title: The First Bullet subtitle: MAGA is fragmenting. The weapon that built it is not. author: Rebecca Rowan publication: For the Republic date: 2026-04-14

The First Bullet

Trump is at 41% approval -- a second-term low. Gen Z dropped 42 points in a single year. Tucker Carlson delivered a 43-minute monologue calling Trump "evil." Alex Jones called him a "dementia risk." Marjorie Taylor Greene -- Marjorie Taylor Greene -- demanded his removal via the 25th Amendment. Two days ago, Orban got crushed in Hungary: 138 of 199 seats for the opposition, the highest voter turnout since the fall of communism. Bolsonaro is serving a 27-year sentence.

That data looks like an ending. Feels like one, too. Authoritarian populism losing everywhere, all at once. The MAGA era winding down. Exhale.

Now look at what's still standing.

The Weapon That Didn't Disappear

A Media Matters study from March 2025 analyzed 320 online shows and found right-leaning content commands 480.6 million followers across platforms. Left-leaning content: 104 million. That's a 5-to-1 ratio. On YouTube alone, right-leaning channels have racked up 65 billion views. Nine of the top ten highest-follower political shows are right-wing. But here's what should unsettle the celebration: 42% of those right-wing shows are categorized as comedy, sports, or entertainment. Among all shows that self-identify as nonpolitical but carry an ideological bent, 72% lean right. Political content dressed as entertainment, reaching people who'd never click on a politics tab.

None of this infrastructure emerged because Trump was president. And none of it disappears because his approval cratered. Podcast ad spending on right-wing shows is up 4x since 2023 -- roughly $5 million a month -- sustained by commercial incentives, not centralized political coordination. The engine runs on commerce. It doesn't need a figurehead to stay running. It just needs engagement.

MAGA was the bullet. The attention economy is the gun.

The bullet is losing velocity. Weber had a term for what Trump has: charismatic authority. What matters in that phrase is the second word -- it's a bond between a specific leader and a specific base. It doesn't transfer. DeSantis spent $158 million and lost Iowa by 30 points. Ramaswamy held 300+ events across all 99 Iowa counties -- twice -- for 8% of the vote. Flavio Bolsonaro's endorsement failed to unify even Bolsonaro's own coalition. The pattern holds from Coughlin in the 1930s through the present -- though it's worth noting that some authoritarian-populist movements have outlived their founders. Peronism survived Peron. The BJP survived its pre-Modi leadership. The difference is those movements built institutional structures -- parties, bureaucracies, ideological canons -- independent of one personality. MAGA has not. It has a figurehead, a media ecosystem, and a vibe.

But the gun -- the 480 million followers, the algorithmic amplification, the commercial infrastructure -- isn't sitting in a drawer. It's being maintained and improved on a commercial test range, generating engagement from MAGA's own internal fracture. Yahoo News observed it plainly: "The biggest media personalities on the American right are turning their internal disagreements into bingeable content, and the audience is rewarding them for it." The weapon works whether it's pointed at Democrats or at Trump himself.

The Hardest Objection

But here's the problem with "repurpose the weapon." Maybe you can't.

Radio was architecturally indifferent. A calm explanation of the banking system traveled through the same wires at the same speed as a Coughlin tirade. The attention economy is not indifferent. It's an optimization engine built on a $700 billion advertising infrastructure that treats human cognition as inventory. False news spreads six times faster than truth, and is 70% more likely to be retweeted (MIT, 2018 -- and that study measured falsehood versus truth, not partisanship versus neutrality, which is actually a more damning finding). Outgroup language increases sharing by 67% -- on both sides, but authoritarians define and target outgroups as core strategy. These are the revenue model. As Georgetown Law's Denny Center put it: "Traditional solutions like fact-checking, media literacy, and content moderation assume a content problem when the issue is actually an infrastructure problem."

Maybe the gun can't be repurposed because it was built to fire in one direction.

