Fusion or Super-Fusion?
What Institutional Reform Would Most Depolarize Politics?
The stresses confronting American democracy increase almost every day, it seems. An assassination attempt upon a leading presidential candidate, less than a week before that candidate officially became his party’s nominee at the party’s convention. Then, just a few days later, the presumptive nominee of the other major party dropped out of the race, because of mounting concerns within his party that he lacked the capacity to wage a vigorous campaign, as evidenced by his seemingly failing health. This is all occurring in the midst of an election that has been described by both sides as “existential,” meaning that each side feels that the Republic will be lost if the other side prevails.
In the context of this national trauma, it is worthwhile to continue considering—as Common Ground Democracyhas done all year—what alternative electoral institutions would better serve the nation as it navigates these increasingly polarized and perilous times. Here, Common Ground Democracy examines a proposed reform, “fusion voting,” that is gaining considerable traction among policy analysts. The essay evaluates this proposal on its own terms and compares it to other options aimed at achieving the same objectives.
The Surging Interest in Fusion Voting
To say that support for fusion voting is on the rise is an understatement. It’s been the subject of effusive praise in various opinion columns. Earlier this month, a large group of prominent scholars published an “open letter” advocating a return to the practice of fusion voting, which permits candidates to appear on a general election ballot as the nominee of more than one political party. For example, in New York, where fusion voting still exists, ten years ago in the governor’s race Andrew Cuomo appeared on the ballot as the nominee of four different parties (Democratic, Working Families, Independence, and Women’s Equality), while his main gubernatorial opponent was on the ballot as the nominee of three different parties (Republican, Conservative, and Stop Common Core).
Fusion voting is a relic of nineteenth-century electoral practices, when political parties rather than the government printed the ballots (or “tickets”) for voters to cast. In 1872, for instance, Horace Greeley was the presidential nominee of both the Democratic Party and a newly formed Liberal Republican Party, which splintered from the Republican Party because of disappointment that President Grant’s administration had become too corrupt. Consequently, the Liberal Republican Party separately from the Democrats printed its own tickets for voters to cast.
During the twentieth century, after American states adopted the so-called “Australian ballot,” whereby the government rather than the party prints the ballots that voters cast, fusion voting died out except in the very few states—like New York—that continued the practice by permitting candidates to appear on the government-printed ballot multiple times as the nominee of different political parties. In Timmons v. Twin Cities Area New Party, 520 U.S. 351 (1997), the U.S. Supreme Court upheld the constitutionality of state laws that limit candidates to appearing only once on the state’s ballot, thus requiring candidates to choose which party’s nomination they want listed in this single appearance on the ballot even if the candidate has received the nomination of more than one party. After Timmons, states are permitted to use ballots of the form that New York has, but they are not required by the federal Constitution to do so.
The scholars who now advocate for the return of fusion voting are essentially asking states to enact laws like New York’s. Their argument is that NY-style fusion voting will be an effective antidote to the increasing hyper-polarization of American politics. They claim that if laws permit candidates to be listed as the nominees of multiple parties, a centrist or moderate party will form that will act as a powerful incentive to each of the two major parties to depolarize or moderate their messaging, so that their nominees can earn the nomination of this pivotal moderate party. As the scholars put it in their open letter: “In the current moment of divisive polarizing politics, fusion voting would be particularly empowering for the pivotal but increasingly homeless political middle.” Specifically, they invoke “the many ‘constitutional Republicans’ who can’t bear voting on the Democratic Party ballot line but also cannot bear voting for Donald Trump.”
As readers of Common Ground Democracy know, I very much share the goal of depolarizing American politics. I very much want states to adopt electoral reforms that will have a “centripetal” effect of causing electoral competition to return towards the center of the electorate, rather than pulling candidates more and more to partisan extremes. A well-designed electoral system should favor candidates with the broadest range of appeal within the electorate, rather than divisive candidates with intense but narrower support.
Reasons for Skepticism about Fusion Voting
Despite sharing these goals with advocates of fusion voting, I doubt that fusion voting can be an effective means of achieving these goals. I don’t see fusion voting working in New York in the way that its advocates describe, and I don’t envision fusion voting operating elsewhere in the manner that they hope. In particular, it would be unlikely to make a difference in saving the nation from the predicament caused by Donald Trump’s candidacy.
