There is a wonderful moment in any political debate when an advocate for a controversial reform finally admits what critics have been saying all along.
Paul Haughey, a CalRCV.org board member, recently offered a remarkable defense of Ranked Choice Voting. His basic argument was that critics keep finding examples where RCV doesn’t live up to its promises—but that’s okay, because those examples are merely “exceptions.”
Alaska? Exception.
Condorcet winner? Exception.
Ballot exhaustion? Nothing to see here.
New York City’s election problems? Not RCV’s fault.
Civility? RCV works in the “majority” of elections. You just happened to find an exception.
Well, there you have it.
Apparently, the new RCV standard is not that the system works as advertised. The standard is that when it doesn’t, simply call the failure an exception. That is certainly one way to run a reform campaign.
Haughey’s Alaska defense is particularly entertaining. He argues that Alaska “is not really a Red state” because it has a large independent population, and suggests that with a normal Republican candidate instead of Sarah Palin, Republican voters would have consolidated around one candidate. But that is precisely the problem with the Alaska example. The 2022 Alaska special election featured Democrat Mary Peltola, Republicans Sarah Palin and Nick Begich, and a Libertarian candidate. Peltola actually led the first-choice vote count with 48.77%, while Palin received 25.74% and Begich 23.33%. Under Alaska’s RCV rules, if nobody reaches 50% plus one in the first round, the lowest candidate is eliminated and votes are redistributed.
So Haughey’s argument essentially becomes: RCV didn’t produce the outcome Republicans wanted because Republicans nominated two candidates who split their vote—and that’s somehow evidence that RCV is working. Except under a conventional plurality system, Peltola also would have won because she received the most votes. That’s an important point. The Alaska example isn’t necessarily evidence that RCV “stole” an election from Republicans. It is evidence that RCV didn’t magically prevent a candidate from winning who already led the first-choice vote.
And Haughey’s suggestion that Republicans should simply nominate “normal” candidates is an interesting prescription from someone defending an electoral system. Apparently RCV works best when political parties make better candidate choices. Breaking news: so does every other voting system.
Haughey’s response to the Condorcet criticism is even better. The Condorcet method asks which candidate would beat every other candidate in one-on-one matchups. Haughey acknowledges that RCV can produce a winner who isn’t the Condorcet winner. His response? “You found the one exception.” That’s it. The argument isn’t that the problem doesn’t exist. The argument is that the critic found it. This is becoming a pattern.
Critic: “Here is an election where RCV doesn’t produce the Condorcet winner.” RCV advocate: “Exception to the rule.”
Critic: “Here is an election where ballots are exhausted.” RCV advocate: “Exception to the rule.”
Critic: “Here is an election where civility wasn’t particularly civil.” RCV advocate: “Exception to the rule.”
Critic: “Here is an election where results took a long time.” RCV advocate: “Not RCV.”
At some point, “exceptions” stop being a rebuttal and start being a research agenda.
Haughey’s response to ballot exhaustion is equally clever. He asks what we should compare it to. And that’s a fair question. Ballot exhaustion is still a feature of the RCV process, not a figment invented by opponents. If a voter ranks only one or two candidates and all of those candidates are eliminated, that ballot no longer participates in the later rounds. RCV advocates can reasonably argue that this is preferable to what happens in a separate runoff, where some voters don’t return to vote. But that doesn’t make exhausted ballots disappear. And it certainly doesn’t justify pretending that every voter who cast a ballot necessarily participated in the final decision.
RCV’s own mechanics acknowledge that rankings matter from round to round. Alaska’s official explanation says the process continues until two candidates remain and the candidate with the most votes wins. That’s a legitimate electoral system. But it isn’t magic.
This is perhaps the favorite defense mechanism of RCV advocates everywhere. Something goes wrong? “That’s not RCV.” A technical problem with testing ballots? Not RCV. Delayed results? Not RCV. Problems administering the election? Not RCV. Candidate behavior? Not RCV. Ballot exhaustion? Well, that’s technically RCV, but don’t compare it to anything inconvenient. Of course, some problems really aren’t caused by RCV. Haughey is correct that a voting-system controversy shouldn’t automatically be blamed on the voting method itself. But that standard has to work both ways.
If RCV advocates get to claim every successful RCV election as proof that RCV works, they don’t get to exile every unsuccessful or problematic election to the Island of Things That Don’t Count. That isn’t analysis. That’s brand management. That’s spin.
Haughey says RCV produces civility in “the majority of elections,” but admits it doesn’t happen in all of them. Again: exception. FairVote’s own research does provide evidence that voters in some RCV cities perceived less negative campaigning than voters in comparable plurality cities. That’s worth acknowledging. But even FairVote’s research describes voter perceptions and campaign tone—not a guarantee that RCV transforms politicians into civic-minded angels. And that’s really the larger problem with the RCV sales pitch. RCV can be defended as a different way of counting votes. It can be debated as a potentially useful alternative to plurality voting. What it cannot honestly be is a political religion in which every success proves the theory and every failure is an exception.
