Skip to content

535 Final forecast · Nov 3
Dispatches

More

← All dispatches

Movement note AK-1

Our AK-1 model drops Democrat Eric Hafner's chance from 10% to 0% in a day

The shift reverses the model's recent move toward Hafner, and nothing in our data pins down what drove it between the two runs.

Written by the newsroom · anthropic/claude-opus-5.5 · November 2, 2026 at 2:32 AM EST

Our model now gives Democrat Eric Hafner a 0 in 100 chance of winning Alaska's at-large House seat. In the run of November 1 at 12:32 AM ET, it gave him 10%. That is a move of 9.5 points of Democratic win probability toward Republican incumbent Nick Begich III over one day. It reverses the shift toward Hafner that we reported on November 1.

The projected margin moved with it. Our model's mean margin is now R+20.2, with a 5th-to-95th percentile range of R+32.3 to R+8.1. In the earlier run, the mean was R+9.5 and the range ran from R+21.6 to D+2.6. The current range no longer includes a Democratic margin. Our model gives Begich a 100% chance, up from 90%.

Our data does not show what caused the change. The most recent polling we have is a set of New York Times/Siena University surveys, all fielded October 23-30. Our data does not say when they entered the model, so we cannot tie them to this run.

In those surveys, Hafner is not the main challenger. In the ranked-choice first round, among 576 likely voters, Begich took 47.7%, independent Bill Hill 40.4%, Hafner 6.0% and Jim C. McDermott 5.9%. In the final ranked-choice round, among 562 likely voters, Begich took 51.8% and Hill 48.2%. By that measure the Begich-Hill contest is close. Our model's Democratic probability tracks only Hafner.

Polls cited

  • The New York Times/Siena University · Oct 30, 2026 · n=562 · AK-1
  • The New York Times/Siena University · Oct 30, 2026 · n=576 · AK-1

Every constant behind these numbers is on the methodology page.