On ranked-choice voting
North Dakota’s governor recently signed a bill prohibiting ranked-choice voting in the state. That started a conversation and disagreement in this open thread on the subject of ranked-choice voting. In red or blue states, ranked-choice voting often pits people from the same party against each other in the final round, and because people from the opposing party have no candidate, they can vote for the candidate the members of the dominant party don’t want. It also causes confusion and delay in getting final results.
I don’t especially like it. It’s not that the two-party system – with a primary for each party and generally a single nominee for each party – is perfect. Certainly not. But I prefer it to a system where the final candidates sometimes end up being from the same party, and/or where second-choices can determine elections.
Take the example of the re-election of Lisa Murkowski to the Senate in 2022:
This was the first U.S. Senate election in Alaska to be held under a new election process provided for in Ballot Measure 2. All candidates ran in a nonpartisan blanket top-four primary on August 16, 2022, and the top four candidates advanced to the general election, where voters utilized ranked-choice voting.
Murkowski had been a vocal critic of Donald Trump during his presidency and opposed several of his initiatives. Murkowski was one of seven Republican senators who voted to convict Trump in his second impeachment trial in 2021, and was the only one up for re-election in 2022. On March 16, 2021, the Alaska Republican Party voted to censure Murkowski and announced that it would recruit a Republican challenger in the 2022 election cycle. Kelly Tshibaka, a former commissioner of the Alaska Department of Administration, was endorsed by Trump and the Alaska Republican Party. Republican Senate leader Mitch McConnell and the National Republican Senatorial Committee supported Murkowski.
In addition to Murkowski and Tshibaka, Democrat Pat Chesbro and Republican Buzz Kelley also advanced to the general election. On September 13, Kelley suspended his campaign and endorsed Tshibaka but remained on the ballot. Murkowski received a plurality of first-place votes; however, because no candidate received a majority of the votes in the first round, an instant runoff was triggered. Murkowski won reelection in the third and final round, winning most of the second-choice votes from Chesbro’s voters.
Perhaps Murkowski would have won anyway, even under the rules of a traditional election. But her majority vote came from Democrats who preferred her to the other Republican (see the chart at the link; rounds 2 and 3).
Another Alaska election that same year was the House race in 2022 in which Democrat Mary Peltola was elected in a Republican district. In the open primary dictated by the rules (rather than primaries to select a candidate from each party), sole Democrat Peltola had gotten about 36% of the vote and the two GOP candidates combined to get about 56% of the vote.
However, the two GOP candidates (Begich and Palin) stayed in the race and spent much of their money and time attacking each other. If one had been eliminated in single-party primaries, that would not have happened (unless one ran as an Independent). In the final vote, Peltola won in this manner:
Peltola won the election with 54.94% of the vote in the third round of ballot-counting, after two other candidates, Begich III and Bye, were eliminated and their supporters’ votes were reallocated to the remaining candidates, according to the Alaska Division of Elections. Peltola had won nearly 49% of the vote in the first round, putting her close to victory from the beginning. Even though only about 10% of the eliminated Republican candidates’ supporters ranked Peltola as their next choice, rather than Palin, it was enough to secure her win.
In this case, as far as we can tell it wasn’t Peltola’s being chosen as second choice by Democrats that put her over the top (she was the only Democrat in the race), it was being chosen as second-choice by a small number of voters who had supported the losing GOP candidate Begich and who placed her rather than the controversial Palin as their second choice. The fact that a ton of money had poured into the state to support Peltola didn’t hurt, either.
In 2024, however, Begich ousted Peltola. The Republicans were wiser that time, despite ranked-choice voting still being in operation in Alaska:
Peltola finished first in this summer’s top-four, nonpartisan primary, followed by Begich and Lt. Gov. Nancy Dahlstrom (R). But Dahlstrom, backed by President-elect Trump, then bowed out of the race as the party sought to avoid fissures that were seen as paving the path for Peltola in the midterms.
House Republicans’ campaign arm had targeted the Alaska seat as an “offensive pickup” opportunity and named Dahlstrom to its “Young Guns” list, but its chair praised Dahlstrom for her exit. Begich then picked up Trump’s endorsement.
