Reform for Realists: The False Promise of Condorcet Voting

Abstract

As Americans grow disenchanted with democracy, many scholars suggest that election reforms may offer a path forward. The fastest growing of these reforms is ranked choice voting (RCV). In RCV elections, voters rank candidates in order of preference: first, second, third, and so on. The candidates with the least support are eliminated, and ballots that ranked those candidates count for their next choice instead. This method has over a century of use in public elections and provides proven benefits well-suited to many of the problems facing our country.

Yet some social choice theorists now argue that reformers have backed the wrong horse. They propose a different family of methods: Condorcet-compliant methods (CCMs). These scholars note that on rare occasions RCV can elect a winner other than the “Condorcet candidate” (i.e., the candidate who would have won against every other contender in pairwise contests), and they contend CCMs would fix this “flaw.” 

In this Article, we step back from the theoretical conversations about formal, mathematical election-system criteria that have dominated the discourse and offer a more functional framework for analyzing proposed reforms. This “reform for realists” approach seeks to situate the existing literature in a richer scholarly context, to surface and center key normative questions, and to ground future study in a thicker (albeit messier) account of mediated, pluralistic politics and competing reform options.

Through this wider lens, we evaluate the latest CCM literature and argue that much of it misses the forest for the trees, ignoring fundamental principles of democratic design and first-order questions at the foundation of election law. CCM advocacy to date builds on unrealistic assumptions about voter behavior; elides critical, complex, and contested normative questions; overlooks essential relationships between candidates, parties, campaigns, elections, and governance; and ultimately cannot deliver on its core promise: the guaranteed election of a genuine Condorcet candidate. Under real-world conditions, CCMs may even risk electing the candidate opposed by most voters.

In the end, a functional framework suggests there are limits to the value of comparing the real-world performance of known reforms to the hypothetical performance of untested reforms based on how rational voters or candidates “should” behave. CCMs may benefit from the study of private use cases and then initial municipal adoptions before wider applications are considered. For now, CCMs do not appear to offer benefits that can be predicted with the confidence to justify immediate use in government elections to public office, especially at the state or federal level.