Currently, there is no centrist or moderate party in New York for the “homeless political middle,” despite the existence of fusion voting in the state. An “Independence Party of New York,” formed at the time of Ross Perot’s proto-populist bid for the presidency in the 1990s, arguably attempted to occupy the political center in the state for several decades, but it lost its status as a political party under New York law as a result of legislative changes that tightened eligibility requirements for ballot access. Since then, no new party in New York has organized to represent the political middle. Accordingly, the only additional parties that participate in New York’s fusion voting system are further from the center than the two major parties: the Working Family Party, historically to the left of the Democrats, and the Conservative Party, historically to the right of Republicans. Thus, the current practice of fusion voting in New York tends to further polarize, rather than depolarize, electoral competition in state and local races—as the nominees of the two major parties seek to appeal to the two more extreme parties in order to secure their nominations as well.
An Ohio Example
Looking at my own state of Ohio, I’m skeptical that fusion voting would make a significant difference in the current electoral competition between Democrats, occupying the left side of the partisan divide, and MAGA-dominated Republicans, currently veering sharply to the far right. Take the 2022 race for the state’s U.S. Senate seat, after the more traditionally conservative Republican incumbent, Rob Portman, decided not to run for reelection in a MAGA-dominated partisan primary. The race ended up being between Tim Ryan, a moderate Democrat from blue-collar Youngstown in the mold of Senator Sherrod Brown, and J.D. Vance, who rebranded himself as a MAGA enthusiast in order to secure former president Trump’s nomination (and who, as we all know, has since become Trump’s running mate). In the November general election, Vance walloped Ryan, 53% to 47%.
Would it have made any difference if Ohio had adopted fusion voting for the 2022 election and some sort of centrist party had formed in the state and put Ryan’s name on the ballot as its nominee in addition to his being on the ballot separately as the Democratic Party’s nominee? I very much doubt it. Even without fusion voting, Ryan won the votes of independents and Republicans who would have voted for a traditional GOP candidate, like Portman, but who refused to vote for an authoritarian-sounding extremist like Vance. We know this because on the same November 2022 ballot as the Ryan-Vance Senate race was the gubernatorial race between the non-MAGA traditional Republican incumbent, Mike DeWine, and his moderate Democratic challenger, Nan Whaley, then mayor of Dayton. DeWine obliterated Whaley, 62.5% to 37.5%—a 25-point margin, in comparison to Vance’s impressive 6-point victory over Ryan.
There were, in other words, plenty of independent and more moderate Republican voters who happily voted for DeWine but, when it came to the choice between Vance and Ryan, couldn’t stomach Vance and cast their ballots for Ryan. There just weren’t enough of these crossover voters to prevent the election of the extremist Vance. Would Ryan have picked up a few more crossover voters if he had been on the ballot twice, as both a Moderate and a Democrat? Maybe. But certainly not nearly enough to overcome his 6-point deficit against Vance. Ryan is widely credited with running about as good a campaign as he possibly could have, and he essentially maxed out the percentage of Republican voters willing to vote for a Democrat when their own party’s nominee has become too extreme.
Ohio’s statewide races in 2022—for attorney general, secretary of state, treasurer and auditor—confirm this point. Like the gubernatorial election, but unlike the U.S. Senate race, they were all roughly in the 60%-40% range. Ryan did better than his fellow Democrats because, even without fusion voting, Republican and independent voters who cast their ballots for all these other GOP candidates (but were unwilling to vote for ultra-extreme J.D. Vance) didn’t just abstain from the Senate race; rather, they affirmatively voted for Ryan. In short, the availability of fusion voting would not have changed this dynamic significantly because the voters it would have appealed to voted for Ryan without it.
