n the landscape of contemporary film, Citizen Vigilante arrives less as entertainment than as a cultural artifact that captures a fracture already visible in official records across Europe. Directed by Uwe Boll and starring Armie Hammer, the movie follows a wealthy American veteran who becomes a masked avenger in European cities. He hunts violent criminals, rapists, and the corrupt officials who shield them. The narrative centers on gang rape and predation tied to unchecked migration from Islamic societies. It portrays indigenous citizens who never consented to the demographic and cultural transformation of their countries. These citizens watch perpetrators from migrant backgrounds receive lenient treatment or escape justice while natives who complain face prosecution or ruin. The film does not invent this world. It dramatizes one that government statistics, independent inquiries, and public polling have documented for years.

The core claim is straightforward. European elites imposed a civilizational transformation without asking the people who would live with the results. Citizen Vigilante gives artistic form to the resulting sense of stolen sovereignty. When the state will not defend the native population, individuals step forward. That premise rests on a factual foundation that begins with specific events and scales to national patterns.

Consider the night of December 31, 2015, in Cologne and other German cities. Coordinated groups of men, predominantly North African and Middle Eastern, many of them asylum seekers or illegal aliens, sexually assaulted approximately 1,200 women. Cologne alone recorded around 650 assaults, including 22 rapes. Police communications initially instructed officers to downplay the ethnic profile of the perpetrators. Media outlets followed suit for days. Only sustained public pressure forced fuller acknowledgment. The attacks were not an isolated outburst. They served as an early, public demonstration that the 2015 open-door policy carried immediate, foreseeable costs to women’s safety. Authorities had been unwilling to discuss those costs honestly beforehand, during or even after.

The United Kingdom presents a longer and more systematic record. Between 1997 and 2013 in Rotherham, hundreds of children, overwhelmingly white British girls as young as eleven, were groomed, gang-raped, trafficked between towns, beaten, and in some cases threatened with guns or petrol by predominantly Pakistani men. Taxi drivers and takeaway workers used alcohol, drugs, and intimidation to control victims. Police and council officials received repeated, detailed reports yet refused to act decisively. The independent inquiry led by Alexis Jay concluded that a widespread perception that authorities should downplay the Islamic dimensions of the abuse, for fear of racism accusations, paralyzed effective response. Similar organized networks operated in Rochdale, Telford, Oxford, Oldham, and elsewhere. Thousands more victims suffered under the same pattern.

Rupert Lowe’s Rape Gang Inquiry Report, released in June 2026, extends this record across the entire country. The 219-page document identifies evidence of identical grooming gang operations in at least 149 local authority areas. It traces the pattern back to documented cases in Bradford as early as 1955. Extrapolating from established local patterns, the report estimates at least 250,000 predominantly white British girls have endured repeated rape, gang rape, trafficking, torture, and in some instances forced Islamic conversion attempts since the middle of the last century. The gangs operated with the active or passive consent of public authorities. Police, social services, the NHS, and local councils treated victims as promiscuous or problematic rather than as targets of organized crime. Warnings were ignored. Files were not pursued. Perpetrators from recent migrant backgrounds were rarely deported even after conviction. The report describes this history as a rotting stain on Britain’s past and states that many citizens no longer trust the government to investigate its own failures without external pressure.

Germany’s Federal Criminal Police Office data reveals a consistent national pattern since the 2015 migrant influx. Non-German nationals, who comprise roughly 15 to 16 percent of the population, account for approximately 35 to 42 percent of crime suspects overall in recent years, excluding pure immigration violations. The disparity sharpens in violent and sexual crime. In 2024, of 11,329 identified suspects in rape and sexual assault cases, 4,437, or about 39 percent, were non-Germans. Asylum seekers and recent arrivals have been overrepresented by factors of 3.9 to 4.2 times or higher in sexual offenses. Certain North African and sub-Saharan groups have recorded rates ten to twenty-one times the German average for rape and sexual assault in some datasets. Rape and serious sexual assault cases rose steadily, reaching 13,320 in 2024 with further increases into 2025 and 2026.

These numbers sit alongside a two-tier system of enforcement. Native citizens who post on social media about crime patterns or criticize migration policy have faced fines, job loss, or prosecution under hate-speech provisions. Meanwhile, many perpetrators from favored migrant groups receive de facto protection through institutional reluctance to name ethnic or cultural patterns. The asymmetry produces the precise sense of inverted justice the film dramatizes. Equal protection under law yields to selective enforcement that shields one set of groups while criminalizing dissent from another.

A reader might reasonably ask whether these disparities simply reflect the youth and maleness of recent migrant cohorts or broader socioeconomic conditions. Official German analyses and age- and gender-adjusted studies still show substantial overrepresentation. The organized grooming cases in Britain display a consistent offender profile tied to specific communities and, according to multiple inquiries, to cultural attitudes toward non-Muslim girls and concepts of consent that clash with host-nation norms. These are not claims that every individual from a given background offends. They are observations of measurable patterns that policy must address rather than deny. Suppressing the ethnic and religious dimensions of the crimes for fear of racism accusations did not make the problems disappear. It multiplied victims and destroyed public trust.

Public opinion data confirms the democratic breach. A 2017 Chatham House survey across ten European countries found an average of 55 percent agreeing that all further migration from mainly Muslim countries should be stopped. Majorities appeared in Poland at 71 percent, Austria at 65 percent, and Germany at 53 percent. Subsequent years of rising support for restrictionist parties across the continent show the sentiment has not dissipated. Citizens did not vote for this outcome in any binding referendum that authorized the transformation of the demographic and cultural character of their countries. When they object, they are labeled, censored, or in some cases prosecuted while the underlying patterns persist.

Viktor Orbán has articulated the civilizational stakes without euphemism. He has stated that Hungary does not see these people as Muslim refugees but as Muslim invaders. Drawing on his nation’s historical experience under Ottoman rule, he warns that large-scale Muslim immigration produces parallel societies that cannot peacefully coexist with Christian Europe. The evidence from crime statistics, grooming inquiries, and the emergence of areas where host-nation norms no longer fully apply lends weight to that assessment. Countries that resisted the 2015 wave, such as Hungary and Poland, maintained far lower rates of these specific crimes and preserved greater social cohesion. Germany and Sweden, which accepted the largest shares relative to population, recorded the sharpest rises in sexual violence and gang-related crime.

The reception of Citizen Vigilante itself illustrates the dynamic at work. Germany’s ratings board denied the film any age classification, effectively blocking theatrical and major streaming release on the grounds that it incited violence against migrants. Director Uwe Boll called the decision deliberate censorship and pursued legal action. On June 25, 2026, the full film became available for free on 𝕏 for forty-eight hours. The platform’s amplification produced millions of views, a sharp spike in German search interest, and a 96 percent audience score on review aggregators. Professional critics offered cooler assessments. The attempt at suppression generated the classic Streisand effect. It confirmed both public appetite for the narrative and the futility of elite information control in the digital age.

This pattern carries practical consequences. Suppressing legitimate grievances about crime, cultural incompatibility, and demographic change does not resolve the underlying issues. It erodes trust in institutions, radicalizes segments of the population, and creates the precise conditions in which vigilante fantasies migrate from screen to street. Lowe’s report includes an explicit pledge that if government fails to take necessary steps, private prosecutions will follow to obtain justice. The film and the report converge on the same point. When the state forfeits its core duty of protection, citizens and elected representatives alike turn to remedies outside official channels.

The moral argument is more fundamental still. The first duty of any legitimate government is the physical protection of its own citizens, especially the most vulnerable. When that duty is deliberately subordinated to ideological experiments in mass migration and multiculturalism, and when criticism is punished while predation is excused, the government forfeits its moral claim to authority. Leaders who expose their own women and children to predation in order to signal virtue commit the ultimate betrayal of both the living and the unborn who will inherit the resulting societies. The data from Cologne, Rotherham, national statistics, and the Lowe inquiry document the costs in concrete terms. The resonance of Citizen Vigilante, despite official resistance, demonstrates that the public recognizes the breach of the social contract.

The film therefore functions as a warning rather than mere entertainment. It shows what happens when the state stops defending the native population and instead shields favored migrant groups while criminalizing native dissent. It dramatizes the artistic expression of stolen sovereignty. The choice now facing European governments is whether they will restore equal protection, honest enforcement, and genuine consent over borders and culture, or whether they will continue down the path that makes the vigilante not an aberration but a predictable response. The record is clear. The public already knows. Further denial only deepens the fracture.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly. Data in sponsored partnership with Polymarket.

In the late 1990s, before the term 'Ai' meant anything to anyone, I built a network of data centers. We lit up floors in New York and Los Angeles, in St. Louis and Chicago, in Dallas and Miami. They hummed with the servers that carried the first great wave of the commercial internet. And not one of them consumed a single drop of water to keep those servers cool. Not a gallon. We rejected heat the way a car rejects heat, with air and with glycol circulating in a sealed loop, the warmth carried out to the rooftop and handed off to the sky. The water bill for cooling was zero, because there was no water in the cooling.

I tell you this not for nostalgia but because it settles an argument that has lately been dressed up as a crisis. You have read the headlines. The data centers are coming, the story goes, and they will drink your rivers dry. The artificial intelligence boom will guzzle the reservoirs while families are told to let their lawns go brown. It is a vivid picture. It is also, in its central premise, false. And I can say so with confidence because I solved this supposed crisis a quarter century ago, on a budget, with technology that was not even new then.

Here is the first thing to understand, and everything else follows from it. A data center is a heat machine. Servers take in electricity and convert nearly all of it into heat, and that heat has to go somewhere. Cooling, properly understood, is simply the removal of heat. Now, you can move heat with water, by letting it evaporate and carry the warmth away, which is what a cooling tower does and what your own skin does when you sweat. Or you can move heat with air and refrigerant and a closed loop of glycol, dumping it straight into the atmosphere with no evaporation at all. Water and air are two interchangeable media for the same job. Water is not a fuel the machine burns. It is one of two tools, and the other tool uses none of it.

This means that water consumption in a data center is a design choice. It is not a law of physics, not an unavoidable cost of computation, not a debt that civilization owes the watershed in exchange for email. It is a knob, and the operator decides where to set it. Microsoft’s own engineers measure this with a metric called Water Usage Effectiveness, and the industry’s stated ideal for that number is 0.0, achievable by any facility that rejects its heat through air rather than evaporation. Zero is not a fantasy on a whiteboard. Zero is what I ran in six American cities while Bill Clinton was still in office.

Why, then, would anyone choose to use water at all? Because the knob has two sides, and engineers are not fools. Evaporative cooling is more efficient with electricity. When you let water carry the heat away, you spend less power running chillers and fans. So if you build in a place where water is cheap and plentiful and electricity is expensive, turning the knob toward water lowers your power bill and your strain on the grid. If you build in a desert, where water is precious and the sun is free, you turn the knob the other way, toward air, and you pay a little more for electricity instead. This is the entire debate, stated honestly. As the economist Thomas Sowell put it, and the line governs this whole subject, “There are no solutions. There are only trade-offs.” Water buys cheaper power. Air buys dry operation. A grown-up picks the setting that fits the place.

And the price of choosing air, of going completely waterless, turns out to be remarkably small. We have a clean measurement of it. In Nevada, a single operator ran two comparable facilities side by side, one leaning on evaporation and one running essentially dry. The waterless building’s total overhead power ran only 6.4% to 10.2% higher across the year. Microsoft, which in August 2024 mandated zero-water cooling for every new data center it builds, calls the added energy cost of eliminating water “nominal,” and drove its own water metric down from 0.49 to 0.30 in three years while doing it. The company chose to pilot this design in Phoenix, in the Arizona desert, the single hardest place on the continent to make the case, the very spot critics held up as proof that data centers must guzzle water. Each such facility avoids more than 125 million liters a year. The place that was supposed to prove water cooling is mandatory became the showcase for proving it is optional. The crisis, in other words, has an off switch, and the largest operator on earth is flipping it across its entire fleet.

So much for the claim that data centers must use water. Now consider the louder claim, that the water they do use makes them gluttons draining the nation dry. This is where the panic depends entirely on hiding a number from you, the only number that matters, which is the denominator.

All the data centers in the United States together consumed roughly 17.4 billion gallons of water directly in 2023. That sounds enormous until you set it beside the national total. America uses about 322 billion gallons of water every single day. Run the division and the entire data center industry accounts for about 0.015% of national water use, fifteen thousandths of one percent. Every alarmist headline you have read survives only by quoting the big absolute number and omitting the total it sits inside. Anyone who shows you “billions of gallons” without showing you the national figure is either innumerate or counting on yours.

The honest comparisons are devastating. American irrigation uses 118 billion gallons a day. Thermoelectric power plants use 133 billion gallons a day. Watering residential lawns consumes nearly 8 billion gallons a day, which is to say American grass drinks well over a hundred times what the computers do. The country’s golf courses applied about 531 billion gallons in a single year, roughly thirty times the direct water of every data center combined. California’s almond orchards consume around 4.2 billion gallons a day, dozens of times the data center figure, so that we can have almond milk. Nobody is holding congressional hearings about fairways or nut groves. And consider what the water buys. By one widely circulated calculation, five gallons routed through a data center generates about $132 of economic output, against about two cents for the same five gallons sprinkled on almonds. One of those gallons underwrites American productivity, national security, and technological supremacy. The other waters a snack.

Share

Then there is the viral centerpiece of the whole scare, the claim that a single question to an AI chatbot drinks a 500ml bottle of water. It is the most repeated figure in this debate and the most thoroughly false. It traces to a 2023 estimate built on an obsolete model and measured per long multi-page response rather than per query. OpenAI’s own disclosed figure is about 0.000085 gallons per query, roughly one fifteenth of a teaspoon. A single hamburger takes hundreds of gallons to produce. The bottle-of-water-per-question line is not a statistic. It is propaganda.

And that word is the right one, because the obvious question is why. If the engineering is solved, if the scale is trivial, if the marquee statistic is fabricated, why does this panic march on with such discipline and such funding? Ask who profits from a crippled American technology sector. We are in a race for artificial intelligence dominance, a race with the most direct consequences for military and economic power in the coming century, and our adversaries understand that they cannot win it at the server rack, so they would dearly love to win it in our zoning boards and our op-ed pages. The water scare is a weapon aimed at exactly that target. It is a foreign malign influence campaign, seeded to slow the construction of the infrastructure that keeps America ahead, and it is amplified, witlessly or willingly, by the environmental and anti-capitalist organizations that have spent decades looking for a reason to say no to growth. They were handed a reason. They did not check the denominator because they did not want to. The campaign tells Americans that their faucets are under threat by the machines that will define the future, and it is, root and branch, a lie.

The answer to a manufactured scarcity is not surrender. It is to build. Build the data centers, and build them with the cooling that fits the place, dry in the desert and efficient where water runs free. Build the power plants, the gas and the nuclear and the transmission, so that choosing air over water costs us nothing we cannot easily generate. And where a local system genuinely bears a new burden, make the wealthy company that caused it pay full freight for every pipe and every permit, which is simply how honest commerce works. None of that requires believing the lie. I built six waterless data centers when the technology was crude and the internet was young. The notion that we cannot do now, at will, what I did then is not an engineering claim. It is a story told by people who would rather America lose. Do not buy it.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.

The recent Los Angeles mayoral primary placed California’s election mechanics under an unforgiving light. On election night Spencer Pratt held a clear path to the runoff against Karen Bass. Late mail ballots then arrived in batches that favored Nithya Raman so heavily that she overtook Pratt and finished with a 3,113-vote lead. NBC Los Angeles recorded one Friday update in which Raman received twice as many votes as Pratt, followed by continued narrowing on Saturday and the final overtaking on Sunday. Bass’s share stayed roughly stable at 34.68 percent while Raman climbed to 27.12 percent and Pratt fell to 26.69 percent. Observers noted that the arithmetic required for Raman to erase Pratt’s lead demanded an unusually large share of the remaining ballots, a distribution bordering on a mathematical impossibility under normal variation. The early leader’s margin collapsed only after the delayed counting of mail ballots from skidrow voters that California law permits counties to process for up to thirty days after election day.

This sequence did not occur in a vacuum. It unfolded inside a system built since 2020 on a series of deliberate policy choices. Assembly Bill 37 made permanent the practice of mailing a live ballot and return envelope to every registered voter before every election. In the 2024 general election California reported 22,595,659 registered voters and 13,034,378 mail ballots that were ultimately counted. That left roughly 9.56 million ballot packets that were printed, mailed, and never returned as counted votes. Those packets move through ordinary mail, apartment mailrooms, and forwarding addresses that may be years out of date. California law allows any person to return a completed ballot so long as the person is not paid on a per-ballot basis. The sole front-end control is a signature comparison performed on the identification envelope.

That comparison rests on standards that deliberately favor acceptance. Senate Bill 503 instructs officials to begin with the presumption that the signature is the voter’s own, to accept similar characteristics rather than an exact match, and to reject only when two officials determine beyond a reasonable doubt that the signature differs in multiple, significant, and obvious respects. No witness attestation is required. No photograph or other documentary identification is demanded at the point of return. If a question arises, the cure process allows the voter or a third party to submit a replacement signature by mail, email, fax, or other remote means, and some cure signatures may update the voter’s record for future elections. Once the envelope is accepted, the ballot is separated from it to protect secrecy. From that moment forward, any error or impropriety in the acceptance decision cannot be corrected without destroying the secret-ballot guarantee.

One might suppose that signature review still provides meaningful protection. Yet the process verifies only that a mark resembling a signature on file was presented; it cannot establish who physically held the ballot, who marked it, or who placed it inside the envelope. In high-turnover apartment buildings, ballots addressed to former residents continue to arrive long after those residents have moved or died. Anyone with access to the mailroom can retrieve them. The same administrative culture that has produced documented large-scale fraud in California’s public-benefits programs now oversees this ballot stream. Federal prosecutors reported millions stolen from California EBT beneficiaries between June 2022 and February 2024. When the same agencies apply subjective, high-volume verification standards to both benefit claims and ballot envelopes, the transfer of exploitative techniques from one domain to the other is not a speculative leap but a predictable adaptation to weak controls.

Close contests make the exposure concrete. California’s Thirteenth Congressional District was decided by 564 votes in 2022 and by fewer than 200 votes in 2024. City council, school board, and special-district races are routinely settled by margins measured in the low hundreds or low thousands. In such settings even modest collection of unclaimed ballots from a single building or neighborhood can shift the outcome. California law preempts local jurisdictions from requiring identification at ballot-submission sites, removing a safeguard many communities might otherwise impose. Recent statutes have further narrowed observer access to address verification on mail envelopes and restricted certain law-enforcement review of voter lists and voting technology absent a court order or specified fraud investigation. The result is a system in which prevention is structurally subordinated to volume and in which after-the-fact prosecution is hampered by the very secrecy rules that follow acceptance.

