With things somewhat up in the air about Wisconsin, Alaska is the next state to complete its voting.  Alaska is one of the states that moved from a caucus to a party-run primary since the last cycle.  In response to Covid-19, the Alaska Democratic Party has moved to a vote-by-mail election.  Ballots are supposed to be received by April 10, and the state party will release the results on April 11.

Alaska is using a ranked-choice ballot with candidates being eliminated (starting with the candidate with the least votes) until all of the remaining candidates are viable.  With only two candidates still running, even if some small segment of voters who voted early have another candidate as their first choice, the final count after preferences are distributed should leave only Senator Sanders and Vice-President Biden with votes on the final count.  Under the rules for the primary, because only viable candidates will be left in the final count, there is no difference between the final total votes and qualified votes.

What makes delegate math unusual in Alaska is the fact that it is a single-district state.  Under the rules, the allocation of delegates is calculated separately for each of the three pools, but the state-wide results are used to allocate all three pools.  The fact that the pools are calculated separately gives a different result than if all the delegates were allocated as one pool.

For Alaska, there will be nine district-level delegates, two party leader delegates, and four at-large delegates for a total of fifteen delegates.

The math for the two party leader delegates is simple.  Assuming that both candidates are viable, each candidate will get one delegate.  (And if only one candidate is viable, you can forget the rest of the math as that candidate will get all fifteen delegates.)

For the four at-large delegates, the key split is 62.5%-37.5%.  If one candidate finishes over 62.5%, that candidate gets three delegates to one for his opponent.  If neither candidate reaches 62.5%, each candidate will get two delegates.

For the district-level delegates, it will take 16.67% to get a second delegate.  At 27.78%, a candidate will win a third delegates.  A candidate will assure themselves of a fourth delegate at 38.89%.  A candidate will get five delegates if they get the majority of the votes.  At 61.12%, a candidate will earn six delegates.  A candidate needs 72.23% to assure themselves of seven delegates.  Finally, a candidate can get eight district delegates with 83.34% of the vote.

With all of the votes being at the state level, the following chart shows the weird choppiness of the delegate distribution

15.00% — 3 delegates (1 district, 1 party leader, 1 at-large)

16.67% — 4 delegates (2 district, 1 party leader, 1 at-large)

27.78% — 5 delegates (3 district, 1 party leader, 1 at-large)

37.751% — 6 delegates (3 district, 1 party leader, 2 at-large)

38.89% — 7 delegates (4 district, 1 party leader, 2 at-large)

50.001% — 8 delegates (5 district, 1 party leader, 2 at-large)

61.12% — 9 delegates (6 district, 1 party leader, 2 at-large)

62.501% — 10 delegates (6 district, 1 party leader, 3 at-large)

77.23% — 11 delegates (7 district, 1 party leader, 3 at-large)

83.34% — 12 delegates (8 district, 1 party leader, 3 at-large)

85.001% — 15 delegates (9 district, 2 party leader, 4 at-large)

The key parts pf the chart are in the area around a 62-38 split and at 84-16.  In the area around 62-38, a swing of 1.5% will impact two delegates.  At 63-37, the split is 10-5.  At 61-39, the split is 8-7.  Likewise, in the area around 84-16, a swing of around 2% will impact four delegates.  At 83-17, the split is 11-4.  At 85.001-14.999, the split is 15-0.

There are no polls of Alaska.  That makes guessing about what might be happening rather difficult.  Given the national numbers, I expect both candidates to get at least 30%. (The 2016 spilt was Sanders beating Clinton 79.61-20.23.)  So the winner will probably get between 8 and 10 delegates and the loser between 5 and 7 delegates.  The key questions will be who wins and can that candidate get to the 62.501% needed to get the 10-5 split.  A net of one for either candidate basically preserves the status quo.  Given that Sanders won Alaska in 2016, a net of five for Biden is one more nail in the coffin of the Sanders campaign.  A net of five for Sanders is a necessary first step to closing the delegate gap into something that Sanders could bridge with the June primaries.

If things go as currently planned, after the Saturday results in Alaska, we will get the results on Monday in Wisconsin, and another set of results the following week from Wyoming.  Puerto Rico has announced that the April 26 primary is being postponed (no new date has been set yet).  The last of the April contests will be the conclusion of Ohio vote (now entirely by-mail).  I am assuming that with Wyoming and Ohio being vote by-mail that the vote will be concluded as currently scheduled.  Counting, particularly in the case of Ohio, might be somewhat delayed.  (For example, California is still counting and the deadline for completing the count has been extended until April 24 although over 99.8% of the ballots have been counted with less than 70,000 ballots remaining in eight counties as of the close of business on April 3.)

There is an apocryphal Chinese curse of “May you live in interesting times.”  The 2020 cycle — with a public health crisis dropping into the middle of an election campaign — certainly meets the criteria of interesting times.  For almost forty years (ever since candidates and state election officials recognized that the voters would actually choose the nominees of the two parties), primary dates have attempted to crowd the front of the election calendar under the belief — which has mostly been true — that late states would have little influence on who wins.  Covid-19 has hollowed out the middle part of the calendar and is creating a backloading of a significant number of primaries.  Whether that has the effect of extending the race to the end remains to be seen.  But, for now, the race continues although at a much more glacial pace than anyone would have expected last November when the first cases of Covid-19 were occurring and nobody had yet heard about it.

UPDATE:  Now that we have the results, ranked choice voting had little impact on the delegate allocation.  After the first round of voting, both candidates were over 39% and Biden was at 49.9%.  So Biden only needed a handful of votes to get the majority and the 8-7 split (or for enough votes to drop out in subsequent rounds).  If Alaska had not used ranked choice voting, the percentages of the qualified vote would have been roughly similar  to the percentages after the preferences were distributed and the split remained the same.  (Sanders did 0.5% better under ranked choice voting.)  What is unknown, of course, is how much Sanders announcement on Wednesday that he was suspending his campaign impacted the results.  With mail-in ballots, most ballots were probably already in the mail at the time of the announcement.  Approximately 8% of the ballots cast for the “other” candidates dropped out before the final count — either failing to give additional preference or neither of the top two candidates making the voter’s top five.