The Phantom Base

Is a civil war tearing apart the American Right? The question, which doubles as a threat, has haunted Donald Trump’s second term. Ironically, it stems from the same tectonic shifts in media that Trump once leveraged to win the presidency. The digital insurgency he cultivated and, at times, maneuvered so cannily has now turned on him.

Trump returned to the White House in 2025 as the first Republican in two decades to win the popular vote, having expanded his support among key voting blocs across the electorate. He began his presidency with the highest approval ratings of his political career and bipartisan backing for central planks of his domestic agenda. His victory appeared to signal a realigning election of the kind that occurs only a few times a century—reshaping the balance of power between the parties and ushering in a new political order. At the time, it did not seem fanciful to envision the birth of a national consensus that might endure for a generation.

Instead, support for the president began fracturing from the moment Trump took office. Surprisingly, much of the early pressure on the White House did not come from Democrats and liberal institutions, still reeling from their losses; it came from a power-seeking faction inside the Trumpian Right. Often falsely portrayed as a revolt of the masses, the right-wing populist opposition to Trump has been driven by a cohort of party elites made up of media figures, political operatives, and their opaque benefactors, who see Trump’s singular connection to his supporters as the main obstacle to their takeover of the MAGA movement.

Adapting Trump’s campaign playbook, his challengers on the Right are exploiting a sweeping technological shift. The twentieth-century political order, shaped by print and broadcast media, rested on hierarchical ties among politicians, parties, journalists, and the public. That interlocking structure unraveled with the rise of the internet. Five centuries earlier, the printing press helped usher in a new scientific worldview and the liberal nation-state. The internet has done something comparable in a much shorter span of time by eclipsing the old print-based world and spawning a new culture—and with it, a new form of governance. I call this emerging system the Information State. In it, power shifts from written laws and procedure-bound institutions to digital platforms governed by hidden code, operating at speeds that exceed human perception.

In the Information State, the distinction between public and elite blurs. Who, after all, is really the elite on social media—the once-esteemed institutional figure, or the “influencer” leading the mob to heap scorn on him? Power can still be measured in offline assets, of course, but the confusion between hard resources and fleeting attention, real authority and viral clout, is central to this new politics. Decades ago, at the dawn of the information age, Marshall McLuhan anticipated this change. As “means of moving information increase,” he wrote, “there occurs a fluidity of the categories of natural resources. . . . [A]lmost any natural resource can be substituted for any other as levels of information rise.” Online attention can become money; money can buy influence, muscle, or still more attention.

The newly anti-Trump Right exploits this confusion. Claiming to speak for “the base,” it stokes social-media swarms—digital stand-ins for “the public”—to dominate attention and pressure politicians into ceding control of the party apparatus. By invoking a phantom base, whose outrage they purport to channel, they practice a kind of stagecraft that defies straightforward accounts of politics.

The controversy surrounding disgraced financier Jeffrey Epstein, particularly regarding his ties to Trump, offers a striking illustration of how this dynamic works. Last July, the FBI released a memo that “confirmed that Epstein harmed over one thousand victims” (a figure that was apparently based on an unusually broad definition of “victim” and that, contrary to widespread misconceptions, overwhelmingly refers to women who were adults when they came into contact with Epstein). But the memo also rebutted in detail central claims of Epstein lore. Investigators concluded that Epstein had maintained “no incriminating ‘client list’ ” and found “no credible evidence” for the oft-repeated allegation that he “blackmailed prominent individuals.”

The memo set off a media firestorm, boosted by self-appointed MAGA spokespersons and online influencers claiming that Trump’s core supporters were revolting against him. Yet by early August, when CNN’s pollster Harry Enten checked in with voters, he found that interest was already fading. Google searches for Epstein had fallen 89 percent in three weeks. Trump’s approval ratings had not meaningfully shifted and remained historically high among Republicans. Enten called the episode a “nothingburger.”

But the saga persisted. Three months later, Epstein coverage remained a staple of both mainstream outlets and MAGA-branded independent media. Weighing in again, Enten labeled it “Trump’s worst issue by far with voters.” Even so, he conceded, there was no solid evidence that it had dented Trump’s overall approval rating. The one thing he could say with certainty was that “the more that this case is in the news, the worse it is for Donald Trump.”

Thus, the question “Is there a civil war on the Right?” can be restated in terms of the Epstein affair. If Trump’s core supporters truly see his handling of Epstein as an existential betrayal, then a MAGA crackup is real. Trump, however, dismisses the Epstein files as a fake-news narrative engineered by political opponents and a hostile media to derail his agenda. If he’s right, then talk of a civil war is itself fake, part of the same ploy.

