The Duverger Syndrome is democracies' most critical illness. Both the causes and the fixes are known. Solutions must be applied as a matter of priority.
The lesser of two evils — Symptom 4 of the Duverger Syndrome — describes a general election where both remaining candidates are disliked. The voter has no genuine preference between them. Strategic Imprisonment is something different, and in some ways worse: the voter has a genuine preference, a candidate they actually want is in the race, and they still cannot vote for them.
The obstacle is arithmetic. In a single-choice voting system, casting a ballot for a candidate who is unlikely to win does not merely fail to elect that candidate — it may actively help elect the candidate the voter most fears. The sincere vote becomes a weapon against its own holder. This is the spoiler effect: the mechanism that converts a voter's honest expression into a gift to their least preferred outcome.
The prisoner in this scenario is not choosing between two evils. They are watching their actual preference — present, viable in principle, genuinely wanted — become undoable by structural logic. The system has made honesty irrational and calculation mandatory.
The spoiler effect occurs when a losing candidate changes the identity of the winner simply by being in the race. The mechanism is vote splitting: two candidates who appeal to overlapping voter bases divide the support that, combined, would have prevailed. Each holds the other back. A third candidate — one who shares less common ground with either — wins by plurality.
The spoiler need not intend to spoil. The effect is structural, not motivational. A candidate running on principle, or on a single issue, or simply on the belief that the field deserves more voices, can determine the outcome of an election without receiving enough votes to come close to winning it. The complaint "you spoiled the election" is mathematically precise and morally unfair simultaneously. What it really describes is a system so fragile that a third option existing at all destabilises the binary.
Under single-choice voting, the voter who expresses their true preference in a multi-candidate field takes on a structural risk. If their candidate cannot win outright, their vote may divide the vote of someone who could — and hand victory to the candidate they like least. The system therefore punishes honesty and rewards calculation. The voter who wants to vote sincerely and the voter who wants to vote strategically are in permanent conflict — because they are the same person.
The language that emerges from this conflict is revealing. "A vote for X is a vote for Y." This phrase — repeated in elections across every country that uses single-choice voting — is not a logical fallacy. It is an accurate description of an arithmetic trap. The voter receives it as a threat. It is: if you vote for the person you actually want, you will get the person you most fear.
The result is voters who do not vote for what they want. They vote for what they can afford to want. Over time, having suppressed their sincere preference across election cycle after election cycle, many voters reach a point where they are no longer certain what their sincere preference is. The system has not merely trapped them. It has begun to reshape what they believe themselves to prefer.
The United States operates one of the most unforgiving single-choice systems for third-party candidates. Two elections made the spoiler dynamic visible to a generation of voters. In 1992, a Reform Party candidate drew significant support from voters dissatisfied with both major parties, ultimately winning no Electoral College votes despite receiving nearly one in five popular votes. Whose support had he taken? Post-election analysis was divided, but the question itself became the story.
In 2000, a Green Party candidate in Florida received more votes than the margin between the major-party nominees. The final margin in that state was 537 votes. Whether the Green candidate "caused" the Democratic candidate's defeat was debated intensively — but the structural logic was clear: in a plurality system, a candidate with genuine support but no path to victory can determine who wins without coming close to winning themselves.
For American voters with preferences beyond the two-party options, these elections established a rule that has governed every subsequent cycle: third-party votes are structurally costly. The voter who wants to support an independent or minor-party candidate must weigh that preference against the arithmetic of loss. The calculus is repeated at the ballot box, election after election, until the preference itself may begin to feel impractical.
France's two-round presidential system is sometimes presented as a solution to the spoiler problem: run many candidates in the first round, let the top two proceed to a runoff where the majority speaks. In practice, it creates a different form of strategic imprisonment.
Because only two candidates advance, the first round becomes a national tactical calculation about who is viable. The concept has a name: vote utile — the useful vote. The voter who prefers a smaller-party candidate must estimate not what they want but what the electorate will produce — and whether their genuine preference will survive to the second round. If the calculation suggests it will not, the rational action is to abandon that preference in round one and vote for the candidate most likely to represent their values in the runoff.
