First and anecdote: In a former life I was a project manager in the tech space. This was around the time J2EE (java) was emerging as the predominant programming framework for enterprise applications. The company I worked for had built several applications in a competing language called Forte (aka TOOL), and we had a developer who was really good with that language. Let’s call him Brad. The industry had clearly picked a winner, as Java was quickly taking over the planet. Forte. even though it was a powerful 4th gen language, never took off. The company made the enterprise decision to rebuild applications in JAVA and entirely move off Forte. Brad however for years would talk about how this or that would be better with Forte: “that process would be using 0.4% less processing time with Forte….. Forte can do the same function in 2 less lines of code” etc. etc.
It got to the point where I couldn’t let meetings be derailed or waste valuable time with other developers talking about the arcane benefits of a (by now virtually dead) language, that only one of twenty programmers was still in love with. The New Yorker in me just laid it out bluntly for Brad; “Brad for the love of God shut up about Forte.”
What does this have to do with Approval Voting (AV) versus Ranked Choice (RCV)? Well, in the current discussion about how to reform our electoral system; our corrupt two-party duopoly plurality rules system is 1st generation BASIC, Approval Voting (while an improvement, virtually anything is) is like Forte, and Ranked Choice is the Java here. Ranked Choice is the clear winner with voters. Implemented in a growing number of entire states and municipalities. (AK, ME, NYC, SF, NV etc etc) , while Approval voting has all of Fargo ND, and a bunch of academic and non-governmental organizations. (St. Louis recently revoked its Approval Voting system after using it only once)
If you’re not aware, Approval Voting is a system where every voter casts multiple votes for all candidates they approve of. It’s essential “mark all the candidates you find acceptable, and we’ll tell you who’s least objectionable.” In practice, most major party partisans only pick one candidate as to not “dilute their vote” and only voters for smaller parties bother to register their “safety choice” (aka “lesser of two evils”). In Fargo 60% of voters only approved of one candidate.
Consider a hypothetical matchup for NY Senator featuring (R)Trump, (D)Schumer, (Labor)AOC (Social Justice) Al Sharpton. Under AV, this race would be largely about name recognition. The major parties would be favored and only the “long shot” new party voters would be incentivized to select more than one approval. The MSM, stocked with former Dem and GOP operatives will be declaring AL and AOC have no change so those voters had better also approve of Schumer. Essentially it locks in place the name recognition, negative partisanship, and “lesser of two evils” voting paradigm that exists now under plurality rules.
Ranked Choice Voting asks voters to pick their favorite but then select a second third or even fourth choice should their preferred candidate be the lowest vote-getter. Another name for RCV is Instant Runoff. In fact, the GA Runoff cost over $200 million in political donor money and another $40 million in cost to the GA taxpayers. All that money could have been saved if GA simply had gathered voters' 2nd choices the first time they reported to the polls. www.betterballotgeorgia.org
So while Approval Voting does open the ballot up to allow 3rd parties it does little to combat negative partisanship, and heavily favors the establishment candidates. The fatal flaw in AV is this though: People hate it. Voters want to go to the polls and register an affirmative vote for their favorite candidate. They do not want to place equal support behind the candidate they truly like and the candidate they just barely tolerate.
Any system that purports to measure voter support must acknowledge that people support candidates at different levels, and truly incentivize voters to select all the candidates they have some level of support for. Ranked Choice voting does this better than Approval Voting. You are not diluting your vote in RCV by selecting your 2nd & 3rd choices. Of the Millions of RCV ballots already cast in America, the vast majority have selected a 2nd and 3rd option.
We are in severe danger of falling into autocracy in our current corrupt two-party duopoly. We are one serious economic meltdown from voters turning back to the now fully corrupt and autocratic GOP. We desperately need to reform our voting systems to allow third parties to safely (with no spoiler effect) form and grow when a party goes off the rails, as the GOP clearly has. In this regard, RCV is taking off like a rocket and spreading like wildfire. thefulcrum.us/...
Approval Voting….. not so much. All it seems to be doing is mudding the waters of discussion around election reform, and getting revoked in the few places it’s been adopted. This does real damage to the election reform movement.
So please Brad stop talking about Forte.
_______________________UPDATE _________________________________________
Chris Hayes is often asking on his show: "are we thinking about this from our academic bubble?" when discussing how a policy actually translates to the lives and preferences of ordinary workaday American voters.
A commenter posted
"approval voting intuitively doesn't seem as accurate, but intuition isn't science. when you actually measure the effectiveness of these voting methods mathematically with voter satisfaction efficiency calculations, approval voting is much better. it's also similar, precinct summable, has a lower risk of ties, uses a smaller ballot, etc. Approval voting is superior in every way we can measure."
They should have added "-superior in every way we can measure, in the hallowed halls and rarified air of academia" The Microsoft Zune was superior in every way measurable to the iPod (cheaper, more storage, look! more buttons!) but you've still got to sell it, and that's where intuition comes in.
Commenter: "approval voting intuitively doesn't seem as accurate, but intuition isn't science."
Science can try to model and collect data with subjects voting for their prefered fruits or pizza toppings and come up with metrics to measure "satisfaction". In the real world people far more emotionally attached to candidates(that their willing to knock on doors and donate money to) than any hypothetical test scenario. They are also emotionally attached to the promise their right to vote affords. It is the solemn moment they have to express their opinion.
The sales job to the American voter the Approval Zuning advocates have is this: "Yes we understand your support and preference for candidates ranges in levels and you are fully capable of ranking them, but forget about that, trust us. If you just equally mark off all the candidates you can tolerate in a way that expresses no distinction of your preference, in the agregate we'll come up with the least objectionable candidate and more people will be happy. Well, maybe not happy but at least not upset and vaguely satified they got the least common denominator of a candidate."
This will never intuitively feel as good in the hand as the Ranked iPoding sales pitch : "You've got strong feelings for one candidate. Great express that by ranking them first. Even if the MSM is telling you they have no change rank them first. No worries! If your candidate is last in vote total, and no one already has 51% support, you can pick a 2nd choice, even a third choice if necessary until someone has a true majority of support. Third, fourth parties can grow safely with no spoiler effect, and negative partisanship is reduced as candidates compete to be each others 2nd choice."
But the real problem for election scientists pushing approval voting is that we no longer have to hypothesize. RCV and AV are both out in the wild now. Two competing vaccines for the corrupt two-party duopoly were released, and we have real-world results to measure. RCV is doing great things in entire states. https://thefulcrum.us/Elections/Voting/ranked-choice-voting-survey
Millions have used it and like it. In Seatle, voters were given the choice of Approval vs Ranked Choice they chose RCV.
Approval voting is working in all of Fargo ND for 2 election cycles. Who knows if they'll switch to RCV? St Louis had AV for one election before repealing it.
So the AV supports are left to tear down any and all hiccups in implementing RCV in the many, many more places it's being implemented. NYC had some test data left in the election DB that was found and corrected. This type of error would happen while switching to ANY new system. In one CA county, election officials applied the wrong algorithm to some races while tabulating the Ranked Choice elections. This affected one race significantly enough to change the result. https://oaklandside.org/2022/12/28/alameda-county-registrar-miscounted-ballots-oakland-election-2022/ Errors like this are easily prevented with better transparency measures where raw ballot data can be made public to open source analysis. This is no indication that RCV is so complicated it can't be implemented, it's that election officials need to be well trained any time ANY new system is being implemented.