The problem is that the Epstein affair occupies a kind of political superposition, like a quantum particle existing in two states at once. It is a lurid spectacle—impossible to disprove, drifting into fantasies about child-sacrifice rings—and largely detached from voters’ stated concerns. At the same time, it carries an intense moral charge, with the potential to explode and damage Trump (who has been accused of stonewalling and, worse, of depraved acts). As in quantum physics, an observer effect applies. The more attention the controversy receives, the more solid and consequential it becomes—not necessarily because underlying opinion has shifted but because attention itself can be converted into political leverage.

In an Information State, the struggle centers on who can generate and assume control over these bubbles of attention. The aim is to become expert at producing them so that when one bursts, another can take its place. This has become the work of a strange alliance: nominally pro-Trump figures like Tucker Carlson and Steve Bannon joining forces with liberal media outlets. Once shunned for their ties to Trump, Carlson and Bannon are now treated as credible and brave sources by publications eager to amplify stories that cast him in a damaging light.

The result: the enduring trope of a “MAGA base in revolt,” which entered the news cycle even as Trump was winning by a historic margin in 2024 and has never left. Notably, in light of factional skirmishing among right-wing elites, coverage of this supposed civil war relies less on field reporting than on breathless accounts built around overt partisan messaging and leaked quotes from anonymous administration officials.

On a single day in mid-June 2025, for instance, Politico ran one story touting “the MAGA split over Israel,” citing Tucker Carlson’s claim that Israel was dragging the United States into war with Iran, and another headlined: “MAGA Warned Trump on Iran. Now He’s in an Impossible Position.” In a lengthy post on X, Carlson warned that “the first week of a war with Iran could easily kill thousands of Americans.” He called a strike a “profound betrayal” that would end Trump’s presidency and predicted that the United States would lose to Iran’s supposedly superior military. Bannon said that military action would “tear the country apart.” His protégé Jack Posobiec asked followers what a new Middle East conflict would do to summer gas prices—after Carlson had already forecast $30-a-gallon fuel and a “collapse” of the U.S. economy.

To point out that these predictions were inaccurate is too generous. They functioned as threats, issued by the Carlson-Bannon faction and echoed by sympathizers within the administration, aimed at asserting a veto over the president’s policy. When Trump nevertheless ordered strikes on Iranian nuclear sites, his base overwhelmingly backed him. According to a CBS News /YouGov poll, 85 percent of Republicans supported the action, including 94 percent of self-identified “MAGA Republicans.”

Trump’s base faced its ultimate stress test this March, when the U.S. and Israel jointly launched a war against Iran. This time, Carlson, Bannon, and others moved past dire predictions into an open conflict with the president and his party. With the war underway, they were joined by former National Counterterrorism Center director Joe Kent, a decorated combat veteran, failed congressional candidate, and member of the Carlson media circle. Kent resigned from his post with a flamboyant open letter in which he blamed Israel for dragging America into the current war; for the death of his first wife, killed in Syria in 2019 by an Islamic State suicide bomber while deployed as a U.S. Naval officer; and for the U.S. decision to invade Iraq in 2003. Within hours of resigning, Kent embarked on a press junket that had clearly been coordinated beforehand. First stop: an interview with Carlson.

If the premonitions of civil war were valid, this was the moment when the simmering discontent on the Right should have erupted into a full-scale rejection of Trumpism. After all, Trump had betrayed his promise to end  “stupid wars” in the Middle East. Yet even as a majority of Americans expressed disapproval of the war, the military action received overwhelming support from Republican voters and proved exceptionally popular with self-identified members of the MAGA base. One poll conducted by CBS News and YouGov between March 17 and March 20 found that 92 percent of MAGA Republicans supported the military action against Iran. Of course, pollsters are often wrong—but so are podcasters. If MAGA sentiments shift, a possibility that becomes more likely if ground forces are deployed in a protracted struggle, that would only confirm the truism that unsuccessful wars are unpopular.

This is not the place to dissect the claim that Trump has betrayed his foreign policy principles. The point is that the persistence of the “MAGA in revolt” meme, despite contrary polling evidence, is not the result of an error but a matter of deliberate strategy. Durable partisan identity—what voters say when asked in surveys—has proved far more stable than the daily churn of online outrage. But in an era when political warfare has become indistinguishable from information warfare, what matters operationally is not only settled opinion but the highly pressurized bursts of attention that can be manufactured around particular controversies. Attention is elastic but finite; once fixed on one issue, it is diverted from others. The value of that resource was underscored by a recent remark from a nominee for a senior Trump administration post, who told The New Yorker: “If we have something that’s popular in right-wing Twitter, the White House is acting on it ninety-plus per cent of the time.”

