On June 30, 2026, the Supreme Court struck down the last federal limit on what a political party may spend in coordination with its own candidates. The vote was 6 to 3. The case was National Republican Senatorial Committee v. Federal Election Commission, and the most remarkable fact about it was not the outcome.

It was that the United States government agreed with the plaintiffs. The Justice Department told the Court that the 2001 precedent upholding those limits, Colorado II, “no longer retains vitality,” and declined to defend a statute it is charged with enforcing. When the government stops defending its own campaign finance law, the law is not being narrowed. It is being vacated.

That ruling is the reason this piece exists. For sixteen years the argument over Citizens United has been conducted as though the next election, or the next statute, or the next justice would settle it. June settled something else: the direction of travel. The Court has now removed the corporate expenditure ban, the aggregate contribution limits, the party coordination limits, and the public financing trigger that once answered a wealthy self-funder. Each decision narrowed the space in which a legislature may act. None of them restored any.

So the real question is not whether Citizens United should be fixed. It is which fix can survive the institution that produced it. There are three routes, and only one of them the Court cannot undo.

What the decision actually held

Precision matters here, because most of the shouting does not have it. Citizens United did not legalize corporate checks to candidates. The federal ban on direct corporate and union contributions to candidates survived the case and survives today, with the individual limit set at $7,000 per candidate per cycle. What the Court struck was the ban on independent expenditures: money spent by a corporation or union to advocate for or against a candidate, without coordinating with that candidate’s campaign. It upheld the disclosure and disclaimer requirements by 8 to 1. The famous sentence from the majority is that independent expenditures “do not give rise to corruption or the appearance of corruption.”

Two months later, the D.C. Circuit took the next step in SpeechNow.org v. FEC, holding that if independent spending cannot corrupt, then contributions to a committee that only makes independent expenditures cannot corrupt either. That decision created the super PAC.

So the structure since 2010 is this: a candidate may take $7,000 from a donor. A committee that does not talk to the candidate may take unlimited sums from the same donor, spend them on the same race, and disclose in aggregate while the original source stays hidden behind a nonprofit. The limit is real on one side of a wall and absent on the other, and the wall is a coordination rule that nobody enforces because the agency that enforces it has no quorum.

That is the machine. Fixing it means changing one of three things: the constitutional rule, the enforcement, or the corporate form.

The ruling that changed the arithmetic

NRSC matters because it closed the last seam. Before June, a party committee could spend only a capped amount in concert with its candidate, which meant a large donor who wanted to help that candidate faced a ceiling on the coordinated route. Justice Kavanaugh’s majority held that the cap was no longer “necessary” or “narrowly tailored,” because Congress has less restrictive tools: earmarking rules and disclosure.

Read that reasoning twice, because it is also a roadmap. The Court did not say Congress is powerless. It said Congress must use disclosure and earmarking rather than caps. Justice Kagan, dissenting with Justices Sotomayor and Jackson, called what remained “a remnant of a remnant.”

The practical effect is that a party committee and a candidate’s campaign can now plan and pay for the same advertising together, without limit. The $7,000 contribution limit still exists on paper. The route around it is now a paved road.

Route one: an amendment

Article V sets the arithmetic: two thirds of both chambers, which is 290 representatives and 67 senators, then ratification by three fourths of the states, which is 38. There is no presidential signature. There is no filibuster in the ratification count.

The idea is not fringe and it is not new. Twenty-five state legislatures have passed resolutions calling on Congress to propose such an amendment, the twenty-fifth being Idaho on April 8, 2026. Those states are not a bloc: they include Massachusetts and Montana, Alaska and Nevada, West Virginia and California. In Congress, the current vehicles span the ideological range within the Democratic caucus and reach into the Republican one:

  • H.J.Res. 191, introduced by Representative Tom Barrett, Republican of Michigan, on June 3, 2026. Its operative section is one sentence: nothing in the Constitution forbids Congress or the states from “reasonably regulating and limiting contributions to and spending in campaigns,” including by “distinguishing between natural persons and artificial entities and prohibiting artificial entities from raising and spending money.”
  • H.J.Res. 122 and S.J.Res. 78, introduced in September 2025 by Representative Joe Neguse and Senator Adam Schiff, which also authorize public financing systems and viewpoint-neutral limits.
  • H.J.Res. 119, from Representative Jim McGovern, which is the prescriptive version: it writes specific limits, a corporate spending ban, and a public financing requirement into the text.
  • H.J.Res. 221, from Representative Hillary Scholten on November 2, 2026, which focuses on contributions and paid advertising and adds an express bar on foreign money.

Those are not the same amendment, and the difference is the whole game. A structural amendment, the Barrett and Neguse-Schiff shape, restores authority: it says elected lawmakers may set the rules, and it leaves the rules to ordinary politics. A prescriptive amendment writes the rules into the Constitution, which means every future disagreement about a contribution limit becomes a constitutional question litigated for a generation. The first can be ratified and then improved by statute. The second freezes one Congress’s judgment at the highest possible level, and hands its interpretation to the same Court that has been reading the First Amendment expansively for fifty years.

