ELECTION WON’T BE CERTIFIED UNTIL SIX DAYS BEFORE SWEARING IN

The new election system approved by voters in 2020, known as Ballot Measure 2, has created a system that may lead to further distrust in the election, if counting and reporting of the numbers is not handled carefully, the state’s own director of the Division of Elections implied during a Senate hearing on Tuesday.

Election Director Gail Fenumiai, a veteran manager of many elections in Alaska, told the Senate State Affairs Committee last week that the division will release results of the ranked choice voting general election in just two stages — once on Election Night, Nov. 8, when all the ballots received at that point will have their first choice candidates counted — and then no more results will be released until the deadline has passed for mail-in and absentee ballots, which is Nov. 23.

That is a full two weeks after the election, and unlike with normal elections, where results are updated daily, with Ranked Choice Voting, the public won’t be able to see a trend, and many of the initial results may flip.

Fenumiai said that if her office released updated results every day between Nov. 8 and Nov. 23, the results could vary so widely that it could lead to mistrust in the counting process.

The counting process for ranked choice voting is done by a computer system with an algorithm, or computer rules that are internal to the counting software. Those voters whose first choice candidate was not successful in reaching the majority of votes will have that vote crossed off and their second choice vote will move up to be their first choice vote.

Watch Elections Director Gail Fenumiai explain the risk involved with releasing tallies daily at this link.

Sen. Mia Costello said that the problem is structural with the way Ranked Choice Voting is designed.

“You really can’t go beyond the first round [in counting] until you have every single ballot because you have to know if somebody is the outright winner. So it seems in this age of technology and modernization and progressive way to vote with ranked choice voting what we find is that we are just adding more time before the voters actually know the results of the election,” said Costello at the end of Fenumiai’s presentation. “I understand that now. If we are going to allow ballots to appear 15 days after the election, then we can’t progress past round one until all of those are counted, because you could already have a winner that is just out there in the mail. There are people serving in this Legislature who have won elections by a coin toss.”

Watch Sen. Costello’s comments at this YouTube link.

After an hour and a half of discussion in Senate State Affairs, it was clear that lawmakers have a hard time understanding Ranked Choice Voting and explaining it. Many suspect the public will have difficulty as well.

Sen. Scott Kawasaki, however, underscored the need for the Division to educate the public on how to vote with the new system, so they can “vote for us.”

The Fairbanks Democrat emphasized that Ranked Choice Voting is now the law of the land and its merits should not be debated any further, but rather lawmakers should focus on getting the people comfortable with it.

Watch Sen. Scott Kawasaki talk about how the focus should be on getting the public comfortable with Ranked Choice Voting here.

Sen. Mike Shower, who chairs the committee, said that the people who will be most likely disenfranchised by the new system are Democratic voters, those who are elderly, those who do not have English as their first language, and those who have disabilities. That’s because their ballots are most likely to be “exhausted” for any number of reasons, but primarily because they may not understand how Ranked Choice Voting works and how to correctly mark a ranked choice ballot. Some voters will not necessarily have time under the new system to fix a ballot they’ve made an error on, such as if they mistakenly voted for two people as their first choice.

Other interesting aspects of the new voting system:

• Write-ins are not allowed in the primary.
• If there are only four people in a race in the primary, all will advance to the general election ballot, where Ranked Choice Voting takes place.
• The order of where candidates are seen on the ballot will be random.
• Write-in candidates must file with the Division of Elections to be qualified.
• There appears to be a clear advantage to going to the polls in person, rather than voting by mail, so officials can give voters a new ballot to mark if they make a mistake. But the ballot they will be handed is likely to have the names of the candidates in a different order.
• Hand counts won’t happen with Ranked Choice Voting because it is too complicated.
• On Election Night, only first vote candidates will be counted and announced.
• Ranked Choice Voting requires centralized tabulation.
• Votes can continue to arrive at the Division of Elections up to 15 days after the election.
• The final tabulation on Nov. 23 should go quickly since it is all done by computer algorithm. That is one day before Thanksgiving Day, Nov. 24.
• The election will not be certified until Nov. 29.
• Dec. 2 is the deadline for requesting a recount for the governor and lieutenant governor’s race. Dec. 4 is the deadline for requesting a recount for all other races.
• The governor, by order of the Alaska Constitution, will be sworn in on Dec. 5, just six days after the election is certified, and three days after a recount challenge can be filed.
• The new governor won’t know he or she is governor until the last minute and won’t have time to start transitioning into office, preparing a budget for the Dec. 15 deadline, and assembling a cabinet.

Shower said that national studies prove that poor people are disenfranchised by the system that is used by the State of Maine and the City of San Francisco.

Shower said that a Princeton professor showed that the average disenfranchise rate is 11 percent of the vote, which means of 400,000 voting Alaskans, more than 40,000 are likely to have their ballots tossed. But as much as 20 percent of the ballots could be marked invalid.

According to the Princeton report by Nolan McCarty, Ph.D., Professor of Politics and Public Affairs at Princeton University:

1. RCV resulted in a substantially lower “full participation” rate in Maine in 2018 as compared to plurality and runoff systems, where “full participation” means casting a ballot that could not be exhausted and thus is guaranteed to count toward the final outcome. This is particularly true in jurisdictions like Maine with more elderly and less- educated voters.
2. These results, as well as the high number of ballots cast that lack any clear rational explanation, demonstrate that the low “full participation” rate in Maine cannot be explained by deliberate voter choice alone. Indeed, the results demonstrate that voter confusion causes many voters not to fully participate. The inherent complexities of the system are preventing voters from fully participating and thus effectively disenfranchising large numbers of voters.
3. The purported benefits of RCV have not manifested in jurisdictions where RCV has been utilized over long periods of time.

“As I outline in my report, an RCV system comes with a significant number of vices, many of which manifested themselves in the 2018 Maine elections. Chief among them is that the system provides many significant impediments to full participation of the voters who choose to cast ballots,” McCarty wrote.

“Central to this issue is the phenomenon of exhausted ballots. In an RCV election, ballots may become unusable in later rounds of tabulation when the voter has failed to rank any of the candidates that remain in contention. When such a ballot is cast aside after the first round of voting for this reason, it is said to be exhausted, and it is no longer counted for purposes of determining the ‘majority’ winner. The academic literature and the analyses in my report demonstrate that ballot exhaustion is pervasive in RCV elections, sometimes leading to the discarding of over 20% of the ballots during the final round of tabulation. It also appears to be persistent, as rates of exhaustion do not decline over time. Jurisdictions that have used RCV for decades suffer from ballot exhaustion at similar rates as new adopters of the voting system,” McCarty wrote.

The Princeton study is found here: