Electoral Rules for a Low-Trust Era
The case against Ranked Choice Voting
I’m a centrist. But not that kind of centrist.
In my next post, I’ll have more to say about the kind of centrist I am.
Today’s post is more about the kind of centrist I’m not. I’m not the old-fashioned David Broder type of centrist who looks at what each party is doing and saying at any given moment and then plants himself equidistant between them. Neither am I especially impressed with Andrew Yang’s new-fangled centrist Forward Party, convinced as I am that such third-party gambits usually backfire by increasing the likelihood that the party furthest away in ideological terms ends up winning. I’m also pretty skeptical of Ranked Choice Voting (RCV), which is quite popular among centrists looking for an institutional fix for political polarization.
No wonder, then, that I was especially unenthused by this recent tweet from Yang:
Yang tweeted that around the time Alaska announced the results of its special election, held two weeks earlier, to fill the state’s at-large congressional seat, which has been vacant since the surprise death of Rep. Don Young in March. Democrat Mary Peltola prevailed in the election over Republican Sarah Palin when votes for third-place finisher, Republican Nick Begich, were reallocated, with 29 percent of Begich voters opting for Peltola over Palin with their second choice.
One reason I didn’t like Yang’s comment is that it’s wildly overstated. RCV isn’t remotely going to “save our democracy,” and neither will any other change to our electoral system. Certain reforms might help to discourage extremism, or remove incentives for candidates to stake out radical positions. But the polarization that’s been building in our politics for decades comes from many causes, and by now it is driven to a considerable extent by public opinion, not the rules of our electoral system. That doesn’t mean we should avoid undertaking electoral reforms, but we shouldn’t exaggerate the likely benefits of doing so.
But even looking at the range of possible electoral reforms, I’m not a fan of RCV, which is also known as “instant-runoff” voting. In an old-fashioned (non-instant) runoff, elections are typically held in two stages: in the first round, every candidate for a given office takes part; then, if none receives more than 50 percent of the vote, the top two finishers move onto a runoff round a few weeks or months later, with the winner of that two-person race being declared the winner of the election overall.
RCV combines these two steps into one by asking voters to rank order their preferences for a list of candidates. Votes are then counted, with distant finishers eliminated and the second-choice picks of those who voted for the losing candidates reallocated to other candidates. This process continues through several rounds of counting, with third, fourth, fifth, or more choices potentially coming into play, until one candidate tops 50 percent and is finally declared the majority winner.
Though it took two weeks for Alaska to count the votes in the special election that was the occasion for Yang’s tweet—and New York City’s RCV Democratic mayoral primary in 2021 took roughly as long and included an embarrassing screw up in tabulating votes—there’s no reason why computers couldn’t calculate results much more quickly, living up to the method’s promise of an “instant” runoff.
For and Against RCV
As my friend and fellow Substacker Noah Millman notes, there are two somewhat contradictory cases for adopting RCV. The first is that it gives voters a wide range of candidates without the accompanying concern that a vote for an unusual or extreme option will be “wasted” on someone who can’t win. With RCV, voters are free to vote for anyone they want in the first round and then choose more mainstream options in later rounds, thereby ensuring their vote contributes to the final result.
And that points toward the second benefit of RCV: It ensures that the winner will be a consensus candidate. Even if he or she wasn’t the top choice for a majority of voters, the winner is guaranteed to be someone a majority can live with. That’s what leads Yang and others on the center-left and center-right to embrace it.
So what’s not to love? Judged in the abstract, nothing. On paper, RCV sounds like a splendid method of gauging public opinion and translating it into representation. The problem is that elections aren’t run on paper. They are run in the real world, and the reality is that the United States in the 2020s is marked by incredibly low and continually sinking confidence in public institutions.
In my opinion, this may be the single most important fact about our political moment—the fact, more than any other, that explains how Donald Trump managed to take over the Republican Party and become president, a position from which he drove trust in our institutions even lower. That certainly includes trust in elections.
I’m fine with blaming Trump’s lies for a good part of this lack of trust in how elections are conducted and votes counted. But regardless of its origins, it exists now, and we need to recognize it as a new baseline—preferably by making electoral rules as simple and as transparent as possible. The alternative is for election confidence to sink even lower, to a point where large numbers of Americans begin to view the process of awarding political power to be a rigged game lacking in legitimacy.
In my view, RCV is not a particularly good match for a country struggling with low trust in elections. Whereas a first-past-the-post system merely depends on authorities counting up votes and announcing who received the most, RCV multiplies that process across several rounds of reallocation and vote counting. Not only do voters need to trust that the total number of first-choice picks is counted accurately, as in any first-past-the-post election, but also that second-, third-, and fourth-choice picks are accurately counted as well. That’s a much bigger ask.
This is especially so given the complexity of the vote-reallocation process. I’m sure many champions of RCV will roll their eyes at me describing it this way, since the process is quite clear and consistent—but I bet many of them would or did do quite well on the Analytical Reasoning section of the LSAT. That’s the section of the law-school entrance exam that involves “logic games” in which a number of variables have to be ordered or grouped according to rules the test-taker must follow and from which he or she is asked to draw inferences.
The process of vote reallocation across several rounds of counting is one form of such a game. Grasping how an outcome was reached requires following its rules, just as accepting the validity of the result requires trusting the fairness and accuracy of how the rules were applied behind the scenes.
In a country with low confidence in public institutions as well as mediocre performance by schools in teaching elementary skills in reading and mathematics, this may well be asking and expecting too much. The last thing the country needs is to adopt electoral rules that make the process of declaring winners and losers more opaque.
A Better Way
That’s what I fear the use of RCV would do—and in a way that other, simpler reforms would not. Like, for instance, holding open primaries rather than partisan ones. That tends in most cases to weaken extremists on either side of the spectrum, since such candidates have to compete against more mainstream options for the votes of the entire electorate rather than just committed Republicans or Democrats.
Open primaries work especially well when they are combined with the non-instant-runoff model that France uses in presidential elections. Instead of asking voters to rank various candidates in order of preference, simply hold an open primary with all candidates—and then, if none of them crack 50 percent, hold a second, runoff election between the top two finishers from the first round. This guarantees that one of the final two candidates receives a majority—and without voters having to defer to election officials reallocating votes behind the scenes.
Easy, simple, transparent. I can’t think of anything more needful for American democracy in 2022 than that.
What I don't get about open primaries is how they differ from a general election. If all candidates run in an open primary, why would you then need a general election. Does it mean that everyone gets to vote in the primaries for each party? RCV makes sense in a general election, and it also makes a 3rd, or 4th party more competitive. Would you be able to clear up my confusion on open primaries? Thanks.
I feel like you’re essentially saying Americans are too dumb for RCV to improve things. I’m intrigued by RCV, myself, but it’s hard to dismiss your point! (You put it more diplomatically, of course.)