What the heck happened in Alaska?
Specifically, the August ’22 special election for the US House of Representatives, which was Alaska’s first use of Ranked Choice Voting statewide.
The Background
In 2020, by a razor-thin margin, Alaskan voters adopted a new election process — replacing the partisan primary and “choose only one” plurality general election to a “Top 4” open primary and a general election using the Ranked Choice (also known as Instant Runoff) voting method.
Advocates argued persuasively that the status quo was seriously deficient, saying, “Under our current system, voters are trapped in an endless cycle of voting for what they consider to be the lesser evil. Voting their conscience or for a candidate they truly prefer is not an option for most, as they are told they are ‘stealing’ votes from one of the two old parties. This causes a forced choice between two undesirable candidates — therefore many choose not to vote at all.”
The new system, they promised, “would allow voters to rank candidates in order of preference and ensure that the winner is the candidate with a majority of voter support. Ranked-choice voting is a commonsense reform that saves money, makes elections more inclusive, and discourages negative campaigning.”
The claim that RCV is a commonsense reform that always elects the candidate with a “majority of voter support” is chiefly promoted by FairVote, the leading national advocacy group for RCV, and that claim rests on another assertion — that in Ranked Choice, you can honestly express you true preferences because if your first choice is eliminated, your vote counts for your backup choice:
To be clear, these are very desirable properties in an election system — for voters to be able to express their true preferences on the ballot, and to elect the candidate that best represents the will of the electorate overall. Unfortunately, these claims are false when applied to RCV, and Alaska’s first foray demonstrated this with stunning clarity in an election with national consequence.
So what happened in the Alaska ’22 Special Election?
In the summer of 2022, Alaska held a Special Election due to the death of US Representative Joe Young. According to the Alaska Division of Elections, 48 candidates vied for the spot in the June Open Primary. The top four candidates far outshone the rest of the field: Sarah Palin (Republican) with 27.01%, Nick Begich (Republican) with 19.12%, Al Gross (non-affiliated) with 12.63%, and finally Mary Peltola (Democrat) with 10.08%. No other candidate garnered even 5% of the Open Primary vote. Shortly after the primary tally, Al Gross quit the field, leaving three candidates to advance to the General Election in August.
According to an analysis of the public cast vote record, there were 16 distinct groups of voter preferences expressed, summarized below in order of number of voters in each group. A sum of miscast ballots is also included. The expression A > B means that the voters in that group put A above B in preference order on the ballot.
A few notes:
- The source code used for computing the unique rank orderings above from the Cast Vote Record and the CVR itself are publicly available here. The official Alaska Ranked Choice tabulation report is available here.
- Astute readers will note that the totals for blank and miscast ballots differ from Alaska’s official RCV results. This discrepancy is assumedly the result of Alaska’s Division of Elections working hard to determine voter intent of miscast ballots, but is immaterial to the analysis that follows.
- Voters did not express preferences strictly along “party” lines. Some voters preferred Peltola (Democrat) first and Palin (Republican) second, others preferred Begich (Republican) first and Peltola (Democrat) second, and so on.
The table above can be compressed further by noting that voters who left their third preference blank are casting an equivalent rank expression in a three-candidate race to those who expressly marked their last choices:
For the moment, pay no heed to the background colors in certain rows in the table above… we’ll get to that later :-).
When the voter preferences above are summed to compare the candidates head-to-head, the following results emerge, including as a percentage of the 192,289 total ballots cast:
- Peltola was preferred to Begich on 79,340 ballots (41.26%)
- Peltola was preferred to Palin on 91,138 ballots (47.40%)
- Begich was preferred to Peltola on 87,853 ballots (45.69%)
- Begich was preferred to Palin on 101,155 ballots (52.61%)
- Palin was preferred to Peltola on 85,998 ballots (44.72%)
- Palin was preferred to Begich on 63,532 ballots (33.04%)
Said more simply, Alaskan voters preferred Begich over Peltola by a plurality (45.69% to 41.26%) and preferred Begich over Palin by a majority (52.61% to 33.04%). Further, the voters’ majority preference for Begich over Palin (20% lead) was significantly larger than their plurality preference for Peltola over Palin (47.40% to 44.72%, a 2.7% lead).
The only actual majority voters expressed on the ballot was for Begich over Palin. Begich also beat Peltola on a plurality of ballots cast (45.69% to 41.26%, a 4.4% lead) — indeed Begich was preferred over Peltola by a greater margin than Peltola was preferred over Palin.
We can look at the preferences in another way — of all the voters, how many expressed support of any kind for each candidate? — ie how many voters put each candidate in either first or second position in this three-candidate race?
By this measure, a majority of voters expressed no support at all for Peltola, and a majority of voters expressed no support at all for Palin. By contrast, a super majority of Alaskan voters expressed support for Begich as either their first or second choice.
The Ranked Choice Voting promises
Remember — in order to convince Alaska’s voters to adopt the system in 2020, its advocates claimed (and still regularly claim today) that RCV guarantees a “majority winner.” They also make the claim that voters can honestly express their true preferences because if a voter’s favorite is eliminated, their vote will count for their second choice.
Given those claims, and the fact that Begich was the only candidate for which voters expressed a majority preference of any kind, and whom a super majority of Alaskans put in either first or second position on the ballot (when neither of the others had even majority support as either first or second choice), and that he was also preferred head-to-head to Peltola, it would be natural to assume that Begich emerged the winner.
