by Dan Fitch
If we have to vote for people to represent us, could we at least do better than the garbage we are doing now? Our plurality, first-past-the-post systems of electing people for office are bad. We know no voting system is perfect, thanks to Arrow. So, should we consider systems like sortition or liquid democracy, which attempt to fix the problem by doing away with elections in various ways? Maybe, but what are our options if we must vote?
After Red Madison recently discussed ranked choice voting (RCV), it’s time to look at a simpler alternative voting system: Approval voting. It works just as it sounds: you vote for everyone you “approve” of. Your ballot looks exactly the same as it does in familiar first-past-the-post. A bunch of circles with names by them — but now, you can fill in more than one circle. That’s the only difference. We get rid of the rule of only voting for one. (If you learn better with visuals, try this video that compares plurality, ranked choice, and approval voting.)
Advantages of Approval Voting
This is very easy to describe to voters: vote for everyone you approve of. Got it? Done. Far fewer ruined ballots, because now it’s near-impossible to screw up. (You can make your ballot useless by being pro-everyone. But maybe you’re… pro-everyone? Well, you’re weird.) It’s an old idea, used for selecting popes in the Middle Ages, and in Greece’s legislative elections, from 1864 to 1923, but it didn’t get the name “approval voting” until 1971.
Each party can have a candidate on the ballot, just like they can now, but “third party” candidates can get a better sense of voters’ honest support. There are no longer “spoiler” candidates, and it’s easy to see why: you can vote for them and the other one. It’s even possible to eliminate primary votes (and cut costs!) if there are small numbers of candidates for office, and just put everyone from all parties on an approval ballot.
How to count the votes and find a winner of an approval election? It’s almost too simple: just count them. All of them, on every ballot. Tally them all up, and boom. Done. The candidate with the most votes is the winner. Want to do multi-member districts? The top N candidates get seats.
Another major advantage is that approval is cheaper than other voting system changes. It requires no new voting machine hardware, and no new infrastructure. It stays easy to subdivide, so precincts can report their summary votes in a legible, easily-combinable way — unlike ranked choice. Finally, it’s easy to audit and recount, where counting and audits can be difficult with ranked choice.
The best way to sum it up is that approval voting is simpler in almost every way than ranked choice, range voting, and other systems. Simple should make everyone happy. (Well, everyone except the people who benefit from our current two-party domination. Looking at you, defense firms and health insurance companies!) So why should socialists be in support of approval voting, other than the benefits of increased democracy?
If we believe people’s true preferences lean toward democratic fairness and socialist policies, approval voting is the simplest system that gives third parties a fair chance to compete without being labelled as spoilers. For example, Nader voters in 2000 or Green Party voters everywhere are blamed for “spoiling” elections. Approval would eliminate that blame game. It also provides a better way for voters to express themselves in primaries: a place where it’s important that socialist policies get a fair shake.
Comparing Approval with Ranked Choice
If approval voting is better in almost every way than our first-past-the-post, how does it stack up to ranked choice? Ranked choice has more mindshare, is actively being used by cities like Minneapolis and states like Maine, and is being pushed by some advocacy groups like FairVote as better than approval voting. If big money is currently pushing ranked choice, that should perhaps give socialists and other outsider parties pause.
Let’s look at FairVote’s two main complaints about approval voting in detail.
First, they talk about the issue of majority rule: “If voters truly are free with their approvals in an approval voting election, it’s quite possible two or more candidates could earn more than half the vote. Indeed, it’s possible that a candidate whom well over half of voters see as a top choice could lose to someone who nobody sees as their top choice.” That sounds like it might not be a downside at all, and instead the most voters would have “approved” of that candidate for office. Sounds downright… democratic?
The second issue FairVote raises (which incidentally contradicts the “majority rule” claim) is that “[in] approval voting elections, you can’t indicate support for more than one candidate without support for a lesser choice potentially causing the defeat of your first choice.” They claim this kind of “bullet voting”, where you use your approval ballot to only vote for one candidate, simply reduces approval to a plurality system. While it’s true that ranked choice and range voting get around this issue to some extent, ranked choice also has tactical voting problems where ranking your true first choice first is not always best for that candidate, and it may not even solve the problem of spoiler candidates. Honestly, ranked choice’s failure modes are extremely confusing. For the nerds out there, you can dig into the Gibbard-Satterthwaite theorem which proves tactical voting problems will pop up in any voting system that’s not a dictatorship or lottery (hey, sortition again!)
So why not keep it simple? Approval voting is still far more expressive than plurality voting, which is possibly the worst of all worlds; and approval is at its strongest to help voters make decisions in highly contested elections with multiple candidates who don’t reach a majority at all. Simpler is better for another pro-democracy reason: more-complex ballots may decrease voter turnout.
In practice, under approval voting, in elections like the 2020 presidential election, voters who want to strongly “disapprove” of one or more candidates could do so by voting for everyone else. After approval voting is used for a while, DSA members who wanted to approve a DSA candidate and not approve the Democratic party candidate would be free to do so. But this does not mean that approval reduces to a plurality system as FairVote claims: voters would continue to have the free choice of voting for more than one candidate, and there are cases (like 2020) where it may express voter’s desires better to simply vote for more than one candidate.
In addition, there is some evidence that ranked choice still leads to two-party lock-in. See, for example, Australia… where ranked choice has been in use in their lower house for a century, and they still have a two-party lock. However, switching from plurality voting may have encouraged their center-left Labor Party to rely on left-wing voters for Green and other parties choosing Labor as a “second best”, and this has pushed them to incorporate environmental policies. Approval would likely result in a similar “party pull” mechanic, to pull to where the actual public center is.
Third parties should be in favor of approval voting, where they may start to gain additional support from normally Democratic-party-line voters, and a snowball could start rolling downhill. The mainstream parties should be in favor as well, because it will push a competition to have positive policies that more voters actually support instead of just anti-policy like “orange man bad.” Of course, that’s likely not what the Democratic Party powers are interested in, and that is why approval voting is going to need major grassroots support.
If approval voting is better and still very simple, why is nobody voting this way? Well, The Center for Election Science helped Fargo vote to switch to approval voting for city positions in 2018, and they had their first election in June. This week in St. Louis, voters were asked to consider adopting approval voting for future city elections. The proposition passed with a healthy 68% margin, so their primaries will be conducted with approval voting with the top two candidates going into a runoff.
We should strongly consider approval voting for internal group elections. We should then push for it at the local level, for city and county elections. Then we could try to push for state primaries, and if it works there, eventually all state elections. Finally, if it proves an improvement in those cases, it sure seems like we could someday stop trying to bake a cake out of rotting garbage with our federal presidential elections and the Electoral College, and get simpler. Approval voting: it’s simple.