I genuinely don't have a clean answer to that. Content that depends on institutional trust, procedural legitimacy, multi-step reasoning -- the kind of democratic discourse that actually matters -- faces a structural disadvantage that authoritarian content does not. The algorithms don't penalize democracy on purpose. They just reward engagement, and fear is faster than nuance. Pretending otherwise is dishonest.

Grant the evidence its full weight for a moment. A system that amplifies falsehood six times faster than truth, that rewards outgroup hostility with 67% more shares, that runs on a $700 billion advertising engine optimized for emotional arousal -- that system isn't neutral ground where the better argument wins. The playing field tilts, and the tilt is baked in. The people who designed these systems weren't trying to undermine democracy. They were trying to sell ads. Same result.

But the argument proves too much. If the attention economy made democracy structurally impossible, we'd have no explanation for Magyar winning a supermajority two days ago, or Zelensky's wartime communication reaching millions without a media empire. And I'm not naive about what those cases are -- they involved exceptional circumstances, a former insider with unique credibility, a wartime leader with existential stakes. They don't disprove the structural disadvantage. They prove it's beatable. Harder than it should be. Not impossible.

Not without precedent, either -- though the precedent isn't what most people think.

Build During the Fight

The standard version of the Coughlin story goes like this: Coughlin weaponized radio, reached 30 million listeners -- one in four Americans. He was eventually deplatformed. FDR seized the window after Coughlin fell and repurposed radio for democracy.

That timeline is wrong.

FDR's first fireside chat was March 12, 1933. Coughlin wasn't deplatformed until 1938-1942 -- a process that took years, requiring the coordinated action of the National Association of Broadcasters, individual station owners, the Catholic Church, and the FBI. FDR didn't build after the figurehead fell. He built while the weapon was actively firing. For five years, the fireside chats competed directly with Coughlin's broadcasts on the same medium.

This matters. You don't have to wait for the figurehead to fall before you start building. Waiting is complacency.

FDR did two things at once. He built better content -- the fireside chats used radio's intimacy for democratic accountability rather than spectacle. He didn't make radio boring; he made it intimate. And he built better rules -- the Communications Act of 1934, which established the airwaves as a public resource with public obligations. Both were necessary. Neither alone would have worked. And -- this part matters -- FDR had a New Deal to talk about. The fireside chats worked partly because they explained policy that was already helping people. A content strategy without content is just a medium.

Peter Magyar just proved the content model is replicable. A former Fidesz insider (and yes, more conservative than Western progressives might hope -- his politics aren't the point here, his media strategy is), Magyar built his campaign without access to Orban's traditional media monopoly. He used social media and grassroots organizing. A viral interview on Partizan reached 2 million people in a country of fewer than 10 million. Social media wasn't suddenly fair -- he was just more authentic than what people had been getting from Fidesz. Two days ago his party won the most votes any Hungarian party has ever received.

The parallel breaks here, though. FDR had regulatory leverage -- radio spectrum was physically scarce and required government licensing -- that nobody has over privately owned algorithmic infrastructure. The Communications Act of 1934 has no obvious equivalent for the attention economy. I won't pretend that gap is small. But the direction is the same one FDR pointed in: better content and better rules, built simultaneously, on the same infrastructure the authoritarians used. What the rules look like for algorithmic platforms -- I'm honestly not sure anyone has a good answer yet. But "wait for the figurehead to fall and then figure it out" has been tried. It doesn't work.

The Countdown

The infrastructure isn't waiting. It's being commercially maintained right now, today, while people celebrate. AI is reshaping how content gets discovered, consumed, and monetized. As the Shorenstein Center puts it: "The standards aren't set yet. The architecture remains fluid." The rules of the next media architecture are being written right now -- by the people who profit from the current one.

The window between bullets is not a gift. It's a countdown. And "window" is generous -- the gun is operational the entire time. The only thing temporarily absent is the charismatic figure who aimed it.

Another bullet is coming. It always does -- and it won't look like this one. So the only thing that actually matters right now is whether anyone builds on the weapon's infrastructure before whoever comes next picks it up.

The celebration is earned. The data is real. But every time a populist leader falls and the pundits declare victory, ask the harder question: what happened to the weapon?