On September 17, 2026, AARP published its Alaska Voter Survey, providing data collected from 800 random Alaskans likely to vote in the general election. The survey focuses on providing insights into issues valued by Alaska’s seniors. 63% of the survey participants are aged 50+, 25% aged 35-49, and 12% aged 18-34. According to the survey, Jonathon Kreiss-Tomkins (D) received 55% of the votes in the final round. 198 participants (approximately 25%) did not rank their ballots. Of those 198 participants not represented in the final round, 144 of them did not vote because they were undecided, and 54 of them had exhausted ballots by the final round.
The initial survey results for the gubernatorial election reveal 18% of participants undecided (but intending to vote in November), 7% voted for Treg Taylor, 14% for Dave Bronson, 20% for Bernadette Wilson, and 40% for Jonathon Kreiss-Tomkins. The undecided votes were then removed, and percentages were recalculated, revealing 49% for JKT, 25% for Wilson, 18% for Bronson, and 8% for Taylor among participants who chose at least one candidate in Round 1 (equal to 656 voters after undecided voters were removed).
According to ranked-choice voting rules, the candidate with the fewest votes drops off the ballot, and second choice votes come into play. According to the AARP survey, 82% of participants who chose Taylor as their first choice ranked a second choice— 42% ranked Bronson second, 37% ranked Wilson second, and 4% ranked JKT second. 18% of voters who chose Taylor as their first choice did not rank a second choice, causing their ballots to be exhausted. 8% of 656 (number of votes after undecided voters removed) equals 52 votes for Taylor and of those 52, approximately 9 voters did not rank a second choice, lowering the total number of voters from 656 to 647.
The redistribution of Taylor’s votes left JKT with 50% of the votes, Wilson with 28%, and Bronson with 21% in Round 2.
In Round 2, Bronson dropped off the ballot. 67% of Bronson’s voters ranked either Wilson or JKT next. Wilson received 63% of Bronson’s redistributed votes and JKT received 4%. 33% of Bronson’s voters did not rank either Wilson or JKT, and their ballots were exhausted. 21% of 647 (the new total votes after Round 1) equals approximately 136 votes for Bronson in Round 2. Of those 136 votes, 33% were exhausted— equal to 45 votes— lowering the total number of voters from 647 to 602.
Of the 602 votes still active, JKT won with 55% of the votes (equal to 331 votes) in the final round. Due to ballot exhaustion, approximately 54 participants who voted for at least one candidate were not counted in the final tally.
The survey counts the initial 144 undecided participants as “not ranking.” Adding the 144 undecided voters to the 9 voters who chose Taylor but did not rank and the 45 who chose Bronson but did not rank equals a total of 198 participants who did not rank, or approximately 25% of the initial 800 participants. The majority of the 25% not-ranking participants did not vote at all because they were undecided. Of the 7% participants who chose at least one candidate but did not rank one of the finalists— Wilson or JKT— some may have ranked Bronson #1, Taylor #2. Despite ranking Taylor #2, their ballots were still exhausted because Taylor had already been eliminated.
The survey does not specify how many people bullet voted, meaning picked only one candidate. At least 9 people bullet voted by picking only Taylor and then having their ballot exhausted. The survey does not specify how many of the 45 voters with exhausted ballots after Bronson’s elimination had ranked Taylor #2 or only bullet voted Bronson. The survey also does not specify how many of Wilson’s votes were bullet votes, though 105 of her votes came from voters ranking her either #2 or #3 (19 from Taylor’s voters, 86 from Bronson’s voters). Wilson received a total of 271 votes out of the 602 active votes (45%) in the final round, meaning approximately 39% of her votes came from voters ranking her #2 or #3.
JKT received approximately 7 votes from voters ranking him #2 or #3 (2 from Taylor’s voters, 5 from Bronson’s voters).
The 18% undecided voters could significantly alter the race depending on who they choose for #1 and whether or not they rank. If undecided voters pick Wilson #1 or rank her #2 or #3, she could beat JKT in the final round. Or undecided voters could bump either Taylor or Bronson ahead of Wilson in Round 1— initiating a whole new tabulation scenario. It is also possible that undecided voters land Kreiss-Tomkins a win in Round 1.
AARP’s Alaska Voter Survey: Alaska Voter Survey
