Highest-Median Voting Neutralizes the Spoiler Effect by Scoring Candidates on Shared Qualitative Grades

In 2007, outside polling stations in Orsay, France, mathematicians Michel Balinski and Rida Laraki tested a radically different ballot on voters leaving the presidential election. Instead of asking citizens to pick one name, the ballot asked them to grade every candidate on a shared qualitative scale from "Excellent" to "Reject".[1][2]

The experiment proved a mathematical theory that would culminate in their 2011 book, Majority Judgment: Measuring, Ranking, and Electing. By changing the ballot from a ranking to a report card, they neutralized the most destructive force in democratic elections: the spoiler effect.[1]

The spoiler effect occurs when a third-party candidate draws votes away from a major candidate with a similar platform, inadvertently handing victory to their shared ideological opponent. It is the mathematical flaw that forces voters to vote strategically rather than honestly.

For decades, political scientists have attempted to solve this by changing how votes are tallied, most notably through Ranked Choice Voting. However, Balinski and Laraki demonstrated that the flaw lies not in the tally, but in the fundamental nature of a ranked ballot itself.[1][3]

The Zero-Sum Trap of Traditional Ballots

Under standard plurality voting, a ballot is a scarce resource that can only be spent once. If a voter supports both a moderate and a progressive candidate, they must choose one, mathematically punishing the other.

"Plurality voting is highly vulnerable to spoiler candidates," notes a 2024 analysis by the electoral reform group FairVote.

When two similar candidates run, they cannibalize the same pool of supporters. This zero-sum division allows a less popular candidate with a unified base to win with a mere fraction of the total vote.[3]

Ranked Choice Voting (RCV), or Instant Runoff Voting, attempts to fix this by allowing voters to list candidates as first, second, and third choices. If a voter's first choice is eliminated, their vote transfers to their second choice, theoretically freeing them to support a minor candidate.

Yet RCV still suffers from a hidden spoiler effect due to its elimination rounds. If a third-party candidate performs slightly too well, they can knock a broadly popular second-place candidate out of the race early, altering the final head-to-head matchup.[3]

In this scenario, the new entrant acts as a spoiler by changing the elimination order, even if they cannot win the runoff themselves. Because RCV still treats candidate rankings as relative to one another, the presence of one candidate inherently alters the mathematical standing of the rest.[3]

Grading Instead of Ranking

Highest-median voting, specifically the Majority Judgment system, abandons relative rankings entirely. Instead of asking voters who they prefer over whom, the ballot asks them to evaluate each candidate independently against a shared, absolute standard.[1][2]

Voters assign a qualitative grade—such as Excellent, Very Good, Good, Acceptable, Poor, or Reject—to as many candidates as they wish. Because the grades are independent, giving one candidate an "Excellent" does not prevent a voter from giving another candidate an "Excellent" as well.[1][2]

This independence is the mechanical key to neutralizing the spoiler effect. Whether a voter rates their preferred candidate as "Good" or "Excellent," that evaluation has absolutely no impact on the grades they choose to assign to the rest of the field.[1][3]

Once the ballots are cast, the system does not calculate an average, which could be skewed by a few extreme scores. Instead, it finds the median grade for each candidate—the exact middle evaluation when all a candidate's grades are lined up from best to worst.[1][2]

The candidate with the highest median grade wins the election. If an absolute majority of voters grade a candidate at least "Good," then "Good" becomes their final, unassailable score, regardless of who else is in the race.[2]

Mathematical Isolation

This mechanism mathematically isolates candidates from one another during the tabulation process. Under Majority Judgment, a candidate's final grade depends entirely on how the electorate evaluates them, not on who they are running against.[1][3]

If a new, similar candidate enters the race, voters can simply assign both candidates high grades. The new entrant does not siphon "first place" votes away from the established candidate, because the ballot does not force voters to ration their approval.[1][3]

Consequently, a new candidate can only change the outcome of the election by achieving a higher median grade themselves. They cannot act as a spoiler by altering the relative standing or survival of the existing candidates, because there are no elimination rounds or zero-sum vote divisions.[1][3]

Balinski and Laraki proved that highest-median rules are the most strategy-resistant systems available. By minimizing the incentive for tactical voting, the system encourages citizens to express their sincere opinions without fear of accidentally electing their least preferred option.[1][2]

Breaking Ties and Building Consensus

Because qualitative scales are discrete, multiple candidates can occasionally tie with the same median grade. Majority Judgment resolves this by removing one median grade from each tied candidate's tally until one candidate's median shifts higher than the other's.[2]

This tie-breaking procedure ensures that the depth of a candidate's support is fully measured. It rewards candidates who have broad, consensus appeal rather than those who rely on a narrow, highly polarized base of passionate supporters.[1][2]

Critics often argue that median-based systems are less familiar to voters and policymakers than simple plurality rules. Transitioning to a qualitative ballot requires public education, as voters must learn to evaluate candidates like students on a test rather than competitors in a race.[3]

However, the cognitive burden of grading is arguably lighter than the burden of strategic voting. In a plurality system, voters must constantly guess how others will vote to avoid wasting their ballot; in Majority Judgment, they only need to decide how they feel about the person on the paper.[1][3]

The Future of Electoral Reform

The push for alternative voting systems has gained momentum as political polarization deepens. While Ranked Choice Voting has captured the majority of reform efforts in the United States, its mathematical vulnerabilities leave the door open for superior models.[3]

Highest-median voting offers a paradigm shift from comparing candidates to evaluating them. By treating a ballot as a measurement of merit rather than a scarce currency, it aligns the mathematical outcome of an election with the actual will of the electorate.[1][3]

As jurisdictions look for ways to increase participation and reduce negative campaigning, qualitative grading presents a compelling solution. It allows third parties to demonstrate their true level of support without threatening the stability of the broader electoral system.[1][3]

The next hurdle for highest-median voting is moving from academic theory to binding statutory elections. Until a major municipality adopts a qualitative ballot, the system remains a mathematical proof waiting for a real-world electorate to test its limits.[1][3]