Attention is also reflexive rather than deliberative; it has no goal beyond participation. External actors can harness this latent but unchanneled force. Though swarms express immediate demands (release the Epstein files!), they are inherently transient, dissolving unless an outside tribune claims to speak for them. This may not be readily apparent, however, to the dentist suddenly besieged by anonymous accusers after his name is offhandedly mentioned in a random Epstein document, or to the politician confronting what appears to be an enraged mob on X. Good luck convincing them that the swarm is a passing mirage. Its targets feel compelled to appease it, and the simulation of public pressure becomes real currency, redeemable for policy concessions or personnel changes. That is how the imitation of reality known as the hyperreal becomes real. 

A recent Manhattan Institute study illuminates another aspect of the problem. Examining the composition of Trump’s supporters, it poses what it calls “the central question confronting the modern Republican Party”: Can Trump’s coalition remain cohesive once he leaves the stage? The report pays particular attention to the rise of openly anti-Jewish, racist, and anti-Israel voices on the Right, prominently represented in the camp fomenting a “MAGA revolt.” The study finds that such views remain well outside the mainstream among Trump voters, though they are held by a growing minority. Seventeen percent of current GOP members, for instance, fall into a category that the report labels “anti-Jewish Republicans.” That figure does not represent the party’s center of gravity, but it is large enough to supply the energy for attention-driven swarms that pressure political actors by claiming to speak for “the base.”

The findings show a sharp divide between older, traditional Republicans, who hold conventionally conservative positions on most issues, and newer entrants to the GOP coalition. This second group is “younger, more racially diverse, and more likely to have voted for Democratic candidates in the recent past.” Though they identify as Republican and backed Trump in 2024, they are “often substantially more likely to hold progressive views across nearly every major policy domain.” At the same time, they account for much of the rise in overtly racist and anti-Jewish sentiment within the party.

The ideological fault line is not just generational but technological, reflecting the uneven impact of the digital media environment. Younger Trump voters, instinctively attuned to social media and podcasts, are more likely to participate in attention-driven swarms and are especially susceptible to their pull. In its affective dimension, their world increasingly resembles the medieval culture of orality more than the print and broadcast culture that shaped Trump and most of his core supporters.

Unsurprisingly, then, younger people show the greatest enthusiasm for the closest thing we now have to a medieval cult belief, complete with demonic Jews and child-sacrifice orgies. “The people that are inordinately interested in Epstein are the new members of the Trump coalition,” White House Chief of Staff Susie Wiles told Vanity Fair. These are people, she said, whom she “think[s] about all the time—because I want to make sure that they are not Trump voters, they’re Republican voters.” The Trump White House, in other words, faces a bind reminiscent of the Biden administration: whether to resist or pander to the ideological demands of its most radical internal factions.

“Though digital swarms express immediate demands, they are inherently transient, dissolving unless an outside tribune claims to speak for them.”

In a study published shortly after the Manhattan Institute report and building on its findings, political scientist Eric Kaufmann examined the influence of right-wing media on its audience. He opens with a caution: “It’s time to press pause on the panic about antisemitic and racist influencers taking over young conservatism.” The radicalizing impact of openly anti-Jewish figures such as Groyper leader Nick Fuentes, Kaufmann contends, has been overstated. “Fuentes and others are infotainers,” he writes, “with very little impact on public opinion.”

Kaufmann’s advice to avoid panic and apply sober analysis is sound. But dismissing Fuentes as merely an infotainer misreads a figure who operates less as a mass mobilizer than as a cultural vanguardist. It also assumes a stable relationship between public opinion, identity, and political action. In reality, these are volatile categories, where small but strategically significant shifts in sentiment can create pressure points that cause large-scale policy cascades.

Consider the march of transgender causes within the Democratic Party’s coalition. Though the cause never commanded majority support among Democratic voters, within a few years a small activist vanguard had turned it into a defining party issue—to the point where, during her 2020 presidential run, Elizabeth Warren pledged that her choice for education secretary would need to meet the approval of a nine-year-old transgender child. Though this development is often described as ideological capture, its rapid pace suggests that it was also a downstream effect of digital technology.

Public opinion is increasingly an effect, not a cause, of the political battles now unfolding. In that context, it’s premature to dismiss Fuentes’s influence. History shows how small, disciplined cadres can reset moral baselines within far larger institutions, producing cumulative effects well beyond their numbers.