So although this was a ranked-choice race with an open primary, the GOP made sure that the general was like a conventional race rather than a ranked choice one; it featured a single candidate from each party. The results were that Begich won – this time despite Peltola’s candidacy being supported by about seven times more money than his (probably mostly from out-of-state), and despite the fact that she was endorsed by Murkowski and also had gotten more votes in the initial primary than the two GOP candidates combined.
Or, take the 2022 election of Oakland’s mayor Sheng Thao who won – barely – because of 2nd-place votes, giving her the victory over the person who actually got the most 1st-place votes. Thao ended up being recalled after a year in office, by the way.
I mention that ranked-choice often sows confusion; this article explains some of what I mean:
With a ranked-choice ballot, however, if there are five candidates running for each of those offices, then a voter is supposed to “vote” 30 times, ranking all five candidates for each of the six offices.
This requires a longer, more complicated ballot with more instructions, more pages and more ways to make mistakes. The process takes longer, which means more ballots are left incomplete. Many voters simply don’t have an opinion about who is their third, fourth or fifth choice in many elections. Yet leaving rankings blank creates the possibility of a ballot being excluded from the final results.
Counting ranked-choice ballots must be centralized and can only proceed after all ballots are returned and adjudicated. Initially, only first-preference votes are counted. If a candidate has a majority, he or she wins (and the whole ranked-choice process becomes irrelevant). If not, then the least popular candidate is eliminated, ballots with that candidate first are “adjusted” to move up the second preference to be counted as a first preference, and there is a new round of counting. Any of those ballots that have no second preference are eliminated.
This means that some ballots are counted for the same candidate in every round, while voters who prefer the least popular candidates may be counted for several different candidates as their choices are eliminated. If a voter’s preference is eliminated with no more rankings, then that voter’s ballot is considered “exhausted” and is not included in any further counting or in the final results.
Lastly, I believe it’s telling that it’s mainly the left pushing for ranked-choice voting:
Progressive groups and their donors spent more than $100 million last year pushing ranked-choice voting …
Ranked choice is a scam and too easily manipulated by the unscrupulous — which is to say, politicians. If you can’t predict how your vote will be counted, you don’t really have the vote.
I prefer Approval Voting – vote for all the candidates you approve of, most votes wins.
Rank choice voting also destroys any leverage that third party/independent candidates have. That means establishment candidates don’t have to pay attention to any minority viewpoints.
ranked-choice voting is a perversion of democracy.
We defeated it here in CO, surprisingly. A lot of money spent to get it passed – outside money.
But, I am sure it will be on a ballot again.
The rank choice voting repeal in Alaska in 2024 lost by 664 votes out of 340,000. Well, in cheating distance.
The No on 2 campaign attracted nearly $14 million in contributions, largely from outside the state, and outspent the Yes on 2 campaign by a 100-to-one margin.
https://alaskapublic.org/elections/2024-11-20/alaskas-ranked-choice-repeal-measure-fails-by-664-votes
The yes on 2 people were not well organized. I tried to donate to their campaign, but they returned my money.
I wonder how much of the no on 2 money came from the turtle, Mitch McConnell?
Ranked-choice voting does not require physical centralization of ballot counting. It requires that tabulators count ballots in their precinct, report results to the county board of elections. In constituencies which are enveloped by the county board’s jurisdiction, the county board collates the results and instructs the precincts about which candidates are due for elimination and the redistribution of their ballots. If it’s a multi-county constituency, the reports from the precincts are collated and then forwarded to the state board for further instruction.
==
IMO, the advantage that Republicans have over Democrats in statewide registration in Alaska is insufficient to justify the institution of jungle primaries in statewide contests. Party primaries should be the order of the day. There may be particular constituencies for which jungle primaries would be apposite, but not statewide contests.
==
Again, candidate screening is insufficient in Alaska. If you’ve set your signature threshold or your monetary deposit threshold at such a level that 34 candidates are running for an office, you’ve set it too low.
==
and can only proceed after all ballots are returned and adjudicated.