Prosecution records confirm the enforcement gap. The Heritage Foundation’s Election Fraud Database contains multiple California cases involving false registrations and ineligible voting, yet the visible state and local response remains episodic and reaction to self-reported incidents designed to expose the weakness of the system’s election integrity. The Costa Mesa prosecution of a woman who registered her dog to vote and cast mail ballots in the animal’s name only surfaced because the conduct was self-reported on social media. Federal authorities have brought additional cases, including allegations of payments for registrations targeting vulnerable populations. These prosecutions are valuable, but they depend on self-reporting, whistleblowers, or federal intervention. Once a ballot has been accepted and separated from its envelope, the system possesses zero practical capacity to identify and subtract it. A republic that treats prevention as secondary to convenience has inverted the proper order of safeguards.

James Madison warned in Federalist No. 10 that no man should be permitted to judge in his own cause, because interest biases judgment and may corrupt integrity. A registration system that relies on self-attestation of citizenship, a signature process that presumes validity and sets a high bar for rejection, and a counting timeline that permits results to shift over weeks place the integrity of the count in the hands of the very actors whose interests may diverge from honest tabulation. Ronald Reagan stated the complementary principle: the right to vote is the foundation of democracy and must be exercised freely and fairly, without fraud or intimidation. When millions of live ballots move outside official custody and the primary backstop is a permissive envelope review, the system fails Reagan’s test even if outright fraud remains difficult to quantify after the fact.

Federal taxpayers possess both authority and obligation to insist on minimum standards. Election infrastructure is designated critical infrastructure. Various federal programs supply funding and technical support to state and local election offices. When a state maintains procedures whose design features predictably generate large pools of unverifiable ballots and actively preempts additional local safeguards, it externalizes costs onto the rest of the country in the form of contested outcomes and diminished confidence. The principle is simple: federal election-security support should be conditioned on objective, auditable controls that demonstrate every counted ballot was cast by an eligible citizen. Documentary proof of citizenship for registration, rigorous list maintenance with timely removal of ineligible entries, transparent chain-of-custody reporting, and meaningful identity verification before acceptance are not partisan demands; they are the minimum requirements of a republican form of government.

The SAVE Act, which passed the House in 2025, would establish precisely such a floor by requiring documentary proof of citizenship and amending registration procedures accordingly. The Senate should advance this legislation without delay. The structural features of California’s system, the prolonged counting window that allows results to change in slow motion, and the specific dynamics observed in the Los Angeles mayoral primary supply concrete reasons for immediate action. A state that mails millions of ballots to addresses that may be stale, verifies them through a presumption-laden signature process, separates them from identifying envelopes, and then resists federal insistence on basic citizenship documentation has forfeited any presumptive claim to continued federal election-security assistance.

If California wishes to retain federal support, it must demonstrate that its procedures can reliably distinguish legitimate ballots from those that are misdirected, harvested, or cast by ineligible actors. Until it does so, the Department of Homeland Security should withhold election-related funding and technical support. The alternative is to continue subsidizing a system whose attack surface is large by design and whose detection mechanisms are deliberately limited. That choice is not compelled by any constitutional requirement to maximize convenience; it is a policy decision that the rest of the country is entitled to decline to underwrite.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly. Data in sponsored partnership with Polymarket.

There is an old principle, which economists call revealed preference, holding that we discover what people truly value not from what they announce but from what they do. A man may say he prefers the symphony, but if every free evening finds him at the ballpark, we have learned something his words concealed. Institutions reveal themselves the same way. A league may praise its brightest star, sell her jersey, and run her highlights, and yet, if we watch what it does when she is knocked to the floor and when its honors are handed out, we may find that its real preference runs the other way. The case of Caitlin Clark and the WNBA is, at bottom, a study in revealed preference. The words say one thing. The conduct says another, and the conduct is what counts.

Let us begin with the value, because the value is not in dispute. Ryan Brewer, a finance professor at Indiana University Columbus who specializes in valuation, calculated that Clark alone accounted for more than 26.5% of all WNBA economic activity in the 2024 season, a figure spanning attendance, television, and merchandise. Of the 24 WNBA broadcasts that drew at least 1 million viewers that year, 21 featured Clark. Her games averaged roughly 1.2 million viewers, about 200% more than games she did not play. Indiana set a single-season league attendance record, opponents relocated home games into larger NBA arenas to hold the crowds she summoned, and league merchandise sales surged above 600%, with Clark atop the jersey rankings. In the summer of 2024 the league signed an eleven-year media-rights agreement worth roughly $2.2 billion, more than triple its prior deal of about $50 million per year, a windfall that arrived on the growth she ignited. The independent economist Victor Matheson of Holy Cross estimated that roughly 1 in every 6 tickets sold leaguewide, home and away, owed to the Clark effect.

Hold that number in mind, more than a quarter of the enterprise resting on one young woman, and ask the question a child could ask. Why would any institution treat the source of a quarter of its revenue as an inconvenience? On the ordinary assumption that a business wishes to prosper, the behavior we are about to examine is not merely ungracious. It is unintelligible. And when conduct becomes unintelligible on the assumption of self-interest, the rational move is to question the assumption: perhaps self-interest is not the motive at all.

Consider first the matter of physical safety, because it is the gravest. In June of 2024 the Chicago player Chennedy Carter lowered her shoulder and drove Clark to the hardwood, away from the ball, after the whistle, on a play that had nothing to do with the game in progress. The officials called a common foul in real time and declined to review it. Only the next day, when the cameras had made indifference impossible, did the league upgrade the call to a flagrant, and even then it imposed no fine and no suspension. Sports Illustrated’s women’s basketball coverage counted four Chicago flagrants against Clark in 2024 alone, a remarkable concentration of violence on one player. Then came the playoffs, and the pattern held. Less than two minutes into her first postseason game, Connecticut’s DiJonai Carrington caught Clark squarely in the eye. No foul was called. Clark, who later developed a visible bruise, played half-blind through a blowout loss. In 2025 an eye poke and a hard shove to the floor produced, once again, a foul upgraded after the fact with no suspension, while Clark herself collected a technical for her trouble.

I dwell on these episodes not to claim each was a deliberate attempt to injure, which is more than the evidence will bear, but because of the response, which is the part the league controls completely. The contact that goes uncalled when Clark is the victim would draw an immediate whistle for nearly anyone else, and the league’s enforcement, when it comes at all, arrives a day late and a punishment short. Common foul now, flagrant tomorrow, no real consequence ever. That is not the conduct of an institution guarding its most valuable asset. It is the conduct of an institution that has quietly decided the asset may guard itself.

The point sharpens when we set this permissiveness beside the league’s eagerness to discipline Clark’s own behavior. This season she received her fifth technical foul, and the offense, almost too absurd to credit, was clapping. Not screaming at an official, not hurling anything, not charging anyone, but clapping toward an opposing bench. With eight technicals triggering an automatic suspension, the arithmetic is its own argument. Clark answered with the dry precision that has become her trademark, observing that the league might as well circle a date on the calendar now for the suspension it evidently intends to manufacture. Asked for an explanation, she was told the technical was for clapping and instigating, and replied that the official must therefore simply dislike competitive basketball. A league that swallows its whistle on a fist and reaches for it on a pair of applauding hands has told us, in the clearest possible terms, where its sympathies do not lie.

Now turn from the floor to the honors, for the snubs rhyme with the officiating. In 2025 Clark drew 1,293,536 All-Star fan votes, a single-season record, and finished third in the vote of the media. Her fellow players ranked her ninth among guards, low enough that several appear to have left her off their ballots entirely. The public beholds a generational talent. A meaningful share of her peers will not concede the point aloud. The broadcaster Dick Vitale, no man’s idea of a radical, named the obvious cause when he called the players’ vote “absolutely PURE JEALOUSY.” A year earlier she had been left off the US Olympic roster, in part, the reporting suggested, out of unease over how her enormous following would react to seeing her benched, which is to say she was too popular to be permitted to play. And then, in the summer of 2026, came the perfect emblem of the whole affair. The league released a $29.99 poster celebrating thirty years of its history, advertised as a tribute to the stars who defined and shaped the game, and it represented the Indiana Fever not with Clark but with Sophie Cunningham. The engine of the league’s modern boom did not appear on the league’s own monument to itself.

What explanation fits all of this at once? The conservative writers at Outkick have not been shy, and their bluntness is a virtue here. Clay Travis compressed the entire thesis into a sentence when he said the league got a golden goose and is doing everything it can to kill it. The explanation that fits is institutional resentment, and we should be candid about the shape that resentment appears to take. The WNBA is a predominantly black league in which openly lesbian players are represented far above their share of the general population, and into it walked a straight white woman from Iowa who became, within weeks, its biggest draw and the face of an audience that had never before belonged to it. The most economical reading of the pattern is that Clark is being treated as she is not in spite of being an outsider but because she is one, that a newcomer who made everyone wealthier is resented precisely for arriving from outside the tribe, and that the resentment surfaces in the elbows that go uncalled, the ninth-place ballots, and the poster that found no room for her face.

A reader of fair mind will pause here and object. Is this not simply hard-nosed defense, the ordinary hazing every rookie endures, dressed up in the language of grievance? The objection deserves an answer, and the answer is a parallel that Americans already know by heart, one that cannot be unseen once named. When black athletes integrated white professional baseball in the 1940s, they were spiked, thrown at, and abused, while officials and league offices looked studiously away, and they were expected to absorb every blow in silence as the toll exacted for their presence. The direction of our case is reversed, and that reversal is exactly why so many are reluctant to name it, but the structure is identical. A player marked as the outsider is subjected to physical punishment the rules plainly forbid, while the institution that profits from her declines to enforce the protections it wrote. One need not prove the private heart of any single opponent to accept the more modest and still devastating claim, that a league which would have built a fortress around a different kind of superstar has built nothing around this one, and that the difference is too conspicuous to be waved away as the rough courtesy of the game.

And here the argument closes on ground that requires no speculation about anyone’s soul. A sports league owes its athletes a duty of care. It is obliged to protect them from dangerous contact away from the ball, and that obligation does not lapse because a particular star has become inconvenient to a referee, a locker room, or a front office. Place that duty beside the balance sheet and the indictment stands on two independent legs. Morally, the league fails to shield a young woman from blows it could prevent. Commercially, it degrades and underhonors the player who supplies better than a quarter of its value, which is not the behavior of an enterprise pursuing profit but of one nursing a grudge. Her own coach, Stephanie White, watching a fist pressed into Clark’s neck, abandoned the usual diplomacy and said only that it was crazy, that it was dangerous, that the cheap shots were unacceptable. When the people nearest the team are reduced to narrating their own star’s peril to reporters because the officials will not act, the institution has stopped pretending.

The WNBA was handed the thing it had awaited for three decades, a player who could make the country care, and the country did. The arenas filled, the networks bid against one another, and the money came in torrents. At the precise moment the league should have been raising walls around the source of all of it, it kept reaching, instead, for the knife. Revealed preference is a stern teacher. Watch what the league does, not what it says, and you will know what it has decided Caitlin Clark is worth. So you may be asking what the real issue is and the answer to me is clear - the league is struggling with the fact that it took a straight white woman to make it relevant.

UPDATE: The Phoenix Mercury deleted the post mocking the attack suffered by Indiana Fever guard Caitlin Clark at the hands of their player Alyssa Thomas.

UPDATE: The WNBA announced it had belatedly penalized Alyssa Thomas for attacking Caitlin Clark at last night's game.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly. Data in sponsored partnership with Polymarket.

When I set out to take the measure of the two leading candidates for chair of the Dallas County Republican Party, I did the simplest thing a curious voter can do. I picked up the phone. I called Tami Brown-Rodriguez, and she called me right back. We agreed to have lunch so that I could get to know her, and when the day came she brought her husband and sat with me for more than an hour. We talked about her plans for the county, her plans for the party, and the unglamorous mechanics of running a volunteer organization that lives or dies on relationships. I came away understanding her, which is the entire point of asking.

I tried to do the same with Monty Montañez. I reached out for a meeting or a call, and he was too busy. After my lunch with Tami, I invited him to join Chris Salcedo and me on our weekly livestream, The Third Rail, either during Tami’s interview or after it, so that our viewers could hear his story in his own words. He declined. When I saw that he was at the Texas GOP Convention and reached out again, he never responded. I tried once more after returning home, offering yet another chance to come on the program, and again heard nothing. Only after I had made my endorsement did I receive a text, and it offered a meeting after the election, with a request that I route future contact through someone named Ajua Mason rather than reach him directly (I reached out to Ajua but have not heard back). There is a small lesson in that sequence, and it is not a personal one. It is a lesson about how a person treats the work of persuasion when the work is inconvenient.

Let me state my thesis plainly, because the rest of this piece is simply its defense. A county party chair is not a candidate for office in the ordinary sense. The chair is the steward of a volunteer army, and the qualification that matters most is not eloquence on a stage but trust earned in the trenches. Tami Brown-Rodriguez has spent years earning that trust. Monty Montañez, whatever his virtues, has not yet begun. On Thursday, Dallas County Republicans should choose the person who has done the work.

Consider first what the job actually is. The Dallas County Republican Party is 100% a volunteer organization. Nothing it accomplishes, not a single block walked, poll watched, or election worked, happens because the chair issues an order. It happens because the chair has friends, and because those friends trust that the chair knows what she is doing and will stand by them when the day is long and the pay is nothing. This is why a serious party operative starts at the bottom. You volunteer at elections. You become a precinct chair. You attend conventions. You volunteer for Republican candidates and you donate to them. You build, over years, the web of obligation and goodwill that lets a chair pick up the phone and get a hundred people to show up on a Saturday. There is no shortcut, and there is no substitute.

By that standard, the contrast between the two candidates is not close. Tami served as Vice Chair under Allen West and remained loyal to him through his entire term, a loyalty that matters because loyalty under pressure is the truest test of character in politics. When West selected her as his Vice Chair, the County Executive Committee confirmed her unanimously, which is to say the people who know the work best, and who are notoriously hard to please, looked at her and said yes without dissent. She has been a precinct chair for six years. She chaired the Training Committee for two years and has personally trained more than 3,000 precinct chairs and over 1,000 election judges, clerks, and poll watchers. She has poll watched at Central Count since 2017, served as an election judge in 2020, and led the 2024 Rapid Response Election Hotline. She ran a separate primary runoff in 2026 with record Republican turnout. She has served on the Bylaws Review Committee and the election integrity efforts that conservatives rightly care about. She has worked on more than 40 local campaigns and attended three Texas state conventions. This is not a resume assembled for a press release. It is a record of presence, the kind of presence that accumulates only when a person shows up, again and again, for years.

Share

Set beside that record, Monty Montañez’s involvement with the Dallas County party is difficult to locate. By the most reliable data I have found, and the one I have used for multiple Dallas County campaigns, he volunteered to be an election judge twice, and on one of those occasions, he cancelled at the last minute. There is no record of his serving as a precinct chair, a member of the Senior Executive Committee, a CEC officer, a committee chair, a volunteer coordinator, a poll watcher, or an election judge besides that one time. There is no evidence of his attending Dallas County Republican Party meetings before 2025. A puzzled reader might ask whether this is merely the absence of a paper trail rather than the absence of involvement. It is a fair question, so consider the surrounding facts. The moment Montañez arrived in Dallas, he ran for Congress. The first election he ever voted in here in Texas was his own primary. Before that, he voted in Miami-Dade County from 2016 to 2024, where he was identified as an independent and cast a ballot in only one primary. The Republican chair of Miami-Dade has no recollection of his involvement in the party there either. The pattern is not local to Dallas. It is a pattern of arriving and seeking the top of the ladder without having climbed it.

The public records sharpen the point. According to L2 Data, the most reliable voter file I have worked with, and one I trust enough that I checked my own entry against it and found it accurate, Tamara Brown is registered as a Republican, registered in 2008, with a vote frequency near the top of the scale. Montañez is listed as Democratic, registered in Texas in October 2025, with a vote frequency of zero. Residency tells a similar story. Tami has lived in Dallas County for 26 years and owns three properties here. Montañez has rented a house in Dallas County for one year. These are not insults. They are the plain facts of who has roots in the community a chair is asked to lead, and who has just arrived.

There is one more duty of a chair that deserves its own paragraph, because it is the duty most easily measured and most often decisive. A chair must raise money. The party cannot function on goodwill alone, and the person who leads it must be able to ask for resources and receive them. Here the record is not a matter of interpretation. During his own congressional race, Montañez raised $39,224, of which $22,286 was self-funded, and he finished seventh. To grasp what that figure means, set it next to his primary opponents. Ryan Binkley raised $1.9 million. Paul Bondar raised $1.9 million. Jace Yarbrough raised $424,554, and Jace won. A candidate who could persuade donors to part with only $16,938 is telling us something important about his ability to perform the single most quantifiable task the chairmanship requires. Tami, by contrast, has spent a lifetime raising money, more than $7 million for nonprofits over her career and over $300,000 for her current organization since January 2026 alone, and she arrives with a network of elected officials, candidates, and grassroots leaders across Texas. Fundraising is not a vanity metric. It is the oxygen of a county party, and one candidate has shown she can supply it.

I did not rely only on records and my own impressions. I asked people who know the terrain. I spoke with Doug Deason, a major Republican donor, who said Montañez was a nice guy but admitted he was surprised to see him running for chair after running for Congress in the primary. When I told him Montañez would be running against Tami, Deason said immediately that Tami would be awesome, a much better pick for the chair. I spoke with Chris Putnam at the Dallas Express, who had genuine concerns. Putnam was troubled that Montañez had not been involved with the county party in Dallas or in Miami-Dade, and that he had arrived and gone straight for a congressional seat. These are not partisans for Tami. They are serious people offering candid assessments, and the assessments pointed the same direction.

I want to be fair, and fairness requires that I say this clearly. Monty Montañez may be a perfectly good man. He is, by all accounts, an excellent public speaker, and ambition is no sin in a republic. But ambition is not the same as readiness, and a good speech is not the same as a built relationship. At the county level, half the job is simply knowing the people in the county party, and the other half is having earned the right to ask them for help. I would be the first to admit that I am not even slightly qualified to be chair, and I say that as someone who has been more involved in the Dallas party than Montañez has. That is not a knock on him. It is an honest accounting of what the position demands and what it does not forgive.

