2024. When Trump ran against Biden, the share of Muslims who said that they were Republicans rose to 20% - up 8 points since 2016. Then, that same thing happened again. 27% of Muslims were Republicans in 2024 and just half were Democrats.

When comparing those results to 2008, the Democrats have lost thirty points among Muslims while the GOP has picked nearly 20 points. The share of Muslims who are independent more than doubled to 22%. My thought process is this - the voting results are pretty hard to parse because of the aforementioned reasons. And, by itself, the data from 2020 on partisanship could have been a weird survey quirk. But seeing that same thing pop up in 2024 has convinced me - Muslims have gone from a deep blue to a much more purple hue.

But there are two dimensions to the political worldviews of folks - partisanship and ideology. There can be moderate Republicans or liberal Independents. Looking at both dimensions can help draw a more accurate portrait of what’s happening with the Muslim vote.

Given what we know from the prior analysis, it should come as no surprise that the top row of these heat maps (the Democrats) are where the biggest percentages can be found. But note which is the most popular square - in 2012, 2016, and 2020 it was moderate Democrats. They were 35-40% of the sample. That’s a key point to remember - there’s not a huge contingent of liberal Democrats hanging out in mosques. That gives me echoes of what we see with Black Protestants. They are overwhelmingly Democrats, but of the moderate flavor (I wrote a whole chapter about this in 20 Myths about Religion and Politics in America). But do you notice what happened in 2024? The moderate Democrats dropped to just 21% - down 17 points from 2020.

Where did all those Muslims go? Well, you can see a noticeable jump in moderate Independent Muslims - up to 13%. But there’s also a big chunk in the bottom right corner - those are Republicans. In 2012, just 6% of Muslims said that they were moderate or conservative Republicans. It was 4% in 2016 and 10% in 2020. But in 2024, that figure was 26%. So this is not just a shift away from the Democrats, there’s also some evidence of a growing number of conservative Republican Muslims.

Now let me temper this by saying something very simple, yet incredibly important: there aren’t that many Muslims in the United States. According to the estimates from Pew, 0.4% of the population were Muslims in 2007, 0.9% in 2014, and 1.2% in the most recent data. There’s no reason to think that there’s going to be exponential growth of Muslims, especially given the changing immigration policy of the United States. In fact, I would be very surprised if Muslims were 3% of the population in 50 years. Thus, it’s highly unlikely that there will ever be a national election where Muslims make the difference.

That’s not to say that Muslims don’t matter in the largest tapestry of American religion and politics. I think they are telling a pretty interesting story about something I’ve written about a lot - religion in the United States is seen as increasingly conservative. That’s true among Protestants, Catholics, and Latter-day Saints. But, if this data is any guide, it’s also happening among American Muslims. That’s certainly something to keep an eye on.

Code for this post can be found here.