Washington Secrets: Pollster sets out the wild cards that could upend the 2026 midterm elections

ADVERTISEMENT Subscribe to the Magazine View this as website Forward this email          BY ROB CRILLY Welcome to Monday's snow-bound edition of Washington Secrets, your guide to who's doing the shoveling and who is being shoveled. Today we speak to one of the nation's most accurate pollsters on the forces that will shape the midterm elections and the unknows that could throw off polls, we sift through thousands of complaints sent to broadcast regulators about Jimmy Kimmel, and we talk you through Trump's new and old daily activity … James Johnson, co-founder of J.L. Partners, nailed the 2024 presidential election. While many pollsters underestimated the level of support for Donald Trump, his final forecast before polling day gave the Republican candidate a 3-point lead in the popular vote and the biggest Electoral College advantage of any election model. Trump eventually won by about 1.5 points. While other pollsters had to explain away their mistakes, Johnson's approach catapulted his firm up the rankings. So what does he make of the midterm elections, and what are the factors that will sort the also-ran polling outfits from the elite? Some years pollsters get it right. Some years it doesn't work out so well. How do you think it looks for the midterms? Will pollsters get it right? Pollsters had a terrible record in 2024 largely because they failed to pick up the more disengaged part of the electorate that voted for Trump. They had the same experience in 2020 and 2016. It's hard to pick up these more disengaged voters, sometimes because they haven't voted before. The difference at the midterms is that the electorate is a bit more engaged, is a bit more online, is a bit more likely to have voted before. So on paper, it means it's a bit easier to poll, and pollsters tend to do better. That doesn't mean that it's going to be straightforward. A trend that makes me worry about this year, that we saw in 2024, is that there are a lot of online polls darting around now because they're much cheaper to do, and they skewed some of the polling averages in the wrong direction in 2024. As you think about launching your own polls, what are the calculations and assumptions you're making about what the electorate will look like this time around? This is the big thing to get right. It's not so much how people are voting, but who's going to vote. What you're going to see is a lot of pollsters building what they think is the "2026 likely voter universe." What I mean by that is, who do they actually think is going to turn out in the election? At the moment, we have got a "2026 likely voter model" based on demographics. We know that men are slightly more likely to vote in midterms. We know that older people are more likely to vote in midterms. We know that graduates are slightly more likely. But what we're not currently doing is going so far as to say we think there'll be this many Democrats or this many Republicans. That will come nearer the time. And there are some unknowns in this. If you have a midterm happening when there's a huge news story, then that will motivate certain people. If Trump places himself front and center of the 2026 campaign, which he's saying he's going to do, that might increase the number of Republicans. Is there a previous cycle you reference for a starting point? 2018 is the nearest comparison to what we're doing, which is a midterm under Trump. You expect that the opposition party will be better at getting their people out, because it's easier to motivate someone to vote against something than it is for something. So you therefore expect to see a few more Democrats in the sample than you would have seen in the 2022 midterms. What is the potential unknown that could upend forecasts? What if the Trump administration issues stimulus checks based on the tariff income? It's been a much mooted thing. I still get people in my focus groups today talking about Trump's COVID stimulus checks, and that's six years ago now. It turns out paying voters works! Look, we'll see whether that's enough, but that's one of the wild cards I have out there. The other one is just how much Trump puts himself at the heart of the campaign. There are a lot of Trump voters who may not bother to turn out if his name is not on the ballot, and that is the logic behind the idea of a midterm convention where the president will be front and center. You had the most accurate presidential forecast in 2024. What was the secret to your success, and can you use the same secret sauce in the midterms? The key thing was meeting voters where they were, and by that, I mean we polled them in the information ecosystem they operated in. So we called them on their cellphone, we called them on their landline, as pollsters have always done, but we also polled them when they were playing games on their phone, when they were shopping online. That allowed us to pick up those more disengaged voters that were absolutely key. The danger in midterms is the reverse. We've got to be careful not to be picking up too many disengaged people who might tell us that they'll vote, maybe to get us off the phone, but then they've got no real intention of doing so. And what's changed since the presidential election? The number of people getting their information from AI. That's going to be very important for us to think about as a place to meet voters in the future. All right, let's not mess around. Give me numbers. What is the House of Representatives going to look like after the midterms? It's 10 months away. Any pollster telling you what's going to happen is bulls***ing you. However, from what we know about past midterms, we know this is not going to be a good cycle for Republicans. But this is not a slam-dunk election for the Democrats either. If you compare it with 2018, Democrats had 10-, 12-, 15-point leads going into those polls, and they are not there yet. So rather than looking into the crystal ball, I would tell your readers to keep an eye on the key numbers: Trump approval and turnout. And the really big one, although it's hard to track in the public polls, but we'll be tracking these and giving Secrets the inside scoop, are the mid-propensity Trump voters. These are the people who don't usually come out at elections but did come out in 2024 and voted for Trump. If they start to move, then things get very interesting indeed. Viewers complained about Kimmel — but not why you think When Jimmy Kimmel used his monologue last September to score points after the death of Charlie Kirk, it triggered a string of complaints to the Federal Communications Commission. But nothing like the deluge of angry messages when his show was taken off the air days later. "Allow Kimmel back on air you spineless cowards," reads one of more than 1,600 complaints obtained by Secrets using a freedom of information request. Our rough tally suggests that for every complaint about what Kimmel said, essentially accusing Trump's allies of trying to make political capital from Kirk's death, there were about 100 accusing the regulator of kowtowing to the president, abandoning the First Amendment, or acting like Nazis. One complaint, sent from Richfield, Minnesota, printed the word "shame" 268 times. Kimmel returned to the air about a week after being suspended. He said it was never his "intention to make light of the murder of a young man." "This show is not important," he said. "What's important is that we live in a country that allows us to have a show like this." And the right to complain, he might have added. You can read the bonus edition of Secrets here. Executive time is back at the White House Do you remember when an insider leaked three months of Trump's private schedules in 2019? They revealed swaths of his days set aside for "executive time." It quickly became clear that it was a euphemism for the unstructured time the president spent watching Fox News, telephoning friends, and posting on social media. His staff was unapologetic. Sarah Huckabee Sanders, his press secretary at the time, told Axios it was part of the "more creative environment that has helped make him the most productive president in modern history." And it seems his staff is leaning into it again. His official daily program, published on Friday evening, showed that he would be engaged in "executive time" from 8 a.m. Saturday, without specifying quite how long it would last. And again, the president will "participate in executive time" on Monday from 1 p.m. There's no secret about what's going on. Secrets understands that the president has asked his staff to juice the guidance sent out to reporters and posted online each evening listing his public events the following day. So just as we now see "signing time" or private "policy time" listed, "executive time" is being advertised, which aides say better reflects a busy day even when the events are not public. Lunchtime reading MAGA Right sours on Thune over SAVE Act fight: Senate Majority Leader John Thune (R-SD) was never a natural ally of Trump or his fire-breathing base. But he has found common cause with the president over their shared goals for the past year. But cracks are now starting to emerge over a voter ID bill and the future of the filibuster. Why Iran is betting on war: Why won't Iran just make a deal with Trump when the forces arrayed against it are so vast? Because they think war is inevitable. This is a useful way to understand just what is happening in Tehran, but it is behind a paywall, I'm afraid. You are reading Washington Secrets, a guide to power and politics in D.C. and beyond. It is written by Rob Crilly, whom you can reach at secrets@washingtonexaminer DOT COM with your comments, story tips, and suggestions. If a friend sent you this and you'd like to sign up, click here. ADVERTISEMENT