Monday, December 12, 2011

Winnowing, Really

National Journal's Alex Roarty looks at national polling and polling from South Carolina and Florida and writes of Bachmann, Santorum, and Perry:
But most importantly, their struggles are crippling Romney's chances. After Herman Cain's collapse, Bachmann, Santorum and Perry were the three Republicans left in the race capable of attracting the party's hardline voters. Thus far, they've only continued to cede support to Gingrich...Romney needs one of the three to cut into that voting bloc.
In Iowa? Could happen. After that? If Bachmann, Santorum, and Perry don't finish top three in Iowa, they're not going to be around for South Carolina (okay, I suppose it's possible one of them could rally into a strong fourth and win the spin game and therefore "win" Iowa, but basically they need to be top three).

Quick review: look at the percentage of the overall vote taken by the top two candidates, Iowa vs. South Carolina (all numbers from wikipedia):

             Iowa    SC
1980      62%    85%
1988      62       70
1996      49       74
2000      72       95
2008      59       63

So in each case the percentage of votes for other candidates went down, sometimes just a bit but in some cases dramatically. And, of course, it continued to decrease after that. Most losing candidates don't just go from 10% to 10% to 10% as the campaign goes on; if they have no shot at the nomination, they drop out. Indeed, of the current contenders, only Ron Paul seems likely to stay in after proving he doesn't have the electoral support to have a realistic shot at the nomination.

It's just not realistic to expect the also-rans to soak up a lot of votes beyond the first handful of events.

And of course the flip side of this, as I've started to point out regularly, is that there's no guarantee that Romney will wind up surviving the winnowing process. If he finishes 4th (or worse) in Iowa, he may well fall to second or third in New Hampshire, and be gone before Supertuesday. So if I were running Mitt Romney's campaign, I wouldn't exactly be rooting for a Perry or even a Bachmann surge right now.

At any rate, given this cycle's cast of characters and schedule, I find it hard to believe that we'll have more than two candidates plus Ron Paul after Florida, so most of the delegates will be chosen in states in which well over 40% will be needed to win.

The bottom line is that if Romney really has a hard cap below 50% he's almost certainly not going to win the nomination.


  1. Has your position on Newt changed at all since the most recent debate? In the past week you've basically said there is zero chance of him being the nominee yet you think its possible Romney drops out after a bad performance in Iowa/New Hampshire? That leaves Perry as the sole viable nominee in your view but he's treading water and odds right now are he'll underperform Mitt in Iowa. I just can't see any conceivable scenario where Perry bounces back after all the notoriety he's received from the debate performances and gaffes. And I certainly don't see the 20-25% of Romney voters migrating to him either. Where do you think they go should Mitt drop out early?

  2. Nope, haven't changed at all.

    FWIW...Perry has been fewer than 10 points behind Romney in polling over the last week. That's certainly a plausible shift in three weeks.

    Pretty much I'm sticking to what I've said: Romney + Perry have a ~95% chance of winning the nomination, and at the moment Romney seems more likely, but I have no estimate on breaking that down -- other than to say % Romney > % Perry > 2% (which is what Intrade has Perry at).

  3. Why wouldn't most of Romney's supporters migrate to Perry if Romney dropped out? Most of Romney's elite endorsements would almost certainly shift to Perry over all the others. They share similar resume cred. Who else would a Romney supporter turn to?

    Not saying Perry would gain enough to become frontrunner, but he's still get quite a substantial bounce.

  4. Basically they need to be top three in Iowa to remain viable, and in the new ARG poll released today Romney has a commanding 17% to 13% lead over Perry for third place?


    I agree with Thomas that Perry is the one the 'establishment GOP' would turn to if Romney flames out early. He's also the one with the best chance to raise money and ramp up a national campaign fast in that scenario. I think it's harder to say that all of Romney's supporters would go to Perry though.

    Speaking of polling, there was one other I noticed today, Rasmussen came out with a result that has Gingrich viewed as "the strongest general election candidate" over Romney, 49% to 24%. Now I don't know if you consider that phrasing any different than 'electability', (and it's still possible for Newt to fatally gaffe himself one of these days) but isn't that a rather telling metric? The 25-point gap is actually larger than Newt's lead over Mitt in the national polls.

  5. Romney plus Perry at 95%? Too bad you can't structure a bet like this on InTrade. Maybe somewhere else?

    The model of party power about which you've been educating us this many a month sure *seems* to be coming under pressure, as by all accounts the whole Republican establishment hates Newt. Good for you for doubling down.

    I must say, that while I get the sense that you'd find a Perry resurgence professionally reasurring, since he fits the profile of a plausble candidate, I find the prospect terrifying. I'd prefer the unstable narcissist fraud Gingrich to fascist Perry. I think.

  6. JB: that data would be more useful if you included the date spread between IA and SC. I can't claim to know much about SC's dates going back more than the last couple of cycles.

  7. Matt -- Yeah, fair enough. From 1980 on it was basically in the same place it is now, but the interval after Iowa and the number of minor (and sometimes not so minor) states between NH and SC has varied. It's really just to say that the field will winnow, and that happens rather quickly once the voters get involved...the idea that a bunch of candidates will stick around and allow someone to win with <40% isn't very likely.

  8. ASP,

    Perry vs. Newt? Perry's been a passable governor of a big state for a decade...yeah, it's not the toughest job in the world, and procedurally I'm sure there's some stuff that I wouldn't be thrilled with, but it's nothing like the disaster that was Newt's Speakership. Along with Newt's entire approach to governing and politics. I don't really see either of them as well-equipped for the job, but Newt's the one that I find scary.


Note: Only a member of this blog may post a comment.

Who links to my website?