But a small number of people on the street, again almost everywhere will probably instead respond something along the lines of "Oh, you mean instant runoff voting." It will be a very rare person who thinks there might be any other alternative voting system besides IRV. And without much doubt, if you think that IRV is the only alternative to plurality voting, then you will think that IRV is a great system for voting. As we will see later in this article that IRV really is something of an improvement over plurality voting, but it has its own deficiencies -- even beyond the fact that it only allows a voter to express support for one candidate or another but never any opposition.
Proponents of IRV probably owe thanks to the organization Fair Vote for its popular awareness. The Center for Election Science promotes approval voting and some other alternatives, but their efforts appear to have had dramatically less effect.
There is a movement in Maine, where I live, to adopt IRV for a number of important elections. Anyone who has followed the earlier articles in this series will probably realize that I don't particularly favor IRV, but I do intend to vote in favor of the initiative. IRV would be an improvement that would somewhat improve Maine elections, but even more important, passing the initiative will awake voters, even in other states, to the reality that we need not be forced to forever put up with the deficiencies of plurality voting (or, in time, those of IRV).
Plurality voting is actually a quite satisfactory voting system, but only when there are just two candidates for an office. When there are three or more candidates (as Maine has in had in recent elections), plurality voting is an invitation to an unsatisfactory election that elects a very unpopular candidate to office. This flaw in plurality voting is often called the spoiler effect, though I have become cautious about using that term because of how some people react to the phrase.
People involved with third-party efforts have become very defensive about the phrase and tend to react with an emotional, my candidate is not a spoiler, maybe your candidate is the spoiler! Some in this movement will even deny that there is any such thing as the spoiler effect. They have jumped to the conclusion that if there is a spoiler effect then there must be someone who is the spoiler. We probably should come up with a better name for the phenomena, but I don't know one to suggest.
Spoiler effect is the phrase we use to describe the fact that in a three-way (or more-way) race decided by plurality voting, the election is apt to be spoiled. The typical example of a spoiled election is a three-way race. The two most-similar candidates can be viewed as splitting the vote that just one of them would have received in a two-way election. The result is that the third candidate with the most unique (and unpopular) ideas wins despite the actual opposition of perhaps two-third of the voters. That's how plurality voting works, but it is really not how an election in a democracy should behave.
This example only shows that plurality voting can fail to select the best candidate from among three or more. But still, we might ask whether, in a multi-way race, plurality voting might perhaps be a good method for eliminating the candidate who receives the least votes.
Suppose we wanted to eliminate just one candidate from a pool of five candidates, could we depend on plurality voting to deliver a good decision or is there an example in which that goes wrong as well? Let's assume that one candidate, Susan, out of the five is every voter's second choice, but that the voters split pretty evenly on the other four for first-choice. With plurality voting, only first-choice matters so Susan gets no votes at all and is dropped from contention. But Susan, the second choice of everyone, is the consensus candidate who pretty clearly should, in an ideal election, be the winner. This example shows that plurality voting can be unreliable in a multi-way election, even when it is used to eliminate a candidate.
In several earlier articles I have addressed IRV at some length:
- Self-Expression Versus Actual Effectiveness in Voting
Looking at this list it surprises me that I might still have more to say on the topic, but IRV has great popularity and I think I can now help bring some clarity of understanding to what is fundamentally the problem with IRV.
It seems quite likely that the spoiler effect was what motivated the invention of IRV. Imagine someone pondering on the fact that plurality voting works fine in an election with only two candidates but it is known to falter, and sometimes fail quite badly, whenever there are more than two candidates. Also imagine that plurality voting is the only system you know, but you are trying to dream up some better way to vote.
You might reason that a better approach would be to conduct a series of elections (using plurality voting, what else?) eliminating one at a time until the final elections is between just two candidates. Actually conducting such a series of elections like this would be tedious and expensive, not to mention disruptive, so the inventor of IRV came up with a way to accomplish nearly the same thing but asking voters only going to the polls once (but filling out a more complicated ballot). There is an important basic flaw in this approach, however.