So I will end where I began, with the phone call. When I reached out, one candidate made time, brought her husband, and talked with me for an hour about the future of the party she has served for years. The other was too busy until the election was nearly over. Character in this work is revealed not in the grand gesture but in the willingness to show up when showing up is inconvenient. Tami Brown-Rodriguez has been showing up for years, and Dallas County Republicans would be wise to elect someone who has been in the arena rather than someone who has only just walked through the gate. On Thursday, I am proud to endorse Tami Brown-Rodriguez for chair of the Dallas County Republican Party, and I urge you to join me. If you want to meet her you can tonight at 7PM.

If you enjoy my work, please subscribe https://𝕏.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.

Begin with a belief almost everyone holds. When more people want a thing, its price goes up. We learn this lesson early and we learn it well, because it is true of nearly everything we buy. More shoppers bidding for the same tomatoes, and the price of tomatoes climbs. More families competing for the same house, and the sale price rises. The intuition is so dependable that we apply it without a second thought, and most of the time we are right to.

Electricity is the exception, and the exception matters enormously right now. The public has been told that artificial intelligence data centers, those vast humming buildings packed with servers, are the reason household power bills keep climbing. The story is intuitive, which is precisely why it deserves a second look. The best causal evidence we now have says the intuition is not merely incomplete in this case. It is backwards.

To see why, picture a toll bridge. The bridge cost a fixed sum to build, and it costs roughly the same to maintain whether 100 cars cross it each day or 100,000 do. If only a handful of drivers use it, each one must pay a steep toll to cover the bridge. If thousands use it, the same fixed cost is spread across far more drivers, and the toll per car falls. The cost of the bridge did not change. The number of payers did. A power grid works the same way. The wires, the substations, the transmission towers, and the generating plants represent an enormous fixed cost that must be recovered no matter how much electricity flows over them. Your monthly bill is mostly a share of that fixed cost, not a charge for the next kilowatt-hour produced.

This is the hinge of the whole argument, so it is worth slowing down. In an ordinary commodity market, the price you pay tracks the cost of producing one more unit. In electricity, retail rates track average cost, because the system is a regulated network dominated by fixed investment. The gap between the two is large. For the period from 2014 to 2016, the average wholesale price of power in the Northeast ran around 4 cents per kilowatt-hour, while the average retail rate sat above 10 cents and the national figure near 12 cents. Marginal cost, in other words, was roughly half of average cost. When the extra cost of serving new demand is well below the average cost baked into everyone’s bill, adding demand mechanically pulls the average down. A steady, around-the-clock customer is the toll bridge filling up with cars.

Data centers are that customer. They run near continuously, at utilization rates of 80% to 90%, while the American grid as a whole runs at an average load factor of only about 53%, according to a Duke University analysis. The grid is an airliner that flies nearly empty most of the year. Filling those empty seats with steady load is exactly what lowers the cost per passenger.

Now to the evidence, because an elegant theory is worth little if the numbers refuse to cooperate. A

2026 study by Asa Watten, John Bistline, and Geoffrey Blanford

, published on arXiv, set out to measure the causal effect of data center growth on retail prices. The authors used an instrumental-variables design built on the 1947 Eisenhower interstate highway plan, since fiber-optic corridors tend to follow old highways, which lets them isolate data center siting from the prices that might otherwise have attracted developers. Their finding is plain. Data centers caused average retail electricity rates to fall modestly from 2015 to 2024. Every 10% increase in data center capacity lowered average residential prices by about 0.4%. Translated into lived experience, the typical residential customer lived in a state where data center capacity grew 160% over 2019 to 2024, and that growth cut their rates by 6%. The preferred specification implies that a doubling of capacity lowers residential prices by 3.5%. The statistical test the authors rely on, which is robust to weak instruments, rejects any positive price effect at the 0.1% level.

Notice who produced this result. Two of the three authors work at the Electric Power Research Institute, the electricity industry’s own nonprofit research arm, not an advocacy shop with a thumb on the scale. And they are not alone. Charles River Associates concluded in February 2026 that the data center buildout did not trigger retail rate increases over the past decade. Researchers at Lawrence Berkeley National Laboratory, a federal lab, working with the Brattle Group, found that the states with the greatest load growth saw real retail prices fall over the past five years. When a federal laboratory and the utilities’ own consultants concede the point, the debate is effectively over.

The geography seals it. Virginia is the data center capital of the world, with these facilities consuming more than 20% of the state’s electricity, yet Virginia’s residential rates run roughly 9% below the national average. North Dakota saw electricity demand grow about 35% over five years while its real price fell. If data centers raised bills, these two states should be cautionary tales. They are advertisements.

So why does nearly everyone believe the opposite? Here we reach the heart of the matter, and it is a lesson in human nature as much as in economics. When a bill rises, people look for a cause they can see. Data centers are large, new, and conveniently unsympathetic, so they become the answer to a question they did not cause. The anti-AI activist blames them because he dislikes the technology. The anti-capitalist blames them because he dislikes the firms that own them. The politician blames them because blaming a faceless tech company is easier than explaining his own energy policy. And the utility, most revealingly, blames them because the real story implicates the utility itself.

Share

Consider Georgia Power. Where rates near a project do climb, the cause is frequently a decision the utility made, not a demand the data center imposed. A data center operator would often prefer to build its own generation on its own site, financed with its own capital, drawing nothing from the public grid. That arrangement would add load without adding a penny to anyone else’s bill. Yet the utility has every incentive to block it, because a self-powered data center is a sale the utility loses. Georgia Power would rather supply that power itself, which means stringing transmission lines across miles of countryside to reach the site. Building those lines can require condemning property and tearing down homes that stood in the path. The disruption is real, and it is expensive, and it lands in the rate base. But it is the consequence of the utility’s choice to monopolize generation, not of the data center’s choice to consume electricity. The operator asked to power itself. It was told no.

Texas tells the same story in a different accent. Utilities there are pressing to build new transmission lines out to the Permian Basin, and the reason is straightforward. They want to sell their power into that market, and selling more power requires more wire. To finance that infrastructure they will raise rates, which means consumers pay more in the near term to fund a buildout that may lower costs much later. None of this was demanded by a data center. The data center, once again, would happily generate on site and stay out of everyone’s way. The lobbyists who insist that only the utility may supply the power are the same parties who then point at the data center when the bill arrives. They authored the cost and assigned the blame to someone else.

Strip away the misdirection and the actual drivers of higher bills come into focus. They are the premature retirement of reliable baseload plants, two decades of grid neglect, renewable mandates that raised costs while adding little firm supply, and, in California, wildfire liabilities that pushed average prices up nearly 40% over a few years. The Heritage Foundation and the Manhattan Institute have argued for years that the answer to surging demand is to expand and modernize supply, not to ration it. Energy Secretary Chris Wright has put it more bluntly, calling data centers “the answer to drive down electricity prices,” not the problem.

He is right, and the conservative prescription follows naturally. Do not ban the most valuable new customer the grid has seen in a generation. Require data centers to pay their own way, which the Georgia regulators and the White House Ratepayer Protection Pledge already demand, and then get out of the way and let them build the generation they are asking to build. Abundance lowered the price of power across the whole of the 20th century, as demand exploded and real prices fell decade after decade. It will do so again, if we let it. The villain in this story was invented because the truth was inconvenient to too many people at once. The data center did not raise your bill. The people blaming it did.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.

Every premiere has a dress rehearsal. The actors run the full script in a smaller room, before a friendlier crowd, so that the timing is right when the curtain goes up for real. On June 13, in Grand Ballroom B at the Republican Party of Texas convention, I watched a dress rehearsal. The premiere is scheduled for July 18, when the same network stages a ‘conservative grassroots’ “Nationwide Day of Protest” against AI data centers in at least 13 locations. The panel and the protest are sold to conservatives as two independent expressions of a genuine conservative grassroots awakening. They are nothing of the kind. They are one production, with one script, one director, and one source of money, and the convention was the dress rehearsal you were allowed to watch.

Start with the question that should stop any thinking conservative cold. Why would money tied to the AI industry, and to the tech fortunes built beside it, pay to organize opposition to AI infrastructure? The premise seems self-refuting. People do not fund their own enemies. The puzzle dissolves the instant you grasp what the fight is actually about. It is not a fight over whether AI gets built. It is a fight over who is allowed to build it, and over who writes the rules that decide. The economist Gabriel Kolko explained this pattern a half century ago. The great incumbents of the railroad and meatpacking eras did not fear regulation. They wrote it, because rules drafted to their specifications crushed the smaller rivals who could not absorb the cost of compliance. Safety becomes the moat. The incumbent does not drain the swamp around his castle, he deepens it, and he names the project public protection.

Read with that lens, the funding trail stops being mysterious and becomes legible. The group behind the July 18 protest is called Humans First. The Washington Post reported that Humans First was incubated by the Center for AI Safety, which supplied seed money as a loan and whose co-founder appears on the California incorporation papers as the new group’s chief executive. The Center for AI Safety is not a conservative institution, and its money is not conservative money. A detailed public accounting puts its funding near $23.3 million, of which more than $21 million came from a small circle of effective-altruism donors. Open Philanthropy, financed by Facebook co-founder Dustin Moskovitz, supplied roughly $12.49 million. Jaan Tallinn, an early Skype engineer who also sits on the Center’s board, routed about $7.5 million through his Survival and Flourishing Fund. Sam Bankman-Fried’s FTX Future Fund put in $6.5 million before the fraud collapsed, after which the bankruptcy estate clawed back some $5.2 million of it. Even OpenAI is in the documents, with a $333,333 grant on its 2023 tax filing, and its chief executive among the signatories of the Center’s headline extinction statement.

The staffing confirms what the money implies. The movement strategist for Humans First, Jeremy Ornstein, helped build the Sunrise Movement, and in 2018 he was among the activists who occupied Speaker Nancy Pelosi’s office alongside a newly elected Alexandria Ocasio-Cortez. His coalition counterpart, Alexander McCoy, is a self-described member of the Democratic Socialists of America who co-founded a veterans group built to impeach President Trump and who, as the Washington Post separately confirmed, campaigned for Vice President Harris in 2024. These are not incidental hires. They are the people who write the script that conservatives are then recruited to perform.

Now watch the same company take the convention stage. The host was the Alliance for Secure AI, run by Brendan Steinhauser, whose conservative resume is authentic and whose usefulness depends on exactly that. The tell sits in one line of his own biography, where he serves on the board of the AI Policy Institute, one of the policy shops the trade press places squarely inside the effective-altruism orbit. The panel’s moderator, Robert Arnakis, is the Alliance’s own Director of External Affairs, which means the supposedly neutral host and the supposedly neutral moderator are a single organization. Dr. Vael Gates, introduced from the stage as arriving from Berkeley, is no Texas grassroots activist at all. She is a career effective-altruism AI-safety researcher whose own talk materials list Open Philanthropy among her funders, and she delivered the verbatim catechism of that movement, warning that frontier labs are building a “second intelligent species” that humanity may not survive. The faith segment came from a Future of Life Institute speaker, an organization founded by Max Tegmark and the same Jaan Tallinn whose money already sits in the Center’s ledgers. The youth-jobs testimony came from the Young People’s Alliance, a group whose founders, by the cataloguing of InfluenceWatch, began by searching for ways to oppose Republican initiatives in the North Carolina legislature, now rebranded as a bipartisan generational voice.

Then there is Amy Kremer, the recurring face bolted onto the front of the apparatus. She told the room she had never advocated for more regulation in her life and found herself doing so now, a confession that should have made the audience sit up rather than nod. Kremer co-founded Tea Party Patriots in 2009 and was ousted within roughly a year, after which her own co-founders sued her, and a judge found her in contempt, remarking that she was straddling a very fine line and had fallen over it. The woman now warning conservatives that elites are capturing their movement was once accused, by the movement’s own founders, of capturing theirs. She does not stand still, as the moderator said admiringly, and she has not stood still here. She has simply found far-left benefactors with deeper pockets than the Tea Party ever had.

Senator Angela Paxton presents the harder and more important case, precisely because she is a genuine elected Republican rather than a rented face. That is what makes her performance so valuable to the network and so costly to the rest of us. Her central message was opposition to federal preemption of state AI law, delivered in the unimpeachable language of the Tenth Amendment. The argument collapses the moment you ask what it produces. If every state writes its own AI code, the country does not get 50 laboratories of liberty, it gets 50 incompatible compliance regimes, which is the single most effective way to strangle an emerging industry. A patchwork is a moat. Only the largest incumbents can afford 50 sets of lawyers, and the open-source challengers and Texas startups cannot, which is exactly why the safety lobby wants the patchwork and dresses it as states’ rights. Worse, this is a national-security question wearing a federalism costume. The US is in a race with China to set the terms of the most consequential technology of the century, and a thicket of conflicting state rules is the surest way to lose it. Senator Paxton, whatever her sincerity, stood on a Republican stage and delivered the precise policy outcome that a Berkeley-funded lobby has spent millions to engineer. That is not a smear. It is a description of what happened.

The factual claims fared no better than the constitutional ones. Heather Fazio warned of forever chemicals and heavy metals in data-center wastewater, then conceded under questioning that closed-loop cooling recirculates its water, which is the design most large facilities now use. Kremer volunteered a “second water footprint” for power generation as if she had uncovered a conspiracy, when steam-cycle cooling is ordinary engineering that applies to every thermal plant on the grid. The water numbers offered from the floor ranged from 50 billion to 160 billion gallons, a spread so wide it is not a measurement but a mood, and a tripling of one’s own estimate is the signature of fear, uncertainty, and doubt rather than fact. The cost-shifting alarm ignores that Texas already passed Senate Bill 6 to reallocate large-load infrastructure costs, and that hyperscale developers routinely fund their own interconnection. Every grievance, examined, turned out to be either already addressed, easily mitigated by existing technology, or simply asserted.

I want to be precise, because precision is what makes the charge stick. Not everyone on that stage is a witting operative, and saying so would be both false and lazy. The point is structural. The convening machinery and the issue framing belong to the effective-altruism AI-safety network, and sincere conservatives with real anxieties about water, jobs, and their children are being incorporated as validators, their credibility lent to a script written in California and funded through a Facebook fortune, a Skype fortune, and the residue of a crypto fraud. The tell is the uniformity. The same preemption line, the same chatbot tragedies, the same word-for-word hedge that the speaker uses AI daily and wants only common-sense guardrails, recurs across groups that claim no connection to one another. Independent movements do not generate matching messaging architecture. Coordinated campaigns do. The moderator even named the model aloud at the close, calling it the Bernie-to-Bannon coalition, which is an honest description of an operation that runs on Bernie’s money and Bannon’s audience.

The dress rehearsal is over. The premiere is July 18. You have now seen the cast, read the playbill, and traced the financing, and you are entitled to decide whether the anger you feel about your water bill and your children’s future should be spent advancing the regulatory agenda of the very tech billionaires you were told to fear. The grievances are real. That is what makes the operation work. But the hands on the marquee have not changed, and they are not your hands.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.

The formation of citizens capable of self-government stands among a republic’s most consequential responsibilities. Education does not merely transmit skills. It cultivates the habits of mind, the understanding of history, and the loyalty to constitutional principles that allow a free people to remain free. When agents of an Islamic power gain authority to shape that formation, to preview lesson plans, to observe classrooms, to coach teachers, and to supply the materials through which young minds first encounter the wider world, the republic has placed part of its future in hands not its own. Qatar Foundation International, the US entity wholly funded and directed by Qatar’s state-sponsored Qatar Foundation, executed precisely this kind of operation. It distributed at least sixty-five million dollars across more than two hundred programs nationwide, reaching over one hundred thousand students in documented K-12 initiatives. In Texas Qatar spend millions to indoctrinate over 10,000 students and hundreds of teachers, concentrated in Austin and Houston independent school districts. The arrangements created an unaccountable parallel Islamic authority inside public schools. They substituted Qatari narratives for American ones. They exposed children to materials containing antisemitic conspiracy theories and religious content aligned with Islam and Sharia Law. A democratic republic has a fundamental duty to prevent such capture.

Consider first what occurred in Houston. At the Arabic Immersion Magnet School, a public pre-K through eighth-grade Islamic institution, core subjects including history, art, and physical education were delivered in Arabic under

grant agreements

signed by former Superintendent Terry B. Grier. Those agreements required the district to grant QFI-designated observers unfettered access to classrooms on through each school year. Qatari monitors received lesson plans in advance, conducted monitoring visits, and issued post-observation critiques that triggered one-on-one coaching for teachers whose instruction did not reflect the Islamic values being promoted by Qatar. Teachers were obligated to attend QFI-sponsored summer institutes and to participate in professional development that included not just Islamic and Sharia content but Diversity, Inclusion, and Equity (DEI) trainging. Both teachers and students signed publicity waivers allowing the foreign government to use classroom images for promotional purposes. QFI also commissioned studies that collected student-level data on race, religion, gender, free or reduced lunch status, and academic proficiency. Parallel terms applied in Austin Independent School District, where hundreds of thousands of dollars supported Arabic instructors and materials at Austin High School, Burnet Middle School, and International High School. When public records requests finally produced the files, Austin officials stonewalled the requests and demanded fifty thousand dollars on translation services before releasing Arabic-language content to scrutiny. Only intervention by the Texas Attorney General secured the documents without that expenditure. These mechanisms did not merely supplement local resources. They created ongoing dependency and reduced local oversight of what entered American classrooms.

The content delivered under these terms makes the structural problem even more acute. Materials included a textbook published by the Kingdom of Saudi Arabia Ministry of Higher Education and used in Texas programs instead of state-approved curriculum. Its introduction states that a man’s ignorance of Arabs and Arabic may lead him to hate Arabs, avoid them, or stand in the ranks of their enemies, and that this is precisely what occurs when Westerners whose ignorance of Arabs allowed Zionists to manipulate them. Classroom maps labeled the State of Israel as Palestine and omitted it entirely. Students received assignments directing them to create greeting cards decorated with mosque imagery, to praise the Muslim Prophet Mohammed, and to attach cutouts of the Kaaba. Vocabulary and lessons were explicitly tied to Qatari values, Sharia Law, and the Arab and Islamic culture, introducing terms such as Allah and emir at early grade levels. QFI’s broader Al Masdar curriculum project distributed lessons such as “Express Your Loyalty to Qatar” and scenario-based prompts titled “Whose ‘Terrorism’?” that framed Israeli military actions in ways inviting students to equate self-defense with terrorism while providing no comparable framing for designated terrorist groups. A workbook developed under a U.S. Department of Education StarTalk program included the poem “Identity Card” by Mahmoud Darwish, a senior official of the Palestine Liberation Organization, without disclosing that organizational history of terror attacks or the PLO’s record of targeting civilians. The Foundation for Defense of Democracies examined the grant contracts, observation emails, and physical materials from Austin and Houston and concluded that kids were being exposed to lesson plans and textbooks stained with Islamic and Sharia propaganda and antisemitism. The Anti-Defamation League has catalogued the persistent antisemitic and conspiratorial content in Saudi educational materials of the exact type later imported into U.S. classrooms through these channels.