Tactical voting in France is therefore not a failure mode. It is a structural feature — built into the architecture of the system, practised openly, and named. The voter who votes utile is doing exactly what the system demands. They have simply accepted that what they can afford to want is not the same as what they actually want.
In the United Kingdom, the mismatch between voters' genuine preferences and what first-past-the-post allows them to express safely has produced an industry. Tactical voting guides, websites, and campaigns advise voters on a constituency-by-constituency basis which candidate to support to maximise the chance of defeating the most feared party — regardless of the voter's actual first preference.
This infrastructure is entirely rational within the logic of the system. It is also a monument to how completely the system has failed. A democracy that requires voters to conduct their own electoral game theory before casting a ballot — to look up which candidate the strategic calculus requires, rather than which candidate they actually prefer — has transferred the cognitive and moral burden of systemic failure entirely onto the individual voter. The system is broken; the voter is instructed to compensate for it.
Canada uses first-past-the-post for federal elections. For decades, the centre-left vote has been divided between two parties: a social-democratic party and a centrist liberal party. In election after election, polling shows the social-democratic party with meaningful support that does not survive to polling day. Voters who prefer it receive the structural warning: a vote for your genuine choice could split the progressive vote and deliver the election to the party you most fear.
The migration of sincere preferences into strategic ones is measurable — visible in the gap between polling numbers and final results. Strategic voting infrastructure has grown to match the United Kingdom's in sophistication: websites, campaigns, and constituency-level guides advising voters which candidate the arithmetic requires, regardless of their actual first preference.
The social-democratic party has governed several provinces. At the federal level it has never formed government. The structural arithmetic that prevents it from translating genuine national support into power is not a failure of ideas. It is the geometry of single-choice voting in a multi-party field.
Brazil uses a two-round presidential system. In 2022, a prominent centrist candidate — a former governor and finance minister with decades of national experience — entered the campaign with genuine support and a serious platform. Throughout the campaign, his voters were warned that backing him in the first round was structurally irrational: a vote for the centrist would weaken whichever of the two leading candidates they found least objectionable in the runoff.
His support collapsed between the polling period and election day. His final first-round result was roughly a third of his earlier polling numbers. He named what had happened publicly and explicitly: his voters had been imprisoned by the arithmetic, forced to choose between expressing their genuine preference and exercising any influence over the eventual outcome.
France's vote utile has a name because France has practised it long enough to domesticate it. The Brazilian version of the same mechanism was newer and rawer — the imprisonment visible in real time because the candidate it silenced spoke about it while it was happening.
Proportional representation is frequently cited as the solution to vote-splitting and strategic imprisonment. Germany's federal elections demonstrate that the mechanism survives in a different form. German PR operates with an electoral threshold: a party that fails to clear a minimum share of the national vote receives no seats, regardless of how many votes it received.
The threshold converts proportional representation into a binary at its margins. A voter who genuinely supports a small party must estimate whether that party will clear the threshold. If the answer is no, their vote is structurally wasted — in exactly the same sense as under plurality voting. The rational response is the same: abandon the sincere preference and vote for the larger party whose values most closely approximate the one you wanted.
In 2013, a liberal party that had been part of the governing coalition for four years fell below the threshold and was eliminated from parliament entirely. Voters who had supported it in previous elections migrated in response to the warnings. The party received enough votes to have mattered under a pure proportional system. Under threshold PR, those votes vanished. Strategic imprisonment does not require first-past-the-post — only a rule that divides the field into viable and non-viable zones.
Japan uses first-past-the-post for single-member constituencies. A single party has governed almost without interruption since 1955, not because it commands majority support but because the opposition vote consistently fragments among several parties. In multiple recent elections, the combined opposition vote exceeded the governing party's share — and the electoral system still converted that fragmented majority into a minority of seats.
The strategic imprisonment operates silently. Progressive voters face a choice with no good answer: which of several opposition parties to back in their constituency, knowing that splitting the vote hands the seat to the incumbent. There is no national infrastructure of tactical voting websites. There is no named concept equivalent to the French vote utile. The imprisonment is simply assumed — absorbed into the background noise of democratic life as an unexamined fact, below the threshold of political consciousness.
