Sarah Palin lost fair and square under the rules, but her loss illustrates why Alaska’s new election rules are ridiculous.
Sarah Palin lost fair and square under the rules, but her loss illustrates why Alaska’s new election rules are ridiculous.
A laska held a special election August 16 to replace Republican Don Young, who died in March after nearly four decades in the House. Wednesday night, 15 days after the votes were cast, the state finally declared a winner: Democrat Mary Peltola. Peltola got 39.7 percent of the vote; the two Republican candidates, Sarah Palin and Nick Begich III, combined for nearly 60 percent of the vote. Thus, a 60–40 Republican electorate ends with the election of a Democrat.
This is not an illegitimate or rigged result: The race was conducted under new election rules, and under those rules, which all candidates knew in advance, the Democrat won fair and square. The outcome reflects a number of dynamics, including Palin’s polarizing nature and personal unpopularity outside the Republican base, Begich’s unusual partisan identity (running as a Republican, but with a family name synonymous with Alaska Democrats), Alaska’s perennially quirky politics, and the apparent failure of Republicans to adequately adjust their strategy to Alaska’s new election system. Republicans didn’t need to lose this race and would not have lost it with different candidate choices. Still, the outcome nonetheless reflects the ridiculous perversity of Alaska’s new system, which combines a “jungle primary” with a unique ranked-choice voting system.
The Lisa Law
Some background: The longest-running partisan drama in Alaska is the perennially tenuous political position of Lisa Murkowski. The Murkowski family — the archenemies of Palin — have held Republican office in the state for decades, and Lisa Murkowski continues to hold a Senate seat (to which she was originally appointed by her own father) despite being badly out of step with Republican base voters both nationally and in Alaska. A moderate in the Susan Collins mold may be tolerable in blue Maine, but it is deeply frustrating in reliably red Alaska, even given the state’s tendency toward big-government, socially libertarian Republicans. In 2010, Republican primary voters rejected Murkowski, so she ran as a write-in and won in the general election — and was promptly welcomed back into the Senate Republican caucus as if nothing had happened. It would be better for conservatives if she was replaced in the Senate.
Murkowski is up for reelection again in 2022, and Republican voters are again unhappy at her pro-choice stance and at a number of her votes in the Senate, including breaking with the party to vote against the confirmation of Brett Kavanaugh, for the confirmation of Ketanji Brown Jackson, and for the second impeachment of Donald Trump. Trump, being Trump, has sworn vengeance.
The solution favored by Murkowski’s allies? Change the rules. Alaska voters in 2020 narrowly approved Ballot Measure 2, an initiative that overhauled statewide elections (more on how in a moment). The Byzantine new rules, billed as “Alaska’s Better Elections Initiative,” ran 26 single-spaced pages.
The changes were explicitly billed as an effort to boost moderate and independent candidates — such as Murkowski and the state’s former governor, Bill Walker, who was elected as a Democrat-aligned independent — at the expense of those favored by the parties’ bases. The ballot initiative was written by a self-identified Republican who worked as Walker’s chief of staff. Nearly $3 million was spent in support by an outside group advised by a former Murkowski staffer.
The state’s Republican governor, Mike Dunleavy, opposed it. So did a number of leading Democrats who feared that it would protect Murkowski and potentially shut Democrats out of statewide races, in the way that the jungle primary system does to Republicans in California. Ballot Measure 2 was so closely identified with Murkowski’s fortunes that some Democrats derided it as the “Lisa Law.” Until yesterday, punditry about the law focused heavily on its projected impact on Murkowski. It ended up passing in November 2020 by the narrowest of margins, 3,781 votes out of 360,852 cast — a margin of barely over one percentage point.
Welcome to the Jungle
The new Alaska system is unique in the nation, combining two features. One is a “jungle” primary, different versions of which currently exist in California, Washington, and Louisiana. In a normal two-party system, each party holds a separate primary to choose its candidate, and they then square off in the general election, alongside any third-party candidates or (like Murkowski) write-in candidates. Some states even have “sore loser” laws to prevent candidates from running in party primaries, losing, and then doing what Murkowski did: run again in the general election, either as a spoiler or to appeal over their heads of their own party. By contrast, in a jungle primary, all the candidates from both parties — or any third-party or independent candidates — run in a single primary open to all voters, and the top finishers go on to the next round.
What is unique about the new Alaska system is that it uses an all-parties jungle primary and then runs a general election among the top four vote-getters in that primary, regardless of party. None of the other states that use a jungle primary have a four-candidate second round. California and Washington send the top two vote-getters to the general election; Louisiana holds its jungle election on Election Day, and then holds a two-candidate runoff if nobody wins a majority. Thus, each of these systems ends up forcing a binary choice in which the winner is the first choice of a majority of the state’s voters, once they are given only two options. With the exception of a majority winner in the first round in Louisiana, who in most cases is an incumbent, the voters make that choice knowing who the two finalists are.
