Oversized majorities and RCV appeal

When you can’t agree on the one you want, but you can agree on the one you don’t.

I mean instant-runoff voting, which goes these days as “single-winner ranked-choice voting.” As readers of this blog know well, IRV manufactures a majority. If no candidate has a majority of first-choice votes, the last-placed candidate is eliminated. Ballots for the eliminated person flow to next-ranked picks on each. Rinse, repeat.

Two developments now catch the eye.

In Maine, the Republican Party has resisted IRV with all its might — admittedly enabled by a few breakaway Democrats. We will see how this faction behaves in the run-up to June’s roll-out.

In Minnesota, the Republicans want to ban IRV, now used in Minneapolis and St. Paul, and wherever else it might take root. State House and Senate bills appeared a few weeks ago, each with a Democratic cosponsor. Due to one or more unknown factors, those Democrats withdrew their cosponsorships.

I think we will see more IRV adoptions — and Republican resistance. Consider:

  1. Democrats struggle these days to agree on a single candidate.
  2. Democrats do agree that any of them would be better than a Republican.
  3. IRV outsources their problem to an algorithm.

This dynamic may not be unique to the left. In other states, it is the Republican Party that cannot agree on a single candidate — yet can agree that they don’t want a Democrat.

If my model is correct, a similar situation propelled the last big spate of adoptions around the election of 1912. In that race, and having failed to secure the Republican nomination, Theodore Roosevelt bolted on his party, effectively handing victory to Democrat Woodrow Wilson.

One view of the 1912 election holds that Roosevelt was an egotist.

A second view of 1912 might note deep divisions within the rapidly growing Republican Party. The Democrats were becoming a white-nationalist party of voting restriction — the “Jim Crow” Dixiecrats, with a few machine outposts in the North. It may be that younger voters found the Democrats repulsive, instead preferring to affiliate with Republicans. Meanwhile, replacement on the Democratic side was not sufficient to offset the passing of the Civil War generation. Wilson’s victory was the only Democratic presidential win between 1896 and 1932.

But swelling Republican ranks did not foment party loyalty. The party had been divided between its “standpat” and Progressive wings. This divide appears to have been playing out at the local level, in the runup to 1912, with more than 30 cases of something known as Buckin voting. Bucklin was a two-choice cousin of today’s IRV, a.k.a. ranked-choice voting.

The 1912 election brought the cleavage to a head. Consider this article from January 1913, in which one C.F. Taylor spells out “the preferential ballot for ensuring election by a majority.” Taylor opens his piece with the story of a fractious Democratic primary in New Jersey — a classic way to show that both sides could benefit. But the rest of the piece is built around a toy, four-way election. There are three “progressive” candidates and one “reactionary.” He writes (verbatim):

The preferential system permitted them to vote first for the man of their choice and then to mass the progressive field against the common enemy—the reactionary machine candidate—with the result that the progressive voters, who are in a vast majority, are represented by a progressive senator.

Here we are in 2018. Many kinds of people claim to be Progressive.