It is my understanding that the residency rules for our federal offices are set by the constitution. We may not be able to change that. But the fact that a candidate gets to choose their party label at the time they register to run (no matter their prior registration) is a choice we can change.

More thoughts from my door knocking. And more moose. Being involved and supporting neighbors is important.

These area moose were more interested in each other than the fall colors. 🚪✊🫎👀

As I head out to knock on doors in the rain this afternoon, I thought I’d share the moose I encountered while door knocking yesterday.

Door knocking in Alaska.

She is awesome. She kicked *** on our behalf as our midwife for the delivery of our first daughter. Our household are big fans!

In college working tourism in Fairbanks and Denali, the interaction was: Q - So, when does it get dark around here? A - August Q - No, really, I’m having trouble sleeping. When does it get dark? A - August.

Kids are our future - so glad they are paying attention. This specific group of teen authors caught my eye because I know most of them. Two were elementary school classmates of my kids, one is the child of friends I studied with at UAF, and one is a current classmate of my middle daughter.

I was out door knocking in the rain tonight - and look who I ran into! Matt Schultz has been an amazing advocate for justice and compassion in our community. Anchorage is blessed to have so many people working hard to for a better future.

The extra 10 votes -plus how ever many people bullet-voted a write in (an unknown number between 0 and 21) - came into the first round of RCV because their 1st rank column was blank, but they had a vote in their 2nd rank column.

Total initial vote count is 3032. 1st round RCV is 3042. 21 write-in votes existed in the initial 3032 count. They were allocated in the 1st round of RCV. That allocation cannot cause the total vote count to go up. It can only stay the same or go down if anyone bullet voted write-in.

It goes up because the 1st round of RCV also counts the 2nd ranks on ballots where their 1st rank is blank - because that’s what our statutes call for. In 2022, this pattern existed on most of the RCV races. It existed less in 2024, but check District 40.

The summary table shows that a ton of people choose not to make a second rank. This logic applies to the write-ins as well. With this logic, the total vote count should have gone down with the distribution of the write-ins. But sometimes it goes up.

The state’s basic election-night system literally only counts the votes in the first rank column (the state statue does not read this way, but it’s close enough). When they put it in the RCV system, round 1 of RCV has already re-distributed the write-in votes to their second rank.

Since the weeds seem interesting for this discussion, here is a gem: the way you can discern the existence of ballots with the first rank blank but votes in rank 2 is because the total vote count goes up between the original non-RCV count and the first round of RCV in some races.

The JSON has all the computer legible data. It has metadata re: where & how a ballot was cast. It has all the ranks for all the races, including over-votes & blanks. The fact that the summarized table has no skipped ranks or write ins suggest some counting logic has been applied before summarizing.

I haven’t parsed this specific race before, but at a quick glance, this table is missing two categories of votes: write-in votes at any rank, and votes where #1 is blank, but #2 has a name. There were several thousand of the latter vote types in the statewide races in 2022.

The raw data is downloadable from the state. It is a HUGE data set, and parsing it is not easy. I did a deep dive into the data in my district. I did not have time or resources or interest to do the same in other races, but the full data is available.

18% of Hafners voters went to Howe next. 19% of Howes voters (including Hafner voters who chose him 2nd) went to Mary. In the RCV calculation logic from the state, a Howe -> Hafner -> Mary/Begich vote shows as Howe-> Mary/Begich and a Howe -> Hafner -> blank vote is exhausted in the Howe pie chart.