Walter Lippmann, one of the twentieth century’s most influential intellectuals, effectively founded the modern study of public opinion with his 1922 book Public Opinion. Contrary to the romantic hopes of democratic reformers, Lippmann argued, self-government is largely impossible in complex modern societies. Advanced industrial nations like the United States had become too intricate and far-flung to be fully grasped, let alone directed, by ordinary citizens. “Modern society is not visible to anybody, nor intelligible continuously and as a whole,” he wrote.

Yet people need some sense of the whole in order to act within society. They achieve it by representing the world as a story—and that story, more than the underlying reality, guides behavior. We act, Lippmann observed, according to “the pictures in our heads”—abstractions that “do not automatically correspond with the world outside.” Opinions form through the interplay of those stories with the new information a person encounters.

The public, a term that seems to designate a fixed and venerable subject in a democracy, is really another kind of abstraction. It serves as shorthand for an undefined subset of the population but has no independent social reality. As Lippmann wrote, the public “is merely those persons who are interested in an affair and can affect it only by supporting or opposing the actors.” By “actors,” he meant the elite technocrats, officials, and other “doers” who operate society’s complex machinery that functions beyond the grasp of ordinary bystanders. Unable to operate that machinery directly, the public acts indirectly—by fixing its attention on an issue until the doers respond to its will.

The underlying formula might be expressed as: public = attention + opinion. A gawking crowd remains politically inert until it expresses a desire about the object of its attention. That desire is opinion. The formula, which describes a core dynamic in the process of how political change occurs, is simple enough to automate with bots that simulate both attention and opinion. It can also be “Astroturfed” by purchasing political messaging and activist campaigns. And finally, it can be engineered by revising platform code, a form of algorithmic social engineering that steers users toward certain stories or rewards particular views with likes and reposts, while suppressing others. All such practices are common online, though their full extent is unknowable, since tech companies exercise proprietary control over the underlying rules that structure the “public” part of the “digital public square.” 

Only a few decades ago, public opinion was filtered through civic institutions. Today, much of that mediation has shifted to global social-media platforms, where anyone can publish. As Elon Musk declared after buying Twitter and renaming it X: “You are the media now.” One might expect opinion in such a setting to become more direct and authentic, and X does host an extraordinary amount of valuable information about the world. But it is also a maelstrom of rumor, bot-driven AI slop, propaganda, and coordinated campaigns pushed by paid influencers. It’s no accident that X became a central hub of Epstein coverage, a hysteria that Musk has helped amplify and continues to elevate in the platform’s feed.

The machinery for conducting mass information campaigns to shape public opinion, now a hallmark of the Information State, was assembled a decade ago during the Obama administration. It was publicly described by Ben Rhodes, Obama’s close confidant and speechwriter, who served as deputy national security advisor. Seeking to generate the appearance of broad support for the administration’s Iran deal—Obama’s second-term signature initiative, which faced strong opposition from voters and Congress—Rhodes developed a coordinated media strategy to drive the narrative.

Working in his favor, Rhodes saw that the press had been hollowed out and could no longer act as a skeptical check on the Iran policy, which the White House had shrouded in layers of deception and technical detail. “The average reporter we talk to is twenty-seven years old,” Rhodes told a journalist in 2016. “They literally know nothing.” The institutional memory and reporting depth built in to the twentieth-century print model had given way to a new informational pipeline that ran through social media and was considerably easier to manipulate.

On social media, Rhodes coordinated dozens of Democratic officials, think-tank analysts, and media allies to act as a chorus for the Iran deal. Nonprofit staffers from impressively named organizations supplied quotes to sympathetic reporters, who then presented the approved messaging as objective analysis. Together, they formed what Rhodes himself called an “echo chamber.” “They were saying things that validated what we had given them to say,” he admitted. The echo chamber served a dual purpose: to persuade the public while simultaneously impersonating public opinion. By creating the impression that public consensus already existed, it pressured the undecided to get in line with what appeared to be the prevailing view.

The functioning of the Information State in the Obama era was technocratic, relying on expert administrators and appeals to scientific authority. Messaging flowed from the top down through the party hierarchy. Talking points that originated with people like Rhodes were disseminated through the constellation of Democratic-subsidized institutions. The result was ideological cohesion within these interconnected spheres and the projection of a hegemonic consensus to the broader public. 