==
Whether you use ranked-choice voting or not, postal ballots should be subject to signature checks as they return. The ballots should be in an envelope marked with just the precinct, that envelope in a larger envelope with identifying information including signatures and thumbprints, and that larger in a postage-guaranteed mailer. Once the signatures are checked, rejected ballots should be placed in a lockbox for storage while the approved ballots should be sorted into pigeonholes in another set of lockboxes, with the mailers stored separately from the envelopes holding the ballots. Ballots arriving after the day before the election are properly deemed invalid and placed in a separate lockbox for eventual return to sender. If you get postal ballots in the mail early enough, the recipients should be expected to have them returned on time. You can begin tabulating approved postal ballots on election day. If you hold elections on Saturday, there isn’t much justification for postal balloting bar for those who have an abiding impediment to voting; such people do not account for more than about 13% of the population).
==
With a ranked-choice ballot, however, if there are five candidates running for each of those offices, then a voter is supposed to “vote” 30 times, ranking all five candidates for each of the six offices.
==
Ranked choice does require more tick marks on a ballot. Two more salient problems you have here is an excess of elected offices and a poor electoral calendar. A way around this is to have four year terms as the rule, with judges serving 12 year terms. That way, each berth in a quadrennial cycle has the same set of offices up for election. You can have referenda on ballot propositions, judicial elections, and elections to offices adjacent to the judiciary in May, other elections in November. In November, elections to federal offices might take place in year 1; municipal councils, county councils, mayors, and county executives in year 2; governor and state legislature in year 3; school board and other offices in year 4. By default, “other offices” would have to instituted via referendum and their status as elected offices periodically confirmed by referendum.
==
The voter isn’t ‘supposed’ to do what you say. He is invited to rank his preferences. If he has no rank-order in mind, he can mark his first preference and leave it at that (or turn in the ballot blank).
Rank choice voting also destroys any leverage that third party/independent candidates have.
==
It does no such thing. Quite the contrary. It allows 3d parties to build a constituency to which major party candidates might hope to appeal as a 2d choice.
I’m a precinct worker in Kentucky, and my main objection to any sort of ranked or multiple choice voting (other than the fairly common “vote for x number of these candidates) is this
“Counting ranked-choice ballots must be centralized and can only proceed after all ballots are returned and adjudicated.”
This is also the fundamental issue with mail-in voting. Ballots should be tabulated as close to the voter as possible to give people a chance to correct spoiled ballots as well as to restrict many of the fraudulent practices (introducing fake ballots, multiple counting of ballots) that can happen in a big centralized counting location. Tabulation across multiple sites using scanned paper ballots also produces much faster results because the only centralized operation is accumulating the individual precinct totals.
@Art Deco – You don’t appear to actually understand how ranked choice voting works physically if you think some election board “instructs the precincts about which candidates are due for elimination and the redistribution of their ballots.” Ballots are not redistributed. In most RC systems, the ballots are counted multiple times as candidates are eliminated and lower ranked choices are tabulated. And I can assure you that nobody is going to hang around in every precinct in a large jurisdiction such as a state for hours, (or days. IIRC it took multiple days to establish the winner in AK) for multiple rounds of ballot tabulation. The demands for ballot processing virtually assure that RC tabulation is going to happen in centralized facility, and the multiple rounds of counting and handling ballots increases the chances and opportunities for both errors and fraud.
You don’t appear to actually understand how ranked choice voting works physically if you think some election board “instructs the precincts about which candidates are due for elimination and the redistribution of their ballots.”
==
If you’d bothered to read carefully, I was taking exception to the contention she quote that tabulation of ballots had to be physically centralized and describing how it could be done.
==
Ballots are not redistributed. In most RC systems, the ballots are counted multiple times as candidates are eliminated and lower ranked choices are tabulated.
==
You have multiple rounds of tabulation. There is no need to count a physical ballot twice unless it is being redistributed. If you’re doing that in Kentucky, you could use a more efficient procedure.
==
And I can assure you that nobody is going to hang around in every precinct in a large jurisdiction such as a state for hours, (or days. IIRC it took multiple days to establish the winner in AK) for multiple rounds of ballot tabulation.
==
Have aspirant precinct workers in a given county place bids online for what they will accept to work a given shift in a given precinct. You award shifts to the low bidder. If they work the shift, they get paid what they bid. If the shift is cancelled, they receive 20% of what they bid.