Nope.
Ranked Choice Voting elected Peltola. To be clear, this is not a political position evaluation of any of the candidates in that election. This is an examination of what Alaska voters expressed and what the election system produced, as well as an examination of the repercussions of RCV’s failure.
Why was Peltola the “winner” under RCV ?
While Ranked Choice Voting allows the voters to express a preference order amongst the candidates, it only counts the secondary preferences of some of the voters whose first choice couldn’t win, leading to obviously skewed results in contests with more than two strong candidates.
The Ranked Choice Voting (Instant Runoff) method is counted in steps. In each step, RCV looks at every ballot’s top non-eliminated candidate, and eliminates the candidate with the fewest top rankings. Said more simply, while RCV allows the voter to express support for “backup” options, it ignores all of those backup expressions in each step of counting. Due only to the balance of political positions of candidates in the field (two Republicans, one Democrat), Begich was eliminated first, as could have been expected from the Open Primary vote once Gross dropped out.
Ranked Choice eliminated first the only candidate with any majority preference, who was also preferred by a plurality over the “winner”, who also had a super majority of support as either first or backup choice when no other candidate had even a majority of first or second place support, then advanced Palin, the candidate who lost to both of the others head-to-head, and finally elected Peltola from a plurality of ballots.
Now refer back to the second table above — voters who put Begich first (green rows) had their second choices counted. Voters who put Palin first (red rows) did not.
Let that sink in: the second largest bloc of voters in the election — nearly 20% of the voters — didn’t get their first choice, and their second choice was never considered at all.
The Fallout
1. FairVote, the leading advocacy group for Ranked Choice, claimed this result as a stunning victory for democracy by pretending that the counting system was a multi-round contest. They wrote:
- “Peltola built on a strong first-round lead” — false. All the ballots were cast — there was no second or third round of general election voting. All three candidates had strong actual “first choice” support from the actual first round — the Top 4 Open Primary — in which Peltola came in fourth place.
- “Begich voters split on their second choice” — false. The only “Begich voters” in this ranked election were the ones who just chose Begich. Voters who express a preference order on a ranked ballot between multiple candidates are not owned by any of the candidates. The vote is an expression of desired outcome, and voters who expressed multiple candidates on the ballot were “split” in secondary preferences amongst all three first choices.
- Most tellingly, FairVote neglected entirely to remark upon the voters who put Palin in the first spot. Their first choice couldn’t win, and their second choices were never counted. Oops.
2. As a result of the outcome, Palin supporters smelled a rat. Despite being promised that if their first choices couldn’t win that their second choices would be counted, those expressed backup preferences were entirely discarded. (The candidate who loses head-to-head to all of the others is the gold standard for a candidate that should never be able to win btw.). What are the options for those voters going forward?
- A) Vote dishonestly against their favorite candidate in future elections in order to prevent their worst perceived outcome (ie put a “lesser evil” “Begich” in first position) or
- B) Work to repeal the system.
Spoiler alert: they chose B. Led by Palin supporters, Alaskans petitioned a repeal this year and will vote on it this November.
3. National Republicans smelled a rat. The election of a Democrat to the US House from Republican-leaning Alaska because of an archaic system that obviously chose the wrong winner has helped lead to outright bans of Ranked Choice in 10 states as of June. And this blowback isn’t limited to Ranked Choice — in Oregon, as just one example, Republicans blasted STAR Voting and RCV with the same brush, despite STAR’s step change superiority over RCV in terms of simplicity, accuracy, expressiveness, equality of weight, and ability of voters to be honest on the ballot.
Most tellingly, in this present election cycle, once again two Republicans and one Democrat advanced through Alaska’s Open Primary. Due in no small part to their awareness of RCV’s failure, the lesser vote-getter in the primary has now exited the race, leaving the general election a contest between only two “viable” polarized candidates. What is it we are trying to fix again?
The Path Forward
The urgent need for voting method reform becomes more apparent with every passing election cycle. Our current election process and the plurality voting method are obviously broken, and with every use, the wedge is driven ever deeper in our body politic.
Passionate reform advocates across the nation have latched on to Ranked Choice Voting due to the marketing messages its leading advocates have promoted for several decades. This is not a criticism of the vast majority of RCV advocates! (personal note- I was an advocate of RCV/Instant Runoff for more than a decade before learning about how far short of its promises it falls). Still, the continued amplification of these false messages does real long term damage to true reform, presently turning voting method reform into yet another partisan fight to the detriment of us all.
Despite the current refrain that Ranked Choice/Instant Runoff is an innovative new reform with “momentum”, in truth, its long history of adoption, failure, and repeal over more than a century are record enough that we ought put it out to pasture for good. Newer science-backed reforms need to take center stage, but even more critically, we need an honest conversation that digs beyond the sound bites in order to deliver on Madison’s demand for the equal weight vote for all the voters.
Disclaimer
I’ve been engaged in the problem of voting equality since my teenage years in the ‘90s. You can read more about that journey, and the unlikely story of STAR Voting’s inception here. Also, this article is a bit of a sequel to “What the heck happened in Burlington” — an examination of a nearly-identical Ranked Choice Voting failure at the city level in 2009.
