The scoreboard is extraordinary, and it is being read backwards.

Elon Musk put $445 million into federal races this cycle, according to OpenSecrets, the largest sum any individual has spent on an American election. Of the 43 Republican House candidates he backed, three won and 33 lost, with seven races still uncalled when Reuters tallied it. Of the seven Senate candidates, six lost and the seventh was trailing. In Texas, where he put almost $230 million into the Senate race alone and another $5.5 million into the races below it, every candidate he backed lost. WIRED’s accounting found that of every traceable dollar he put into a decided race, more than 99 cents went to a race he lost. On election night he posted, “Guess my election-interfering space lasers didn’t work this time.”

The conclusion being drawn from that is that money does not buy elections. It is the wrong conclusion, and it is wrong in a way that will keep the problem in place. Money did not buy these elections. It rented the terms on which they were fought, and it has already bought something that will outlast Tuesday: a demonstrated willingness to spend, at a price everyone in Washington can now quote.

What the filings show

Start with the plumbing, because it is more instructive than the result.

Texas PAC, the super PAC supporting Ken Paxton, filed a quarterly report showing exactly $158,000,000 in receipts between July and September. Its itemized Schedule A, the part that names donors, was empty when I queried the Federal Election Commission’s data on November 5. The summary reports the money; the filing does not yet say who sent it. Reporting has described the sum as Musk’s money routed through the Senate Leadership Fund, and the figure’s precision fits that description. But as of this week the public record shows a committee receiving a round $158 million from somewhere, and spending $181.9 million in independent expenditures through October 26, $112.2 million of it against James Talarico and $69.7 million for Paxton.

That is the structure the Supreme Court built and the FEC can no longer police: money enters a committee, leaves it the same size, and arrives in a race with the payer’s name temporarily missing. The post-general reports, due December 3, will fill in more. They will not fill in the dark money that never reaches a filing at all.

The totals will move too. OpenSecrets’ $445 million covers federal races and is a projection completed before the last filings land; the Guardian’s tally is $446 million; counts that include state races run higher. Any number published this week is provisional, and the piece that quotes one without saying so is the piece to distrust.

The three arguments that a loss refutes, and the one it does not

Argument one: he tried to buy a seat, and the price was refused. True as far as it goes, and it is the version everyone is repeating. Talarico answered one Musk post with a line that will outlive the cycle: “There is something deeply wrong with a political system that allows one trillionaire to buy his own personal US Senator.” Texas voters then rejected the purchase by twelve points, 55.1 to 43.0, in the most expensive non-presidential race in American history, at roughly $573 million. Talarico put it plainly from the stage: “We just proved that all the money in the world is no match for the people of Texas.”

Argument two: the money was wasted, so he will spend less next time. This is a prediction, not an observation, and the evidence points the other way. Musk spent nearly $290 million in 2024 and then spent more in a midterm, a cycle with no presidential race to justify it. The record across two cycles is not a man learning that spending fails. It is a man learning that spending is affordable.

Argument three: the electorate is resilient, and the immune response worked. This is the comforting reading, and it mistakes a backlash for a safeguard. Voters rejected Musk in part because he made himself the issue, which is not a rule on which the republic can rely. A system whose defense against concentrated money is that the moneyman overplayed his hand has no defense at all; it has a lucky streak.

The argument a loss does not refute is the one that matters: the spending changes the election whether or not it changes the winner.

Five ways it worked anyway

It set the agenda. A candidate who faces $112 million in attack ads does not get to run the campaign he planned. He spends his time answering. Talarico’s own closing argument was about the man spending against him, which means the money chose the subject of the election even as it lost the vote. That is not nothing. That is the most valuable thing money can do.

It shaped the field before anyone voted. In Kentucky, Musk put $10 million into a group formed to win the Republican nomination for Mitch McConnell’s open seat. The group spent $9.6 million promoting Nate Morris, who quit two weeks before the primary, and $3.8 million attacking Andy Barr, who won the seat by 12 points. The money did not decide the primary. It did decide which primary the state had, and it taught every Kentucky Republican what a challenge from that direction costs.