Rank Chaos
Not so in Alaska. Instead, Alaska combines the jungle system in the first round with ranked-choice voting in the four-candidate second round. Alaskans are asked to sort their picks from first choice to fourth. If a candidate wins an outright majority of the first-place votes, he or she wins, as Collins did in Maine’s ranked-choice general election in 2020. If there is no majority winner — and a four-candidate field makes it harder to have one — the system then requires multiple rounds of vote-retabulating. What happens is that the lowest-ranked candidates are eliminated one at a time, and the remaining candidates are then re-ranked, with the ballots of the eliminated candidates converted into ballots for the second-choice candidate on that ballot. This process continues until somebody reaches a majority. Conspicuously, given Murkowski’s history, Ballot Measure 2 did not eliminate write-in candidates, although it did raise the numerical bar for them to win.
The complexity of the system forced one change on Alaska: It eliminated independent primary elections for lieutenant governor. Previously, the state had contested primaries for the office, but the state constitution requires the lieutenant governor candidates to run on a ticket with the gubernatorial candidate. Thus, Palin ran for lieutenant governor in 2002, finishing second; had she won, she would have been the running mate of Frank Murkowski (Lisa’s father), whom she ended up defeating in the gubernatorial primary in 2006. Due to the interaction between Ballot Measure 2 and the state constitution, however, lieutenant governor candidates are compelled to join a ticket with the gubernatorial candidate before the primary.
In this year’s races, the primary was scheduled to be held on August 16. On the Senate side, that resulted in a final four of Murkowski, Trump-backed conservative Republican Kelly Tshibaka, Democrat Patricia Chesbro, and a third Republican, Buzz Kelley. On the House side, the final four is Peltola, Palin, Begich, and Libertarian Party candidate Chris Bye — thus setting up nearly a rematch of the special election.
The special election, however, held its primary on June 11. It was a circus, with 48 candidates, including a white-bearded city councilman from North Pole, Alaska, who goes by the name of Santa Claus. Palin got 27 percent of the vote, easily the most in the field; Begich got 19 percent, former Democratic senate candidate Al Gross (running as an independent) got 13 percent, and Peltola got 10 percent. Gross dropped out of the race on June 20, leading to litigation in which the Alaska supreme court ruled that the fifth-place Republican could not take his place on the August 16 general election ballot. Thus, the general election was a three-way race between Peltola, Palin, and Begich. The state’s Democrats united behind Peltola — she was even endorsed by Begich’s uncle, former senator Mark Begich — while Palin and Begich divided the Republican vote.
Maine is the only other state that uses ranked-choice voting in statewide elections. The system is also used in some cities such as San Francisco and New York. In none of these places does it have a long pedigree: It was used for the first time in Maine in 2018 and in New York City in 2021. The ranked-choice voting system in San Francisco was instrumental to the election of left-wing extremist district attorney Chesa Boudin, who was not the first choice of a majority of the city’s voters and who ended up getting recalled by 60 percent of them once the voters saw him in action.
In Maine in 2018, it took over a week to decide who won the House race for the state’s second congressional district. It wasn’t the candidate who got the most first-place votes, so his lawsuits were not resolved until Christmas Eve.
In New York, which already had serious problems conducting timely and accurate vote counts, the introduction of the additional layer of complexity of ranked-choice voting resulted in chaos and delay in the city’s Democratic primary for mayor, held on June 29, 2021:
The New York City Board of Elections called for “patience” from candidates and the public after discovering a “discrepancy” in the latest vote tally of the Democratic mayoral primary . . . before abruptly removing the data from its site. . . . In a subsequent statement posted online — with the data removed, and confusion and frustration surrounding the tally mounting — the board explained that it had failed to remove sample ballot images that were used to test its ranked-choice voting system, so when the board ran its tally, it included “both test and election night results, producing approximately 135,000 additional records,” the statement said.
Eric Adams, who won 31 percent of the first-place votes — the most of any candidate — issued a statement charging that the screwup “rais[ed] serious questions” about “irregularities” in the election. Maya Wiley, who had the second-most first-place votes with 21 percent, howled that this level of “mismanagement . . . has resulted in a lack of confidence in results. . . . The BOE must now count the remainder of the votes transparently and ensure the integrity of the process moving forward.”
New York City allowed a week for absentee ballots to arrive. It released unofficial tallies of the first-place votes, but waited until the absentee deadline, July 6, to run the ranked-choice system for the first time, without the absentee ballots. For a time, the city hung on the prospect that late-arriving absentee ballots in a close race might mean that the primary would be decided in mid or late July by people who voted from Florida or the Hamptons.
