On Tuesday night and into Wednesday, a crowded Democratic primary in Massachusetts’ Third Congressional District blew up my Twitter feed. There were ten declared candidates, and 52 votes now separate the top two, each of which has 21.6 percent support. Because this is Massachusetts, the winner of the primary will win the general election (unless the party splits). That person will claim a congressional district with barely more than 18,000 votes.
What blew up my feed were calls for ranked-choice voting. Not surprising, nor a bad idea. As the MA-based reform group notes, “RCV is like insurance. You don’t need it… until you need it.” This is very different from saying, “RCV lets you vote for whomever you darn well please.” Not that anyone really says that. We all know there has to be a bandwagon effect for second-choice votes at least. (Cox says so.).
But the election we saw was in a single-seat district, filled by plurality rule. If you don’t see coordination in one of those, what makes you think we’d see it for second-place votes in an RCV system? In order for voters to bandwagon, they need to know who the frontrunners are. It doesn’t matter what the system is — first-past-the-post, RCV, even Approval Voting (as Nagel points out in his excellent piece on the Jefferson-Burr election).
There is no question that RCV would produce a winner with more than 22 percent support. The only way for this not to happen would be widespread non-use of rankings below the voter’s first — basically what one Republican wanted to happen in Maine’s first RCV primary last June. This seems unlikely. The most egregious case I know is somewhat less extreme. It was the 2011 San Francisco mayoral election. There were 16 candidates and 12 rounds of counting. According to Craig Burnett and Vlad Kogan, 27 percent of first-round ballots did not continue to that 12th round. But the winner did get 43 percent of ballots from the first-round count. If you think that vote totals mean something, anything greater than 22 percent is better than 22 percent. “RCV is like insurance.”
To get to a “true” majority, though, (by which I mean a majority of first-round ballots) voters need to know who the latent frontrunners are. Or maybe the Invisible Hand needs to stop some candidates from running — if you think true majorities mean something. Also true in a plurality election.
For these and other reasons, strategic voting is important. So is organization.
And more people than political scientists need more second-choice polling.