When the public’s opinions aren’t what you would hope

How can Trump's immigration policies still be relatively popular?

A (narrow) majority of voters still approve of Trump’s handling of immigration policy.

YouGov-Economist Survey, April 5-8

In fact, his immigration stances are more popular than his stances on trade, the economy, or Russia, per a recent AP-NORC survey. If you’re a progressive and watching the news, this is probably shocking. It certainly was for me. There are clear, obvious horrors unfolding. How can this still be so popular?

Immigration is a complicated issue once you get into specific policies rather than a generalized “handling” question, see see my post on “What’s Popular” and G Elliot Morris here. I don’t agree with the recent line that immigration policy (the Abrego Garcia case specifically) is a political trap for Democrats. I haven’t seen any direct polling on Abrego Garcia’s deportation, but my guess is that it would be mixed in popularity. It’s also been a nexus for Republicans overreaching in their statements- talking about deporting citizens (extremely unpopular) or reveling in the cruelty of it all. I don’t think you have to follow the polling all the way to ignoring or failing to condemn the use of El Salvador prisons by the Trump admin, although I do wish our single-issue-focused environment had more space for continuing to talk about tariffs and prices at the same time. Politicians ought to have the courage to say something is illegal and monstrous when it is, and I want to stress that there is a firm line between *presenting the popularity of an issue* and *endorsing a particular view on it*.

What does keep me up at night is the large chunk of voters who very much still approve of Trump’s actions on immigration. This is the 44% who support “Mass deportation of everyone who is in the country without legal status”, per an Ipsos poll. The 35% who support letting ICE raid churches and hospitals. The 31% who support ending birthright citizenship. I vehemently disagree with these people, but I can’t deny that they exist.

Under the toplines of the most extreme immigration policies not being very popular is dramatic consistent support among Republicans. Now, some of this is likely just partisan polarization, following their party leaders on the issue, and it does erode when you start talking about children or those who haven’t committed any crimes. It’s still pretty stark.

To illustrate this point, here are some questions on immigration pulled from a Feb 13-18th Ipsos poll (because their crosstabs were easy to access). I’ve compared net support overall (a gen pop frame) vs net support among Republicans specifically.

This is *so stark*. And it’s not just this single poll finding firm Republican support for even the most dramatic immigration policies, it’s visible in all available data. Those net unpopular numbers are driven by near-unanimous Democratic opposition.

What do you do with this? If you’re a Democrat and you want to win elections, and ideally win them by a wide margin, how do you grapple with this?

“We don’t need to win these people”

This is a common reflex that I see presented often.

Unfortunately, you really really do need to win those people. Not all of them- elections are won on small margins, and these positions are far from unanimous even among Republicans. But when you’ve got a chunk of existing Democrats supporting these positions, more Independents, and a ton of Republicans, you’re going to need some of those voters in close or light-red seats. If it was just immigration, you could maybe assemble a coalition of only people who believe in expanding protections for undocumented immigrants, but that isn’t the only issue where large chunks of the electorate hold very conservative positions. It’s that plus trans issues, climate issues, crime and policing, the whole set. And it’s not that one portion of the electorate is very conservative on all of these- it’s different overlapping subsets. In the end, you cannot build a winning coalition without people who hold some conservative positions.

A lot of turnout-only analysis comes from a fundamental distaste for a broad coalition. If conceptually you can drive up turnout among people who hold only liberal and palatable positions, you don’t have to deal with the fact that some Democratic voters (and a ton of swing voters) hold these really conservative opinions. Earlier in 2024, there were some hard conversations about elements of the existing Dem coalition being more religious and more socially conservative, which drove them to Republicans as those became more salient. I don’t think most folks working in politics spend a lot of time grappling with the range of beliefs that exist in their coalition. Which like, fair, I also find this stressful and miserable to dwell on. Unfortunately, I just don’t find the math of relying exclusively on turnout to work, and I don’t find good evidence for particular candidate choices notably increasing turnout.

What about convincing them? This doesn’t work. For all you can talk about pushing the Overton window, mass opinion change tends to be slow. When you are talking about issues (like immigration) in which you’re directly opposed by a party that is dedicated to keeping its voters on side, “just explain it” is not enough. This isn’t a blank field where you can frame the issue to your liking, it’s a contested field. Convincing voters to alter their positions is extremely difficult, and not a role that campaigns and parties are well suited to. Plus, will you wait until you’ve convinced someone to be a progressive on *every* issue before being willing to include them? Just some issues? The line is, imo, unclear.