After the party issued a new edict on DEI or gender-affirming care, pressure to adopt it radiated across the establishment. Social-media swarms threatening to cancel people for racism, transphobia, and other moral crimes functioned as disciplinary enforcers, accelerating the adoption of new standards. Prestigious universities, august medical bodies, former presidents, and decorated generals—collectively carrying centuries of accumulated trust—publicly bowed to the demands of social-media hashtag campaigns and affirmed the new oaths. Many Americans came to resent the censorship and perpetual loyalty tests. Yet the technological machinery that enforced compliance proved remarkably effective.

The same digital infrastructure for shaping mass opinion, now a hallmark of the Information State, is being used, albeit in a different register, both by and against the Trump administration. The infrastructure itself is neutral: it can support centralized, technocratic coordination or decentralized, swarm-like agitation. If the earlier Democratic model evoked a towering, centralized computer, the current Republican one resembles something more chaotic: a violent pinball machine. Buttons and levers are frantically slammed, carnivalesque lights flash, and alarms blare, until the intended message is delivered to its audience and the ball rolls back to its starting position.

Lacking the Democrats’ organizational depth and party discipline, the Trump administration has pursued a digital strategy relying heavily on social-media influencers as its emissaries to public opinion. It has effectively elevated a cadre of people whose chief credential is online reach to supplant the legacy institutions it wishes to marginalize. Those institutions, having gutted their own credibility in the recent past, find few defenders now. Still, this is not a simple story of progress. The Democrats’ centralized and tyrannical coordination of opinion has given way to a decentralized network built around charismatic, and sometimes corrupt, individuals linked by overlapping interests and social connections.

Working with influencers “is probably the best use of time that you can spend as a policy person or campaign person,” Trump’s former press secretary Sean Spicer told the Wall Street Journal earlier this year. A Pew Research Center study from November 2024 identified roughly 2,000 active “news influencers,” a figure that may have grown since Trump returned to office. In a reversal of traditional media patterns, the report found that right-leaning influencer accounts were more prevalent than left-leaning ones by a margin of 27 percent to 21 percent.

What influencers actually do is harder to quantify. Federal laws regulating lobbyists and political action committees don’t apply to social-media personalities. Influenceable, a new firm cofounded by former Trump campaign manager Brad Parscale, connects clients seeking to promote messages with influencers willing to post them. No clear metric captures how widespread the practice is, but signs on social media suggest that it’s rampant. Right-leaning echo chambers routinely materialize, as networks of influencers pivot in unison to a policy issue, using strikingly similar language.

In some respects, the influencer economy is a digital update of traditional lobbying—less regulated at the moment and thus more vulnerable to abuse, but also smaller and seemingly less consequential than what takes place on Capitol Hill. Yet the industry also presents an open invitation for self-interested factions and foreign governments to meddle in American politics. For relatively modest sums, a moderately clever foreign intelligence service can coordinate a messaging campaign through a network of influencers, amplify it with a battalion of bot accounts, and loose a swarm capable of sending anxious staffers in Washington scrambling to revise a targeted policy.

More fundamentally, the influencer model elevates a form of representation at odds with the constitutional order and institutions like Congress. Even the most corrupt politicians once owed some minimal accountability to voters. The same was broadly true of newspapers and other pillars of the old media world. Influencers, by contrast, have no obligation to an idealized public and no incentive to moderate in the name of a shared civic interest. They can, and often do, cater to narrow, hyper-ideological audiences, feeding them a steady diet of controversy and outrage.

The parallels between the mechanics of woke politics that cannibalized the Democratic Party and the forces now at work on the Right should not be ignored. The so-called right-wing civil war is, in part, the product of a technological transformation beyond Donald Trump’s control. But it is also partly a consequence of how his administration has handled that transformation.

Trump and his allies cultivated support from influencers and allied media figures, such as Tucker Carlson, by amplifying conspiratorial themes like the Epstein mythology as a campaign tactic. Then, after winning, they failed to sever those ties when segments of that network veered into open anti-Semitism and began turning on Trump himself. Wary of backlash from a histrionic figure like Carlson or of alienating influential, Epstein-focused podcasters, the White House has too often tried to pet the hyenas.

Having just defeated a party that consumed itself by confusing the voting, taxpaying public with its distorted online reflection, Republicans now risk making the same mistake. Agencies that should be focused on executing broadly popular policies competently are instead busy producing “based” memes. That inversion of priorities won’t end well.

The signs are already visible: an advanced purity spiral in which attention-seekers perform for other attention-seekers, mistaking their cracked mirror of the world for reality—and then encoding it into policy. This is how a political class drifts into absurdity.