This year’s presidential election
If we turn to the current presidential race, it seems unlikely that fusion voting would make a difference there. Who are the voters in battleground states—Michigan, Pennsylvania, Wisconsin (and Nebraska’s second congressional district)—who would cast their ballots for Kamala Harris if she were the nominee of a newly formed centrist party as well as the Democratic nominee, but would not vote for her otherwise? There is already a massive effort underway to get “constitutional Republicans”—like Liz Cheney and Mitt Romney—to vote for the Democratic ticket this year, and not just abstain from voting for Trump, in order to maximize the chances of preventing Trump from returning to power. If “constitutional Republicans” really fear for the country’s future in the event that Trump is reelected, then in the current plurality-winner electoral system they ought to cast their ballots for Harris whether she’s on the ballot once (as the Democratic nominee) or twice (as the nominee of both the Democrats and the new centrist party).
Harris is the same candidate no matter how many parties nominate her in addition to the Democrats. Given the plurality-winner system, the choice between Trump and Harris is inevitably a binary one, unless a voter wants to throw away their vote by casting a ballot for a minor-party or independent candidate who can’t come close to winning (like RFK, Jr., or Cornel West). Moreover, even without changes in state law to adopt fusion voting in the way that New York practices, moderate or centrist parties are free to form and endorse whichever major-party candidate they prefer.
In fact, the Forward Party was formed after the 2020 election by Andrew Yang, the former Democratic presidential candidate, and Christine Todd Whitman, the former Republican governor of New Jersey, to provide a “centrist” home for “moderate” voters. The Forward Party has endorsed numerous down-ballot candidates in this year’s elections but so far has refused to make an endorsement in the presidential race. Even assuming that the Forward Party eventually endorses Harris “as the lesser of two evils” from its perspective, how many more voters in Pennsylvania (for example) would vote for Harris only if the Forward Party’s endorsement of Harris appeared on the ballot and not just on the Forward Party’s own website? Maybe a few, and in exceptionally close elections—like Florida’s presidential race in 2000, decided by 537 votes—even a very small number of additional votes conceivably could make a difference. But fusion voting is a very weak reform proposal if the goal is to alter the electoral system to make it more likely that especially extreme and polarizing candidates, like Trump (and Vance), will be defeated.
Convergence Voting—a more powerful tool to achieve the same goals
By contrast, Convergence Voting is a much more powerful way to prevent polarizing extremists from being elected and to give the “politically homeless middle” an effective means of expressing their electoral preferences. Convergence Voting, as I’ve explained elsewhere, is a set of electoral methods that have the property of electing the one candidate whom different overlapping coalitions of voters converge upon as the single majority-preferred choice of the whole electorate. The simplest form of Convergence Voting is the Top-3 system that I’ve described in previous Common Ground Democracy essays: the three candidates who receive the most votes in an all-candidates nonpartisan primary advance to the November general election, where voters express their direct preferences between each pair of the three candidates, and the one candidate who beats both opponents in these three direct head-to-head matchups wins the election. Using our two examples of the 2022 U.S. Senate election in Ohio and this year’s presidential election, we can easily see that Convergence Voting would be far more efficacious than NY-style fusion voting in avoiding the election of polarizing extremists and giving the “politically homeless middle” a way to vote their beliefs.
The Ohio Example, Again
In the case of the 2022 U.S. Senate race in Ohio, if the state had used the Top-3 system I’ve described, we can imagine that Rob Portman would have run for reelection because he would not have had to win the MAGA-dominated Republican primary as a prerequisite to having a chance to win in the November general election. Moreover, in a Top-3 race with Portman running in the middle between Ryan on the left and Vance on the right, Portman most likely would have prevailed. We can see from the results of the actual Ryan-Vance election as well as Ohio’s other statewide races in 2022, as discussed above, that about 40% of the voters would have been for Ryan against either Vance or Portman (since about 40% of the state’s voters are loyal Democrats who would vote for any statewide Democrat). We can surmise that another 45% of the electorate are hardcore MAGA voters, who would vote for Vance against either Ryan or Portman. The remaining 15% would have been the independents and “constitutional Republicans” who would vote for Portman against either Ryan or Vance. On the reasonable assumption that Ryan’s voters would have preferred Portman to Vance, and conversely Vance’s voters would have preferred Portman to Ryan, Portman would have won his two direct head-to-head matchups—beating Vance roughly 55%-45%, and beating Ryan roughly 60%-40%.
