Strength In Numbers Q&A for June 2025, part 2

Who supports bombing Iran now? How much do people know about politics? How should analysts update their priors?

This is the second part of the Strength In Numbers June Q&A! The two parts do not need to be read in order. You can read part one here.

For this issue, I responded to questions about public opinion on Iran, ice cream, the drawbacks to political polling, and my philosophy on updating my political beliefs with new data.

Send me any questions you have for the July mailbag here!

---

Newsy

Cecile asks: I’m sure you’ll be covering this, but since the missile attack on Iran, has any reputable polling on the subject changed, and in what regard? I thought I saw some polling that showed Republicans, including MAGA, have somewhat very recently changed their position on this.

I'm going to have a longer answer to this after we get more data, probably in a column on Friday. The gist of it is that you should generally expect an increase in support for military action in Iran, given the follow-the-leader effect in public opinion. But that bounce may not last long depending on the consequences (if Iran retaliates, if U.S. troops get involved, if the strike did not accomplish what the president said it would/did, etc).

Early polls suggest this will be the case with Iran. The polls before the strike showed just 16% of voters wanted the U.S. to be "involved in the conflict between Israel and Iran," including 22% of Republicans. But in surveys conducted just after the strike, Republican support for Trump's actions has surged. In YouGov's polling, 68% of Republicans now support Trump bombing Iran. Meanwhile, support among Democrats didn't budge at all, and Independents moved just 9 points toward support. Note that small changes in question wording can lead to different results across surveys.

Here's a poll from the Washington Post and George Mason University showing the same thing:

Note that while overall support for bombing Iran has increased as a result of Trump's actions, the public still remains opposed overall. Opposition is still much higher than for prior U.S. involvement in the Middle East. Here's an updated version of my chart of this data:

So we've seen between a 30 and 50-point swing to Trump among his party. Talk about following the leader!

Maria asks: What percent of these illegal immigrants are really truly violent or are criminals. Are there factual numbers out there to support what Trump/Republicans keep harping about? How do we find these data?

I'll start by saying we do not have an official tally to answer this question. But we do have various heuristics available to us, which come with a margin of error. We need two pieces of information:

1. An estimate of the number of undocumented immigrants living in the U.S.

2. An estimate of the rate at which undocumented immigrants commit violent crimes

For the first piece of information, we can rely on data from the Pew Research Center. Pew calculated their estimate by taking official U.S. Census Bureau data on the number of all immigrants living in the U.S. and subtracting the number of lawful immigrants living here, according to records of migration from the U.S. Government.

Pew says 11 million unauthorized immigrants lived in America in 2022. That was before a surge under Joe Biden's presidency, though. An analysis from PBS using a similar methodology with fresher data pegs the undocumented population in 2025 at 13.7 million people. So let's say between 13 and 14 million people live in America without proper authorization.

We need to account for kids, though. About 22% of the total U.S. population is children, but we know that immigrants skew younger. As a conservative estimate one might assume 25-30% of undocumented immigrants are under the age of 18. So our denominator becomes a cool 10 million.

For crime rate, one study by researchers at the Cato Institute estimated an incarceration rate of 782 per 100,000 people among undocumented immigrants (compare that to 1,422 for native-born Americans). Violent crime among undocumented immigrants is much rarer, at 144 convictions per 100,000 (compared to 231/100k for native-born Americans!). Murder is in the low single-digits.

Extrapolating from these estimates, that would put the total number of criminal undocumented immigrants living in America at 78,200, and violent criminals at 14,400. For comparison, the number of currently incarcerated people in America is close to 2 million people, and the total criminal native-born population is somewhere around 3.3 million, according to the math spelled out above — including 530,000 violent criminals. This works out to about 37 native-born criminals for every undocumented immigrant who commits a crime.

Of course, that is a different question from the one of how many of these people are loose, living in America amongst us, ready to commit crimes. By definition, most of these people are already incarcerated (since they are included in the denominator of a study of incarcerated criminals!), so now we are back to making assumptions about the number of people who might commit a crime or have committed a crime and gotten away with it. That number is very likely much, much smaller.

And there are certainly caveats to this back-of-the-envelope math. For one thing, not everyone who commits a crime gets incarcerated for it, so the estimate of "criminals" could be slightly higher. Or, we could vary by a few percent in our estimate of children in the overall group of immigrants. Additionally, the Cato Institute study is of residents living in Texas, which has a higher crime rate than the average state. So take these estimates with a margin of error.

But the broader point here is that the actual number of criminal undocumented immigrants is much smaller than the number that Donald Trump claims. It's also much smaller than the number that Trump and his policy advisor Stephen Miller want to deport — around 4 million, if Immigration and Customs Enforcement hits its 3,000 a day target for the rest of Trump's term.

Even if you assume ICE's targeting was perfect and they snatched up every criminal undocumented immigrant in the country, they'd still necessarily be deporting somewhere around 3.9 million immigrants who aren't criminals. Deporting people like that has proven politically toxic for the administration so far. But you see a huge increase in deportations of non-criminals as a result of the administration’s policies. Given the goals of the policy enacted, that is a necessary consequence of the math.

More questions on polling

Christopher asks: What is the best ice cream flavor?

It's mint chocolate chip, obviously. Nothing more to really say here. This is just one of those self-evident, ancient truths in the world.

Chris follows up with:

**How do we get polling firms to put questions in their surveys to show us what percent of respondents appear to have no clue about reality? I think that’s important because news orgs like to run stories on polls they commission that appear to show the public firmly believes one thing or another, when people may not care that much.

I guess you should ask a pollster! We can certainly run political knowledge questions on future surveys. There is a pretty rich history of this in the academic political science literature, and the results are generally pretty disheartening. For example, in a recent survey, only 75% of Americans knew they had a constitutional right to free speech, and 15% couldn't name any branch of government.

---

The rest of this post is paywalled for members of Strength In Numbers. Paid subscriptions help pay the bills and grow our business, so we can engage in more ambitious projects and bring on freelancers and eventually staff members!

Become a member of our community for independent, data-driven journalism today, and get access to the rest of this article and other perks.