The measure of any reform proposal is twofold. Is it a good idea, and is it possible?

Drutman’s call to restore fusion is both. It is a good idea because it is a party-centric reform with real potential to leverage meaningful democratic returns. We must reconcile ourselves to this core idea: “Modern representative democracy is party democracy.” There is no way out of the party system, but this does not mean that it couldn’t be better. Our contemporary party system is broken both because it is nationalized and because it only has space for two sides. Fusion politics offers a promising path out of this “doom loop.” When third parties can only act as spoilers, their electoral appeal is limited, and they have no incentive to be serious. Allow them a real seat at the table and they will do the associative work of a party, connecting citizens to the government, building power, and delivering legislative goods by negotiating deals.

Restoring fusion is also a refreshingly modest reform. Unlike proposals to abolish the Electoral College or permit the regulation of money in politics, it does not require a constitutional amendment. It also does not require Americans to look to Congress, and no one will need to be educated about how votes are counted, or seats apportioned in proportional systems. Perhaps most importantly, a narrow path to the relegalization of fusion exists. Restoring fusion politics requires only minimal amendments to existing state election laws. The two most significant obstacles are legislative self-interest and the Supreme Court’s 1997 decision in Timmons v. Twin Cities Area New Party. Consider each in turn.

All reform proposals must account for the fact that current officeholders are just not reliable partners when it comes to meaningful democratic reforms; their first instinct is to fortify their political power. Anti-fusion laws are a case in point. The very first anti-fusion law, passed in South Dakota in 1893, was a Republican-led legislative effort. Seeking to thwart the electoral threat posed by political alliances between Democrats and Populists, Republican legislatures around the country soon followed suit. State legislatures, controlled by the two major parties, will have little incentive to reinstitute fusion.

Restoring fusion politics requires only minimal amendments to existing state election laws.

There have been important exceptions—occasions when partisan incentives have aligned with the interest of democratic reform. In 2007 the Democratic-led Connecticut legislature amended the state’s cross-nomination rules to make fusion functionally usable. Before the change, the state allowed only “major parties” (a legal status) to fuse, imposing cumbersome rules for “minor parties” seeking to do the same. The rules were relaxed largely in response to the emergence of the Working Families Party (WFP), despite the cumbersome rules. The WFP had the support of many of the state’s largest labor unions and generally preferred to cross-endorse major party candidates who embraced its policy goals. In Connecticut’s competitive districts, candidates cross-endorsed by the WFP were overwhelmingly Democrats.

One result was that the WFP was able to cross-endorse Democrat Dan Malloy in the 2010 governor’s race. The votes Malloy received on the WFP line far exceeded his narrow margin of victory. Malloy strongly supported the WFP’s top legislative priority—paid sick days—and in 2011 Connecticut became the first state to guarantee paid sick days to workers.

Trump’s bid for a second term may create a similar political opening; the interests of the old guard of the Republican Party and the Democrats may converge to reintroduce fusion legislatively. Alternatively, one party may decide to embrace fusion in response to evidence suggesting the real possibility of partisan realignment.

But I wouldn’t bet on it. The more likely scenario is that of Minnesota. In 1994, the newly formed Twin Cities Area New Party, a coalition of labor unions and community organizers, sought to repeal Minnesota’s ban on fusion with the help of its ally and future nominee, Andy Dawkins, a much-loved incumbent state representative. Representing the Democratic-Farmer-Labor Party (DFL), Dawkins sponsored a bill to relegalize fusion and introduce weekend voting, but the bill was killed in committee. Incumbents from the two major parties worried that striking the state’s ban on fusion would weaken their hold on power.

The classic solution to this conundrum is to turn to the courts. As Justice Harlan Stone suggested in 1938, the power of judicial review is precisely suited to situations where the law makes it difficult to “repeal . . . undesirable legislation” or oust unpopular elected officials. The Warren Court fully embraced this idea, viewing the Court’s primary role in our system of checks and balances as reinforcing the openness of the democratic process.

In 1996 the Twin Cities Area New Party tried to make such an appeal to the courts. The case arose when the party formally nominated Dawkins, who represented a multiracial working-class district that benefitted little from Bill Clinton’s Third Way policies and was keen on the cross-nomination. The secretary of state of Minnesota rejected it. The New Party sued, confident that the federal courts would strike down Minnesota’s ban on fusion candidacies as an infringement on a fair political process. It lost.

In a widely criticized decision, a divided Supreme Court upheld Minnesota’s anti-fusion law, maintaining that it did not severely burden the New Party’s First Amendment rights. The party, it insisted, remained free to nominate an alternative, eligible candidate for the office or to electioneer on behalf of its preferred candidate under the DFL banner. The Court resoundingly rejected the New Party’s argument that “Minnesota’s interest in maintaining a stable political system . . . [did] not give the state license to frustrate consensual political alliances.” Instead, it asserted that whatever burdens the law might place on the New Party were justified by the Minnesota legislature’s view that “political stability is best served through a healthy two-party system.”

The ruling remains a significant obstacle to a federal judicial path to relegalizing fusion. But it does not bind state courts, and many state constitutions provide broad protections for free and fair elections. Advocates of fusion in recent years have seized this narrow opportunity and instituted a state constitutional litigation strategy.

The first of these suits was filed in New Jersey. It makes an affirmative state constitutional case for striking down New Jersey’s, particularly burdensome law while also offering numerous reasons to reject the U.S. Supreme Court’s reasoning in Timmons. (Like several other legal academics, I participated in the New Jersey case as an amicus, highlighting errors in the Timmons ruling’s First Amendment analysis.) These state constitutional lawsuits are promising and likely critical to a successful effort to reintroduce fusion.

Political change is difficult. It requires reformers to find and exploit limited opportunities within an array of structural constraints—political and legal. It should surprise no one that the opportunities for revitalizing fusion politics are severely constrained. However narrow the path, the critical point is that the prospects for reviving fusion politics are far more promising than many of the alternatives currently under consideration.