This model raises questions that extend beyond any single district or any single language. Some may argue that Arabic instruction serves American interests in commerce and diplomacy and that cultural exposure broadens young minds. The argument however collapses once the distinction between domestic control of foreign-language pedagogy and the cession of curricular authority, classroom access, and teacher oversight to representatives of a foreign state is recognized. Ordinary language programs do not grant foreign government employees recurring rights to enter taxpayer-funded classrooms, to correct American teachers in real time, to require advance approval of lesson plans, or to use curriculum and textbooks produced by Islamic states. They do not tie continued funding to the adoption of specific political and religious narratives developed by the donor government. They do not collect detailed demographic and performance data on American students for analysis by that government. Qatar’s documented support for Hamas, CAIR, and Muslim Brotherhood makes any such access and data collection a national security concern. Wealthy districts such as Houston ISD, operating with billion-dollar budgets, possessed no economic need to outsource teacher salaries, observer rights, and imported textbooks to a foreign government. Domestic alternatives exist and should have been prioritized. Most fundamentally, a democratic republic cannot delegate the moral and civic formation of its young to a state whose governing ideology integrates religious and political authority in ways that conflict with the American constitutional order and western values. The moral violation is direct. Elementary and middle-school children encountered antisemitic ideation that blames Zionists for manipulating Western opinion and religious assignments that function as proselytizing within public school hours. American children in public schools deserve protection from foreign propaganda that traffics in such content, regardless of the language being taught.

The historical parallel to another foreign influence effort clarifies what sustained scrutiny can achieve. Beginning in the mid-2000s, the Chinese government’s Hanban established a network of Confucius Institutes on U.S. university campuses and some K-12 settings. Those programs delivered state-approved narratives, avoided discussion of topics sensitive to Beijing such as Taiwan, Tibet, Xinjiang, and human rights, and embedded personnel with access to classrooms and academic programming. Concerns over academic freedom and documented ties to Chinese intelligence prompted bipartisan legislation in the 2018 National Defense Authorization Act that restricted federal funding to institutions hosting the institutes. Most were closed by the early 2020s following FBI warnings and congressional pressure. The QFI model replicates the core mechanism of foreign government funding of language and culture programs that deliver state-approved narratives and embed personnel with classroom access. It adds the aggravating factors of explicit antisemitic textual content and contractual ties to a state that hosts Hamas leadership. As the Heritage Foundation observed in its December 2025 analysis, QFI’s educational programs function as a Trojan horse to introduce anti-Western and pro-Islamist perspectives to American children. The same policy response applied to Chinese influence operations must now apply to Qatari influence operations. Sam Herrera, a Houston activist quoted in the Wall Street Journal in 2017, captured the pattern when he noted that they hide under school districts wantonly taking the money and are not going to overtly come out and tell you what they are doing.

Share

Qatar’s announcement that it would wind down the program this month following the Institute for the Study of Global Antisemitism and Policy’s May 2026 report and accompanying calls for federal review changes nothing for the students and teachers already exposed. It leaves unanswered the practical question of who will supply replacement instructors, materials, and professional development once the primary funder withdraws. Another Gulf state with comparable objectives? Domestic organizations advancing similar interpretive frameworks? Or a quiet reversion to local or federal support without the foreign overlay? Each possibility underscores the original error of allowing foreign state direction inside American public education. The Department of Justice possesses both the authority and the duty to investigate the full extent of these arrangements nationwide. That investigation must determine compliance with disclosure statutes, including potential obligations under the Foreign Agents Registration Act, the complete scope of materials distributed, the precise content of teacher training and classroom observations, the nature of any one-on-one student interactions, and whether the programming produced measurable effects on student attitudes or subsequent choices. In Texas, Attorney General Ken Paxton should direct an equally rigorous examination of every contract, every classroom observation report, every proficiency test administered, every communication between Qatari trainers and American educators, and every instance of student data collection in the affected Austin and Houston programs. The curriculum actually delivered must be reconstructed in full. The tests must be analyzed for embedded assumptions. The agreements signed by teachers and students must be reviewed for the scope of consent obtained under conditions of funding dependency. What Qatar required of each teacher and student must be recovered and made public.

A democratic republic has both the authority and the obligation to prevent foreign governments, particularly Islamic states, from purchasing curricular influence over its children. The principle admits no compromise. Education in a republic is never merely skill transmission. It is the deliberate cultivation of a particular form of citizen oriented toward the regime’s fundamental commitments to ordered liberty, equality under law, and the distinction between religious conviction and political authority. When a state whose foundational commitments stand in systematic tension with those principles obtains recurring access to the formation process, the receiving republic has compromised its own continuity. Abraham Lincoln observed that the philosophy of the school room in one generation will be the philosophy of government in the next. That philosophy must remain American in origin and accountable to American parents and voters. The United States must therefore move from investigation to prohibition. No Islamic nation should again be permitted to fund, staff, observe, or direct curriculum, teacher training, or classroom activities within American public elementary and secondary schools. Language acquisition of genuine educational value can and should continue under fully domestic control and transparent standards. Voluntary cultural exchanges may proceed without embedding foreign state personnel inside taxpayer-funded institutions or tying support to external ideological frameworks. The alternative, repeated across more than two hundred initiatives and multiple states, is a slow transfer of authority over the American mind to actors whose interests and worldview diverge from those of the republic they seek to shape. That transfer must stop. The children already affected deserve the protection of full disclosure and accountability. Future generations deserve the assurance that their classrooms will remain under the sovereign control of their own polity. America should never again allow an Islamic country, or any foreign power, to fund and dictate curriculum and to indoctrinate teachers in its public schools.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.

There is a claim buried inside every proposal to tax a great fortune out of existence, and it is rarely stated out loud. The claim is that the people who run the government could put that money to better use than the person who earned it. This is usually dressed as a moral assertion, a matter of fairness, but at its core it is an empirical one. It says, in effect, hand us the capital and we will produce more value with it than its owner ever could. The lovely thing about an empirical claim is that it can be tested. We do not have to settle it in the abstract. We can look at what the state actually builds when it controls billions, set it beside what a private builder actually builds when he controls billions, and then compare the result. We can place the wreckage next to the rocket and let the reader decide.

California has handed us the cleanest test imaginable, and it did so almost by accident. In 1996 the state created the California High-Speed Rail Authority and charged it with linking the state by fast train. Six years later, in 2002, a man named Elon Musk founded a small rocket company called SpaceX. Hold those two dates in mind, because the six-year gap matters. The public agency had a head start. It was older, it was backed by the full faith and taxing power of the largest state economy in the country, and it was promised a river of money that would eventually swell beyond anything the rocket company ever raised. By every measure that bureaucrats use to predict success, the train should have been running long before the rocket reached orbit.

Now look at what each one did with its years. SpaceX raised roughly $12 billion in private capital across its entire history, money that investors handed over voluntarily, knowing they would lose all of it if the company failed. With that sum it built the first orbital-class rocket that could land itself and fly again, landing a booster in December 2015 and reflying a used one in 2017. It carried NASA astronauts to the International Space Station beginning in 2020. It flew a launch cadence that reached 138 orbital flights in a single year. It drove the cost of reaching orbit down by something like 65% to 70%, delivering payload at roughly $3,000 per kilogram against the Space Shuttle era’s $54,500. In June 2026 it went public, raised about $75 billion in one of the largest offerings in history, and saw its market value climb to roughly $2.5 trillion, the sixth most valuable public company on earth. That is what $12 billion of privately allocated capital produced.

Set against this the record of the train. Voters were sold Proposition 1A in 2008 on a specific promise, an 800-mile system carrying passengers between San Francisco and Los Angeles in two hours and forty minutes, for a total cost of $33 billion, with service beginning in 2020. The 2020 deadline passed with nothing to ride. The price estimate climbed past $77 billion, then $80 billion, then $128 billion, and the Authority’s own 2026 business plan now contemplates a figure as high as $231 billion for the full build. The scope, meanwhile, collapsed in the opposite direction, shrinking from a statewide network to a single 171-mile segment between Merced and Bakersfield, with interim service now hoped for somewhere between 2030 and 2033. After roughly three decades and more than $13 billion spent, the number of miles of high-speed track carrying paying passengers stands at zero. The US Senate Commerce Committee put the matter in a headline that needs no commentary, observing that after 25 years and billions in federal subsidies, not a single train is operating in California.

I want to be precise about the arithmetic, because the comparison is even more lopsided than it first appears. The $231 billion the state may spend is 19 times the $12 billion that built SpaceX, which is to say nearly 1,900% more capital. Even the Authority’s preferred, friendlier figure of $126 billion for the first phase is more than ten times the rocket company’s entire lifetime funding. So the empirical claim that animates the confiscation crowd, the claim that the state spends better, fails its own test by a factor of ten or twenty. One side took a small pool of private money and built a fleet of reusable rockets, a space based internet constillation, and the beginnings of a Mars program. The other side took a far larger pool of public money and built viaducts in a valley and a procession of revised business plans. There is no metric, not cost, not schedule, not output, not capability, on which the bureaucracy wins.

A puzzled reader might ask why this should be. Are the engineers at SpaceX simply smarter than the engineers in Sacramento? That is not the explanation, and assuming it is misses the real lesson. The difference is structural, and it concerns accountability. SpaceX answers to three merciless judges, namely investors who can withdraw their money, customers who can buy launches elsewhere, and the laws of physics, which do not negotiate. Each of those judges punishes failure immediately and visibly. When a Falcon rocket exploded, the company found the flaw and fixed it within months, because the people responsible bore the cost of being wrong and reaped the reward of being right. The Rail Authority answers to none of these. It cannot go bankrupt. Its customers cannot defect, because there are no customers. Its managers draw their salaries whether the train runs or not. Thomas Sowell described the mechanism better than I can, observing that it is hard to imagine a more dangerous way of making decisions than by putting them in the hands of people who pay no price for being wrong. That is the Authority in one sentence. It never stops failing precisely because failure costs the deciders nothing.

This is not a new discovery, and it is not unique to trains. Milton Friedman warned that if you put the federal government in charge of the Sahara Desert, within five years there would be a shortage of sand, and the $33 billion train that became a $231 billion non-train is simply that joke rendered in concrete. Ronald Reagan reduced the whole problem to nine words he called the most terrifying in the language, I'm from the government, and I'm here to help. The genius of these lines is that they identify a pattern rather than a personality. The pattern does not depend on which party holds Sacramento or which contractor holds the bid. It depends on whether the person spending the money suffers when the money is wasted.

History sharpens the point, because Americans did not always build this way. In the 1860s two private companies laid 1,776 miles of railroad across deserts and over the Sierra Nevada in roughly six years, armed with black powder, hand tools, and mules, meeting at Promontory Summit in 1869. In 1930 private developers raised the Empire State Building, 102 stories, in just 410 days, finishing ahead of schedule and under budget as the Depression closed in. The country that performed those feats is the same country whose largest state cannot now connect two of its cities by rail across three decades unless, as it happens, a private citizen does the building instead. The clearest parable arrived in 2024, when two NASA astronauts rode Boeing's Starliner to the space station on a mission meant to last eight days, only to be stranded for 286 days when the legacy contractor's capsule proved unsafe to bring them home. It was SpaceX, the upstart, that flew them back in March 2025. When the old incumbent failed and left Americans stranded in orbit, the accountable private firm went and got them. Swap the names and you have the whole argument about who should be trusted with capital.

Which brings us to the confiscation itself, and to why it would be not merely unwise but something closer to a moral inversion. The proposal on offer is to take tens of billions of dollars from the man who built the rockets and hand it to the political class that built the ghost train. Phrase it that way and its absurdity is plain. It is unwise, because it would transfer capital from the highest-returning use yet observed in American industry to demonstrably among the lowest. It is wasteful, because we have watched in real time what this political class does with money it did not earn. It is unfair, because it punishes the steward who delivered and rewards the steward who did not. And it is fundamentally un-American, because the country was founded on the conviction that what a man builds is his, and that the state is a trustee of the public’s money rather than the owner of the public’s labor. Stewardship of other people’s money is a genuine moral duty, and the people now eyeing Elon Musk’s fortune are precisely the people who violated that duty when they extracted billions from California taxpayers under a specific promise and delivered them nothing. A class that cannot be trusted with $13 billion it already had should not be handed $13 billion more, still less the far larger sum it covets from 𝕏’s owner.

The deepest reason to oppose the seizure is therefore not that Musk is admirable, though by the measure of what he has produced he plainly is. The reason is that the comparison answers the only question that matters. The confiscators say they can spend the money better. The evidence says they spend it on PowerPoint and pour the rest into the dirt. A man will very likely set foot on Mars before a Californian can board a high-speed train from Los Angeles to San Francisco, and when that day comes, the contrast will no longer be an argument. It will be a verdict.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.

Consider what social scientists almost never get to run: a clean experiment. You take a system, hold most things constant, change one large input, and wait long enough that the result cannot be dismissed as noise. In medicine this is the gold standard, and when the input changes by enough and the output refuses to follow, we stop arguing. The federal Department of Education has, without intending to, given us exactly such an experiment, and it ran for half a century. The input, real per-pupil spending, roughly tripled. The output, what an American 17-year-old actually knows, did not move. That single pairing is the entire indictment, and what makes it lethal is that both numbers come from the government’s own record keepers, not from its critics.

I want to walk you through why this happened, because the flat line is not an accident or a run of bad luck. It is what the Department was designed to produce, once you understand who it was built to serve. The answer is not children. It was adults, and the men and women who made the deal said so plainly at the time.

The Department of Education was not born of an educational emergency. It was born of a campaign promise. During his 1976 run for the White House, Jimmy Carter pledged the National Education Association that he would create a stand-alone, Cabinet-level education department, and in exchange the NEA gave the first presidential endorsement in its history. This was a union that had wanted a federal department for roughly a century, and that had never before put its weight behind a candidate. In 1976 it did, and the payoff in raw political muscle arrived immediately. Some 172 NEA delegates sat at the 1976 Democratic convention, the largest single bloc in the hall. Walter Mondale, whose own brother worked for the union, carried the promise to the membership. The transaction was not subtle, and the principals did not pretend otherwise.

Listen to them. Terry Herndon, the NEA’s executive director, said of the agency, “There’d be no department without the NEA.” That is not an opponent’s characterization. It is the union’s own top official describing the Department as the union’s creation. Senator Daniel Patrick Moynihan, a liberal Democrat and no conservative partisan, described Carter’s reelection campaign as having become “a wholly owned subsidiary of the NEA.” An NEA official later put the prize more proudly still, boasting that the NEA was the only union in America with its own Cabinet department. When a Harvard education scholar, Marty West, looks back on the episode today, he concludes that the Department was created in an ultimately unsuccessful attempt by Carter to win the 1980 election. Even Albert Shanker, who led the rival American Federation of Teachers, opposed the new department because he saw it for what it was, a power grab by the NEA rather than a reform for students.

None of this was smooth or consensual, which is telling. The bill creating the Department limped through the House in 1979 by a vote of 215 to 201, and Carter signed it with an eagerness that embarrassed even sympathetic observers as a difficult reelection loomed. Along the way he fired his own secretary of Health, Education, and Welfare, Joseph Califano, who had resisted carving a third of his department away to satisfy the pledge. A president willing to purge his own Cabinet to keep a promise to a union tells you precisely where the children ranked in the order of priorities. They ranked below the endorsement.

Now follow the money, because the scale is the point. When the Department opened in 1980 its budget, in constant 2024 dollars, stood at roughly $56.9 billion. By 2024 that figure had climbed to $268.4 billion, nearly a fivefold increase in real terms. In nominal dollars the jump is starker still, from about $14 billion at the founding to $268 billion today. Summed across more than four decades of outlays, cumulative federal education spending runs into the trillions, more than $4 trillion by reasonable accounting. And this is only the federal slice. Taken as a whole, American education spends at levels no other nation approaches. In 2024 the US led the world with $1.353 trillion in government education spending, with China a distant second near $906 billion. Our K-12 systems spend an average of $20,387 per pupil, third highest among 40 developed OECD nations. At the college level we spend $37,400 per student, behind only Luxembourg and more than double the OECD average. By every measure of inputs, we are not merely generous. We are first.

Against that mountain of spending, set the learning. The National Assessment of Educational Progress maintains a Long-Term Trend series designed for exactly this purpose, measuring the same basic reading and math skills, in the same way, since the early 1970s, so that one generation can be compared honestly to another. Its verdict after 50 years is brutal in its plainness. Today’s 17-year-olds read no better than the high school seniors of the early 1970s. For 13-year-olds the story is the same, with the 2023 average reading score sitting a single point above its 1971 level and math just five points above 1973, before recent declines erased even those modest gains. The economist Eric Hanushek of Stanford, who has studied this longer than almost anyone, found that inflation-adjusted spending rose roughly 150% between 1970 and 2010, and that no state today spends less in real dollars than it did in 1970. His summary of the great spending experiment is the sentence that should hang over every appropriations hearing: the money bought smaller classes and better-paid teachers, “but there were no concomitant improvements in student achievement.”

If domestic stagnation does not persuade, the international comparison should. On the PISA 2022 assessment, the most cited cross-national benchmark, 25 education systems scored higher than the US in mathematics, the subject that matters most for a modern economy. Our average math score was not measurably different from the middling OECD average. One composite ranking placed the US 18th overall, well behind Singapore, Macau, Taiwan, Japan, and South Korea. So here is where five decades and trillions of dollars have left us. We are first in the world in what we spend and somewhere in the middle in what our children learn. No competently run enterprise on earth could post that ratio and survive.

A fair reader will raise an objection here, and I want to meet it directly. Perhaps the problem is that we still do not spend enough, or do not spend it in the right places. The 2009 stimulus tested precisely this. Washington poured roughly $77 billion into schools, more than doubling its annual K-12 outlays in a single rush of cash, and achievement did not follow the dollars. The lesson, repeated now across every level of analysis from Hanushek’s national series to that one enormous infusion, is that how schools spend matters far more than how much. Money was never the missing ingredient. A structure built to reward adults rather than to educate children was the missing ingredient, and money poured into that structure does what water poured into sand does.

This is where the philosophy turns practical. Ronald Reagan saw the problem within two years of the Department’s birth. “Better education doesn’t mean a bigger Department of Education,” he said in a radio address, before adding that the Department should be abolished outright. He campaigned to eliminate it and, in office, tried, proposing to cut the federal education budget sharply. He failed. The reason he failed is the thesis restated in a single line. A program built to serve an interest group does not survive on its results, because it was never meant to produce results. It survives on the political muscle of its patron, and that muscle does not weaken when test scores stay flat. It weakens only when the patron is finally confronted.

