In 2010, California adopted a “top-two” open primary system, in which the primary is nonpartisan and open to every voter. The top two vote-getters—regardless of party—advance to the general election. This was the result of voters’ approval of Proposition 14, which was designed to overcome the extreme gridlock from which the state government had suffered. (The state of Washington adopted a similar proposal in 2008.)
Because candidates who don’t compete in the first round are ineligible to appear on the final ballot, this system is roughly equivalent to a two-round runoff system, such as that used in the presidential elections of France. The difference is that in the French system, the second round is only necessary if no candidate wins an outright majority in the first—which is usually the case given the number of parties competing. In California, both rounds are held regardless.
Despite its popularity in some circles, the top-two system suffers from a major problem: it does not promote a truly competitive political landscape where voters have several choices, presuming instead that they should be given an either-or choice. It can also eliminate potential winners in the first round, leading to a less relevant second round—as was exemplified in the French presidential election of 2002.
In that year, no fewer than sixteen candidates for president competed in the first round. Conservative Jacques Chirac, the incumbent, and Socialist Lionel Jospin were considered the frontrunners, with (far right) National Front candidate Jean-Marie Le Pen running a close third. But the electorate was fragmented: these three accounted for only about half of all voters, in polls conducted just before the election.
In the official first-round tally, Le Pen managed to edge out Jospin to get into round two. Chirac got 19.88%, Le Pen 16.86%, and Jospin 16.18%. The nation then rallied around the defeat of Le Pen, and Chirac won the final balloting with over 82%.
That’s what I mean by an irrelevant question. Given the choice of candidates, for most French voters the relevant two-candidate question would have been between Chirac and Jospin. But due to the mechanics of the top-two system and the number of parties contesting the election, they didn’t have the chance to answer that question. Basically, the system fell victim to vote-splitting. Lionel Jospin, who observers thought would run very close to Chirac in the second round, fell victim. Could a different system have prevented that from happening?
What if fewer candidates had been eliminated after the first round? In instant-runoff voting (IRV), only the worst-polling candidate is dropped in each round, until one candidate crosses the 50% threshold. This means that given N candidates, it may require as many as N-1 rounds of tallying to identify the winner. Voters rank their choices on their ballot, eliminating the need for a series of expensive polls to be held—thus the term “instant”, as opposed to standard, runoffs.
When a voter’s top choice is eliminated from contention, their vote is transferred to the next candidate on their ballot for future rounds. In this way, voters can express their true preference, even if it’s a minor-party candidate who is unlikely to win, without giving up the tactical value of their vote. In later rounds, if the preferred candidate has been eliminated, the vote will default to an acceptable mainstream candidate who can unite the opposition to an unacceptable one.
For an example of how IRV (also known as ranked-choice voting) works, watch this short video recounting its use in the 2012 Oakland city council election. In this case, the first-round plurality winner, Dan Kalb with 28% of the votes, was also the overall winner. But it doesn’t always work out that way. In the District 3 race, the eventual winner placed second in the first round, because her constituency was split. But as some of her ideologically friendly competitors were eliminated in successive rounds of tallying, she gained votes and became the majoritarian winner.
Runoff systems, whether two-round or instant, are primarily designed with this majoritarian goal in mind—to make sure that a plurality winner who is disliked by most voters doesn’t win because of vote splitting. But that’s most effectively done by dropping only one candidate per round, as in IRV. The two-round system takes a “shortcut”: drop everyone but the top two, all at once. Unfortunately, taking this shortcut can backfire, throwing the baby out with the bathwater.
A two-round runoff works fine if there are just two or three candidates, but the more the number exceeds three, the more the chance that vote-splitting will cause a major malfunction. So the question becomes, how many choices do we Americans need? Are two or three enough? It was clearly nowhere near enough for France—so why should it be for us, the land of free enterprise? Shouldn’t we have a free marketplace of ideas, after all? And if we accept that we should, then how well does the top-two system perform, compared with IRV?
I would bet Lionel Jospin has an opinion.