And that brings us back to Irvine. Joshua Moore argues on NextDoor that a “true democratic mandate” requires a majority. It doesn’t matter that RCV is six-tenths of one percent of all voting jurisdictions nationally and that plurality is 99.4% — that’s immaterial according to Moore. But plurality elections are a recognized democratic form of election. The question is not whether plurality voting is democratic; the question is whether voters prefer plurality, runoff, RCV or another system. That’s a crucial distinction. And Irvine has another inconvenient problem for the “instant runoff” sales pitch: Irvine doesn’t have a runoff problem. RCV is routinely marketed as “instant runoff voting.” Alaska itself describes its RCV system as IRV, or Instant Runoff Voting.
Fine.
But if Irvine isn’t conducting expensive second-round elections for these municipal races, what exactly is the crisis RCV is being brought in to solve? Perhaps the answer is that RCV advocates don’t like plurality winners. That’s a legitimate philosophical position. But it is not the same thing as proving Irvine has a broken election system. And Haughey’s own comments reveal something important.
When critics identify an RCV problem, the response is increasingly:
That’s an exception.
When another problem appears:
That’s another exception.
And another:
You found the one exception.
Eventually, Irvine voters are entitled to ask a very simple question: If RCV is “heads and shoulders above” the existing system, why does defending it require such a large collection of exceptions? Maybe the exceptions don’t prove the rule. Maybe they’re telling us something about the rule.
And perhaps before Irvine replaces a simple system—where the candidate with the most votes wins—with a more complicated system that requires voters to rank candidates, tabulators to count multiple rounds and advocates to explain away every inconvenient outcome than a promise that all the problems are merely exceptions. Because if the best defense of RCV is that it works beautifully—except when it doesn’t—then Irvine voters might reasonably respond: Thanks. We’ll keep the system we have.
Do you understand what people don’t like about the current system?
yes, and most won’t like RCV even more.
some new studies that show the benefits of RCV don’t hold up to scruntiny
Several studies and analyses challenge or qualify the claims that ranked-choice voting (RCV) consistently produces less toxic campaigns, leads candidates to reach out to more voters, and enjoys overwhelming voter understanding/liking. Evidence is mixed overall—some pro-RCV findings exist (including the ones you linked)—but critics point to counterexamples, measurement issues, and subgroup disparities. Below are relevant academic papers, analyses, and reports that dispute or complicate those specific claims.
Less toxic / more civil campaigning: The Cogitatiopress study (and related work by Donovan, Tolbert, and others) often relies on candidate/voter perceptions of tone or self-reports of cooperative appeals. Counter-evidence focuses on actual spending, ad content, or limited effects for leading candidates:
• MIT Election Lab analysis of Maine (Jesse Clark): After RCV adoption, a difference-in-differences analysis of independent expenditures found increased negative spending. Sentiment analysis of 2018 Facebook ads (matched to similar non-RCV districts) showed the Maine campaign was more negative. The author concludes RCV boosts non-major-party vote share in some cases but “does not have most of the behavioral and campaign effects that reformers tout.”
• Maine Policy Institute / Alaska Policy Forum reports: In Maine’s 2018 gubernatorial primaries and 2nd Congressional District race under RCV, independent expenditures opposing candidates rose sharply (e.g., opposition spending up substantially vs. prior cycles, while support spending sometimes fell). This is presented as evidence that negativity may shift to (less accountable) outside groups rather than disappear.
• Donovan & Tolbert, “Civility in Ranked-Choice Voting Elections” (Representation, 2023/2024): Results are mixed. Some RCV candidates and groups made multi-candidate appeals, and some voters/candidates perceived less negativity. However, the positive-tone perception “did not extend to frontrunner candidates.” The authors note only partial support for the civility hypothesis.
• Broader observations (e.g., high-profile RCV races in Maine, Alaska, NYC, Oakland) document continued or redirected negativity via PACs/outside groups, personal attacks, and polarized outcomes, even if candidate-to-candidate rhetoric softens in some local races.
Perception surveys can capture self-reported civility, while spending/ad data sometimes show the opposite pattern, especially in competitive partisan contests.Candidates reach out to more votersThe ScienceDirect study claims broader outreach under RCV. Direct academic rebuttals focused solely on this metric are less common, but related work questions the practical incentives and effects:
• Analyses of Maine and other jurisdictions note that incentives to seek second/third rankings exist in theory, but leading candidates may still focus primarily on their base, and negativity can be outsourced. Some observational data show limited changes in overall contact patterns or continued emphasis on core supporters.
• Experimental and observational work on information search (e.g., Ntounias 2023) finds voters under RCV do not necessarily expand information-seeking about more candidates, which indirectly limits the payoff of broader candidate outreach.
• Critiques argue the strategic logic does not reliably produce wider, less polarized outreach once frontrunners and outside spending are considered.