And the patron is still being paid, which is the part of this story that should anger taxpayers most. The transaction Carter struck in the 1970s did not end with him. It compounds. Randi Weingarten, president of the American Federation of Teachers, has been paid more than $4 million by the union, earning north of $500,000 a year. Becky Pringle, president of the NEA, has drawn nearly $3 million, likewise clearing $500,000 annually. These are not the salaries of public servants laboring in obscurity for children. They are the salaries of executives running a political machine whose single greatest asset is a Cabinet department purchased on their behalf and maintained at public expense ever since. The Department of Education is, in the most literal sense, a standing return on a 1976 endorsement, and the union bosses who inherited the deal are still cashing the dividend.

Share

It is worth remembering that none of this federal apparatus was ever necessary. Education appears nowhere among the enumerated powers of the national government, and for two centuries before 1979 it was run, imperfectly but accountably, by states and localities answerable to the parents whose children they taught. The Department added a layer of cost and compliance on top of that arrangement and, by the only measure that counts, added no learning. That is the rare government program whose defenders cannot point to its scorecard, because its scorecard is the indictment.

So the conclusion writes itself, and it is not a call for one more reform or one more reorganization, the traditional Washington ritual of renaming failure. It is a call to finish what Reagan started. Gut the Department of Education. Return its genuinely necessary functions to the states that ran them competently for generations, and return its dollars to the families and classrooms the agency was supposedly built to serve. We have run the experiment. We tripled the money and we froze the mind. The kindest thing we can say about the Department of Education is that it faithfully served the adults who created it. The truest thing we can say is that it was never built for the children, and after 50 years and more than $4 trillion, the children’s own results have told us so.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.

Imagine you are handed a case that no one wanted. A man stands accused of the continuous sexual abuse of a 10-year-old boy. The local prosecutor, as it happens, is acquainted with the accused, and so he must step aside. The file lands on someone else’s desk by default, not by ambition. No one sought it. But now a child’s future runs through it. This is precisely the situation the Texas Attorney General’s office inherited in the matter of Adam Hoffman, a Waco lawyer. It is worth dwelling on that starting point, because nearly everything James Talarico now says about this case depends on the reader forgetting how it began.

Two veteran prosecutors took it up. Brenda Cantu brings 29 years of prosecution experience. Dorian Cotlar brings 25 years of criminal law experience and is Board Certified in Criminal Law. Between them they have worked thousands of child sexual-assault cases. These are not political appointees parachuted in to do a favor. They are line prosecutors, the kind of public servants who spend careers in rooms most of us would not enter, doing work most of us could not stomach. When two people with that combined background tell you what happened in their own courtroom, the reasonable response is to listen before you accuse.

So let us listen. The first claim Talarico has put into circulation is that Hoffman received a sweetheart deal. This is the opposite of the truth. The office charged Hoffman with Continuous Sexual Abuse of a Child, which is the most serious felony available to them and carries a mandatory life sentence. You do not reach for the gravest charge on the books when you are trying to do a friend a favor. You reach for it when you intend to put a man away for the rest of his natural life. The charge itself is the refutation of the slander built upon it.

Here a puzzled reader might ask the obvious question. If they charged him so aggressively, why is Hoffman not serving life? The answer is not corruption. The answer is a jury. In June 2025 the prosecution tried a four-day jury trial. The child, by every account from the people who were actually in the room, testified with extraordinary courage, facing down the man who had hurt him. However, under cross examination he admitted that he embellished or exaggerated some of his claims. As a result, the jury could not reach unanimity. It hung by a margin of 7 to 5, with 7 jurors voting to convict and 5 unwilling to do so. This is the inconvenient fact that no amount of campaign rhetoric can dissolve. A criminal conviction in this country requires all 12, not 7. Reasonable jurors, having heard the same evidence, disagreed about whether the state had met the formidable burden of proof beyond a reasonable doubt that our system rightly demands before it cages a human being for life. That is not a scandal. That is the jury system functioning exactly as the Founders designed it, with all the friction and frustration that design entails.

Now we arrive at the decision Talarico finds so damning, and here precision matters most. After the hung jury, the prosecutors were prepared to retry the case. They said so to the family. But the child, having endured the first trial, made clear that he would not voluntarily take the stand a second time. Consider what a retrial would actually have required. The Sixth Amendment guarantees every defendant the right to confront and cross-examine the witnesses against him. There is no asterisk in that clause exempting children. To retry Hoffman, the state would have had to compel this boy, by subpoena if necessary, to sit again in the same room as his abuser and submit again to hostile cross-examination. The only instrument the office held was force. Force applied to a child who had already spent his formative years enduring the abuse, and who had told the adults responsible for him that he wanted to heal rather than relive.

Pause on the moral structure of that choice, because it is the hinge of the entire matter. The prosecutors faced two paths. The first was to drag a traumatized 10-year-old, now a teenager, back onto the witness stand against his expressed wishes, in pursuit of a conviction that 5 jurors had already declined to render once, with no guarantee of a different result the second time. The second was to secure what could actually be secured without that child’s compelled testimony, namely an admission of guilt to a lesser offense and real jail time for the abuser. They chose the second. They chose the child over the spectacle.

I want to be careful here, because this is where reasonable people of good faith can feel torn. Is it not a defeat that Hoffman pleaded to less than the maximum? In a perfect world, yes. But prosecution does not occur in a perfect world. It occurs in the world of hung juries and reluctant witnesses and constitutional constraints. The relevant comparison is never the conviction we wish we could get. It is the realistic alternatives actually on the table. Had the office insisted on a second trial without its central witness, the most likely outcomes were a dismissal or an acquittal, either of which would have sent Hoffman home a free man with no record of guilt at all. Measured against that, an admission of guilt and time behind bars is not a failure of nerve. It is a floor, deliberately built, beneath a child the system could not in conscience force back into the breach. Sometimes the responsible choice is the one that secures the achievable good rather than gambling it away chasing the perfect one.

This brings us to the most cynical of Talarico’s insinuations, the suggestion that Paxton intervened because he is somehow friendly with Hoffman. The claim collapses on contact with the simplest facts about how a state attorney general’s office operates. That office employs more than 750 lawyers and handles upward of 20,000 cases a year. The Attorney General is not, and could not possibly be, briefed on the individual charging decisions of line prosecutors in the ordinary course of business. He does not micromanage thousands of files. By the prosecutors’ own account, this matter became known to him only when it was turned into a weapon against him. Paxton does not know Hoffman. He has never met him or spoken to him. The friendship is not a discovered fact. It is a manufactured premise, and a campaign has been built atop it.

So what is actually happening here? Strip away the moral posturing and the structure becomes plain. A US Senate candidate has located a genuine human tragedy, the abuse of a child, and recognized in it a serviceable instrument. He has taken a difficult prosecutorial judgment, one made by career professionals with 54 combined years of experience and made expressly to spare a child further harm, and he has recast it as a corrupt bargain. To do so he must talk about this child, publicly and repeatedly, against the express wishes of the child and the explicit plea of the prosecutors who know him. The prosecutors asked one thing of those taking an interest in the case. They asked that the boy’s privacy be respected and that his decision about how to move forward with his own life be honored. Talarico’s answer to that request has been to make the boy a recurring character in a Senate campaign.

Share

Here is the irony that ought to end the conversation. Talarico styles himself the defender of an exploited child. Yet it is his own conduct that re-exploits the child, pulling him back into public view as campaign material every time the story is useful, long after the boy asked to be left alone. The prosecutors used the law to shield him. Talarico uses him as leverage. One of these is the work of protecting a child. The other is the work of using one. They are not the same thing, and no quantity of indignation can make them the same thing.

There is a reason we entrust charging decisions to experienced prosecutors rather than to candidates polishing their talking points. The prosecutor’s duty is to justice, which sometimes looks like a life sentence and sometimes looks like the hard, unglamorous arithmetic of protecting a witness who cannot bear to testify again. The candidate’s incentive is to the next news cycle. When the two collide, as they have here, we should know which one to trust, and it is not the man treating a child’s worst experience as a prop. The public deserves the facts, and the facts do not flatter James Talarico.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.

Begin with a simple picture, because the law in this case is finally as simple as the picture. Two families lived on the same Evanston street in 1955. They shared a fence, a school district, and very likely a set of grievances against the city that zoned and policed them. Today the city of Evanston will write one of those families a check for $25,000 and turn the other away at the door. The families are identical in every particular the law has ever cared about, with one exception. One family is black and the other is not. That single fact, and nothing else, decides who is paid and who is refused.

When a government sorts its citizens that way, it is doing the precise thing the Fourteenth Amendment was ratified to forbid. This is the heart of the matter, and it is worth stating plainly before the lawyerly fog rolls in. The Equal Protection Clause protects persons, not races. It was written after a war fought over racial caste, and it was written to abolish caste, not to license a friendlier version of it. A program that pays citizens according to the color of their skin, or the color of their parents’ and grandparents’ skin, does not repair the old hierarchy. It rebuilds it with the beneficiaries rearranged.

On June 16, the Department of Justice said so, formally and on the record. Assistant Attorney General Harmeet Dhillon, who leads the Civil Rights Division, moved to intervene in the lawsuit against Evanston’s program and put the point with a clarity that the courts have spent 30 years approaching. There are sound ways for a city to help its poorest neighborhoods, she said, but simply handing out money based on race is not the answer. It is race discrimination, pure and simple, and it is illegal. That is the government of the United States adopting as its litigating position what an honest reading of the Constitution has required all along. Dhillon deserves credit for it. For decades the federal civil rights apparatus enforced the colorblind command in one direction only, against discrimination that injured minorities, while averting its eyes from racial sorting dressed up as benevolence. To enforce the rule evenhandedly, against a city that discriminates by race for reasons it calls noble, is not a betrayal of the civil rights tradition. It is the fulfillment of it.

Consider the doctrine the city must overcome, because it is not some recent conservative invention. The Supreme Court has held for more than three decades that every racial classification by government, at every level, must survive strict scrutiny, the most demanding test in constitutional law. In City of Richmond v. J.A. Croson Co., decided 6 to 3 in 1989 and written by Justice Sandra Day O’Connor, the Court struck down a minority contracting set-aside in a majority-black city governed by a majority-black council. Richmond had a real statistical disparity to point to, minority firms receiving under 1% of contracts in a city half black, and the Court still said no. General societal discrimination, however genuine, cannot justify a racial remedy. The government must identify a specific instance of unlawful discrimination and tailor the remedy narrowly to it. O’Connor warned that the alternative was a future in which the dream of a nation of equal citizens would dissolve into a mosaic of shifting preferences based on inherently unmeasurable claims of past wrongs. That sentence reads today like a prophecy of Evanston.

The Court only sharpened the rule afterward. Adarand Constructors v. Peña in 1995 closed the supposed loophole for federal programs and applied strict scrutiny to every level of government. Students for Fair Admissions v. Harvard in 2023, again 6 to 3, ended race-conscious admissions and reminded the country that distinctions drawn solely by ancestry are odious to a free people. Then, just this past April, the Court decided Louisiana v. Callais, a 6 to 3 opinion by Justice Alito holding that a congressional map could not use race as its organizing principle absent a compelling justification the state did not have. Callais matters here for a reason that has nothing to do with voting maps. It restated, in the freshest possible terms, the two conditions any remedial racial classification must meet. The government must identify the specific past discrimination and the precise scope of the injury, and it must possess a strong basis in evidence that the race-based remedy is necessary. And it reaffirmed the disqualifier directly. There is no compelling interest in generally remedying the effects of societal discrimination across a region or an era. A century-old, generalized grievance is entitled, in the Court’s words, to much less weight. One struggles to imagine a sentence more precisely fitted to Evanston’s case if it had been drafted for the occasion.

Here a careful reader will pause and ask the natural question. Surely Evanston has done its homework. Surely a city would not stake millions of dollars and its moral reputation on a program without first identifying the specific discrimination it claims to remedy. The honest answer, drawn from the federal complaint and from the city’s own documents, is that it did not. The enabling resolutions contain only boilerplate acknowledging harm to black residents. They identify no specific unlawful act, name no precise injury, and rest on no strong basis in evidence. The city’s own policy report, the nearest thing to a factual foundation, is a working document whose authors expressly declared themselves neutral and disclaimed any responsibility for how it would be used. A government does not build a constitutional defense on a paper its own authors refuse to stand behind.

It is worse than an absence of proof. While the program was being written, former White House Counsel C. Boyden Gray submitted a detailed warning that it would be unconstitutional and that the city had given only perfunctory thought to race-neutral options like income-based housing aid. The city adopted the program unchanged. Later, when officials added the unrestricted cash payment, the city’s own Corporation Counsel conceded on the record that they did not have the research to show a cash benefit would be a narrowly tailored remedy. The defendant’s own lawyer admitted the fatal flaw before the plaintiffs ever had to prove it. And the program’s nominal escape hatch, a provision for people who can document post-1969 housing discrimination, turns out to be gated by race as well, because such claimants must first submit proof of race and then wait until every race-qualified ancestor and descendant has been paid. The fig leaf is itself a racial classification.

Share

Look at the mechanics and the figures, which the city funds, fittingly, from its cannabis tax. Eligibility runs entirely on bloodline. An ancestor is a black adult who lived in Evanston between 1919 and 1969, and a direct descendant is that person’s child, grandchild, or great-grandchild. The city has paid at least 141 ancestors a total of $3.525 million, identified at least 454 descendants for payment, distributed well over $5 million already, and would owe more than $11 million to the descendants alone, against a $10 million pledge it has effectively blown past. Now widen the lens, because Evanston is a pilot, not an endpoint. California’s reparations task force, drawing on its own economists, priced a statewide program at figures reaching $800 billion, more than 2.5 times the state’s entire annual budget, with per-person estimates running past $1 million. Those are not a critic’s numbers. They are the proponents’ own arithmetic, and they vindicate O’Connor’s warning that once race becomes the measure of who gets paid there is no logical stopping point and no end date.

The deepest objection is not fiscal. A nation is held together by a shared idea of who its members are to one another. The American idea, the one Justice Harlan defended alone in 1896 when he wrote that our Constitution is colorblind and tolerates no classes among citizens, is that we meet the state as individuals, equal before the law, and not as delegates of our ancestry. Reparations by race quietly replaces that idea with another. It tells citizens that their standing before the government is fixed by descent, that the relevant fact about a man is the color of his great-grandfather, and that public money flows along the channels of blood. Whatever its authors intend, the program teaches the citizenry to see one another as members of racial classes with competing claims on the treasury. That is not repair. It is the manufacture of division, the conversion of neighbors into rival castes, and it is precisely the hierarchy the Equal Protection Clause exists to dismantle.

Chief Justice Roberts gave us the rule in a single unforgettable line. The way to stop discrimination on the basis of race is to stop discriminating on the basis of race. The conservative legal tradition, from Heritage to Cato to the Pacific Legal Foundation, has insisted on this for a generation, not as a partisan slogan but as the plain command of equal citizenship. Two-thirds of Americans agree, including majorities in deep-blue California. Evanston placed itself on the wrong side of the Constitution, the wrong side of the Court, and the wrong side of the public. The Justice Department has now placed the United States on the right side of all three. That is worth celebrating, and it is worth defending until the last race-keyed check is voided.

If you enjoy my work, please subscribehttps://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.

Consider a single entry from the Department of Homeland Security’s own files. A foreign national arrives on a student visa, enrolls, and over the years that follow is granted not one extension, not two, but 14 separate program extensions by school officials. He remains a student from 2009 to 2020 and holds F-1 status reaching back to 2005. Now consider a second case from the same files, a person enrolled as an F-1 student for nearly three decades who, having exhausted that span, transfers into an English language program in late 2022 and is still studying English, still in active student status, as of the spring of 2025. These are not the inventions of a restrictionist think tank. They are descriptions that the government wrote about its own program, in the preamble to a federal rule, and they should make any reasonable person ask a simple question. At what point does a student visa stop being a student visa?

The answer, until very recently, was never. A visa Congress created for foreign nationals who come to America temporarily and solely to pursue a course of study has been allowed to function as an open-ended residence permit. The Trump administration has now moved to close it, and the change has been described in some quarters as harsh. It is the opposite. It is the restoration of an ordinary rule that every other category of temporary visitor already lives under, namely that a temporary admission must have an end date.

To see how we arrived here, look at the mechanism rather than the rhetoric. Most nonimmigrants are admitted until a date certain, after which they must depart or seek a new status. F students were placed instead under what the bureaucracy calls duration of status, a regime that became effective in 1979 and that admits the holder for as long as the educational purpose continues, with no fixed departure date. Before 1979, by contrast, F-1 students were generally admitted for one year at a time and had to seek extensions in increments. In 1991 the system was loosened further when designated school officials were given broad authority to grant program extensions and approve school transfers. The practical effect was to move control over years of a foreign national’s stay out of the hands of immigration officers and into the hands of campus administrators, the very people whose institutions profit from continued enrollment. A clock that anyone with a financial interest can reset is not a clock.

The scale of the resulting drift is documented, and it is large. DHS reports that F admissions have grown more than sixfold since 1980, and it estimates that roughly 2.1 million people are in the F, J, and I categories annually, including about 1.6 million F students. More to the point, DHS identified more than 2,100 aliens who first entered as F-1 students between 2000 and 2010 and who still held active F-1 status today, a precise figure of 2,134 in the government’s tracking system as of that April. A separate comment in the rulemaking record acknowledged some 77,000 individuals who have accrued more than 10 years in student status since 2003. The department’s own conclusion is the cleanest summary available, that some aliens have used the F classification, in its words, to reside in the country for decades. A government that admits a visitor on the promise of temporary study and then loses the ability to say when that study ends has not been generous. It has been negligent.

A puzzled reader may object that a small percentage of long-term students hardly amounts to a crisis. The objection misunderstands the problem, because the defect is structural, not statistical. Under duration of status, DHS concedes it may not detect a violation for years if the student stays enrolled, never leaves the country, and never applies for another benefit. Such students generally do not even accrue unlawful presence, which means the ordinary consequence that disciplines a fixed-date overstay simply does not attach. The system was built without the feature that makes enforcement possible, a moment at which eligibility is re-examined. You cannot police a deadline that does not exist.

