Broken Ballots

No matter how you design a democratic election, a mathematical theorem proves that every single voting method is fundamentally rigged by its own rules. When we cast our ballots, we tend to view the tallying process as a neutral container for the public will. We assume that if everyone expresses their honest preferences, the winning candidate naturally emerges as the true choice of the people. Yet mathematics tells us this expectation is an illusion. In 1951, economist Kenneth Arrow proved that no voting system can convert individual preferences into a collective decision while simultaneously satisfying basic fairness criteria when there are three or more options. Arrow's Impossibility Theorem assumes that agents have well-ordered preferences and stipulates conditions for a valid decision procedure, such as allowing individuals to have rational preference orderings and the complete absence of a dictator. Try as we might to construct a flawless ballot, the very architecture of counting votes imposes a compromise on our desires.

Consider how different systems attempt to solve this puzzle. Ranked-choice voting, also known as instant runoff voting, involves voters ranking candidates by ordinal preference. If no candidate receives a majority of first-place votes, the lowest-ranking candidate is eliminated and their second-rank votes are redistributed. It feels intuitive and modern, offering a way to escape the trap of choosing only between two major parties. Yet even here, the mathematics of elections catches up to us. The Gibbard-Satterthwaite theorem states that no single-winner rank-order voting method exists for three or more candidates that avoids strategic manipulation, meaning lying pays, unless the system is a dictatorship or restricts options. Voters quickly learn that ranking their favorite candidate first might actually help elect their worst enemy, forcing them to vote strategically rather than honestly.

This structural fragility gives rise to familiar ghosts in our political life, most notably the spoiler effect. In single-winner elections, a spoiler is traditionally defined as a losing candidate whose removal affects the outcome by changing the winner. The presence of a third candidate who shares an ideological space with a frontrunner can inadvertently siphon enough votes to hand the election to the opposite camp. It is tempting to look for a technical patch, tweaking the rules or experimenting with alternative formats like approval voting, though the exact technical details of how such systems violate specific criteria reveal that every design carries its own hidden compromises. Democracy is not a solved engineering problem, but a continuous exercise in choosing which flaws we are willing to live with.