Thus, Portman would have won the Top-3 election. The authoritarian extremist Vance would have been defeated, and centrist voters by combining with voters on both the left and the right (in the way that Convergence Voting enables) would have been empowered to express their electoral preferences in an especially effective way to generate majority support for their top-choice candidate over each more polarizing opponent. NY-style fusion voting, on the other hand, would have still resulted in Vance’s victory (as we saw above), thus demonstrating its own inefficacy in achieving the anti-extremism and depolarization goals of its proponents. If you want electoral reform that will provide an effective voice for the “political homeless middle,” so that electoral competition returns to the center and away from the extremes, when comparing Convergence Voting with NY-style fusion voting the conclusion is inevitable: there really is no comparison in terms of efficacy.
The Presidential Election, Again
This year’s presidential election illustrates the same point. As I write, the race is in a state of extreme turmoil, with President Biden having dropped out two days ago and Vice President Harris consolidating her hold on the Democratic nomination yesterday. It is possible that over the course of the campaign between now and November 5 Harris could grow her level of support to where she is preferred by a majority of voters over the entire field of opponents, including third-party and independent candidates as well as Trump. But right now, according to polling averages, both Harris and Trump are under 50%, with Harris at about 46.5% and Trump at about 48.5%.
Moreover, throughout this year it’s been clear that there is a small but significant segment of the national electorate that would prefer a candidate somewhere in between the Democrats on the left and the MAGA Republicans, led by Trump and his running mate Vance, on the authoritarian far-right. Some of these voters in the “politically homeless middle” are the “constitutional Republicans” identified by the advocates of fusion voting, like Liz Cheney and Adam Kinzinger. But some others are disaffected centrist Democrats, like Senators Krysten Sinema and Joe Manchin. All year, Manchin has been looking for a way to run for president to represent the “politically homeless middle” without serving as a “spoiler” who would deliver a victory to Trump. As recently as yesterday, less than 24 hours after Biden bowed out, Manchin briefly considered competing against Harris for the Democratic Party’s nomination before realizing that was an untenable idea and abandoning it.
Let’s suppose that a nationwide Top-3 system were in place for this year’s presidential election. Manchin almost certainly would have run in that situation. Let’s also assume that he would have been one of the top three finalists for the November ballot, along with Harris and Trump. Given the fluidity and uncertainty of this year’s election, it is again conceivable that Harris could have ended up able to beat both Trump and Manchin in her two direct head-to-head matchups against each of them. It seems very unlikely that Trump would have been able to beat both Harris and Manchin in his two head-to-heads against them (given Trump’s inability ever to achieve the approval of a majority of Americans). Most likely, however, Manchin would have beaten both Harris and Trump in his two direct head-to-heads against each of them. He would have done this by picking up support of Trump’s voters against Harris, and the support of Harris’s voters against Trump.
Moreover, we can see that Manchin would win a Top-3 presidential election, defeating both Trump and Harris, even as with the exact same set of voter preferences Trump would win if instead NY-style fusion voting were used for the election. Using today’s polling averages to illustrate the point, suppose 46.5% of voters prefer Harris to either Trump or Manchin, while 48.5% of voters prefer Trump to either Harris or Manchin, leaving 5% of voters who would prefer Manchin to either Harris or Trump in a Top-3 election. With these numbers, as long as Trump’s voters prefer Manchin to Harris and Harris’s voters prefer Manchin to Trump, then Manchin beats Harris 53.5% to 46.5%, and Manchin beats Trump 51.5% to 48.5%.
Assume now, however, that NY-style fusion voting is used, and Manchin is not on the November ballot, but instead Harris appears on the ballot as the nominee of the Forward Party as well as the Democratic nominee. Anyone opposed to the election of the Trump-Vance ticket because of its authoritarian-leaning extremism would hope that enough of the voters who would have supported Manchin over either Harris or Trump would, when faced with the binary choice between Harris and Trump, choose Harris either on the Forward Party or Democratic Party line instead of Trump. But what if enough of them don’t? What if two-fifths of Manchin’s supporters pick Trump instead of Harris, even if three-fifths vote for Harris? That split would cause Trump to defeat Harris 50.5% to 49.5%.
