Building On Ranked Choice Voting¶

A Congressional Elections Modernization Act (CEMA) paper in conversation with the ranked choice movement¶

Prepared by Albert Ramos for The American Policy Architecture Institute

The ranked choice movement has done more to make American elections better than any reform effort of the last generation. FairVote, RepresentWomen, the League of Women Voters, the RCV Resource Center, and the organizers who have carried single-winner RCV into Alaska, Maine, New York City, and dozens of municipalities have moved the conversation about elections farther than it has moved in a century. They have a standing federal bill -- the Fair Representation Act, now in its fifth Congress since 2017 -- that is the only live proportional representation proposal in Washington. They have built the political pathway. No one else in the reform world is close.

The Congressional Elections Modernization Act (CEMA) shares almost all of what the ranked choice movement wants: proportional, candidate-centered, list-free representation that breaks the two-party lock without importing closed European party lists. Where CEMA differs is narrow and entirely technical -- not on the goal, not on the diagnosis, but on the counting method underneath the ballot. This paper is about that one difference, and it is offered the way you would offer it to a friend: with full credit for the work, honesty about a real problem, and a way through it.

Where We Agree¶

Start with the diagnosis, because on the diagnosis there is almost nothing to argue about.

Winner-take-all, single-member districts distort representation. They manufacture lopsided delegations out of close electorates, they suppress competition in the overwhelming majority of districts, and they entrench a two-party system that leaves millions of voters with no one who represents them. Proportional representation is the cure, and it is close to consensus among people who study this seriously. CEMA accepts that diagnosis in full. So does the ranked choice movement. The question this paper raises is not whether to go proportional -- it is which proportional method to build the House on, and that is a far narrower disagreement than it first appears.

It is narrower because proportionality itself is not in dispute between these methods. At any given district size, the leading proportional methods deliver comparable proportionality -- the district magnitude does that work, not the counting rule. A five-seat district makes roughly one-fifth representation achievable no matter which reasonable method fills the seats. So the movement and CEMA already agree on the thing that matters most: multi-member districts of meaningful size. What is left to discuss is everything downstream of that shared choice -- how the ballot behaves, whether votes get discarded, and what the method asks of the election held around it.

The ranked choice movement also deserves credit for something subtler: it understood, earlier than anyone, that Americans had to be brought to proportional representation through a ballot they could already trust. Single-winner RCV introduced tens of millions of voters to the ranked ballot, and that familiarity is a genuine asset to the whole proportional project. The strategy was sound. This paper does not dispute it. It builds on it.

What Ranked Choice Gets Right¶

Proportional RCV -- the multi-winner ranked method the movement advances, known in the technical literature as the single transferable vote -- has real strengths, and they are worth stating without an immediate "but."

It is candidate-centered. Voters choose among individual people, not party lists. In the American context, where a large share of local elections are nonpartisan and closed party-list systems are politically dead on arrival, this is decisive. FairVote and RepresentWomen identified it correctly as the feature that makes proportional representation viable here at all when European list systems are not.

It produces genuine proportionality. A faction with a quota's worth of support elects a representative. That is the core promise, and the method keeps it.

It has a century of operational experience behind it. Ireland and Malta have governed under the single transferable vote for over a hundred years; the Australian Senate has used it since 1949. These are functioning democracies with peaceful transfers of power and broadly representative parliaments. No score-based proportional method can yet point to a record like that, and it would be dishonest to pretend otherwise. This is the single strongest empirical point in the method's favor, and it deserves to be met with respect rather than waved away.

And it has something no alternative can match: a movement. The Fair Representation Act is not a gesture -- its lead sponsor has carried it through five consecutive Congresses, and FairVote's organizing capacity, its state and municipal track record, and its decade of public education have built a real political road to proportional representation. The claim that ranked choice is the most achievable route to PR in America is not wrong. It may be the strongest argument the movement has.

CEMA does not contest any of this. It shares the candidate-centeredness and the list-free design in full. It shares the goal of proportionality. What follows is not an argument that ranked choice is bad. It is plainly better than the system we have now. It is an argument about one mechanical vulnerability the movement has had to live with -- and a way to close it.

The Problem a Friend Has to Name¶

Here is the hard part of the conversation, and it is hard precisely because it is said in good faith.

The ranked ballot has a workable band. It needs enough candidates to make a proportional result possible, but few enough that a voter can actually rank them and the ballot stays legible. Inside that band, ranked choice works well. Push past it -- too many candidates for any voter to order sincerely -- and the counting engine begins to struggle. This is not a flaw anyone built into the method out of carelessness. It is a property of counting by sequential elimination: more candidates mean more elimination rounds, more preferences each voter must produce, and more chances for a broadly liked candidate to be cut before their support is ever counted.

