You're Not Getting Your Republican Reckoning, And Here's Why

I swear I am not intending to spam your mailboxes, but I woke up early this AM and I listened to Rick and Andrew Wilson's discussion on YouTube about the numbers, and I am getting sick of the delusion. The cherry-picking of polls that give them hope that Trump is (metaphorically) dying, and that his administration is cratering. But while there is a slight (but measureable) drop in his polling, it remains stubbornly at 40-41% in aggregate, and any claims that he's losing his grip on the Republican Party are mere fantasy.

Let's talk turkey about polling. I know, I know. The word alone is enough to send people into either a panic spiral or a dopamine rush these days, depending on which newsletter(s) you subscribe to. But hang with me, because the discourse around Trump's numbers has gotten so sloppy it's embarrassing, and somebody has to say it.

First, a brief rehabilitation.

Political pollsters took an absolute beating for roughly fifteen years, and some of it was deserved. The shift from landline to cellular blew up their sampling frames. Response rates cratered. The electorate kept surprising them in ways that made the pre-election models look like they were built by monkeys with a dartboard. 2016 was bad. 2020 was better but still off. 2022 gave some of them collective PTSD.

But here's the thing: they adapted. The methodology has caught up. The weighting models are smarter. The aggregators have gotten sharper at filtering out the garbage polls from the legitimate ones. Are they perfect? Hell no. But the aggregate picture they're painting right now is probably the most accurate it's been in over a decade, and that matters for what I'm about to tell you.

So. What does that aggregate actually show?

As of mid-April 2026, most major tracking groups have Trump's approval hovering around 41%, with roughly 6 in 10 Americans disapproving. Nate Silver's Silver Bulletin has him at 39.7%, with a net approval sitting around -16.6, and his economic approval at -22, with inflation approval at -34. CNN's most recent polling put his overall approval at 35%, and his handling of the economy at a career low of 31%, with roughly two-thirds of Americans saying his policies have made economic conditions worse.

Now, those are bad numbers. Objectively bad. CNN's chief data analyst has called Trump the "weakest president this century" at this stage of a second term when measured by net approval, and noted that he's been underwater every single day since March 12, 2025. That's an historically awful record.

Yet here's the part that the Never Trump podcast industrial complex [1] does not want to sit with: he's not at 32%. He's not even trending there in any clear, sustained way. He's sliding. Slowly. Gradually. Like a glacier that really, really does not want to leave.

Which brings us to Sarah Longwell and what she calls the "Bush Line"[2].

Longwell, who runs The Bulwark and has been doing this thankless work longer than most people have been paying attention, has argued for a while that the magic number is somewhere in the 32-33% range. Her theory, as she's laid it out on the Bulwark podcast, is that George W. Bush left office at 32%, and was so unpopular that not only did the next two Republican candidates lose, but it left the party open to the reshaping that Trump ultimately carried out by repudiating the Bush legacy entirely. Her hope is that Trump hitting that floor would trigger a similar reckoning. That the Republican Party would finally, finally be forced to look in the mirror.

I like Sarah. She does good work. Her focus groups are genuinely illuminating - if infuriating[3]. But I think this particular theory has a structural problem the size of a fucking aircraft carrier, and you can bet your bottom dollar I'm going to tell you why.

The Bush analogy just doesn't hold. And not just a little bit.

When Bush bottomed out at 32%, the Republican Party was still, at some level, capable of shame. They had institutional memory of a pre-Atwater, pre-Gingrich party that occasionally gave a damn about governing. No really, there was a time that they did care, I am old enough to have vivid memories of the Republicans governing. The Tea Party was born out of that shame, I will buy that, but it was also born out of a genuine, if misdirected, ideological energy. They wanted something, even if what they wanted was incoherent and increasingly unhinged.

The response to that reckoning? The party commissioned what became known as the "Growth and Opportunity Project" after the 2012 Romney debacle. The so-called "autopsy." Reince Priebus stood up and said, "Our message was weak; our ground game was insufficient; we weren't inclusive," [4]and the report laid out an extensive plan to reach women, Black, Asian, Hispanic, and gay voters. It called for backing comprehensive immigration reform, abbreviating the primary process, and building real outreach infrastructure.

You know what happened to that autopsy?

Less than four years later, Donald Trump won the White House on landslide-level support from white voters, having done the exact opposite of nearly everything the autopsy recommended.

