A Simple, Expressive, Robust Election System
Support, Accept, Reject, Abstain (SARA) voting
If you’re reading this, you probably live in a country with a horrible voting system. Most of us English-speakers use plurality voting: a system which restricts you to vote for just one candidate per race, and thus guarantees that votes for anyone but the top two candidates have no effect. Spoiled elections, complacent incumbents, low turnout; plurality makes all these problems worse. It even favors the worst kind of politicians, those who know more about grabbing the limelight than about the issues facing the country. (I’m not naming names here, because it would be easy to start but hard to stop.)
Better voting systems exist. The simplest is approval voting: simply allow voters to support as many candidates as they want, and elect the one supported by the most voters. This would remove the need to lie on your ballot; you still might support a compromise, but you could always support your true favorite too. No need for spoilers, better voter satisfaction, improved turnout… the benefits go on.
But what if approval voting isn’t good enough for you? What if you want to be able to make an effective distinction between your true favorite and a compromise?
Voting theorists have proposed dozens of more-complicated systems. But there are various theorems — Arrow, Gibbard-Satterthwaite, etc.—which show that a perfect voting system is impossible. So if we’re going beyond approval, I think that the goal should still be something as simple and robust as possible.
And that’s where Support, Accept, Reject, Abstain (SARA) voting comes in. Here’s how it works:
- You can rate each candidate support, accept, reject, or abstain. Default is abstain. Candidates get 2 points for each percent of “support” and 1 point for each percent of “accept”, for a total of 0–200 points.
- Any candidate that gets less than 50 points is eliminated (unless there would be no candidates left).
- Any candidate that gets rejected by more than 50% is eliminated (unless there would be no candidates left).
- The remaining candidate with the most points wins.
To see why this is a good system, let’s run through a couple of scenarios where other voting systems have failed badly:
Egypt, 2012. The first free election in the country’s history. The candidates fall into 3 basic categories: old guard (eg, Shafik); Islamist (eg, Morsi); and reformers (eg, Foutouh, etc.). In real life, the reformers split the vote and are all eliminated in the first round. In the second round, Morsi (who had under 25% in the first round) beat Shafik (who had even less). Using SARA, most voters would have rejected one or both of the old guard and Islamist groupings, eliminating them instead; and a reformer, with support from perhaps 20% and acceptance from a large cross-section, could have won.
US, 2000. The options were Nader (left), Gore (center-left, in the US context), and Bush (right). In the real election, Nader got under 3%, a majority of that from voters who would have preferred Gore over Bush, which turned out to be enough to put Bush close enough to get declared the winner by the Supreme Court. Using SARA, let’s say that over 3 times as many voters would have felt safe supporting Nader, so he’d have gotten around 10%; but most of those would have rejected Bush (eliminating him) and abstained or accepted Gore (keeping him in the race). Since Gore (under these assumptions) would have gotten more support than Nader, he would have won. (You might not like my assumptions; feel free to make your own, and check how SARA would have worked. I’m confident you’ll find that the outcome would be fair.)
In general, whether the scenario is “center squeeze” (two sides against the middle) or “chicken dilemma” (a divided majority against a substantial united minority), SARA tends to get the most democratic result without requiring (or prompting) tricky strategy from voters. It’s simple, robust, and solves various problems.
Yes, it’s a new idea. But it rests on ideas that have actually been tested in the past: approval voting, which was used in Greece in the 19th and early 20th centuries, and Bucklin voting, which was used in over a dozen US cities in the Progressive era as well as briefly in pre-Napoleonic Geneva.
So, if your organization needs to run an election, consider using SARA. And if you live in the US, tell your state representatives and city councilmembers about approval and perhaps SARA.
You can learn more about voting reform at electology.org.