==
The demands for ballot processing virtually assure that RC tabulation is going to happen in centralized facility, and the multiple rounds of counting and handling ballots increases the chances and opportunities for both errors and fraud.
==
In New York, precincts encompass 1,000 residents. Absent promiscuous use of postal ballots, you could expect 235 voters in your average election. You have four offices contested, that’s 940 ballots to run through a tabulator, which can be accomplished in < 1/2 hour.
==
You could do things differently. You cannot be bothered.
This is also the fundamental issue with mail-in voting
==
The average county in this country has 100,000 residents. If you made circumscribed use of postal ballots, you’d have about 3,600 returned ballot packs to tabulate at a county gymnasium.
Art Deco, we don’t want it period. The end. Catch a clue, we don’t want more complexity.
“ Have aspirant precinct workers in a given county place bids online for what they will accept to work a given shift in a given precinct. You award shifts to the low bidder. If they work the shift, they get paid what they bid. If the shift is cancelled, they receive 20% of what they bid.”
Oh good grief! This is exactly what we don’t want.
“Quite the contrary. It allows 3d parties to build a constituency to which major party candidates might hope to appeal…”
Not in the real world. It allows minor parties to wheel and backroom deal for their preferences and therefore policies with the major party nearest to them on the spectrum and create a coalition of sorts.
As I said in the open thread…If you like entrenched status quo government, you’ll love ranked choice voting.
As I said in the open thread…If you like entrenched status quo government, you’ll love ranked choice voting.
==
You said that. You never show your work. I can give you a fine example of entrenched status quo government in Washington courtesy members of Congress who seldom face ranked-choice voting.
==
Not in the real world. It allows minor parties to wheel and backroom deal for their preferences and therefore policies with the major party nearest to them on the spectrum
==
You see that pretty much everywhere you have coalition governments. Which is to say in a comfortable majority of parliamentary systems.
Art Deco, we don’t want it period.
==
I get that impression in re certain participants. They’re perambulating around searching for reasons. The results are amusing.
==
The end. Catch a clue, we don’t want more complexity.
==
If you gave a rat’s ass about complexity, you’d be asking why terms of office vary, why they are not properly synchronized, why the bloody town clerk is an elected official &c. &c.
==
Oh good grief! This is exactly what we don’t want.
==
You don’t want properly staffed voting precincts. Got it.
Ranked choice is a scam and too easily manipulated by the unscrupulous — which is to say, politicians. If you can’t predict how your vote will be counted, you don’t really have the vote.
==
No it isn’t. No it isn’t. You know how your vote will be counted.
“You see that pretty much everywhere you have coalition governments. Which is to say in a comfortable majority of parliamentary systems.”
And they all pretty much suck for that reason. And it pretty much assures a good plurality of voters get exactly what they didn’t vote for. Does that happen in the current US system? Sometimes yes. But I guarantee it in ranked choice systems because the candidate is simply a cipher for the party.
With all due respect… and you’re running close to the margin… you don’t live with it…so you truly don’t understand it.
And they all pretty much suck for that reason.
==
Well, you’re welcome to come up with an alternative. The last crew who had this fancy met their end in Iberia around about 1975.
==
But I guarantee it in ranked choice systems because the candidate is simply a cipher for the party.
==
You’ve confused rank-choice with boss systems. Boss systems are not a function of ranked-choice. You have no ranked choice in Britain or in Canada. You still have parliamentary candidates imposed on local constituencies by central headquarters. The sort of electoral system most congruent with boss systems is national-list PR.
The root cause is in the Law/Constitution: 1) all elections should be binary. Plurality elections are crap.
2) the binary election should be between 2 opposing political parties. “Jungle Primaries/Elections” are crap.
Why does Congress allow this nonsense? Because it puts more money into their back pockets.
“Lastly, I believe it’s telling that it’s mainly the left pushing for ranked-choice voting”
‘Nuff said.
(IOW, firstly, secondly AND lastly…)
As usual with Ivory Tower theorists, when faced with real world examples of problems when their theoretical method is implemented, Art Deco resorts to altering the requirements of His RCV, and thus solves all the problems. In theory, His theory—real RCV hasn’t been tried. Thus, not falsified.
