Why Arrow's Impossibility Theorem Proves No Voting Is Fair

In 1951, economist Kenneth Arrow introduced a groundbreaking mathematical proof known as Arrow's Impossibility Theorem, which fundamentally changed social choice theory. This article provides a straightforward explanation of why Arrow's theorem proves that no ranked-choice voting system can ever be entirely fair. By examining the specific fairness criteria Arrow established, the problem of cyclic preferences, and the unavoidable mathematical conflict between these rules, this piece illustrates why every voting method is forced to compromise on democratic ideals.

The Setup: Ranked-Choice Voting

Arrow’s theorem strictly evaluates ranked-choice voting systems (also known as ordinal voting) where there are at least two voters and at least three candidates. In these systems, voters do not just pick one favorite candidate; instead, they rank the candidates from most preferred to least preferred. The goal of the voting system is to aggregate these individual rankings into a single, community-wide ranking that reflects the "will of the people."

Arrow requires that the final societal ranking must be transitive. This means if the group prefers Candidate A over Candidate B, and Candidate B over Candidate C, then the group must logically prefer Candidate A over Candidate C.

The Four Fairness Criteria

To test if a voting system can fairly produce a transitive result, Arrow defined a set of reasonable conditions that any "fair" democratic election should meet.

  1. Unrestricted Domain (Universality): Every voter is allowed to rank the candidates in any order they want, and the voting system must always produce a definitive, consistent result without breaking down.
  2. Pareto Efficiency (Unanimity): If every single voter prefers Candidate A over Candidate B, then the final societal ranking must also place Candidate A above Candidate B.
  3. Non-Dictatorship: No single voter should hold absolute power. The system cannot simply mimic the ballot of one specific individual while ignoring everyone else's preferences.
  4. Independence of Irrelevant Alternatives (IIA): The society's preference between Candidate A and Candidate B should depend only on how voters rank A and B relative to each other. The introduction or removal of a third option, Candidate C (an "irrelevant" alternative to the A-versus-B comparison), should not flip the outcome between A and B.

The Core Problem: Condorcet's Paradox

To understand why Arrow's criteria cannot coexist, you must first understand Condorcet's Paradox. Discovered in the 18th century, this paradox shows that individual logical preferences can combine to create an illogical, circular group preference—much like a game of Rock-Paper-Scissors.

Imagine three voters (1, 2, and 3) ranking three candidates (A, B, and C):

  • Voter 1: A > B > C
  • Voter 2: B > C > A
  • Voter 3: C > A > B

If we compare the candidates head-to-head:

  • Two out of three voters (Voters 1 and 3) prefer A over B.
  • Two out of three voters (Voters 1 and 2) prefer B over C.
  • Logically, society should prefer A over C. However, two out of three voters (Voters 2 and 3) actually prefer C over A.

This creates a cycle: A beats B, B beats C, and C beats A. The group has no clear winner, violating the requirement for a transitive outcome.

Why the Theorem Proves It Is Impossible

Arrow's Impossibility Theorem proves that the only way to break the cycle of Condorcet's Paradox while still obeying Unanimity and the Independence of Irrelevant Alternatives (IIA) is to designate one voter as a dictator.

Here is why the math forces this failure:

The Trap of the IIA Criterion

Most real-world ranked-choice systems break ties and cycles by looking at the broader picture—such as eliminating the candidate with the fewest first-place votes (Instant Runoff Voting) or assigning points to rankings (Borda Count).

However, looking at the broader picture violates the IIA criterion. For example, in an Instant Runoff system, Candidate A might beat Candidate B. Yet, if Candidate C drops out of the race, Candidate C's votes might transfer to B, causing B to suddenly beat A. Because the presence of C changed the outcome between A and B, the system fails the IIA test. It becomes vulnerable to the "spoiler effect" and strategic voting.

The Creation of a Pivotal Voter (The Dictator)

If you strictly enforce the IIA criterion to prevent the spoiler effect, the voting system is mathematically blinded. It is only allowed to look at pairwise comparisons (A vs. B, B vs. C) in isolation.

Arrow proved that as you shift voter preferences one by one to see when the societal outcome changes, there is always a single "pivotal" voter whose ballot flips the election. Because the system must avoid the Rock-Paper-Scissors cycle and strictly obey IIA, the mathematical equations force that single pivotal voter's preference between A and B to dictate the group's preference between B and C, and subsequently A and C.

To guarantee that a voting system never produces a cycle, always respects unanimous decisions, and never falls victim to a third-party spoiler, that single pivotal voter ends up deciding the election regardless of how anyone else votes. Thus, to satisfy the other three criteria, you violate Non-Dictatorship.

The Takeaway for Social Mathematics

Arrow's Impossibility Theorem does not mean that democracy is broken; rather, it proves that "fairness" is a mathematically contradictory concept. Because no ranked-choice system can satisfy all four criteria simultaneously, electoral designers must choose which flaw they are willing to accept. In practice, every democratic system chooses to reject a dictator and uphold unanimity, which means every ranked-choice voting system in the world must inevitably violate the Independence of Irrelevant Alternatives and remain somewhat vulnerable to the spoiler effect.