The inherent ambiguity of RCV

One of the problems with ranked-choice voting is how many arbitrary decisions are involved in its implementation. Election administrators must make decisions about whether to count or discard ballots that have no obvious answer but can change the outcome of elections. This makes RCV different from a normal election, where most of the answers are obvious to voters and officials alike.

RCV attempts to run many elections, one after the other, using a single ballot. This is why RCV advocates sometimes call it “instant runoff voting.” Really, they should call it “multiple runoff voting.” It is never instant. And RCV can go require a dozen rounds of computer-managed “runoffs” or more.

This makes voting harder, creating new opportunities for voters to make mistakes or simply register their preferences in unexpected ways. What happens in a voter ranks a candidate first but skips second and ranks other candidates third, fourth, and fifth? Or ranks the same candidate first and second, but then another candidate third? What about write-in candidates?

In a normal election system, the first two situations are impossible. RCV creates these problems. Election officials must decide what to do. And the answers are not obvious, even though they could change the outcome of a close campaign.

The third situation—write-in candidates—has an obvious answer in a traditional election: just count up those votes. Yet examining the rules for various RCV systems shows different answers, perhaps because of the difficulty of programing RCV election machines to deal with one more complication. (Alaska’s rules could eliminate a write-in candidate that could otherwise have won for lack of enough first-place votes.)

Even the process of eliminating candidates is not always carried out in the same way. Often only one candidate is eliminated per round, but Minneapolis eliminates multiple candidates in the first retabulation. Again, decisions like this is close elections have the potential to change the outcome. Importantly, there is no obvious answer and so voters’ assumptions about how the process will work can easily be incorrect.

A benefit of normal, plurality elections is simplicity. This is a practical way in which those elections are democratic. Voters can understand what will happen with their votes. Election officials are not in a position where their arbitrary decisions might change election outcomes. No democratic process is perfect, but RCV creates new, serious, and unnecessary problems.