If duration of status is the first half of the story, Optional Practical Training is the second, and here the conservative case becomes a constitutional one. OPT is a work authorization that lets F-1 graduates remain employed after their studies, for 12 months ordinarily and up to 36 months for those in STEM fields. Congress never passed it. It was created by regulation and expanded by regulation, once under President Bush in 2008 and again under President Obama in 2016, and the Center for Immigration Studies has documented that it now operates as the largest guest-worker program in the United States. The numbers carry the argument. The Congressional Research Service reports 418,781 OPT-authorized students and recent graduates in 2024, and CIS, counting the broader practical-training population, puts the figure above 505,000. Compare that to the 141,205 H-1B petitions approved for initial employment in the same period, a program Congress did write, capped at 85,000 new workers a year and bound by prevailing-wage rules. OPT has no cap and no wage floor. The largest skilled-labor channel in America is therefore the one the people’s representatives never voted on, never limited, and never required to pay a fair wage.

There is a fiscal edge to this as well, and it cuts against the American worker. Foreign students on practical training, and the firms that hire them, are exempt from Social Security, Medicare, and federal unemployment taxes. NumbersUSA estimates this gives employers a discount of roughly 8% for choosing a foreign graduate over an American one, and it places the lost Social Security and Medicare revenue at about $4 billion a year. Set aside the abstraction and the point is concrete. A young American graduate competing for an entry-level engineering job is competing against a candidate his prospective employer can hire more cheaply, under a program with no wage protection, created by an agency rather than a Congress.

Share

Then there is fraud, and the record is not new. The Government Accountability Office warned in 2014 that ICE had not assessed the fraud risks in OPT and that 38% of active OPT records, 48,642 of 126,796, lacked even an employer name. It warned again in 2019 about a backlog of school recertifications and gaps in fraud-risk management. The warnings were ignored across administrations, which is why the enforcement findings of 2026 read like a vindication rather than a surprise. Reviewing only the top 25 OPT employers by headcount, ICE flagged more than 10,000 foreign students claiming work with highly suspect employers, a figure the agency called the tip of the iceberg. The particulars are almost cartoonish. One Texas employer reported three OPT workers while more than 500 students named it as their place of employment. A Georgia firm listed a post office box as its office and ran a website flagged for malware. A New Jersey company claimed more than 150 F-1 employees yet could not explain who they were or what they did. Acting ICE Director Todd Lyons described the program bluntly as a magnet for fraud. Fake jobs, real visas, is not a slogan. It is a fact pattern.

Here the steelman becomes unusually strong, because the critics are not who the open-borders framing assumes. The lead plaintiff in the long-running litigation against OPT is not a conservative group at all. It is the Washington Alliance of Technology Workers, an affiliate of the AFL-CIO, which spent a decade in federal court arguing that the program undercuts American workers. Republican senators, including Ted Cruz, filed briefs contending that DHS devised a workaround in which student visa holders need not really remain students. Senator Chuck Grassley used GAO’s findings to document what he called alarming mismanagement. Joseph Edlow, now leading the agency that adjudicates these benefits, told his confirmation hearing that he wanted authority to end work authorization for F-1 students beyond their time in school. When a labor union and a conservative senator reach the same destination by different roads, the xenophobia charge collapses on contact.

The Heritage Foundation’s Lora Ries states the governing principle in three words, temporary means temporary, and that is precisely what the administration’s rule restores. It would replace open-ended duration of status with a fixed admission period, generally the length of the program and capped at four years, with extensions adjudicated by immigration officers rather than rubber-stamped on campus, and a 24-month ceiling on language training. None of this forbids a foreign student from studying here. It asks only that a student visa expire, that a graduate who wishes to work use the lawful employment categories Congress designed, and that the government retain the basic sovereign power to ask, periodically, whether a visitor still qualifies to stay.

Return, finally, to that man with 14 extensions, and to the one who has studied English for nearly thirty years. Every seat they occupy in a classroom, every place in an admissions cohort, every slot in a campus labor market is a seat an American student does not get. The honor system has been tested for four decades and it has failed. A nation that cannot say when a temporary visitor must leave has misplaced something essential, and getting it back is not cruelty. It is the ordinary work of a country that takes its own laws, and its own citizens, seriously.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.

The people who spent a decade teaching us that drag shows, campus encampments, corporate pride rituals, and celebrity galas represent the highest expressions of American civic life suddenly rediscovered the word “dignity” when President Trump hosted UFC at the White House. Their complaints sounded serious for roughly five minutes. Then the facts arrived, and the complaints did not survive contact with them. UFC paid the central production cost. A federal judge refused to block the event. The public had real access through a free lottery at the Ellipse. The enormous structure that critics treated as a monument to vulgarity was temporary, and it came down the next morning. So the question worth asking is not whether the event was corrupt, illegal, or desecrating, since it plainly was none of those things. The question is why the anger continued long after every factual basis for it had collapsed.

Consider a familiar pattern. A man tells you he cannot come to dinner because he has no car. You offer to drive him. He says the date is bad. You offer another date. He says he is watching his diet. You promise a salad. He still does not come. At some point an honest observer concludes that the excuses were never the reason. The excuses kept changing because they were always downstream of something the man would not say aloud, which is that he simply did not want to sit at your table. The reaction to UFC Freedom 250 followed exactly this structure, and recognizing the structure is the whole point.

Begin with the strongest factual charge, because it is the easiest to answer. Critics claimed that taxpayers were forced to fund a $60 million spectacle for the president’s amusement. The trouble is that the $60 million was not a taxpayer bill at all. UFC paid for the production. Dana White said the company would absorb the entire cost, and by his own account expected to lose roughly $30 million on the night, telling reporters plainly that he could not afford it. A private company voluntarily spent a fortune and ate a substantial loss to stage a celebration tied to the nation’s 250th birthday. The honest concession here is small. Federal security and ordinary government services carry some public cost, as they do for every major event on those grounds, from Easter egg rolls to state arrivals. That residual is real, and it is also trivial next to the production figure critics tried to weaponize.

The legality charge fares no better, because it was actually tested. Opponents filed suit to stop the event, and a federal judge declined to do so. The court found that the challengers likely lacked standing, had not shown irreparable harm, and had waited far too long to seek emergency relief. The administration argued, persuasively to the court, that temporary structures on the White House grounds are unremarkable, that nobody demands an act of Congress before a concert tent or a holiday kiosk is erected, and that the rules governing the South Lawn did not forbid the event. A lawsuit that produces a press release rather than an injunction has not exposed a crime. It has lost.

The access objection deserves the same scrutiny and survives even less of it. Critics described an elite private party walled off from ordinary citizens. The seating on the South Lawn was indeed limited, for reasons that should embarrass anyone who pretends otherwise, since that lawn sits a few hundred feet from the president and cannot hold a stadium crowd. But the public was not locked out. Tens of thousands could watch from the Ellipse, where tickets were distributed free through a random drawing open to anyone willing to enter. An event with a public lottery is the opposite of a velvet rope. And the claim that the great temporary structure permanently scarred the capital was perhaps the most revealing of all, since the court itself noted that the alleged harm was temporary because the structure would be disassembled the very next day. The cage came down. The republic survived.

Here is where the argument turns. Once the taxpayer claim, the legality claim, the access claim, and the desecration claim had each been answered, something curious happened. The criticism did not stop. It simply changed costume. It abandoned the language of law and budgets and adopted the language of taste. The objection was no longer that the event was illegal. The objection was that it was beneath us. Critics hated the cage. They hated Dana White. They hated the crowd. They hated the spectacle of a president presiding over a loud, physical, unapologetically American evening that looked nothing like the curated solemnity they believe should govern Washington. One historian sniffed that the event projected weakness and buffoonery. A commentator complained that it associated the presidency with domination. These are not legal findings. They are aesthetic verdicts, rendered by people who assumed their aesthetic was the only legitimate one.

This is the heart of the matter, and it rewards a moment of cognitive empathy with the critics themselves. Why did the spectacle bother them so much? Not because UFC is fringe, because it is not. The typical fan is a man around 37, a Millennial of broadly working or middle-class means, often living outside the US entirely, and in America more likely rural than coastal. This is not a marginal subculture. It is one of the largest sporting audiences on earth, reaching nearly a billion households across more than 200 countries. Nor have the actual elites of money and media rejected it. Paramount signed a seven-year rights deal worth $7.7 billion. The sponsor roster reads like a Fortune 500 directory, with Anheuser-Busch, Procter and Gamble, IBM, and Monster Energy among the names. The business class has fully embraced UFC. What remains is a narrower and more telling form of resistance, the resistance of the taste-maker, the resident of the faculty lounge and the green room who has decided that this enormous, popular, profitable thing is nonetheless unworthy of the nation’s most symbolic address.

That is the religion the freakout exposed. Not democracy, since the man who staged the event won a national election. Not dignity, since the same voices celebrate far stranger ceremonies when the politics suit them. Not fiscal prudence, since the production was privately funded. The real creed is control, specifically control over what counts as legitimate American culture and over who is permitted to perform it at the center of national life. The unwritten rule of the capital, the one Trump violated, is that conservative populists may occasionally win elections but must never be allowed to define the culture of Washington. They may hold the office. They may not set the tone.

Share

This is why the 2025 comeback is better understood as something larger than a change of personnel. It functioned as a cultural eviction notice. The older Washington class wants the White House to project its own values, which are managed decline dressed up as prudence, curated politeness, credentialed boredom, and moral instruction delivered by people who plainly dislike a great share of their countrymen. Trump used the America250 calendar to project a rival set of values, namely strength, competition, risk, patriotism, masculinity, entertainment, and a frank national confidence. The Marine Band played. Fighters toured the Oval Office and stood before the Declaration of Independence. First responders and active military were honored. One does not have to admire cage fighting to notice that this was a celebration of the country, staged for the Americans who like fighters more than functionaries, flags more than finger-wagging, and strength more than the soft vocabulary of decline.

None of this requires defending every moment of the night. One fighter said something ugly and false about a former first lady, the crowd groaned, and Dana White condemned the remark within hours as nasty and false. A single boorish outburst from a known provocateur is not an indictment of an evening built around the Marine Band and the founding document. To treat it as one is to confuse the loudest voice in a stadium with the institution that hosted it, an error no honest critic would make about any event he actually liked.

So return to the man who would not come to dinner. The reasons he offered were never the reason. The reaction to UFC at the White House worked the same way. The taxpayer charge was false, the legal charge lost in court, the access charge ignored the free lottery, and the desecration charge dissolved the moment the structure was carted off. What was left, after every factual objection had been answered, was the thing that had been there all along, which is a deep and barely concealed contempt for the Americans the president invited onto the lawn. Trump did not degrade the White House by hosting UFC. He revealed how much a certain class resents the country that exists outside its own social circle. The cage came down and the crowd went home, but the meltdown revealed something that will not come down, which is that the people who claim to guard the dignity of the White House mostly resent the Americans Trump invited to see it.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.

A flame looks like a thing. It sits on the wick, bright and steady, and we speak of it as though it were an object we could pick up and carry across the room. It is nothing of the sort. A flame is an event, a continuous process of combustion that persists only because fuel keeps arriving from below and air keeps arriving from around it. Interrupt the supply and the flame does not shrink politely into a smaller flame. It ends. This is the distinction that unlocks what President Trump and the United States military accomplished against Iran over the past year, and it is a distinction most commentary has missed. Military power resembles a flame far more than it resembles a possession. An arsenal looks like a fixed inventory a nation either owns or lacks, when in truth it is a process, an unbroken act of manufacture and replacement that appears stable only because the factories never stop feeding it.

Hold that picture in mind, because two American operations did two very different things to the fire. Operation Midnight Hammer, in June 2025, smothered the single most dangerous flame Iran had lit, striking the enrichment complexes at Fordow, Natanz, and Isfahan. Operation Epic Fury, the joint American and Israeli campaign that ran from late February to early May of this year, did the deeper thing. It did not merely beat down Iran’s conventional fires. It destroyed the fuel and the means of making fuel across the whole of the Islamic Republic’s military, so that what was extinguished cannot simply be relit.

Begin with the nuclear file, where the popular understanding is weakest. Iran had spent, by reasonable estimates, as much as $15 billion building a hardened, dispersed, survivable enrichment infrastructure over the course of decades. That investment is gone. What remains of the enriched uranium is, for any practical purpose, buried beyond reach. Only two nations on earth, the United States and China, possess the engineering capacity to exhume material from beneath that kind of rubble, and the only conceivable reason either would do so is to carry it away for good. The stockpile is therefore no longer an asset Iran can recover. It is a liability Tehran must hope someone else removes. To begin again from nothing would cost billions more and, by the assessment of the Center for Arms Control and Non-Proliferation and others working from the IAEA verification record, somewhere between five and ten years simply to restore a hardened and resilient enrichment base, with full restoration of the prewar position pushing toward a decade or longer.

The skeptic will object that Iran keeps the knowledge, and that knowledge cannot be bombed. This is true and beside the point. Knowledge does not enrich uranium; centrifuges do, and centrifuges require a manufacturing base, conversion plants, power systems, tunnels, and trained operating crews, every link of which was a target. For a sense of scale, recall that the Bushehr power reactor, a far simpler undertaking than a clandestine weapons complex, took nearly four decades to finish and cost close to $11 billion in current dollars. If Tehran were to commit every resource it possesses to a sprint for the bomb, it would still be the better part of a decade from a survivable arsenal, and it would have to build that arsenal in full view of two air forces that have now shown, twice, both the will and the means to flatten it again. The promise to repeat the strike is not a bluff. It may be the most credible commitment in the region.

Why, then, was a second and larger operation necessary at all? The reasoning here is subtle and rewards patience. For years, American and Israeli planners drew quietcomfort from a single assurance, namely that whatever Iran buried underground, they could always strike it from the air before it matured into a weapon. That assurance rested on an unspoken premise, that Iran’s conventional forces could neither prevent the strike nor exact an unbearable price for it. The premise was crumbling. Intelligence indicated that Iran had amassed so vast an inventory of ballistic missiles, cruise missiles, and one-way attack drones, with prewar estimates clustering around 2,500 to 3,000 ballistic missiles and a declared ambition to reach 8,000, that it was approaching what analysts call saturation. The peril of saturation is arithmetic, not valor. No interceptor magazine is infinite, and a large enough salvo need not be accurate or sophisticated to break through. It need only arrive in greater number than the defense can absorb in a single sustained wave.

This is the cost-exchange problem in its barest form. A Shahed drone that costs roughly $35,000 to build can compel the expenditure of a Patriot interceptor that costs about $4 million. Iran understood that it could not rival the West in quality, so it invested instead in weapons engineered to be cheap and quick to replace, wagering that sheer quantity would eventually swamp any defense imaginable. Against that backdrop the comfortable old assumption collapsed. A regime capable of blanketing Israel and the Gulf in one coordinated barrage could not be reliably disarmed by airstrikes alone, because it could refill its magazines faster than those strikes could empty them, and could credibly threaten Israel’s cities with a volume no missile-defense architecture yet built could fully intercept. Epic Fury was the answer to that precise and growing danger. Its object was never to dent the arsenal but to destroy the engine that produced it.

So what, concretely, did America win? Across the navy, the submarine force, the combat-aircraft fleet, the helicopter inventory, the ballistic-missile stockpile, and the drone arsenal, over 80% of Iran’s conventional military was destroyed, together with nearly all of the manufacturing capacity that sustained it. Replacing what was lost, at world-market prices, would run on the order of $40 billion for the conventional forces alone, and closer to $50-$55 billion once the nuclear infrastructure is counted. Set that bill against Iran’s real means. SIPRI placed Iranian military spending in 2025 at about $7.4 billion, which means the reconstitution cost equals roughly 5.3 years of the regime’s entire defense budget, and that figure politely ignores the sanctions premium, the inflation, the smuggling markups, the gutted factories, the dead engineers, and the cratered airbases that drive the true cost higher still.

Share

Yet the dollar figure, vast as it is, is not the prize. The prize is time. A full conventional rebuild is a 10 to 15 year undertaking, and its pace is set not by the cheap categories but by the impossible ones. Iran can, given working factories, mass-produce drones and missiles again within a few years, for these are the kindling of its arsenal, quick to gather and quick to burn. It cannot replace its submarines, which were Russian imports it never had the capacity to build, and it cannot replace a modern fighter fleet it could never manufacture in the first place. These are the slow-growing timber of a military, and timber of that kind takes a generation to mature. The deeper lesson returns us to the flame. Iran deliberately built an arsenal of things cheap and fast to relight precisely because its prestige assets were impossible to relight under sanctions. Epic Fury extinguished both, but only the kindling can return, and even the kindling cannot return until the factories that feed it are rebuilt, not merely the inventory restocked.

This is why the strategic gift of the past year is best counted in years rather than dollars. The United States, Israel, and the Gulf states have been handed a window of a decade or more in which Iran cannot project power as it once did, and that window is not idle. It is time to deepen interceptor magazines, to solve the cost-exchange problem the drone exposed, to weave the region’s air defenses into a single fabric, and to negotiate from overwhelming strength rather than anxious deterrence. The old guard that would have wasted such an opening is gone. The leadership that spent decades spurning every reasonable off-ramp was decapitated in the opening hours of the campaign, and the successor leadership has done what its predecessors swore they never would. Through Pakistani mediation it has entered talks covering nuclear rollback, international monitoring, missile limits, and the reopening of the Strait of Hormuz, weighing in detail the very demands the old regime dismissed with contempt.

That engagement frames the range of plausible outcomes, and both ends of the range favor peace. The worst case is that Iran’s capacity to threaten its neighbors has been pushed back a full generation, which is itself a historic improvement on where matters stood a year ago. The best case is brighter still. Iran is a genuinely rich nation strangling itself by its own choices, and its new leaders may yet conclude that the wealth forfeited to sanctions and obsession dwarfs anything a warhead could buy. The President has already named the terms, telling Tehran it may begin reconstruction and that, in his phrase, big money will be made, while the Vice President has set the condition plainly, that Iran must act like a normal country before it is treated as one. Cut off a regime’s ability to make war, and you have not only weakened it. You have, for the first time in 47 years, given its successors a reason to choose prosperity. That is what was won.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.

There is a comforting fiction circulating in Washington that the Senate Parliamentarian holds a kind of veto over what Congress may and may not do. According to this fiction, when Elizabeth MacDonough offers advice that a provision violates the Byrd Rule, the matter is settled, the provision dies, and elected senators must simply accept the verdict of an unelected staff official appointed during the Obama administration. This is not how the Senate works. It has never been how the Senate works. And John Thune, of all people, knows it is not how the Senate works, because in May of 2025 he proved the opposite with his own hands.

The clean way to state the principle is this. The Parliamentarian advises, the Chair rules, and the Senate decides. That ordering is not a clever workaround invented for the present moment. It is the basic architecture of Senate procedure. The Congressional Research Service explains that Senate rules are enforced when a senator raises a point of order, after which the Presiding Officer decides the question unless he submits it to the body, and that the ruling is then subject to appeal. The procedural scholar Jim Wallner puts the same point more bluntly when he writes that the Chair rules on points of order, not the Parliamentarian, and that the Parliamentarian merely advises the Chair. The office that MacDonough occupies was not created by the Constitution. It was created on July 1, 1935, when the Senate elevated Journal Clerk Charles Watkins to the new post. The Senate governed itself for nearly a century and a half before the office existed at all. An institution that predates its own advisers by 146 years does not answer to them.