This example confirms the relative weakness of fusion voting to combat extremism, as compared to the strength of Convergence Voting to do so. Given the same voter preferences, a centrist candidate defeats an extremist authoritarian ticket when Convergence Voting is employed to effectuate the majority choice of all voters, while NY-style fusion voting yields the election of that extremist authoritarian ticket because the centrist candidate is not on the ballot, and the centrist party’s co-nomination of the Democrat is insufficient to pull enough voters away from Republican ticket when centrist-preferring voters can’t cast a ballot for their top-choice candidate.
Convergence Voting as a Form of Super-Fusion
Furthermore, Convergence Voting can actually operate as an especially powerful kind of super-fusion electoral system. Recall that fusion voting emerged in the nineteenth century as a way to build electoral coalitions among multiple political parties. Smaller parties, like the Greenbacks or Populists, would enter electoral alliances—or “fuse” with—one of the major parties in order to win a majority of seats in a state’s legislature. (I discuss an example of this, a “Fusionist” alliance between Greenbacks and Democrats in Maine in 1879, in Ballot Battles.) The “fusing” parties would urge their supporters to vote for each other’s candidates in races where one party or the other had the stronger contender against their mutual opponent—not unlike the recent impromptu alliance in France between leftists and centrists in order to defeat the far-right.
Convergence Voting would facilitate this kind of coalition-building, making it especially powerful against the opposition outside the coalition. The effect would essentially be fusion on steroids. The reason is that, as between the parties on the left and the right, whichever side managed to fuse with the party in the center would be virtually guaranteed to win a Convergence Voting election against whichever side was outside the fusion coalition.
To illustrate how this kind of super-fusion would work in the context of Convergence Voting, we can return to our two examples of the 2022 U.S. Senate election in Ohio and this year’s presidential race.
The Ohio Example, Yet Again
Imagine again a Top-3 race in 2022 for Ohio’s U.S. Senate seat, with (from left to right) Ryan, Portman, and Vance as the candidates. Suppose Ryan and Portman entered a fusion alliance, telling their supporters to vote for each of them against Vance (because he’s an extreme authoritarian), while letting the voters in the fusion coalition choose whichever of the two candidates they prefer in the direct head-to-head between the two of them. If the supporters of both Ryan and Portman followed this instruction from the two candidates as part of their fusion alliance, it would be guaranteed that one of the two candidates would win the election no matter how Vance’s voters cast their ballots in the head-to-head between Ryan and Portman.
If Vance’s voters cast their ballots based on their true preferences, then Portman would prevail for the reason we saw earlier: Vance’s voters, who prefer Portman to Ryan, plus Portman’s own voters would combine to defeat Ryan—and Portman would also defeat Vance pursuant to the Portman-Ryan fusion alliance. Conversely, if Vance’s voters insincerely voted for Ryan against Portman in an effort to sabotage Portman’s victory, they would only end up electing Ryan (not Vance)—an even worse result according to their true preferences. We can see this using the same numbers as before.
Recall that in the Top-3 election Portman beats Vance 55% to 45%, because the 40% of the electorate that prefers Ryan combines with the 15% that prefers Portman to defeat the 45% that prefers Vance—and this would certainly be true in the context of a publicly announced “fusion” alliance between the Ryan and Portman campaigns. If Portman’s voters adhere to the “fusion” alliance, then Ryan will also beat Vance by the same 55% to 45% margin. This would give Portman and Ryan each one head-to-head victory. The only question then, for determining the outcome of the overall Top-3 election, is whether Portman or Ryan wins the head-to-head between the two of them: whoever does would have the two head-to-head victories necessary to be the overall Top-3 winner. Vance’s voters can affect this outcome (by expressing a preference between these other two candidates), but being only 45% of the electorate they cannot prevent the “fusion” alliance from prevailing one way or the other. The 55% of voters who form the Ryan-Portman fusion coalition are a majority that can make sure that Vance, the authoritarian extremist, won’t win. The only uncertainty is which of the two partners to the fusion coalition will be the one elected.
The Presidential Election, Yet Again
The same point applies to this year’s presidential election. Suppose again that there is a Top-3 race in November with (left to right) Harris, Manchin, and Trump as the candidates. If Harris and Manchin form a super-fusion alliance in the context of the Top-3 system, then Trump cannot win. Trump voters could only determine which of the two partners in the fusion alliance, Harris or Manchin, would prevail.