If the goal is a durable fix, the structural text is the one to pass. It is also the one with the broader coalition, which is not a coincidence.

Route two: what the Court left standing

Any honest account has to say plainly that an amendment will not arrive soon. Twenty-five states is a real achievement and thirteen short of the threshold. The congressional count has never been close to 290. So the near-term work is statutory, and the Court has been unusually explicit about what it will accept.

Disclosure. The one part of Citizens United that survived 8 to 1 was the requirement that spenders identify themselves. The Court has repeatedly described disclosure as the less restrictive alternative. Alaska’s “true source” rules, adopted by ballot measure in 2020, require the original funder behind a chain of political entities to be named, which is why reporters could trace $1.5 million from Elon Musk into the state’s 2026 repeal campaign. A federal version would do what the FEC cannot: publish the chain. Public support is not in doubt. The Brennan Center’s survey of 2,000 registered voters, fielded April 28 to May 6, 2026, found 85 percent support for mandatory disclosure of all federal campaign contributions and spending, including online ads, with 85 percent of Republicans and 88 percent of Democrats agreeing.

Foreign money. In July 2026, Representative Jamie Raskin and Senator Sheldon Whitehouse introduced the Get Foreign Money Out of U.S. Elections Act, aimed at spending by domestic companies with substantial foreign ownership. It drew 69 House and 12 Senate Democratic cosponsors. Whatever its prospects, the underlying authority is not in serious constitutional doubt, because the Court has never protected foreign nationals’ election spending.

The referee. None of this works without an agency. The Federal Election Commission lost its quorum on May 1, 2025 and has not regained it. It cannot open investigations, issue fines, write rules, or give advice. Its enforcement backlog doubled from 154 cases to 307. A disclosure statute that no one receives and publishes is a press release with a filing deadline.

Small-donor incentives. These do not restrict speech, so they survive the doctrine. New York City matches small contributions from residents; the Campaign Finance Board reports that a $10 gift can be worth $90 to a participating campaign, and that more than two thirds of contributions in the 2013 citywide elections were $175 or less. A federal tax credit for small contributions polls at 67 percent nationally, including 63 percent of Republicans. It is the one campaign finance policy that both parties can support without touching the First Amendment.

Route three: the states, and the corporate charter

The most interesting development of 2026 is a legal theory that does not ask the Court to overrule anything.

A corporation exists because a state says it does, and the state decides which powers it grants. Hawaii’s Act 11, signed May 14, 2026, exercises that premise directly: it provides that the powers the state grants to corporations, LLCs, nonprofits, and similar entities do not include spending money to influence candidate elections or ballot measures. Not a speech restriction. A limit on the powers conferred. The law takes effect July 1, 2027, and is already in federal court twice, once from the Grassroot Institute of Hawaii and once from the U.S. Chamber of Commerce, which argues that the state cannot “decide who gets to participate in public debate.”

Montana took the same idea to the voters. Constitutional Initiative 194, the “Montana Plan,” prohibits “artificial persons” from contributing to campaigns, ballot measures, or parties, and it passed 65.5 percent to 34.5 percent. Michigan’s Proposal 2, which bars regulated utilities and contractors holding more than $250,000 in state work from contributing, passed 81.3 percent to 18.7 percent and extends disclosure to internet political communications. Alaska’s Ballot Measure 1, in August, capped individual contributions at $2,000 per candidate per cycle, and passed with 72.2 percent.

Ballotpedia counts seven statewide campaign finance measures across six states in 2026, tying 2018 for the most since 2000. The state route is where the energy is, because it is the only route that does not require a single member of Congress to vote for anything.

And there is a fourth lever that requires no government at all. Shareholders. The SEC stopped issuing substantive guidance on whether companies may exclude shareholder proposals, and 2026’s proxy season shows what replaced it: 29 political disclosure proposals filed, 10 companies agreeing to implement them, 12 going to a vote at an average of 31.7 percent support, down from 41 percent a year earlier amid a harder environment. Six companies excluded proposals unilaterally. Meanwhile the CPA-Zicklin Index finds 391 of the S&P 500 now fully or partially disclose or prohibit at least one type of political spending, against 243 in 2015, and 328 have board oversight of it, against 214. Private ordering is slower and it is not a substitute for law. It is the only lever that is working right now.

Why it is essential

Strip away the decorum argument, which is real but secondary. The case for action is empirical, and it is about whose preferences policy actually tracks.

The best-known test is Martin Gilens and Benjamin Page’s 2014 study, which assembled 1,779 policy questions from 1981 to 2002 and modeled the independent influence of average citizens, economic elites, and organized groups. Their finding: economic elites and business-oriented groups have substantial independent effects on federal policy, while average citizens and mass-based groups have “little or no independent influence.” The authors are careful about their measures, and later work has argued about the size of the effect. But the direction has not been overturned, and it is consistent with what the spending data show.