It bought a threat, and the threat is now public. In the weeks before the election, Musk wrote: “Those who pick fights with me should understand that I will find out who is really pulling the strings and come for them, not just their puppets.” Read that as a statement about the next Congress rather than about this one. A senator weighing a subpoena into SpaceX contracts now knows the size of the response, because the size was advertised at $445 million and settled at a price per race. The deterrent does not require a win. It requires only that the willingness be believed.

It proved the price is payable. No individual had spent at this scale in a single cycle before. The relevant fact for the next wealthy person is not that Musk lost. It is that one man could credibly bid $445 million on a midterm and treat the loss as a bad quarter. The ceiling has moved, and ceilings do not move back.

It bought access to the aftermath. Musk’s companies have large and growing federal contracts, and a Democratic Congress has already said it will look at them. Representative Robert Garcia, the incoming chair of the House Oversight Committee, plans to introduce legislation closing what he calls the “Musk loophole,” the provision that lets an executive spend unlimited personal money on politics while his companies may not. The spending did not buy the election. It bought a fight, and the fight is now the point.

The asymmetry is the problem, not the outcome

The instinct to treat this as a story about one man’s bad investment is itself a symptom. It accepts the donor’s own accounting, in which an election is a position and a loss is a write-down.

Consider the arithmetic from the other side. $445 million is a rounding error against a personal fortune measured in the hundreds of billions, which means the expected value of the bet does not depend on winning any particular race. It depends on establishing, publicly and durably, that the money will be there. That is a rational investment even at a 10 percent success rate, and it is an investment no collection of ordinary citizens can match, because ordinary citizens do not have a spare half billion to lose.

He was on the losing side of Alaska’s ballot measure too, giving $1.5 million to the network that principally funded the repeal of the state’s election system, which trailed 51.9 to 48.1 with nearly every precinct reporting.

The 2026 cycle shows the same pattern on the other side of the aisle. Michael Bloomberg gave $80 million to Democrats in the closing weeks, and the forty largest donors gave $2.4 billion by mid-October. Two billionaires and forty families set the terms of a national election. The fact that one of them lost is not evidence that the system works. It is evidence that the system is expensive. Pew found in April that 74 percent of Americans call the role of money in politics a very big problem, the top item among thirteen, with 70 percent of Republicans and 79 percent of Democrats agreeing. The public has already reached the conclusion against which the scoreboard is being offered.

And the cost is not only financial. It is the cost of a politics in which the central question of a Senate race becomes whether one man may buy it. Texas spent $573 million to answer that question, and the answer took until 11 p.m. Every dollar of it was speech in the constitutional sense, and almost none of it was speech by anyone the candidates represented.

What the loss does prove

One thing, and it is worth stating plainly because it is the strongest fact in the piece: the electorate is not for sale at this price, in this year, on these candidates. That is real. It is also contingent on a candidate quality problem Musk created for himself. Paxton carried a decade-old securities indictment and an impeachment acquittal into the race. The Texas money was spent on a flawed vehicle, and some of the loss belongs to the vehicle.

Which is why the remedy cannot be “hope the next billionaire picks badly.”

The narrow, achievable fixes are already on the table and they are the same three from this week’s companion piece: disclosure that names the original funder, an FEC with a quorum to receive it, and a rule that closes the executive loophole so a company’s money and its chief executive’s money face the same treatment. None of them silences anyone. All of them make the price of the next $445 million visible before the ads run rather than in a filing due three weeks after the votes.

The longer fix is the one that takes a country: an amendment restoring to Congress and the states the authority to set limits, so that the question “how much may one person spend on an election” has an answer other than “whatever he has.”

The joke is the confession

Musk’s election-night post was a joke about space lasers. It was also accurate about what he had done, in the way jokes often are: he had interfered, at scale, with a national election, and it had not worked. The country’s good fortune was that the interference failed.

Good fortune is not a constitutional principle. A republic that responds to the largest political expenditure in its history by noting that the spender lost has confused an outcome with a safeguard. The money did not buy the Senate. It bought the terms of the argument, a demonstrated threat, and a new ceiling on what one person may spend. Those purchases do not expire because the candidates did.

Earlier this week: A Fix the Court Cannot Undo and Who Chooses the Choices. On the man himself, from July: Elon Musk and the Engineering of Chaos.

Sources

PRH | huffmanwrites.org | © Philip Huffman