So why does the Parliamentarian seem so powerful? She seems powerful because Senate leaders, for reasons of convenience and institutional comfort, usually choose to treat her advice as if it were a ruling. Her real authority is the authority of accumulated memory. The compilation of Senate precedents known as Riddick’s Senate Procedure runs more than 1,600 pages and contains more than a million precedents, and navigating that thicket is genuinely difficult. Senators prefer predictable administration of complicated rules, and so they defer. Deference, however, is a choice, not an obligation. Influence is not the same thing as veto power, and a Senate that wishes to act as a body retains every ounce of the constitutional authority it had before 1935.

Now consider how this applies to the SAVE Act. The mechanism here is the Byrd Rule, codified at 2 USC § 644, which protects budget reconciliation from being loaded with policy that is merely incidental to the budget. Democrats will argue that election-integrity language is extraneous policy and therefore must be stripped. The crucial detail, the one that the press almost never explains correctly, is that the Byrd Rule does not strike anything on its own. The statute is explicit that a provision falls only upon a point of order being made by a senator and sustained by the Chair. Advice alone does nothing. If no senator raises a point of order, the language stays. If a senator does raise one, the question lands not on MacDonough’s desk but on the desk of the Presiding Officer.

This is where the tactical picture becomes clarifying. Imagine SAVE language is included in the reconciliation bill. A Democratic senator rises and lodges a Byrd Rule point of order. The question does not land on the Parliamentarian’s desk. It lands on whoever holds the gavel, and the majority decides who that is. Republicans can ensure a friendly presiding officer is in the chair at the decisive moment, and Vance, as President of the Senate, can take the gavel himself to deliver the ruling. Whoever presides then rules the point of order not well taken. Democrats may appeal that ruling to the full Senate, and here is the part that should focus every Republican mind in the chamber. Under the enforcement provisions of the Congressional Budget Act, sustaining an appeal of the Chair’s ruling on a Byrd Rule point of order requires three-fifths of senators, which in a full Senate means 60 votes. The burden of finding 60 falls on the side trying to overturn the Chair, not on the side defending the language. Republicans do not need to assemble a supermajority to protect proof-of-citizenship requirements. They need only have the Chair rule in their favor and then watch Democrats fail to find 60 votes to remove it. If Democrats cannot muster 60 votes to strike a measure requiring that only citizens register to vote, the measure survives, and the bill itself can pass by simple majority, with Vance breaking a tie in his separate capacity if the Senate is evenly divided.

If this sounds aggressive, it is worth remembering that Thune has already done a more contested version of exactly this. In May of 2025, Senate Republicans wanted to use the Congressional Review Act to overturn California vehicle-emissions waivers. They received adverse procedural advice, rooted in a Government Accountability Office opinion that the waivers were not rules eligible for expedited treatment. Thune did not fold. He argued that the Senate faced a novel procedural dispute and that, in his words, it is appropriate for the Senate to speak as a body. The Senate then did precisely that, sustaining his position through majority votes recorded in the Congressional Record at 51-46, 51-46, and 53-46. Molly Reynolds of Brookings, writing in the Yale Journal on Regulation, described the maneuver without euphemism, noting that the Parliamentarian had advised the resolutions were ineligible, that Republicans were nervous about overriding that advice, and that Thune engineered procedural votes allowing the Senate to decide for itself.

The logic here travels effortlessly from emissions waivers to elections. If the Senate may speak as a body to halt a California mandate on automobiles, it may certainly speak as a body to protect the integrity of federal elections from noncitizen registration. The principle Thune articulated does not contain a footnote restricting it to environmental policy. A novel procedural disagreement is a novel procedural disagreement, and whether the SAVE Act qualifies for inclusion in reconciliation is exactly the kind of contested question that, by Thune’s own stated standard, the Senate itself should resolve. The honest summary is blunt. Thune was right in May. He should now apply the same principle to a far weightier subject.

It also helps that Thune has armor against the predictable objection that this amounts to abolishing the legislative filibuster. He built that armor himself. During the CRA fight he insisted the maneuver had nothing to do with the legislative filibuster and pledged that under Republican control the legislative filibuster would remain in place. He can say the same here with equal honesty, because reconciliation already operates under special statutory rules that bypass the filibuster by design. Defending SAVE through a ruling from the Chair does not touch the filibuster at all. It simply uses the reconciliation process as it was built to be used, and asks the elected majority to exercise the authority it already possesses.

The public is not divided on the underlying question, which makes the procedural timidity all the harder to justify. Gallup found in October of 2024 that 84% of Americans favor requiring photo ID to vote and 83% favor requiring proof of citizenship for first-time voter registration. Support for photo ID reached 98% among Republicans, 84% among independents, and 67% among Democrats. Pew confirmed the pattern in August of 2025, finding 83% of adults support a government-issued photo ID requirement, including 71% of Democrats and Democratic-leaning independents. The White House has reduced the moral case to a single sentence that opponents struggle to answer without sounding evasive, namely that American citizens, and only American citizens, should decide American elections. When a position commands 80% support and one chamber’s leadership hides behind a single unelected adviser to avoid acting on it, the problem is not public opinion. The problem is nerve.

Share

The legislative groundwork is already laid. The original SAVE Act, H.R. 22, passed the House on April 10, 2025, by a vote of 220-208, with House Republicans unified at 216 yes votes and zero no votes. The stronger second-term version, the SAVE America Act, has since cleared the House as well, passing 218-213 with the backing of President Trump, Speaker Johnson, and Senator Mike Lee. In the Senate, John Kennedy has already forced the issue, offering an amendment to add SAVE concepts to the reconciliation plan that failed only 48-50. A measure two votes short of a majority is not a fringe cause. It is a near-consensus waiting on leadership, and Kennedy has said plainly that he intends to put the SAVE Act on the reconciliation bill. Conservative institutions have supplied the policy scaffolding, with Heritage Action describing how SAVE America would replace registration by self-attestation with verification against DHS and SSA databases, and the America First Policy Institute framing the shift as a move from honor-system registration toward objective proof.

There is even a backstop, though it should remain a backstop rather than an opening move. Parliamentarians have been replaced before, by both parties. In 2001 Trent Lott pushed out Robert Dove after rulings that complicated the Bush tax-cut agenda, and the Washington Post noted that the Parliamentarian serves at the pleasure of leadership. That history is worth keeping in view, but it is not the argument. The argument is simpler and stronger. Thune does not need to fire anyone. He needs only to do again what he already did once, when the stakes were lower. Put SAVE in reconciliation. Let Vance rule the point of order not well taken. Make those who would strip citizenship verification from the bill go out and find 60 votes to do it. If they cannot, and they almost certainly cannot, then the Senate will have spoken as a body, exactly as Thune said it should.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.

Consider what an election actually is. Beneath the rallies and the slogans, a free election is a measuring instrument. It takes the scattered preferences of millions of citizens and resolves them into a single public number, and like any instrument its worth depends entirely on its accuracy. A thermometer that can be nudged tells you nothing about the temperature. A vote count that can be padded tells you nothing about the will of the people. The whole moral weight of self-government rests on one assumption, that the instrument reports what the citizens actually said. So when a count behaves strangely, the correct response is neither blind faith nor panic. It is the scrutiny any careful person brings to a gauge that has produced a suspicious reading.

The Los Angeles mayoral primary of June 2026 produced such a reading. On election night, the insurgent Republican Spencer Pratt sat in second place behind Mayor Karen Bass, holding a lead of roughly six points over Councilwoman Nithya Raman and positioned for a one-on-one November runoff. Then the late batches arrived. Day after day, the successive tallies leaned more Democratic, and by Sunday, June 7, Raman had passed Pratt by 3,113 votes, 27.12% to 26.69%, after several days trailing in third. The man who led when the polls closed was, once the counting finished, edged out of contention entirely. The mechanism was the late tabulation of mail ballots in a jurisdiction where registered Democrats outnumber Republicans by roughly four to one.

Now I can hear the objection, because it is the first thing every defender of the system says. This is simply how California counts. Mail ballots are processed last, late ballots tend to lean Democratic, and the so-called shift is a benign artifact of sequence, not evidence of anything wrong. Fair enough. No court has found fraud in Los Angeles, and I am not asserting one did. But notice what the objection quietly concedes. It admits that the outcome of a race can hinge entirely on a category of ballots cast outside the supervised polling place, counted days after the public has gone home, in exactly the channel where the documented abuses of American history have always concentrated. The benign explanation and the sinister one share the same anatomy. That is precisely why the pattern deserves a hard look rather than a reflexive shrug.

For the pattern is not new, and this is the heart of the matter. The claim that election fraud is a fever dream invented in 2020 collapses the moment one opens the historical record, because the record is long, it is adjudicated, and it produces convictions. The continuity is the story. The same tools surface in case after case, separated by a century or more. Operatives collect and complete absentee ballots. Votes are cast in the names of the dead, the imprisoned, and the departed. A razor-thin contest is reversed days after the polls close by ballots that materialize from outside the ordinary in-person count. The argument here is not that every Democrat is a fraudster, which would be absurd and unworthy. The argument is narrower and harder to dodge. The machine wing of one party has repeatedly reached for the same instrument whenever a race was close enough to steal.

Begin with Tammany Hall. By 1868 Boss Tweed’s machine controlled the entire New York City electoral apparatus, and a congressional investigation documented repeat voting, a naturalization mill that stamped immigrants into instant Tammany voters, the destruction of opposition ballots, and outright alteration of the counts. One audit found 25,000 of 156,000 city votes fraudulent, a rate near 16%, and across 1868 to 1871 the city’s vote totals ran 8% above its entire voting population. A contemporary described the method with grim economy, the dead filling in for the sick. That is the template every later case rhymes with.

Move to 1948 and the most famous stolen election of the last century. In the Texas Democratic Senate runoff, the popular former governor Coke Stevenson appeared to have beaten Lyndon Johnson. Then, six days after the polls closed, Precinct 13 of Jim Wells County produced 202 additional votes, 200 of them for Johnson, handing him an 87-vote win out of nearly a million cast. The added names appeared in alphabetical order, in the same handwriting and ink, and several listed voters insisted they had never cast a ballot. Luis Salas, the election judge who was there, later told the Associated Press plainly that Johnson did not win that election, that it was stolen for him, and that he knew exactly how it was done. A confession from inside the machine is worth more than any outside analysis, and Johnson rode that fraudulent margin to the Senate, the vice presidency, and the White House.

Then watch the playbook migrate decisively into the absentee process, which is the development that should concern us most today. In the 1994 special election for Pennsylvania’s 2nd Senatorial District, a seat that decided control of the entire chamber, the Republican Bruce Marks led on the ballots cast in person. The Democrat William Stinson won on the absentees. A federal court found fraud so pervasive across the more than 2,600 absentee ballots, including votes cast in the names of people in Puerto Rico, in prison, and dead, that it voided the result and seated the Republican. The Third Circuit’s Judge Morton Greenberg, upholding that decision, did not reach for the soft vocabulary of irregularities. He wrote that the record made clear the election had been stolen from Marks, and that city election officials made it possible by disregarding Pennsylvania law. Stolen, by a federal judge, on the record.

The 2003 Democratic primary for mayor of East Chicago, Indiana, followed the same script. The nine-term incumbent Robert Pastrick survived on the strength of absentee ballots, and the Indiana Supreme Court found a deliberate scheme that perverted the absentee process and compromised the integrity of the election, targeting the poor, the infirm, and the limited-English. The court ordered a new election, and the challenger who had been cheated then won it in a landslide, 65% to 34%, a complete inversion of the fraudulent original. This is the clarifying detail. When the absentee corruption was stripped away and the citizens voted in the open, the result reversed entirely.

Here the skeptic owes an answer to a simple question, and it is the most powerful one in this whole debate. If mail and absentee voting were as airtight as its champions insist, why does nearly every major adjudicated fraud case in modern memory run through that exact channel? The vulnerability is not a conservative hypothesis. Courts have found it, repeatedly, in published opinions. And the channel itself is leaky even before any bad actor touches it. A Massachusetts Institute of Technology analysis of the 2008 presidential election found that 7.6 million of 35.5 million requested mail ballots were never counted, lost in transit or rejected for irregularities, a failure rate above 21%. When one in five ballots can vanish or be rejected through ordinary friction, the surface area available to harvesting, forgery, and selective rejection scales accordingly.

The most devastating witnesses for this case are not partisans, which is why they cannot be waved away. In 2005 the Commission on Federal Election Reform, co-chaired by the Democratic former president Jimmy Carter and the Republican former secretary of state James Baker, concluded flatly that absentee ballots remain the largest source of potential voter fraud. A Democratic president put his name to that sentence. In Crawford v. Marion County Election Board in 2008, Justice John Paul Stevens, the Court’s leading liberal, wrote for the majority that flagrant examples of voter fraud have been documented throughout the nation’s history, and he cited the East Chicago case to prove that the risk is real and can decide a close election. When the liberal lion of the Court concedes the point and footnotes a Democratic machine to do it, the fraud-is-a-myth line simply collapses. And in 1982 a Chicago federal grand jury, in a report a chief judge took the rare step of releasing publicly, called vote fraud a cancer that must be treated lest it destroy our constitutional rights, after a Department of Justice investigation estimated 100,000 fraudulent ballots and produced 63 convictions. That is ordinary citizens, not a think tank, rendering the diagnosis.

The Heritage Foundation has assembled the dry confirmation of all this, a database holding 1,567 proven instances of election fraud, each ending in a conviction, an overturned election, or an official finding of wrongdoing, and Heritage is candid that this is a sample, not a census, since it omits fraud caught but never charged and fraud never detected at all. The real figure is necessarily higher.

So return to where we began, to Los Angeles and the lead that died over five days of counting. I am not telling you a court has found fraud there, because it has not. I am telling you that the surface signature of that race, a Republican leading on election night and overtaken by late ballots from the unsupervised channel, is the identical signature that, in Philadelphia and East Chicago and Jim Wells County, turned out under judicial examination to be theft. A pattern does not convict. A pattern tells you where to look, and a serious people looks.

The moral stakes are not abstract. Every fraudulent ballot cancels a lawful one. The citizen who waited in line and marked an honest vote has had it quietly erased by a vote that should not exist, which makes fraud not a victimless procedural lapse but a direct disenfranchisement, the theft of the one possession every American holds in perfectly equal measure. The defenders of unsupervised mail voting are asking us to expand the precise channel in which the documented thefts have clustered for 160 years, and to do so on faith that this time the seam will hold. The historical record does not earn that faith. It earns the opposite, the demand that the instrument be made trustworthy before we trust it.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.

Begin with a simple question. Who is John Thune's constituency? The civics textbook answer is the people of South Dakota. The textbook is wrong, and the SAVE America Act proved it. President Trump declared the bill his number 1 legislative priority. The House passed it. The Republican base demanded it. Polling showed broad public support for requiring documentary proof of citizenship to register to vote. The South Dakota legislature is among the reddest in America, and its voters delivered Trump one of his largest margins in the nation. Yet when the moment came, the Senate Majority Leader refused to change the rules to pass it, refused even a talking filibuster, and let the bill die on the floor this June. He explained himself with arithmetic: the votes aren't there. But whose votes? Not South Dakota's. Not the American people's. The votes of his fellow senators, the small club whose secret ballot made him leader and whose secret ballot can unmake him. In a 53 member Republican conference, it takes 27 senators to hold the leader's gavel. Those 27 senators are John Thune's true constituency. The people of South Dakota, the legislature of South Dakota, and the voters of the United States all stand in line behind them.

This is not a complaint about one man’s character. Thune is behaving rationally given the incentives he faces, and that is precisely the point. When an officeholder consistently serves the people who control his career rather than the people he nominally represents, the defect is structural, not personal. So we should ask the structural question. What changed the Senate from a chamber of state ambassadors into a chamber of free agents accountable mainly to one another, to national donors, and to national media? The answer has a date. It is April 8, 1913, the day the 17th Amendment was ratified.

Recall what the original design was, because the case rests on the gap between design and outcome. Article I provided that the House would be elected by the people while each state’s 2 senators would be chosen by the state legislature. This was not an accident of 18th century logistics. It was a deliberate machine for keeping Washington tethered to the states that created it. When George Mason rose at the Philadelphia Convention in 1787 to argue that state legislatures must appoint the Senate, lest the new federal government swallow up the state legislatures, the motion carried unanimously. Pause on that word. The most quarrelsome assembly in American history, which fought over representation, slavery, the executive, and nearly everything else, did not divide on this question. Madison then defended the arrangement in Federalist 62, writing that appointment by state legislatures carried the double advantage of favoring a select appointment and of giving the state governments an agency in the formation of the federal government. The states’ agency was the point. A senator who voted to centralize power, spend recklessly, or impose mandates on his state faced a legislature that could simply decline to return him. The leash was short, and everyone could see who held it.

Now consider what holds the leash today. Here the second exhibit is unfolding in Texas. The Senate race between Ken Paxton and James Talarico is on track to set spending records, and the money tells you everything about who a modern senator actually answers to. According to OpenSecrets, after Talarico won his primary his donor base did not merely grow, it went national. The number of itemized out of state donors more than doubled in 6 weeks, and out of state money came to account for nearly half of his itemized fundraising, helping push a single quarterly haul to $27 million. His campaign boasts of donations from all 50 states. Layer onto that the affiliated super PACs, including a newly launched group planning to spend $62 million, staffed by national Democratic operatives, and the picture sharpens. The vast majority of these dollars are from out of state. Donors in California and New York are funding the campaign that could make Talarico the next senator from Texas. A puzzled reader might object that money is not votes, and Texans will still mark the ballots. True. But campaigns run on money, and the candidate’s incentives follow the money’s source. A senator elected on a national small dollar machine and national super PAC air cover owes his career to that machine, not to the Texas legislature, not to Texas institutions, and not in any structural sense to Texas at all. The framers built a Senate in which the state, as a state, selected its ambassador to the federal government. We have built a system in which other states can effectively select who represents us. That inversion is not a quirk of one race. It is the predictable product of the 1913 design change.

