Proportional

[bryant]

As the Republican primary season wears on, there’s a lot of discussion of delegate math. Jed Lewison of Daily Kos keeps making arguments based on raw percentages — Romney now has to win 48.4% of the remaining delegates available to reach the convention with the nomination in hand. I think he’s just doing propaganda, though, because he’s making the implicit assumption that delegate apportions are simple. So I took the delegate count from Real Clear Politics and made a super-stupid, basic spreadsheet.

I assumed that Romney, Santorum, and Gingrich would split the remaining primaries and caucuses with 30% of the vote each; I gave Ron Paul 10% of each state. Pause for outrage, yes, I know. If you split delegates, 30/30/30/10, Romney doesn’t get over the hump. But then I went back and gave Romney all of the delegates from the winner take all states: Puerto Rico, Wisconsin, Maryland, Washington DC, Delaware, California, New Jersey, and Utah. Utah should be pretty easy for him. Most of the other states ought to be easy Romney victories. Wisconsin is coming up soon; at 42 delegates, it’s a big prize and Santorum’s leading by a big margin there. Romney needs to take all the winner take all states to get to the delegate threshold. On the other hand, if he just misses Wisconsin, I bet there are enough unpledged delegates out there to push him over the top. Without Wisconsin, he’d be at 1152 delegates — with it, he’d be at 1110.

So OK, that’s kind of a rough road. Then I redid the numbers, assuming Gingrich drops and gets all his delegates to vote for Santorum. I gave 66% of Gingrich’s future support to Santorum, and 33% to Romney, which I think is a pretty reasonable estimate. In this model, Romney winds up with more delegates (1246) and he can afford to lose Wisconsin. Note that this scenario also works if you think Romney can pull in a mere 40% of the popular vote the rest of the way, even with both Gingrich and Santorum in the race.

Oh, wait, lemme fiddle with the model some more… OK. If Romney can get 34% of the delegates from proportionally allocated states the rest of the way, and win all the winner take all states except Wisconsin, he still winds up with enough delegates to win the nomination outright. He won 39% of the available delegates yesterday, so he made progress towards his goal. Romney’s right to think he can slowly push his way over the finish line. Lewison’s wrong; it wasn’t a setback. Also, Gingrich is not going to drop out because it would kill his ability to be any kind of a kingmaker at the convention.

Edit: this blog is a real professional doing the same kind of math, but much much better.

Mirrored from Population: One.

Comments

[telepresence.livejournal.com]

A couple things. I went to high school with Jed, we were friends back then, though we've been out of touch. He's a gamer, or at least he was back then, both RPGs and especially board games, I think you'd like him.

According to Jed's pull of RCP's numbers, Romney got 34% of the available delegates last night, not 39. Which number is right? The other question is, are all the remaining "winner take all" states true "winner take all" states? If some of them are WTA with a minimum share threshold, and it's possible, even likely that Romney doesn't meet that threshold for some of them, then they'd devolve to proportional allocation.

Then, the unpledged delegates. Who says they go with Romney, if Romney gets/keeps the stench of a candidate who can't close the deal/the base doesn't want about him?

Anyway, the longer this goes, the longer Romney will be forced to pander to the hard right, and the more material he'll have to disavow in the general, the more moderates and women and minorities he'll lose, the more the Republicans will strangle each other going into the convention. Which makes me happy.

[bryant]

I imagine I would like Jed! I like his writing, I just recognize that Daily Kos is in the business of pushing progressive causes.

Percentages: yeah, I think both Jed and I were a bit sloppy there. There were 119 delegates available from the four contests. Only 105 were actually alocated. Romney took 39% of the pledged delegates, and 34% of the available delegates. The unallocated delegates break down as follows:

Alabama: 8
Mississippi: 3
Hawaii: 3
American Samoa: 0

Alabama should have five more delegates allocated, as per this post, so I'm skeptical of RCP's numbers anyhow.

On the other question -- yeah, now that I look, Wisconsin, Maryland, and California at least are WTA by congressional district. That hampers Santorum too, a bit.

But in the end I do think that the stench of a candidate who can't quite close is way better than the chaos of a brokered convention. I also think a long primary season is bad for Republicans, so I'm happy either way.

[eyelessgame.livejournal.com]

Bryant knows at least two people who are/were friends with Kossack front-pagers; Hunter was one of my housemates at Tech.

[bryant]

Small world! I would not bet against the possibility that I know some of the operations crew, for that matter.

[eyelessgame.livejournal.com]

This sort of analysis is really useful - I've seen other people, from Nate Silver on down, run the same numbers and come to the same conclusion. It's counterintuitive (people tend to forget the impact of proportional allocation) and it demonstrates the rationale for Gingrich's decision - to stay in the race despite his clear abhorrence of Romney and the fact that staying in "hurts" Romney's chief viable rival (who isn't going to get a majority of delegates at this point regardless).

Here's the interesting thing. If, in a month, polls change to where Gingrich dropping out would start projecting to switch the winners of the remaining winner-take-all states (e.g. if the results of AL/MS/WI cause enough of Romney's NJ or CA support to bleed to Santorum that Santorum would win the head-to-head in those states), will Gingrich's calculations change? Might he "provisionally suspend" his campaign, for example, and play kingmaker at the convention, faced with an almost certain alternative of Romney wrapping up a majority by the convention? Rhetorical questions, of course, as no one knows what's in his mind, but it presents for him an interesting game-theory problem, and Gingrich is a reasonably shmot guy.