This week, I have a piece in the New York Times singing the praises of contested conventions. It being an op-ed, I didn't have the space to elaborate on some of the details. So I have been elaborating here.
The short version of the argument is this: A convention of unbound delegates would be a great way to choose a nominee. And we could get that, at least sometimes, if the parties used proportional representation and a shortened calendar. Then if we had a divided field, each candidate would be represented at the convention in proportion to his or her popularity. But this would require taking delegate selection seriously.
Part II: Proportional representation and a shorter calendar
Suppose you buy my argument that a real convention is better than what we have now. How could we get there?
A contested convention is already technically possible. Right now delegates are bound to vote for the candidates to whom they were allocated on the first ballot. (In fact, every state has its own specifics; the drama on the floor this week was largely over some delegations wishing their own states rules weren't what they were.) But if no candidate wins an outright majority on the first ballot, the delegates are generally freed for later ballots.
So all we need to do is make it easier for the primary process to end with no candidate with a majority. As a bonus, what we would do to get there would make the system more democratic, not less.
Primary elections are governed by state and national laws and by state and national party rules. If we wanted major changes, we'd need to get all of those actors to go along. But the national parties can do things to tweak the system. In particular, they can incentivize states to move in small directions, and they can make some rules narrowing the states' options.
Two things are well within their power. First, they can encourage proportional representation.
Proportional representation is used in democracies around the world for the selection of legislators. It is well known that PR allows for multiple parties to win, whereas plurality rule tends to winnow the options down to two. Plurality rule does this by exaggerating the support for the winner, misrepresenting the voice of the people.
The current system is designed for winnowing. For the Republicans in 2016, early states did use some form of proportionality, but later states use winner-take-all. This helped Donald Trump win about 60 percent of the delegates with only 45 percent of the vote.
The Democrats use some kind of proportional system in every state, but they have high thresholds for marginal candidates, encouraging them to drop out.
A truly proportional representation system would match delegate shares to vote shares as closely as possible. Assuming everyone voted as they did, that would mean about 45 percent of the delegates for Trump and about 25 percent for Ted Cruz. Marco Rubio and John Kasich would each have more than 10 percent. Trump would be in the driver's seat, but the others could force a compromise if they truly were not happy with him.
But everyone probably wouldn't vote as they did under the old rules. The votes were very affected by the long calendar. Jeb Bush ended with less than 1 percent of the vote, but he did much better in the three states he competed in. If you need to win it all, then it's hard to stick around after doing poorly in Iowa and New Hampshire.
That is fundamentally undemocratic. Fewer than 1 percent of Americans live in Iowa. Fewer still live in New Hampshire.
When he dropped out of the race this year, Bush said, "The people of Iowa and New Hampshire and South Carolina have spoken" — but the people who had spoken make up less than 3 percent of the country.
But if we aren't interested in getting down to one candidate, we don't need to encourage people to drop out. It's a long slog from January to June. A second-choice-of-many candidate may not be able to stick around. A single-day primary with proportional representation would even work, but that's not going to happen. The parties can nudge, but they can't take away states' primaries. And anyway, neither Iowa nor New Hampshire will give up its privileged status.
A shorter calendar, though, would make it easier for everyone to stay in. The long process is thought to help underfunded candidates claw their way to the top. But that doesn't really happen now anyway. And I'm not sure that someone who exploits an idiosyncratic calendar really represents the people.
It is true that a shorter calendar will require candidates to build support in the invisible primary. They are already doing that, though. The main lesson from the past several cycles is that the invisible primary has become more visible, allowing outsider candidates to build their own support networks. So a shorter calendar doesn't disadvantage them anymore.
Both of these changes — PR and a shorter calendar — could be incentivized. They would give us contested conventions. They'd also give us more democratic primaries.