As it turned out, Adams was widely declared the winner on July 6, after eight rounds of eliminations. Wiley, who was in second place until the seventh round, was eliminated from contention before the final ballot. Kathryn Garcia, who trailed Adams by over 100,000 first-place votes, ended up just 8,000 votes short of what would have been a shocking upset.
The Sled Dog Run
In New York, the practical problem with implementing ranked-choice voting was the ineptitude of the Board of Elections, although the late-ballot problem didn’t help matters. In Alaska, the practical issue is the geography of the state, exacerbated by a deliberate policy choice in implementing the new system. Alaska is vast, and many remote parts of the state are all but impossible to reach except by airplane or boat. That’s not a problem in counting in-person votes at the polling places, which can be tabulated locally, after which the totals are reported to the state. But when physical ballots need to be moved, it takes a lot of time.
In an election held under normal conditions, Alaska is like California or New York: It starts reporting in-person votes right away, but it takes weeks to count all the absentee ballots, which are being delivered to the state and can arrive up to 10 days later, 15 if coming from abroad (such as military ballots). That normally has little effect unless a race is very close, in which case the need to count the absentee ballots can cause the outcome to hang in the balance for quite a while. In 2020, however, things were different due to the dramatic increase in vote-by-mail:
Alaska won’t start tallying its remaining ballots — at least 40 percent of the total — until [a week after the election] at the earliest, making the state stand out as a gray island in the ubiquitous red and blue electoral vote maps used by national outlets. [Three days after Election Day, it’s] the only state to have counted less than 60 percent of its votes, according to figures collected by The New York Times.
The timeline is one that Alaska has used before. But in past years, the absentee vote count has typically been an afterthought that affects only the closest of races. This year’s massive, pandemic-driven absentee turnout has changed that.
State officials said the wait stems from Alaska’s huge size and complicated logistics: It has polling places in dozens of villages with no road access. Officials said they also need the extra week to finish the painstaking process of logging the names of each Alaskan who voted on Election Day, then cross-referencing with absentee ballots to make sure no one’s votes are counted twice.
Thus, in 2020’s senate race, Democratic partisans were still holding out hope for Gross two days after the election, even though he ended up losing to Dan Sullivan by 13 points, an outcome called eight days after the election. This time around, the addition of ranked-choice voting made things worse, due in part to the state’s decision not to run the ranked-choice tabulations at all until it had all the absentee ballots in hand 15 days after the vote:
With ballots arriving as late as Aug. 31, elections officials here won’t pull the trigger on the state’s new ranked-choice voting system until all the ballots are in. “The tabulation rounds will not take place until after all the ballots that are eligible to be counted on day 15 (are received),” said Gail Fenumiai, director of the Alaska Division of Elections. . . . Preliminary results . . . will show the first choices of voters.
When Maine implemented ranked-choice elections, the process involved trucking ballots to that state’s elections headquarters. In Alaska, that’s not possible — there’s no road to Juneau. . . . Ballots will be scanned on Election Day, generating preliminary reports that will be released after polls close.
In urban Alaska, the paper ballots will be locked up and transferred to one of the state’s four regional supervisors, then picked up by contracted courier for a flight to division headquarters in Juneau. In rural Alaska, ballots will be sealed in envelopes that make it evident if they’re tampered with, locked in containers and shipped by the U.S. Postal Service using package tracking, signature confirmation and a signed check log. . . . In about one-quarter of the state’s voting precincts, paper ballots are tallied by hand, and those ballots need to be scanned in order to tabulate the votes on Aug. 31.
In other words, the problem with finalizing a result was previously only the need to count absentee ballots, so the outcome was in doubt only so long as there were enough such ballots outstanding that they could make a difference. While the public waited for those ballots to arrive, it could see the numbers compiled so far. This was transparent.
Now, because voters mark multiple choices, and their ballots need to be physically delivered to the state to count them in the ranked-choice-tabulation round, and because the state doesn’t do any of the tabulating until all of the ballots are in place, it will be impossible for any remotely competitive Alaska election to be resolved in less than 15 days.
This is not an improvement. It will not help public confidence in an era in which delay and lack of transparency are major sources of conspiracy theories.
Mischief Unmanageable
Even when perfectly implemented, ranked-choice voting can result in dramatic shifts from the initial vote count to the ultimate winner. Those kinds of shifts inevitably feed paranoia about the integrity of the counting process, and the complexity of the process makes it harder to demonstrate transparency.
Moreover, the goal of ranked-choice voting is to make it harder for factional candidates to win pluralities — but who is likely to be a factional candidate with strong first-choice support but far less second- or third-choice support? Quite often, candidates who appeal to a large racial minority group. That is why the move to ranked-choice voting in New York City met with stiff opposition in November 2020:
The City Council’s Black, Latino and Asian Caucus wants to postpone administering New York City’s new Ranked-Choice Voting system for citywide and council races in 2021. . . . They fear inadequate outreach and education during the pandemic about the complex new voting system will disenfranchise minority voters. Fifteen members of the minority caucus sent a Nov. 19 letter to Council Speaker Corey Johnson recommending a “pause” in the law — with the goal of delaying implementation of Ranked-Choice Voting for up to two years.