The Practical and the Emotional

Even if you don’t fully buy my point, grant me for the moment that a successful Democratic coalition will involve *some* voters with *some* conservative views. If that is true, there are two separate questions here. One is “how do we [as Democrats] win people with conservative views on some issues”, and the second is “how do we [as folks with very progressive views] grapple with that”?

The first question is not mechanically very difficult to answer- you win people with conservative views by appealing them on the basis of issues on which they *do* agree with you. For Democrats, this historically has meant the economy. Plenty of voters will happily shelve most of their political stances in exchange for jobs and low prices. Arguably this was the source of Trump’s victory this year.

It can be hard to understand, as someone with strong moral and emotional attachments to my political stances, how voters can just casually put some issues aside. If you really truly believe that undocumented immigrants are destroying the country, how is “low prices” enough of a motivator to vote for a candidate who is pro expanding immigration? The general answer is that extremity of view and salience of view are not the same. Someone can have a very extreme position (i.e. that we should deport all undocumented people immediately) but not have that view be especially strongly held or important to their decision making. Likely they haven’t even thought about it particularly hard- they heard someone on the radio talk about it, went “sure, sounds good” and never considered it more.

Polling really flattens extremity and salience into one dimension. The nature of polling questions tends to lead you to ask directional opinion only (i.e. support, oppose, don’t know) and ignore how much people care about the issue. Asking voters to rank issue importance or select what issue they care about most shows you that most people are very focused on a small set of issues, with everything else far behind. Here’s Gallup (they’ve recoded open responses where people are allowed to mention more than one issue here, so it’s not a straight forced choice between issues. Also this makes some of the response options a bit odd.):

So again, how do you reach someone who wildly disagrees with you on a given issue? Talk to them about something else where you agree, preferably something they care about more.

The emotional side is harder than the strategic side. For committed progressives, it’s unsettling to think about reaching out to voters who support harmful policies. I make no secret of my own political leanings, and I would absolutely characterize the Trump administration as harmful and cruel. It can be baffling how someone might watch what’s going on right now and still strongly support his immigration policies. But for a lot of these voters, they’re mostly *not* watching right now. They voted hoping for low prices and vaguely “tough” immigration policy, and now they’re focused on their lives as prices rise. They might be willing to vote Democratic as the economy tanks. Personally, I am willing to take those votes where I can get them. I don’t think waiting for some kind of atonement or dramatic shift in beliefs is better than taking people as they are.

Democrats tend to balk at the idea of making policy concessions (moderating on crime, for example, or stepping back from a commitment to increasing immigration) in order to win voters with conservative views. Some of this is principled objection to the policies, and some is a general feeling that it’s icky to make changes in order to appeal to people you don’t like. I would argue that this moderation, or heterodoxy, or whatever you call it, is instead an effort to build bridges to voters you can’t otherwise reach. There’s nothing good about deliberately limiting your own coalition. That’s not useful, and it doesn’t lead to you being in power and able to actually enact policy. Being a political party is about representing the people who vote for you, and expanding that pool of voters is going to involve folding in some that you don’t 100% care for. That’s the reality of being a big tent party, and I think that expanding the coalition is a positive good in and of itself.

Is this a satisfying philosophical answer about why policy positions I find awful exist and are popular? Absolutely not. That is a question for someone better read than I am. You could trace a long path through the history of American politics and liberal morality and probably not come to a solid conclusion.

Is it enough to make me feel okay about the complexity of the Democratic coalition? Yeah, it is. I feel incredibly confident that electing Democrats will lead to a better world than electing Republicans. I want more voters to be a part of that, not least because winning elections is how you get stuff done. I don’t imagine all Democratic voters would be entirely comfortable with me being in their coalition, either.

How do you grapple with voters favoring policies you disagree with on a moral level? I try to hold on to the idea that for most people, that position isn’t incredibly thought through or morally rooted. I also hang on to the fact that there are many, many places where we have common ground. Sometimes that’s even enough for them to vote on.