The report was basically toilet paper. Lovely, earnest, well-researched toilet paper that got used once and flushed. The party didn't absorb the lesson. It didn't even pretend to for very long. The Tea Party energy that was supposed to be channeled into coalition-building got hijacked by grievance entrepreneurs who figured out, correctly, that the base didn't want outreach. It wanted validation. It wanted to be told that the problem wasn't them, it was everyone else.

And Trump, god damn him, understood that better than anyone. His lizard brain really nailed that.

So Longwell's theory requires us to believe that a party which used a serious, substantive post-mortem as scratch paper is going to suddenly experience a genuine come-to-Jesus moment if Trump dips to 32% during an active presidency with no electoral consequence attached to it. That the institutional GOP, which has spent the last decade systematically purging anyone who expresses even mild reservations about the cult of personality, is going to look at some polling numbers and decide to grow a spine.

I'm sorry, but what the fuck is she smoking, and can she please share it. That is some good shit.

The Never Trump ecosystem, which I read and follow because I need to know what the reasonable center-right is thinking, has a tendency to treat every bad poll like it's the opening notes of the second coming. Even Longwell herself, who is sharper than most on this, has called some numbers "a bit of a bright spot at least for me emotionally," which is touching, but also kind of tells you everything. When your political analysis is being driven by what feels emotionally reassuring, you might want to pump the brakes. I thought the Republicans (even former ones) were the "Fuck your Feelings™" party...

The core constituency still isn't moving. While "strong approval" within the Republican Party has softened from about 69% at the start of the term to roughly 58% now, 95% of Republicans still support the president. That's not a coalition in crisis. That's a coalition having a slightly less enthusiastic Tuesday. The floor is sticky as hell, and it has been for the entire duration of Trump's political career.

Here's what I actually think happens:

Trump doesn't reach 32. He's not going to reach 32. He's probably got a structural floor somewhere around 37-38% in the aggregate, because that cohort of the electorate has built their entire identity around him, and no tariff, no war, no scandal, no demented Truth Social post at 3am is going to dislodge them. They are not persuadable. They have not been persuadable for a decade. As Longwell herself said back in 2024, about 70% of the Republican Party sees Trump as its north star, aligned with those who still think the 2020 election was stolen. That's not a constituency you peel away with bad polling data. That's a belief system.

And even if, by some miracle of economic catastrophe and genuine scandal fatigue, Trump did crater to 32%? The Republican Party's response would not be self-reflection. It would be a search for the next Trump. Someone slightly more competent, slightly less chaotic, with the same authoritarian playbook but better impulse control. They'd call it a return to normalcy. They'd find someone who can read a teleprompter without sounding like they're being chased by a bear. And then they'd do it all over again.

The autopsy will not be written, because the patient doesn't believe they're sick.

I want the Never Trump project to work. Seriously, I do, even though I am in the middle of a 4-part rant about their overinflated self-importance. These are people fighting an important fight. But the polling obsession, the constant searching for the number that will finally break the fever, is a form of magical thinking that I've run out of patience for. The Bush Line is not a political theory. It's a coping mechanism. From people who tell us they are not here to sell hope, just unvarnished reality. I need concussion protocols for that level of eye-roll.

The reckoning people are waiting, nay desperate for isn't coming from polling data. If it comes at all, and that's a load-bearing 'if', it'll come from something structural: electoral losses that can't be rationalized, economic pain that can't be blamed on Democrats, a generational turnover that simply swamps the current coalition. That's a decade-long project minimum, if you're lucky. If you're not lucky, it's generations.

So by all means, watch the polls. I do. The aggregate is useful and the methodological improvements are real. But if you're hoping that the number on a polling chart is going to trigger some grand Republican reformation?

That, my friends, is a waste of a perfectly good Saturday.

Musical Coda (h/t to Krugman):

Something good:

Check out Annika on PBS, Nicola Walker is fabulous in it. (I wanted to write about that today instead of this...

1 - I like the sound of that: Podcast Industrial Complex. I think I will trademark that!

2 - She swears it is not a double entendre

3 - I have to hate listen, and be in the right mood, because the people are so incredibly stupid it beggars belief

4 - "Inclusive" from Reince fucking Priebus. You can't make this shit up, their conclusion was essentially DEI. Un-fucking-believable