New York has numerous and deep fissures among its voting blocs along a dizzying array of racial, ethnic, and class lines. It typically has not taken very much in the way of controversy around an election — whether it is a primary or general election — for Democratic politicians and community leaders to start trading charges of racism, voter suppression, and rigged elections. Going into the vote in 2021, Adams was already charging his rivals with suppressing the black vote.
By contrast, what kind of candidate is likely to do well in a complicated voting system that rewards organized strategy? Ideological candidates who appeal to highly educated voters. The 2021 New York Democratic mayoral primary was nearly a perfect storm for the massive, racially polarized blowup that will sooner or later result from ranked-choice voting. The winner among first-place ballots was Adams, a black former cop favored by the city’s black and Hispanic neighborhoods and scorned by educated progressives. Garcia, who jumped from third place to almost overtake Adams in the eighth round of the ranked-choice tabulation phase, is a white woman favored by the city’s upscale Manhattanites and endorsed by the New York Times. If the ranked-choice process had supplanted Adams with Garcia, New York would have gotten very ugly in a hurry.
A particular problem is disparities in which voters are likely to have “exhausted” ballots — what happens when a voter leaves some or all of the rankings blank below the top line. As I explained last June, if you adopt the broad theory favored by many progressives that racial disparities constitute a violation of Section 2 of the Voting Rights Act, disparate rates of exhausted ballots among non-white voters could spawn a troublesome lawsuit.
Begich the Question
In the Alaska special election, exhausted ballots were crucial, but so was the Palin–Begich split. Because the race pitted one Democrat against two Republicans, Peltola got the most first-place votes and would have won anyway due to the jungle primary system without ranked-choice voting. On the other hand, a jungle system that narrows the race to two choices, as is done in California, Washington, or Louisiana, would have resulted in Palin defeating Begich.
Palin’s problem in the ranked-choice tabulation phase was simply that only half of Begich’s voters preferred her to Peltola: 27,659 out of 52,504, and 16,399 Begich voters, 31 percent of the total, affirmatively chose Peltola over Palin as their second choice. But what about the 10,726 Begich voters whose ballots were exhausted? Palin needed fewer than half of them to overcome Peltola’s lead. Had the race been a head-to-head choice, would they all have stayed home? Did some of them misunderstand the system? We can’t know, and Palin surely bears a lot of the blame for failing to persuade or inform them; she will have to do better in November. Either that, or somebody in the party will need to get either Palin or Begich to stand down, as Democrats did with Gross. But in any event, this isn’t really how an American election should be resolved.
Keep It Simple, Stupid
Voting, in our system, is supposed to be deliberative, so for all of the hue and cry from progressives about “voter suppression,” there is nothing wrong with requiring voters to do things like register to vote in advance, show up on Election Day, and bring an ID. That said, democracy is for everyone. By definition, half of the voters are of below-average intelligence. By definition, half of the voters are of below-average education.
The fundamental American system of elections has a basic simplicity that inspires confidence. The essential model is that each party picks a candidate, there is a choice between the two, and whoever gets the most votes wins. All you need to do is look at the totals for the top two, and you can see that. At the presidential level, this is done in 50 statewide races (plus D.C.) rather than aggregating those races nationally, for the same reasons why members of the Senate and House are elected by the people of districts or states, but the principle is the same. Sure, there are third-party candidates, but those are a familiar feature, and they usually lose. After two centuries, Americans are long accustomed to the simple idea of two choices, one vote, the most wins. We should not mess with what works simply because we think that tweaking the system in ways that look clever in a political-science class will have predictably good effects.
A consistent preference for the basic American binary-choice, first-past-the-post model is why I have argued consistently against jungle primaries, against general-election runoffs, against recalls, and against the Maine and Nebraska systems of running presidential elections by congressional district. At most, we should use runoffs in party primaries, and only when the leading vote-getter falls below a low threshold (say, 40 percent) that suggests that the party has not really made a choice. But primary runoffs are about picking the best candidate for the party, not respecting the democratic legitimacy that is central to general elections.
The need for speed and transparency in delivering election outcomes that voters trust is also a reason to support hard and consistent deadlines in ballot submission and early canvassing of absentee ballots to allow as many of them as possible to be counted on or immediately after Election Day. Donald Trump’s six-week tantrum in 2020–21 illustrates how badly things can go awry when the outcomes of elections are left unsettled or contested for long periods.
It’s time to end all of these experiments before they produce something we really live to regret.