Cyclic Collective Preference

Definition

When voters rank three or more candidates, pairwise comparisons of the group's stated preferences can form a cycle rather than a consistent order — the electorate prefers Nixon to Humphrey, and Humphrey to Wallace, yet also prefers Wallace to Nixon. No single "true preference" of the group exists to recover in cases like this, because the contradiction lives in the aggregation itself, not in any voter's individual ranking.

In the Book

Chapter Three works through economist Kenneth Arrow's Impossibility Theorem, tracing its roots to the eighteenth-century Marquis de Condorcet, using a constructed three-state example from the 1968 Nixon-Humphrey-Wallace election: each candidate finishes first in one state, second in another, third in the third, and pairwise tallies across the three states form an unbreakable cycle. Arrow's own gloss: "You can talk about true preference in a three-way race only when you can say that if you eliminate a losing candidate, the winning candidate would be the same. That's not what would have happened here." The chapter pairs this with the U.S. electoral college as a second, separate distorting mechanism — sociologist Brooke Harrington calls it a system that "doesn't so much tally people's preferences as average them" — showing two independent ways an aggregation method can produce a result that misrepresents the individuals feeding into it.

Why It Matters

Any process that claims to distill many individual rankings into one collective choice — an election, a hiring committee's ranked ballots, a product team voting on features — can manufacture an internally contradictory "preference" purely from the aggregation method, which means disputes over voting rules are never merely procedural; they can change which outcome counts as the group's will.