Was the original design corrupt? This is the strongest objection, so face it directly. The Progressives sold direct election as a cure for bribery, pointing to the case of William Lorimer of Illinois, elected in 1909 through the bribery of state legislators. But notice what happened to Lorimer: the Senate investigated and expelled him in 1912. The old machinery of accountability caught the crook and removed him. And what did the cure accomplish? The scholarly verdict, assembled in Todd Zywicki’s Cleveland State Law Review study, is devastating. There is no indication that the shift to direct elections did anything to eliminate or even reduce corruption in Senate elections. Hoebeke found the amendment increased the role of organization and money in Senate races. Terry Smith concluded money now dominates them. Vikram Amar found that the need to raise enormous sums opened the federal government to private interest groups. Jay Bybee, later a federal judge, concluded the amendment destroyed the systems of federalism and bicameralism that had checked expansionist federal activity. The bribery did not end in 1913. It was legalized, scaled up, and renamed campaign finance. The Texas race, awash in out of state cash, is the Lorimer affair conducted in the open with a compliance department.

The fiscal record completes the indictment. The 16th Amendment, ratified on February 3, 1913, gave Washington unlimited revenue. The 17th, ratified 64 days later, removed the institution that would have policed how that revenue was spent. The receipts are in. Federal receipts grew from 3% of the economy’s output in 1900 to 16.5% by 2012, while federal expenditures rose from 2.7% to 24%. The federal tax code mushroomed from 400 pages in 1913 to nearly 74,000 pages a century later. Mandatory spending climbed from 47% of the federal budget in 1973 to 73% in 2023, and CBO projects federal debt reaching 116% of GDP by 2034. The Heritage Foundation preserves the mechanism in Zywicki’s formulation: before the amendment, senators had strong incentives to protect federalism because reelection depended on pleasing state legislators who preferred power kept close to home; after it, senators act all but identically to House members, treating federalism as political expediency rather than constitutional principle. Senator Ted Cruz compressed the whole public choice argument into a single sentence quoted by Heritage: if you have the ability to hire and fire me, I’m a lot less likely to break into your house and steal your television. Even a Democrat saw it. Senator Zell Miller of Georgia, introducing an actual repeal amendment in 2004, called the 17th Amendment the death of the careful balance between state and federal government, lamenting that state governments now stand in line as just another special interest, at an extreme disadvantage because they have no PAC. The cleanest modern illustration came in 2010, when 27 states sued to overturn ObamaCare even as the law passed the Senate with 60 votes; in Virginia, both senators voted for the very law their own state officials were litigating against. Under the original design, that vote was career suicide. Under the 17th Amendment, it was a Tuesday.

Share

What, then, is the remedy? Not nostalgia, but repair with improvement. Mark Levin’s proposal in The Liberty Amendments supplies the text: repeal the 17th, return the selection of senators to the state legislatures, and add a power the original Constitution lacked, the authority of a state legislature to recall its senator by a 2/3 vote. The supermajority threshold matters. Senators would not be removed at the whim of a bare majority, but the possibility would remain foremost in their minds, pulling their attention away from Georgetown salons and national fundraising circuits and back toward the interests of their states. A skeptic will ask whether voters lose their voice entirely. They need not. Before 1913, states experimented freely; under the Oregon and Nebraska systems, legislatures bound themselves to ratify the people’s advisory choice. A state could keep a popular advisory vote tomorrow. The constitutional question is not whether voters speak but who holds the leash, and federalism means each state may answer for itself. This is no fringe project. Utah’s legislature passed a repeal resolution in 2016 quoting Madison himself, ALEC has published model resolution text, 20 states have passed convention resolutions, and Senator Mike Lee has carried the argument inside the chamber since 2010.

Return, finally, to John Thune. He told reporters that he is the person who has to deliver the news that the math doesn’t add up. He is right, and the math is the scandal. The math that governs the modern Senate is 27 colleagues, 540,000 national donors, and $62 million super PACs. Nowhere in that arithmetic do the states appear. The framers wrote them into the equation deliberately, and in 1913 we erased them. Repeal restores the founders’ design. Recall perfects it. Until then, every senator is, in the only sense that matters, a senator from nowhere.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.

Ask any executive recruiter the first question they pose about a candidate and you will hear some version of this: who does he answer to? Not who does he claim to serve, not whose interests does he invoke in speeches, but who, as a matter of cold institutional mechanics, can take his job away. The question matters because people respond to the incentives that actually bind them, not to the ones printed on the letterhead. Apply that question to John Thune and the mystery of the SAVE Act dissolves.

Begin with the structural fact, because everything else follows from it. The Senate majority leader is not elected by South Dakota. He is not elected by the country. He is elected by the Senate Republican Conference, by secret ballot, by simple majority. With 53 Republican senators, the threshold is 27 votes. Those 27 colleagues hired Thune, and only those colleagues can fire him from the leadership. Every six years the 900,000 residents of South Dakota can return him to the Senate, but they cannot hand him the gavel or take it away. The tens of millions of Americans who want documentary proof of citizenship before a ballot is cast have no vote in that election at all. Call this the leader’s true constituency. It is a constituency of 27, and it explains his behavior better than ideology, better than temperament, and far better than the official story about Senate norms.

Now set that structure against the record, because the record is no longer hypothetical. The House passed the Safeguarding American Voter Eligibility Act in April 2025 by a margin of 220 to 208, with 4 Democrats crossing over, and passed it again in February 2026. In the Senate, on March 26, 2026, a cloture vote on the Husted amendment to the SAVE Act vehicle drew 53 votes. Read that number again. The bill did not fail for want of a majority. It failed because the cloture rule demands 60, and 53 is not 60. Then, during the vote-a-rama, the case became airtight. The SAVE Act in its original form drew exactly 50 Republican votes after Susan Collins flipped to support Mike Lee’s text. With Vice President JD Vance available to break a tie, 50 is a functional majority under any procedure that requires only a simple majority. Supporters argue, correctly, that the bill effectively commands the backing of 51 senators. The majority is not a forecast or a talking point. It is a recorded fact, entered in the Senate’s own journals.

A puzzled reader might object here. If the majority exists but the rules require 60, is the leader not simply bound by the rules? The objection sounds reasonable until you examine what the rules actually are and who controls them. The 60-vote cloture threshold is not in the Constitution. It is a standing rule of the Senate, and the Senate has demonstrated, repeatedly and recently, that a determined majority can route around it. Three established roads to passage stand open, and Thune has personally declined to walk any of them.

The first road is the talking filibuster. The modern filibuster bears no resemblance to the endurance contests of Strom Thurmond’s era. Today a senator need not hold the floor, need not speak, need not even appear. Conservatives have taken to calling it the Zombie Filibuster, a blocking device that operates without effort or exposure. Mike Lee said it plainly on the floor: but for the Zombie Filibuster, the House-passed bill would already be on the President’s desk. Lee and others have pushed Thune to restore the older, harder version, to force 47 Democrats to physically hold the floor and defend their obstruction on camera, after which the Senate could move to a final vote at a simple majority. Thune has resisted, reportedly worried that his own members might not stay unified through a prolonged floor fight. Notice what that worry concedes. The obstacle is not the Democrats. It is the management of his own conference, the 27 votes that keep him in his chair.

Share

The second road runs through the parliamentarian. Elizabeth MacDonough ruled that the SAVE America Act fails the Byrd Rule and therefore cannot ride in the reconciliation package. President Trump responded by demanding her removal, calling her a radical left lunatic. Thune declined, telling reporters that her rulings break both ways, that the conference would lose a few and win a few. Here the reader should pause on a remarkable admission from an unlikely source. The New Republic, writing in opposition to the bill, conceded that the parliamentarian’s guidance is not binding and that the Republican majority could simply overrule it. When the opposition press stipulates that the leader possesses the power and is merely choosing not to use it, the columnist’s work is half done. Thune’s own explanation for his refusal, that overruling her would create even more vote issues, is the second quiet concession of the piece. Vote issues where? Not in South Dakota. Not in the country, where the bill polls at 71%. Inside the conference. Among the 27.

The third road is the bluntest: end the legislative filibuster outright, as Trump has repeatedly urged. Thune ruled it out in his maiden floor speech as leader, defending the 60-vote custom by describing the Senate as a counterbalance to the House, a deliberative body designed to check the majority. Take him at his word. That is the cleanest possible statement, in his own voice, that he ranks the institutional custom above the outcome his majority wants. The custom, recall, is the same one that protects every sitting member of his conference from ever having to cast a hard vote, which is to say, the custom is itself a service rendered to the 27.

Perhaps, the skeptical reader replies, the leader is a sincere institutionalist, and sincerity deserves respect even in error. The reply founders on the freshest evidence in the record. In May 2025, Senate Republicans rolled back a California electric vehicle standard despite the parliamentarian advising that the measure was subject to a 60-vote threshold. In September 2025, the same majority moved to overrule the chair and lower the threshold to a simple majority for the en bloc consideration of 48 nominees. NBC News, no friend of the right, reported that across 2025 the conference knocked down precedents and weakened minority power in 3 separate instances when its own ambitions demanded it. Same leader. Same conference. Same procedural tools. Deployed for an emissions waiver and a slate of nominations. Withheld for proof of citizenship at the ballot box. An institutionalist who suspends his institutionalism whenever the conference’s internal priorities require it is not an institutionalist. He is an agent, faithfully serving his actual principals.

And what of the country he is supposedly protecting from rash legislation? Gallup found in October 2024 that 84% of American adults support requiring photo identification to vote and 83% support requiring proof of citizenship for first-time registrants, with roughly two thirds of Democrats endorsing both. Pew Research put photo ID support at 83%, including 71% of Democrats. A February 2026 Harvard CAPS/Harris survey, an instrument co-directed by a former Democratic pollster, found 71% of registered voters supporting the SAVE America Act by name, with the underlying provisions polling between 75% and 81%. Heritage Action found strong majorities across 5 battleground states. In February 2026, Representative Brandon Gill led 34 members of the Republican Study Committee in a formal letter urging Senate action, putting 35 members of Congress on record that the votes and the public are aligned and only the schedule is missing. There is no constituency in America for burying this bill. There is only a constituency in one caucus room.

Thomas Jefferson, in his first inaugural address, declared the will of the majority to be the vital principle of republics, a will that must prevail to be rightful. The SAVE Act presents that principle in its purest modern form. A majority of the House has passed the bill twice. A functional majority of the Senate has voted for it on the record. A supermajority of the public, crossing party lines, supports its provisions. The President stands ready to sign it. Between all of that and the statute books stands one man’s discretionary choice, and that man’s career depends on keeping 27 people comfortable. This is not a conspiracy theory. It is a principal-agent problem, the most ordinary phenomenon in organizational life, operating exactly as the textbooks predict.

Which brings us to the remedy, because the same structure that created the problem contains its solution. The conference that hired Thune by secret ballot can replace him by secret ballot, and the precedent is recent. In November 2022, Rick Scott forced the first contested Senate Republican leadership vote since 1996. The challenge failed, but its existence proved the mechanism is live. The 27 votes that protect Thune are votes that can be withdrawn, and every senator who casts one now owns the burial of a bill his own voters support at 91%. In every other American institution, a chairman who refuses to call a vote on a majority-supported item is replaced by the board. The Senate Republican Conference is the board. The question is no longer whether Thune will move the SAVE Act. He has answered it 3 times, once for each road he declined. The question is whether 27 senators will keep paying him to say no.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.

There is an old rule in epistemology that deserves wider use in technology policy: when a man tells you a category of thing is obsolete, ask what he sells. The answer does not settle whether he is right. Salesmen are sometimes right. But it tells you where to apply pressure, because an argument that conveniently terminates in the arguer’s invoice should be examined at the joints rather than swallowed whole.

A genre of essay now circulates widely in conservative media making roughly the following claim. Data centers, the argument runs, are relics of an older computing era, built to run sluggish corporate software for batch reports and billing systems. Artificial intelligence, by contrast, is about fast decisions made where the data lives, on the battlefield, on the oil rig, in the vehicle. Centralized facilities can never deliver that speed. Therefore the hundreds of billions of dollars now flowing into American AI infrastructure are being poured into stranded assets, and the citizens fighting data center construction in their counties are not NIMBYs but prophets. The essays urging this view are frequently written by people whose companies sell the alternative, distributed edge software that, by remarkable coincidence, renders the entire data center industry unnecessary.

I want to take this argument seriously, because it contains a grain of truth, and arguments with grains of truth are more dangerous than arguments with none. The grain is this: a real and growing class of AI work genuinely belongs at the edge. When a drone must decide in milliseconds whether the object below is a tank or a tractor, it cannot wait for a round trip to a facility in Virginia. The industry knows this. Edge inference is a thriving field precisely because everyone serious already concedes the point. So far, the doomsayers and the builders agree.

The error enters through an equivocation, and it is worth walking through slowly, because once you see it the whole edifice collapses. The word “AI” is doing double duty. It names two profoundly different activities. The first is training, the process by which a frontier model is built. The second is inference, the process by which a finished model is used. The obsolescence argument quietly assumes that because inference can be distributed, training can be too. It cannot, and the reason is not a matter of opinion or vendor preference but of physics and arithmetic.

Training a frontier model requires tens of thousands of specialized chips wired together with high-bandwidth interconnects, computing synchronously on a single enormous problem. The chips must exchange results constantly, billions of times per second, which is why they must sit meters apart rather than miles. You cannot train a frontier model across a mesh of phones and desktop boxes any more than you can build an aircraft carrier in 10,000 backyard workshops and weld the pieces together at sea. The thing being built is unitary. Its construction is unitary. And here is the part the obsolescence school never mentions: every smart edge device runs a model that was trained somewhere first. The slick inference engine on the drone is downstream of a training cluster. Edge computing does not replace data centers. It depends on them, the way a paperback depends on a printing press. The relationship was never edge versus data centers. It was always edge plus data centers, a division of labor as natural as the one between the factory and the showroom.

Consider the favorite example of the obsolescence school, the self-driving car. The story goes that autonomous vehicles are failing because they must stream sensor data to a central hub and wait for instructions, a latency disaster. But this describes no autonomous vehicle actually on American roads. Waymo’s cars do their perception and planning onboard, at the edge, in real time, precisely because the engineers understood latency from the start. And the models making those onboard decisions were trained on vast centralized clusters, refined, and then deployed to the fleet. The example meant to prove that data centers are obsolete is in fact a textbook illustration of the complementary architecture, centralized training feeding distributed inference. When your marquee case proves your opponent’s thesis, the problem is not the marketing. The problem is the thesis.

Share

The speed claims deserve the same scrutiny. The obsolescence essayists boast of software that runs 1,000 or even 1,000,000 times faster by eliminating input/output bottlenecks, the wait states that leave a conventional server’s processor idle 95% of the time. Grant every word of it. The claim is about I/O-bound workloads, database queries, billing runs, record searches, the very legacy corporate computing the essayists correctly describe as the old world. But training a neural network is not I/O-bound. It is compute-bound, limited by raw mathematical operations, trillions upon trillions of them. No amount of I/O optimization conjures the floating-point operations needed to train a GPT-class model, just as no amount of streamlining the loading dock makes the factory floor produce more steel. The advertised speedup, however real, applies to a workload class that is not the one filling an AI data center. The argument refutes a machine that the buildout was never constructing.

Now ask the question I posed at the outset. Who benefits from the claim, framed as it is? The vendor of edge software benefits, obviously, since every county commission that blocks a data center is a future customer persuaded that the alternative is not merely viable but inevitable. But the beneficiary class is larger than any one firm. An entire ecosystem now profits from data center fear. Activist organizations raise money on it. Foreign-funded NGOs, which I have documented elsewhere, organize local opposition with it. Writers build subscriber bases on it. Consultants bill hours explaining it. None of this makes every claim in the genre false. Data centers do use water. They do draw power. But a claim can be locally true and globally misleading, and the framing here, that these facts add up to obsolescence and that opposing construction is therefore costless, is wildly misleading. The water story is the cleanest illustration: modern facilities increasingly use closed-loop or reclaimed-water cooling, and the aggregate draw is a rounding error beside agriculture. The Chinese have even sunk a seawater-cooled facility off Shanghai using zero fresh water, demonstrating that cooling is a solvable engineering problem rather than a civilizational limit. The honest framing is that siting matters. The dishonest framing is that the technology is inherently extractive.

The electricity story is, if anything, the reverse of the popular telling. The fear is that data centers will raise your power bill. But consider how utility economics actually work. The grid is a fixed-cost machine. The poles, wires, substations, and plants must be paid for whether demand grows or shrinks, and those costs are spread across every kilowatt-hour sold. After 2 decades of flat demand, in which ratepayers shouldered the full fixed cost of an aging system, large new industrial customers arrive willing to buy enormous quantities of power around the clock. More kilowatt-hours sold against the same fixed base means a lower cost per unit for everyone, which is why analysts at conservative institutions like the Heritage Foundation have long argued that demand growth, properly priced, is a ratepayer’s friend rather than his enemy. The economics resemble a church that has been splitting the building’s mortgage among 40 families and suddenly welcomes 100 more. The mortgage does not grow. The shares shrink. Moreover, the hyperscalers are not even waiting for the grid. Most now plan to build or contract their own generation, gas turbines, small modular nuclear, dedicated solar, behind-the-meter plants that never touch the residential system at all. A customer who brings his own power plant is not a burden on your bill. He is a co-investor in American generation capacity that the country needs regardless.

And the country does need it regardless, which brings us to the stakes the obsolescence school never weighs. The race that matters is not between two architectures. It is between two nations. China is building training compute at ferocious speed, backed by an electricity buildout that already dwarfs ours, and the layer of the AI stack that decides who leads, frontier model training, cannot be won at the edge by anyone, American or Chinese. It can only be won in large centralized clusters. If the United States internalized the doctrine that data centers are obsolete, if county by county we blocked construction on the advice of essayists selling the alternative, we would not be choosing a cleverer architecture. We would be unilaterally conceding the decisive layer of the most important technological competition since the Manhattan Project, while telling ourselves we had outsmarted everyone. Beijing could not purchase better propaganda. It is getting it free, dressed in the language of property values and farmland.

So let us keep two thoughts in our heads at once, since the truth requires both. Edge computing is real, valuable, and growing, and the men who sell it deserve their market. And the claim that its rise renders centralized AI infrastructure obsolete is false, refuted by the dependence of every edge device on centrally trained models, by the compute-bound nature of training itself, and by the capital now flooding into the very facilities the theory declared stranded. When someone tells you America should stop building the infrastructure of the AI age, check what he sells, check who funds the fear, and then check the only scoreboard that matters, the one with China on it.

If you enjoy my work, please subscribe https://x.com/amuse/creator-subscriptions/subscribe

Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes daily political analysis at amuseonx.com. Primary sources cited in this piece are linked inline; campaign finance figures are drawn from FEC filings, polling data from publicly released crosstabs, and legal claims from filed pleadings. Corrections are posted to the original URL with a dated changelog. Readers who identify errors are invited to contact the author directly.