Instant Runoffs: Drawing More of Us into the Electoral Process

Imagine an election with three candidates. And–so as not to offend anyone–let’s say the main issue in this year’s campaign is the type of desserts the new president will serve in the White House.  And, to further simplify things, let’s say that this issue is so important in this year’s campaign that all other issues are irrelevant by comparison. (I know. Big stretch. But we’ll come back to this.)

Candidate #1 — Sen. Aardvark — promises to serve apple pie.

Candidate #2 — Gov. Bacon — promises bourbon.

Candidate #3 — Rep. Cardinal — promises cherry pie.

So we’ve made it simple: 1 = A = apples; 2 = B = bourbon; 3 = C = cherries

Notice that two of the three candidates like fruit pies. Opinion polls of the national election agree with this assessment, since approximately 60% of the electorate say they would prefer fruit pies to bourbon at the White House.

But candidates Aardvark and Cardinal are splitting the fruit pie vote. On election day, Bacon leads with 40% of the vote, followed  by Aardvark with 38% and Cardinal with 22%. The Aardvark campaign is furious. If only Cardinal hadn’t entered the race as a third party candidate!

Under current law, Bacon might have won because of the Electoral College. Every state (except for Maine and Nebraska) will give Bacon 100% of its electoral votes if Bacon gains only a bare plurality in that state.

Under the instant runoff system–at least the simplified version we are using in this hypothetical case–we ask voters to rank the three candidates on their ballot by order of preference. Then, when the votes come in, we look to see if anyone got a majority of the first place votes.  In this case, the answer is no. Bacon only got 40% of the vote.

Now we take a look at the ballots of all the voters who voted for someone other than the two leading candidates. In this simplified case, there is only set of ballots to look at–those who voted for Cardinal. (If there were more losing candidates, we would look at all of the losing ballots.) What we are asking Cardinal voters at this point–keeping in mind that only their ballots are here to  silently answer our question–is the following: “Now that you know Rep. Cardinal is not the winner, who would you rather have as president–Aardvark or Bacon?” And we know the answer, because voters ranked their choices by preference.

All of Cardinal voters’ second place choices now become “runoff” votes. We add them to the candidate who won them. So, in this case, we can guess that most Cardinal voters listed Aardvark in second place, since they preferred fruit pies to bourbon in the White House.


Hurray! Americans are spared the trauma of accepting a president who was opposed by 60% of the voters. They get fruit pie in the White House instead of bourbon, even if turned out to be apple rather than cherry.

Oh. But I’m celebrating too soon. This was just a hypothetical case. We still have an Electoral College. And we still refuse to believe that third-party voters might have opinions about which of the leading candidates they would prefer if push came to shove.

Which brings me back to the simplified nature of this hypothetical. Does it defeat the logic of instant runoff voting when we add issues and a host of additional parties to the mix?

On the contrary, I think it makes instant runoff voting that much more important. Maybe I like what I’m hearing from a minor-party candidate. Wouldn’t I be more likely to cast my ballot if I knew my vote would not be utterly wasted? After all, my second-place vote might play a role in deciding the race.

And wouldn’t it be nice for the winner to know that the election victory required the assistance of voters whose first choice was another candidate?

Doesn’t this type of Spirited Reasoning draw more of us into the process?