The spending data, as of this week: $11.9 billion on the 2026 midterms, $3.9 billion of it outside spending, $1.47 billion of it dark money that cannot be traced to a person, and $445 million from a single donor. In a country of 340 million people, one man gave $445 million to federal committees. That is not a scandal in the sense of a bribe. It is worse and duller than a bribe: it is a standing market in access, in which the price of a meeting is known and the price of a vote is not.

Three things follow, and none of them is about manners.

Policy stops responding to majorities. When the marginal dollar is more decisive than the marginal voter, legislators optimize for the dollar. Gilens and Page’s near-zero coefficient for average citizens is what that optimization looks like when you measure the output.

Voters notice, and the noticing is corrosive. Pew’s April 2026 survey found 74 percent 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 Brennan Center’s survey found 92 percent say corruption is a significant problem in politics and government. This is the one issue on which the electorate is not polarized. It is polarized about remedies.

The Court is not finished. Every limit struck invites the next challenge, and the challenges are being funded. NRSC was not a case about corporations. It was a case about whether any coordinated spending limit can survive, and the answer was no. A statutory fix passed in 2027 faces the same six justices in 2029.

That is the argument for the amendment, and it is an argument about durability rather than purity. A statute is a hypothesis the Court may reject. A clause is a fact the Court must apply.

The objections that are right

Three serious objections deserve answers rather than dismissals, and the answers shape which version to pass.

The civil society objection. Hawaii’s Act 11 and Montana’s I-194 do not distinguish between a corporation and a nonprofit, and the ACLU filed an amicus brief calling Hawaii’s law “premised on an untenable theory of plenary government power over the speech rights of civil society organizations.” That is not a frivolous complaint. A rule that strips every incorporated entity of election speech would silence the Sierra Club, the NRA, a local housing coalition, and a university press alongside Exxon. Some of those organizations are the counterweight to corporate money, not an extension of it. An amendment that enables such a ban, without distinguishing membership organizations from business corporations, buys a real gain at a real cost. The structural text leaves that line drawing to legislatures, where it can be drawn and redrawn. The prescriptive text freezes it wrong.

The incumbent-protection objection. Contribution caps protect whoever is already in office, because a challenger needs money to introduce themselves and an incumbent needs nothing. This is the strongest argument the Institute for Free Speech makes, and it is correct as far as it goes. It is also why the package should lead with small-donor matching and disclosure rather than with caps. Matching funds help challengers. Caps help incumbents. A reform movement that only passes caps has passed incumbent protection with a good conscience.

The amendability objection. The most candid version came from American Promise’s own counsel, who named the single biggest obstacle as the widespread belief, particularly among the professional class, that the Constitution is no longer amendable at all. The record justifies some of that: 27 amendments in 235 years, the last in 1992, and six proposed amendments that died in ratification. But the same record shows amendments arriving in clusters after long agitation: four between 1961 and 1971. The states that have passed resolutions are not performing a gesture. They are building the ratification map that a future Congress would need.

The order of operations

If the goal is a fix the Court cannot undo, the sequence is not mysterious, and it is not quick.

This Congress. Pass disclosure. Confirm commissioners until the FEC can vote. Take up the foreign-money bill. None of this requires an amendment and all of it is constitutional under existing doctrine, which the Court has said twice.

The states, in 2027 and 2028. Keep passing resolutions, and keep passing ballot measures, with drafting that distinguishes a business corporation from a membership organization. Hawaii’s case will be decided first, and its outcome will tell every legislature whether the corporate-charter theory is a route or a detour.

Shareholders, continuously. The proxy season is the only venue where the question is asked of the spender directly, and the trend line is up even as the votes soften.

Congress, eventually. A structural amendment, introduced every session, reported out of committee once, and passed when the coalition is 290 and 67 rather than before. The point of introducing it now is not to pass it now. It is to make it ordinary, because an idea that is never introduced is never ratified.

The alternative is what the country has been doing since 2010: passing reforms at the state level, watching them litigated, winning some, and losing the ones that matter most. Montana voters approved a corporate spending ban by 31 points. It will be in federal court within the year. Hawaii’s law was in court twice before it took effect. Michigan’s utilities ban passed by 62 points and will be challenged on the same First Amendment grounds.

Those votes are not nothing. They are the clearest evidence available that the public has already decided this question, and has been overruled. A country that keeps voting for a rule it is not allowed to have is not a country with a policy disagreement. It is a country with a broken constitution, and the only repair for that is the one the Constitution itself provides.

The previous essay in this series, on campaign finance reform, ranked-choice voting, and open primaries, is Who Chooses the Choices.

Sources

PRH | huffmanwrites.org | © Philip Huffman