ELECTIONS
Alaska’s special election illustrates an unpredictable system
Right and Left alike are rushing to conclusions about partisan advantage of Ranked Choice Voting
Democrat Mary Peltola has defeated Sarah Palin (R) and Mark Begich (R) in the second round of a special election to replace the late Congressman Don Young (R). She did so in the second-round count of a ranked-choice-voting (RCV) tabulation.
Republican opinion leaders are projecting fury.
Left-leaning commentators are urging wider adoption.
Such assessments are premature. Systems like the one in Alaska are inherently unpredictable.
Alaska RCV is not the ‘moderating’ version from Australia
RCV proponents often point to Australia, where the electoral system has been stable for 104 years. There, we hear, the system has a ‘moderating effect.’
Three features of Australian politics make RCV — and the environment in which it operates — fundamentally different from how it works in Alaska. First, the voter must rank all candidates for their ballot to be valid. Second, the country has a system of multiple parties that control use of their labels. Third, as a consequence, parties can cut deals that make voting behavior predictable.
Turn now to Alaska, which has become the poster child for Final Five (Four) Voting. A ‘jungle primary’ sends four candidates to the RCV round. Simple registration with a party confers the right to use its label. There is no requirement that the voter rank all candidates. (As you might expect, such requirements raise constitutional questions.)
Further, multiparty politics in Alaska have evaporated.
The trend below gives the percent of votes polled by minor parties — Green, Libertarian, and Alaskan Independent — in the at-large U.S. House district since 1958. (Notably, the 2000 spike preceded an unsuccessful RCV referendum in 2002.)
Begich voters held the balance
The screenshot below shows what happened to ballots on which Begich was first choice. Because no candidate won a majority in the first-round count, the last-placed candidate (Begich) was eliminated. Begich ballots — 28.8% of them — got Peltola to 51.5% and a seat in Congress. Hence the claim among Republicans that reform is bad. Hence, also, the claim among some Democrats that more places should reform themselves.
But they were not a monolithic bloc…
Yet, 50.3% of Begich ballots went for Palin. If we exclude exhausted ballots from the calculation (more below), Begich ballots favored Palin over Peltola 2:1.
Is that a lot? Maybe.
Is it enough to say that the outcome was due to a deal in either direction? I leave the answer to the reader.
…and an appreciable number of them didn’t rank choices.
Another thing to note is that Begich’s number of exhausted ballots (11,222) exceeds the final-round difference between Peltola and Palin (5,219). We don’t know why these voters didn’t rank the other candidates. Surely, surveys and precinct-level analyses will give us some ideas. This is a consequential and underappreciated area of inquiry.
None of the above should be surprising. Anyone who followed news today has seen all sorts of speculation.
What we haven’t heard much about are the sorts of RCV ‘add-ons’ that reduce uncertainty. Nor have we heard much about a world — why would we? — in which parties make the deals that voters are being asked to make.