The Fair Representation Act's drafters saw this clearly and built carefully around it. The bill draws the walls of the workable band in statute -- a default ballot asking voters to rank a manageable number, a permitted cap when even that is too many, and a floor ensuring enough candidates advance to keep the result proportional. This is serious, conscientious design. The drafters were working with the ranked ballot as their instrument and did about the best that can be done with it. The craftsmanship is not the problem.

The problem is what holds the field inside that band. Something upstream of the general election has to decide how many candidates reach the voter. In the American system, that something is the primary. The international record confirms the dependency is real, not theoretical: Ireland sustains the single transferable vote through disciplined party nomination, which works but concentrates gatekeeping in party organizations -- the opposite of the voter-empowerment promise the movement makes here. Australia shows what happens without that discipline: Senate ballots swelled so far that voters are now instructed to number a handful of boxes rather than rank the field, the ballot itself conceding the full slate is too large to order. The lesson the FRA's drafters absorbed is the right one. A ranked proportional ballot has to be actively managed to stay usable.

This is where the friendship matters, because this is the part that has already cost the movement dearly, and it was never anyone's intent.

No one who built ranked choice intended Burlington 2009, where a center-squeeze result -- a broadly acceptable candidate eliminated despite being preferred head-to-head -- fed directly into the city's repeal of RCV the next year. No one intended Alaska's 2022 special election, where Nick Begich, the candidate who would have defeated either opponent one-on-one, was eliminated in the first round because he held the fewest first preferences, producing a result a majority would have rejected in a direct contest. These were accidents in the truest sense: outcomes no advocate wanted, produced by a mechanism nobody was trying to exploit. But accidents have consequences, and honesty between friends means naming them. Those two elections became the raw material for a wave of anti-RCV legislation and repeal campaigns across the country. The center squeeze did not just produce a bad result twice; it handed the movement's opponents a clean, repeatable talking point, and the movement has been on the back foot in state after state because of it.

The reason this matters for the conversation about proportional methods is that the squeeze does not disappear in multi-member districts -- it softens, but it returns at full force in the contest for the final seat, which collapses into a single-winner race with the single-winner vulnerability restored. And the larger the field a multi-member district draws, the more often the engine is making exactly the kind of elimination that produced Begich and Burlington. The evidence here is sharper than "it persists." Marcus Ogren's 2024 study of candidate incentives simulated a four-winner, twenty-candidate field and found the single transferable vote produced a more lopsided incentive structure than any single-winner method he examined -- a candidate's incentive to appeal to voters who opposed them sat near 3%, against 38% for a variant identical in every respect except that it eliminated by a Condorcet rule rather than by fewest first preferences. Holding everything else fixed and changing only the elimination rule moved the number from 3% to 38%. The distortion is not a cost of proportional representation. It is a cost of the elimination engine specifically.

None of this is said to reopen old wounds. It is said because a friend who watched the accidents happen, and watched the opposition weaponize them, would want to offer the thing that prevents the next one.

The Piece Worth Adding¶

So here is the offer. It keeps everything the ranked choice movement values and closes the one vulnerability that keeps getting exploited.

CEMA fills multi-member districts using Proportional STAR, the multi-seat form of STAR voting. The ballot asks the voter to score each candidate from zero to five stars -- the same five-star rating people already perform fluently when they rate a driver, a restaurant, a product. Seats are filled in rounds: the highest-scoring candidate is elected, the ballots that supported that winner have their weight reduced in proportion to how much they contributed, and the next seat is decided on the weight that remains. The result is a delegation that tracks the distribution of voter support across the district, with no party lists and no closed slates. A voter who asks which member represents them has a direct answer: the one they scored highest among those elected.

Everything the movement wants is preserved. Proportional STAR is candidate-centered and list-free, exactly like the single transferable vote. It delivers proportionality from district magnitude, exactly like the single transferable vote. It empowers the individual voter rather than the party organization -- and unlike the Irish model, it does so without needing party gatekeepers to discipline the field.

What changes is the one thing that has been hurting the cause. Because a voter scores each candidate independently rather than ordering them, there is no elimination sequence -- so there is no center squeeze. The broadly preferred candidate the engine would cut is the candidate the score rewards: someone rated three stars by 80% of voters outscores someone rated five by a narrow 30% and zero by the rest. The method rewards exactly the breadth the squeeze punishes. Begich is not eliminated under a score ballot; broad acceptability is the thing it measures. The attack surface that Burlington and Alaska opened simply is not there.