Voters overwhelmingly like and understand RCVThe FairVote exit-survey compilation reports high self-reported understanding and support among participants in RCV elections. Counter-studies highlight confusion rates, error patterns, preference for the status quo in experiments, demographic gaps, and lower confidence:
• Atkeson et al., “The impact of voter confusion in ranked choice voting” (Social Science Quarterly, 2024): Exit-poll data from Santa Fe’s first RCV election (2018) found 16% of voters felt very or somewhat confused (higher among Hispanic voters). Confused voters ranked fewer candidates, had lower confidence in ballot counting, and were less supportive of RCV. Raises equity and longevity concerns.
• New America / experimental surveys (e.g., “Ranked-Choice Voting Is an Acquired Taste”): When respondents compare systems, a strong majority rate single-mark plurality higher on fairness and ease-of-use and prefer it over ranked voting (one figure: ~78% preferred single vote). Brief explanations of RCV’s vote-transfer features did not increase support and sometimes reduced satisfaction ratings. Support is higher among younger voters, Democrats, the more educated, and those with prior RCV experience; lower among older voters and Republicans. Status-quo bias is strong among those unfamiliar with RCV.
• MIT Election Lab on Maine: Survey experiments and post-adoption surveys found significantly lower voter confidence, satisfaction, and ease of use under RCV versus plurality. Time to vote increased; some concern about others being confused (possible proxy for own difficulty). Confidence drop exceeded what “winner’s effect” alone would predict.
• Neely & McDaniel (and related work on ballot errors): Analysis of large numbers of RCV ballots shows higher rates of mismarks (overvotes, overranks, skips—nearly 1 in 20 voters in typical races improperly mark in at least one way). These rates are elevated in areas with more racial minorities, lower-income households, and lower educational attainment. Patterns raise questions about equitable understanding and participation.
• Donovan, Tolbert, et al. on self-reported understanding: Voters in RCV cities were less likely than those in plurality cities to say instructions were “very easy” to understand (e.g., 61% vs. 73% in one comparison), though majorities in both still reported relative ease. Older voters reported less understanding.
• Nolan McCarty (Princeton) analyses of ballot exhaustion: In NYC 2021 Democratic primaries and Alaska 2022 elections, exhausted ballots (when all ranked candidates are eliminated) were substantially higher in precincts with high concentrations of minority voters (Black, Asian, Hispanic, Alaska Native in various races). This can reduce those groups’ electoral influence relative to claims that RCV expands participation. (Note: some pro-RCV responses critique methodology or funding.)
Additional context includes higher rates of incomplete ranking or errors among certain demographic groups and experimental findings that many voters initially prefer the familiar single-vote system.Overall caveats
• Pro-RCV literature (including FairVote-commissioned or supportive academic work) often emphasizes self-reports from RCV jurisdictions and local nonpartisan races, where effects can look stronger. Critical work more often examines competitive partisan contests, spending data, experimental comparisons, or subgroup disparities (race, education, age, income).
• Ballot exhaustion, error rates, and confusion are real phenomena under RCV (and some non-RCV systems), but magnitudes and consequences are debated.
• Many jurisdictions have adopted, repealed, or banned RCV; outcomes vary by context (number of candidates, education efforts, partisan intensity).
For primary sources, search the titles/authors above (MIT Election Lab pieces, Donovan/Tolbert papers, Atkeson 2024, Neely/McDaniel, McCarty reports, New America evaluations). Peer-reviewed journals such as Electoral Studies, Social Science Quarterly, and Representation contain much of the academic back-and-forth.
How would you articulate the complaints people have with plurality voting?
what complaints?
You said you understand what people don’t like about it.
about RCV, yes.
Democratic National Committee insiders are poised to ban the use of ranked choice voting (RCV) in selecting their presidential nominee in 2028 – overriding presidential primary laws in Maine and Washington, D.C
Yes, and a lot of Maine and D.C. folks are upset about that.
You probably know that in some states military and absentee voters use ranked ballots. I wonder how many of them are unhappy about being allowed to indicate backup choices.
RELATED: https://virginiamercury.com/2022/12/14/virginia-republicans-are-using-ranked-choice-voting-again-democrats-still-arent/
So you don’t understand what many people dislike about plurality voting?
I do; and you don’t understand what people don’t like about RCV. I have been able to take apart every talking point RCV has and it did not take oodles of time
So we should keep plurality voting, where minority winners, vote-splitting and nasty campaigns are the rule, and not adopt RCV where they are the exception? Brilliant.
Vote splitting, nasty campaigns are not eliminated by RCV. One person, one candidate, one vote. I’m ok with a minority winner
Thank you, Dan, for identifying why exceptional Irvine citizens aren’t falling for this manufactured voting crisis and dangerous solution.
Ranked choice voting enables the deceptive web of unaccountability. Since 2016, Xavier Becerra and Alex Padilla have used our already complicated system to their personal benefit.
Christine Bish has the documents.
https://joehoft.com/exclusive-california-democrats-becerra-padilla-identifed-2016-election/
In our beautiful Irvine, we like our elections simple, clean and verifiable. We like debatable issues, not ridiculous propaganda.