The numbers that illustrate this point are, as before, based on current polling averages—with Harris at 46.5% and Trump at 48.5%—and giving Manchin the remaining 5%. With these numbers, Manchin beats Trump 51.5% to 48.5%, as Manchin receives the votes of the 46.5% who prefer Harris, and this is especially true in light of the fusion alliance. If Manchin’s voters also act pursuant to the fusion alliance, then Harris will beat Trump by the same 51.5% to 48.5% margin. Thus, Manchin and Harris would each have a single head-to-head victory, and the outcome of the Top-3 election would depend on which of those two candidates prevailed in their head-to-head against each other. In this one of the Top-3 head-to-heads, Trump’s voters can make a difference in the outcome (by deciding which candidate, other than their top choice, to vote for). But being only 48.5% of the electorate, Trump’s supporters could not prevent Trump’s defeat in the overall Top-3 election, with the victory going to one of the two partners in the fusion coalition. Either Manchin or Harris would have two head-to-head victories, thereby winning the overall Top-3 3 election, because a majority of voters cast their ballots in favor of the fusion coalition—and against Trump.
Why might Manchin’s voters willingly support Harris over Trump in the context of a Top-3 fusion alliance but not necessarily vote for Harris instead of Trump in the context of a NY-style fusion voting system? This is because, as we discussed before, some portion of Manchin voters would have Trump rather than Harris as their second choice. In the context of NY-style fusion voting, without Manchin on the ballot, these voters might decide just to go ahead and vote their own personal preference between Trump and Harris—and Harris being the nominee of the centrist Forward Party as well as the Democratic nominee wouldn’t affect their decision to vote their personal preference between the two candidates. In the context of Top-3 “super-fusion,” by contrast, Manchin’s supporters will be able to cast their ballots in favor of their top-choice candidate (indeed, they will be able to vote in favor of Manchin against both other alternatives); and having been able to express this top-choice electoral preference, they may be more willing to accept the instruction of their top-choice candidate that as part of the fusion alliance with Harris, they should sacrifice their second-choice preference to advance the fusion alliance that their top-choice candidate has entered. Psychologically, it is much easier to give up a second-choice preference when one is able to assert one’s first-choice preference than when one’s first choice is not an option and thus, in effect, one’s second choice necessarily must function as one’s first choice.
Even if some of Manchin’s voters did not follow Manchin’s instructions that as part of the fusion alliance they should vote the Top-3 head-to-head between Harris and Trump in favor of Harris, Trump still would not be able to prevail over the fusion alliance between Harris and Manchin. That is the power of super-fusion in the context of Convergence Voting, in contrast to the inability of NY-style fusion voting to prevent Trump’s election. To see this, let’s assume again that two-fifths of Manchin’s supporters would insist upon voting their second-choice preference for Trump, rather than following Manchin’s instructions to vote for Harris over Trump pursuant to the fusion alliance. In this situation, Trump would win his Top-3 direct head-to-head against Harris, 50.5% to 49.5%, just as Trump would do in the context of NY-style fusion voting (as we saw above). Thus, Trump would have one head-to-head victory (against Harris), and so would Manchin (against Trump).
Therefore, the outcome of the Top-3 election would depend on the result of the Harris-Manchin head-to-head. If Trump’s supporters voted their true preference of Manchin over Harris, then Manchin would defeat Harris 53.5% to 46.5% (as above). That would give Manchin two head-to-head victories over each other candidate, making him the overall winner of the Top-3 election. That would mean victory for the super-fusion alliance against Trump in that context.
Suppose instead that Trump supporters attempt to defeat Manchin’s election by voting for Harris over him in the Top-3 direct head-to-head between those two. Then, Harris would win that head-to-head 95% to 5%, assuming that all of Trump’s supporters voted strategically this way. That would create a three-way tie with Harris, Manchin, and Trump each having one head-to-head victory (and each having one head-to-head defeat). What then?