And the field-size problem dissolves with it, which is the structural gift. Proportional STAR's demand on the voter does not grow as the field grows -- scoring forty candidates is the same kind of act as scoring four, rate the ones you have a view on and leave the rest. There is no ordering to construct, no rank left blank at a cost, no ballot that exhausts when preferences run out. So there is no workable band to police, and nothing upstream needs to keep the field small. The ranked choice movement has had to tie its proportional method to the primary -- the very institution many reformers most distrust -- because the ballot needed the field pre-trimmed. Proportional STAR cuts that cord. The method tolerates an open field, so the general election can stand on its own, so the primary becomes unnecessary rather than load-bearing.

That is the piece worth adding: a counting method that gives the movement its own goals -- proportional, candidate-centered, voter-empowering representation -- while closing the failure mode that keeps getting turned into repeal campaigns, and freeing the design from its dependence on the primary.

Two honest notes belong here, because a real offer does not hide its costs. Proportional STAR has no century-long national track record; the single transferable vote does, and that is a genuine point in ranked choice's favor that score methods cannot yet match. And in multi-winner form Proportional STAR requires central tabulation rather than precinct-by-precinct counting -- the same requirement the single transferable vote imposes, so no advantage is claimed there. These are real, and naming them is part of being straight with a friend.

An Honest Word About Where We Differ¶

It would be easy, and false, to claim CEMA and ranked choice are simply the same project. They are not. There is a real choice between counting methods, and it rests on a judgment about which costs are worth bearing.

Ranked choice satisfies a property called later-no-harm: ranking a second choice can never hurt your first. Voters value that, and the single transferable vote delivers it. STAR does not -- honestly scoring a second choice can, in rare cases, help that candidate edge out your first in the automatic runoff. CEMA accepts that cost deliberately, because the only way a method can guarantee later-no-harm is by ignoring lower preferences in the early counting, and ignoring lower preferences is exactly the mechanism that produces center squeeze. You can protect the consensus candidate or you can guarantee later-no-harm. No method does both. CEMA chooses to protect the consensus candidate, because the center squeeze is empirically the more damaging of the two -- it is, after all, the thing that got RCV repealed in Burlington.

That is a values judgment, and reasonable reformers can weigh it differently. Someone who prizes later-no-harm above all, or who places decisive weight on the century of real-world experience behind the single transferable vote, can look at the same facts and land on ranked choice in good faith. This paper does not claim a proof that ranked choice is the wrong instrument. It claims that there is a method which delivers the movement's goals while closing the vulnerability that has cost it the most -- and that this is worth a serious look from the people who have carried proportional representation farther than anyone.

Conclusion¶

The ranked choice movement diagnosed the disease correctly, built the only live federal vehicle for the cure, and brought America to the threshold of proportional representation. On all of that, CEMA stands with the movement, not against it.

The single difference is the counting method underneath the ballot, and it matters for one reason: the elimination engine carries a vulnerability -- the center squeeze -- that has twice produced results no advocate wanted and that the movement's opponents have turned into a campaign against ranked choice itself. Proportional STAR delivers the same proportional, candidate-centered representation while closing that vulnerability and freeing the design from its dependence on the primary. It is offered not as a replacement for the movement's goals but as the instrument most likely to secure them.

That is the conversation this paper hopes to start -- between people who, after all, want the same thing, and who would rather build it together than apart.

Works Cited¶

Burnett, Craig M., and Vladimir Kogan. "Ballot (and Voter) 'Exhaustion' Under Instant Runoff Voting." Electoral Studies 37 (2015): 41-49.

Gallagher, Michael. "Monotonicity and Non-Monotonicity at PR-STV Elections." Trinity College Dublin, 2013.

Graham-Squire, Adam, and David McCune. "An Examination of Ranked-Choice Voting in the United States, 2004-2022." Representation (2023).

Ogren, Marcus. "Candidate Incentive Distributions: How Voting Methods Shape Electoral Incentives." Electoral Studies (2024).

Pettigrew, Stephen, and Dylan Radley. "Ballot Marking Errors Under Ranked-Choice Voting." Multi-state analysis, 2024.

Wolk, Sara, Jameson Quinn, and Marcus Ogren. "STAR Voting, Equality of Voice, and Voter Satisfaction." Constitutional Political Economy (2023).

Fair Representation Act, H.R. 4632, 119th Congress (2025).

Revision history available in the raw file.

📥 Download this document (opens on GitHub -- click the ⬇ download button)