Blog

After more than $100 million was spent pushing ranked-choice voting in failed state ballot measures last year, more people are aware of this reform proposal. What many don’t know is that there are several other proposed reforms in competition with RCV.

All the alternative voting ideas that I’m aware of are more complicated than a normal election, but the simplest is probably approval voting. It's actually used for city elections in St. Louis. In this system, voters can vote for any number of candidates. Essentially, they give one point to each candidate, and the one with the most points is the winner.

The trouble with approval voting is that each additional vote weakens the effect of a voter’s other votes. Because all votes are weighted equally, the incentive is to vote for just one candidate, and many voters do that. This means the system benefits more strategic voters, and is unlikely to produce different results than in a normal election. So just like with RCV, most of the time there simply is no benefit to this more complicated process.

Recognizing the shortcomings of approval voting, some reformers advocate range voting or score voting. In this system, voters give a score to each candidate, according to a predetermined scale. On a 0 to 5 point scale, a voter might give two candidates 4 points, another 1 point, and two other candidates no points. All the points are added up to determine the winner.

Score voting only partly solves the flaw in approval voting. A voter who wants one candidate to win is still better off just giving that candidate the maximum number of points and all the others zero. And the system is even more complicated than approval voting. Both systems deviate from “one person, one vote.” They also make it harder to reconcile an election, because the number of votes/points and the number of ballots/voters will be very different.

Most complicated of all is STAR voting, which stands for Score Then Automatic Runoff. The process looks like score voting, with voters rating candidates (advocates compare this to reviewing movies or online purchases). But the scores are used to determine the top two candidates. Then, all ballots are re-assessed to determine whether each voter had a preference between those two candidates (scored one higher than the other) and that preference is treated as one vote. This is the “automatic runoff” that determines the winner.

STAR voting is a great example of how reformers, as they try to solve the problems of earlier proposals, usually wind up making things worse. Each of these proposals, along with RCV, makes the election process more complex. Ballots and instructions get longer, voting takes more time, opportunities to make mistakes increase. Likewise, tabulation is more complicated, with new possibilities for error and reasons for doubt.

Reformers seem almost entirely ignorant of the benefits of a normal election process, which include:

• intuitive rules,
• simple math,
• easy recounts, and
• easily understood results.

More than these, however, our normal, “first-past-the-post” election process incentivizes building very large coalitions, because the only sure way to victory is with a majority coalition. In our state and national politics, this results in massive, awkward, highly diverse political parties. There is a sense in which all these election reforms really come from an anti-social impulse, a desire to escape from compromise and meaningful diversity.