Still no real world case of actual details, Neo provided far more, including Alaska (thanks!). Smarter Republicans & a better party reduce the problems, by selecting only 1 out of many to be the Rep candidate. Done so as to maximize chance of winning.
The real issue is that folks don’t agree on priorities: anti- illegals? Anti-crime? pro-life? pro school (more cash? more choice?) more tariffs? How much support for Ukraine?
In Slovakia and other parliamentary systems, not Two-parties, most governments are coalitions of multiple parties, and thus less stable than the US (almost only) Two-party system. The stability is a huge advantage, most of the time.
When the govt is good, most folks don’t have to worry much about it, and don’t, and it’s ok- simpler voting is better. Much much better. When the govt is so bad that it does more hurting of folks than helping, changing the govt becomes more important, and a more simple way to change is better.
For democracy, with half the folk (eligible but often not actually voting) below the median IQ, simpler is better.
I’d guess there’s little to no RCV among college faculty votes for various offices &promotions, but just not hearing of any doesn’t mean Harvard doesn’t do it some.
I have never seen a clear explanation why RCV should favor the left. That it does, now, in our country, I’ll grant, I guess. But those things are usually contingent. I expect this is. I cannot see why it must be, or ever why it’s likely.
I also don’t see a strong argument one way or the other on the general issue.
The root cause is in the Law/Constitution: 1) all elections should be binary. Plurality elections are crap. 2) the binary election should be between 2 opposing political parties. “Jungle Primaries/Elections” are crap. Why does Congress allow this nonsense? Because it puts more money into their back pockets.
==
There is no constitutional provision which says elections must be binary. Congress has only circumscribed authority to impose strictures on states in the realm of elections administration. These choices are made at the state level. I’m fascinated to know how you propose to rig the designation system so only two candidates are on the ballot at any one time for any one office. (Another mystery is why you fancy that in any collectivity all political preferences should be summarized in two parties).
As usual with Ivory Tower theorists, when faced with real world examples of problems when their theoretical method is implemented, Art Deco resorts to altering the requirements of His RCV, and thus solves all the problems. In theory, His theory—real RCV hasn’t been tried. Thus, not falsified.
==
No one knows you’re a pig on the internet, so biographical information carries little weight. If anyone’s interested, I got into this issue after spending a half dozen years as a petition carrier and committeeman and learning how New York State election law is implemented on the street-level. Different technology at that time, of course. There’s political science literature on the subject, too, of course. There’s nothing wrong with disciplined empirical study of how other countries do business. Tom Grey is not Donald Horowitz. My ‘theory’ is that arithmetic is valid and people can and do rank their preferences.
==
The rest of your post is a meander.
==
Eeyore (8:14 am) began: “I have never seen a clear explanation why RCV [ranked-choice voting] should favor the left.”
Based on my experiences and meager understandings, the left, when encountering a critical issue, is much more prepared to trash norms and implicit understandings than is the right, who will bitch ‘n’ moan and vow to wait until next time around, at which point the left will outmaneuver the right again.
Eeyore continues: “That it does, now, in our country, I’ll grant, I guess. . . . I also don’t see a strong argument one way or the other on the general issue.”
In theory, I agree, RCV should favor no one. However, I offer to Eeyore and to other esteemed readers a quotation, often credited to Yogi Berra, Albert Einstein, and Richard Feynman (among lesser persons), but evidently still up in the air as to ultimate origin:
“In theory there is no difference between theory and practice. In practice there is.”
https://quoteinvestigator.com/2018/04/14/theory/
I live in Maine where ranked choice voting is used. It is a flat out scam used by corrupt (and here, Democrat) politicians to dilute the will of the voters. I am glad other people recognize this. It should be outlawed.
I live in Maine where ranked choice voting is used. It is a flat out scam used by corrupt (and here, Democrat) politicians to dilute the will of the voters. I am glad other people recognize this. It should be outlawed.
==
There’s a difference between an assertion and an argument. Try showing your work.
WTH is a “partial veto”?!
There’s something very rotten about the State of Wisconsin…
“ Wisconsin Supreme Court allows Democrat governor’s 400-year school funding change”—
https://justthenews.com/nation/states/center-square/wisconsin-supreme-court-allows-evers-400-year-school-funding-change