As explained in other essays, a Top-3 voting system would need a tiebreaker rule. The simplest tiebreaker (and the one most consistent with the basic consensus-seeking goal of Convergence Voting) is to elect the candidate with the smallest margin of defeat. In this situation, Harris would have the smallest margin of defeat: 50.5% to 49%, against Trump. Trump’s margin of defeat, against Manchin, is larger: 51.5% to 48.5%. And Manchin’s margin of defeat is a humongous 95% to 5%, again assuming that all of Trump’s supporters voted for Harris against him. Thus, the attempt of Trump’s supporters to defeat Manchin only backfires, electing Harris instead, and is also a victory for the super-fusion alliance against Trump.
Moreover, mathematically, there is no percentage of Trump’s supporters that could vote strategically in the Harris-Manchin head-to-head that would cause the defeat of the Harris-Manchin alliance. Either (a) Manchin wins this head-to-head, in which case his two head-to-head victories makes him the overall Top-3 winner; or (b) Manchin loses this head-to-head by an even smaller margin of defeat than Harris’s narrow (50.5% to 49.5%) loss to Trump, in which case Manchin is still the overall Top-3 winner according to the Top-3 system’s tiebreaker rule; or (c) Manchin loses this head-to-head by some greater margin of defeat than Harris’s narrow loss to Trump (even if Manchin’s loss to Harris is nowhere near as large as 95% to 5%), in which case Harris has the smallest margin of defeat and thus is the overall Top-3 winner.
With these numbers, Trump never can have the smallest margin of defeat, no matter the specific margin in the Harris-Manchin head-to-head. Trump’s margin of defeat (against Manchin) will always be larger than Harris’s margin of defeat (against Trump), no matter what percentage of Trump supporters vote sincerely or strategically in the Harris-Manchin head-to-head. Thus, given these electoral preferences, Trump can never with the Top-3 election, and the super-fusion alliance of Harris and Manchin must always prevail in the Top-3 election against Trump. This result is in sharp contrast to the outcome if NY-style fusion voting is the electoral method instead of the Top-3 system: again, with the exact same set of electoral preferences Trump wins the election if NY-style fusion voting is used, even though with these electoral preferences he cannot win the Top-3 election—no matter how manipulative his supporters attempt to be in casting their Top-3 ballots.
Bottom Line
The implications of all this analysis are uncontrovertibly clear. If the goal is to counteract polarization and avoid the election of extremists while at the same time making it possible for the “political homeless middle” to express their electoral preferences—as advocates of fusion voting profess their goal to be—then Convergence Voting, including the specific Top-3 method considered here, is a far more effective way to achieve that goal than NY-style fusion voting.
The states are “laboratories of democracy,” as advocates of fusion voting correctly observed, and states should indeed be encouraged to experiment with alternative electoral reforms in the effort to combat polarization and extremism and to provide for better representation of all electoral preferences among voters. But rather than encouraging more states to experiment with NY-style fusion voting, since we already have evidence of it failing to achieve its professed goals in that state, I would encourage states to experiment with Convergence Voting instead, and then we could compare with real-world outcomes the relative efficacy of the two systems in achieving their shared goals.
Ohio is a state where a proposed electoral reform can be put on the ballot as an initiative measure if enough signatures are gathered. If I were asked whether it makes more sense to organize a signature-gathering campaign in Ohio to put on the ballot either a proposal to adopt NY-style fusion voting or a Top-3 version of Convergence Voting, I would without any hesitation recommend the latter over the former. Knowing what I do about elections in Ohio, including the 2022 U.S. Senate race that we have focused on here, I have no doubt that Convergence Voting would be a better way to combat polarization and extremism—and to provide a mechanism for the expression of the electoral preferences of the “politically homeless middle—than NY-style fusion voting.
Reforming the presidential election system is obviously a heavier lift than reforming elections in just one state, like Ohio. But even for presidential elections, I would pursue the power of Convergence Voting over the rather relatively weak reform of NY-style fusion voting. This year’s presidential race, even after the switch from Biden to Harris, presents the ongoing risk of an extreme outcome in the absence of electoral reform aimed at mitigating that risk.
As a sensible risk-mitigation strategy, why settle for the relatively weak efficacy of NY-style fusion when it’s possible to pursue the much greater efficacy of super-fusion in the form of Convergence Voting?