The scale of the distortion is considerable. A governing party that regularly attracts minority approval ratings continues to form government election after election. The voters who prefer an alternative cannot coordinate their sincerity without tools that single-choice voting makes structurally impossible to build.
Taiwan's 2000 presidential election ran five candidates. Two pan-blue candidates — the incumbent party's official nominee and a former governor running as an independent — divided their combined vote of nearly 60% between them, allowing the pro-independence candidate to win with 39%. The classic spoiler mechanism, operating at the top of the ballot.
But the race also illustrated strategic imprisonment at a deeper level. A former opposition party chairman ran on a serious cross-strait reconciliation platform. He received less than one percent of the vote — not because his ideas lacked support, but because the arithmetic had rendered his candidacy invisible before polling day. Voters who might have genuinely backed his approach calculated that doing so was futile, and migrated to whichever viable candidate they found least objectionable. Under a rated system, those suppressed preferences could have been expressed without cost.
Four years later, in 2004, the two pan-blue campaigns unified onto a single ticket and came within 0.22% of reversing the result — a near-perfect demonstration that the parties had drawn the Duverger lesson. The full case is examined in the dedicated article on Taiwan's 2000 and 2004 presidential elections.
Strategic imprisonment produces an effect that is less visible than a spoiled election but more durable: it changes what voters believe themselves to want. A voter who has spent twenty years suppressing their sincere preference because it is structurally unaffordable may gradually stop identifying with that preference. The calculus of what is possible becomes internalised as a substitute for what is desired. The voter who wanted something that cannot win eventually internalises a lesser preference so they no longer have to suffer the gap.
This is how single-choice voting reshapes political culture from the inside out. Not by forbidding preferences, but by making certain preferences structurally costly for long enough that they are abandoned. Movements that cannot win elections under single-choice rules do not merely lose — they are gradually unlearned by the electorate. The ideological range that the voting population expresses narrows to fit the range the system can process.
The spoiler effect has a passive form: a candidate runs sincerely and accidentally divides the vote. It also has an active form: a party or faction deliberately recruits or funds a candidate designed to split the vote of an opponent.
This manipulation is rational under single-choice voting because the rules make it possible. A third candidate need not win to change the outcome. They need only draw enough votes from the right competitor. Parties that understand this have used it. The voter who casts a ballot for the manipulated spoiler candidate may be entirely sincere in their choice — and entirely unaware that the option they chose was constructed to serve someone else's interest. The spoiler's manipulation does not require the voter to cooperate. The structural arithmetic does the rest.
Instant-runoff voting — sold in many countries under the name ranked-choice voting — is frequently presented as the solution to the spoiler problem. It is not. The proof exists in practice.
In a Vermont city council election held under instant-runoff voting, the candidate preferred head-to-head by a majority of voters was eliminated in the first round because he received fewer first-place votes than two stronger-branded opponents. A candidate whose presence in the race changed the outcome — the mathematical definition of a spoiler — caused the Condorcet winner to lose. The city subsequently repealed instant-runoff voting. The same structural failure appeared in Alaska's first ranked-choice federal election, where the candidate most voters preferred in head-to-head matchups was eliminated before the final round.
IRV reduces the frequency of spoilers compared to simple plurality voting. It does not eliminate them. The spoiler problem is a consequence of restricting voters to an ordered list rather than allowing them to express the strength of their preferences independently. Only a rated system — one in which a voter can express calibrated support for every candidate simultaneously — closes the structural gap entirely.
Informed Score Voting resolves Strategic Imprisonment at its root. When voters can express independent, calibrated support for every candidate they approve of, there is no structural cost to voting sincerely. The preference that was previously undoable can be expressed without risk. The prisoner is released not by individual courage — by changing the rules of the cell.
But Strategic Imprisonment is only one of the ways that a damaged system turns its dysfunction back on the voters who live inside it. In the next symptom, we turn to what happens over time when the parties themselves become the primary architects of the damage: adversarial politics, and the zero-sum logic that makes governance structurally impossible to sustain.