Tuesday, March 6, 2012

Santorum Can't Get to 1144

[NOTE: Please see FHQ's post-Super Tuesday discussion of the delegate math here.]

...and neither can Gingrich.

FHQ has been saying since our Very Rough Estimate of the delegate counts a couple of weeks ago that Romney is the only candidate who has a chance to get there. But, of course, I have not yet shown my work. No, it isn't mathematically impossible, but it would take either Gingrich or Santorum over-performing their established level of support in the contests already in the history books to such an extent that it is all but mathematically impossible. Santorum, for instance, has averaged 24.2% of the vote in all the contests. Since (and including) his February 7 sweep, he is averaging 34.7% of the vote. That is an improvement, but it is not nearly enough to get the former Pennsylvania senator within range of the 1144 delegates necessary to win the Republican nomination.

FHQ has modified that original model and put together a spreadsheet that not only better captures the rules in each state, but also allows for a constant level of support across all upcoming contests to be to be plugged in. Let's begin by assuming that Santorum enters with 19 delegates and project a 50% level of support across all the remaining contests with bound delegates. This 50% would apply to not only the statewide vote but the congressional district votes as well. In other words, this would trigger a winner-take-all allocation of delegates in most states that have the conditional winner-take-all/proportional rules hinging on a candidate receiving a majority of the vote.

This is extremely generous. It assumes that candidate X would win nearly all the delegates in states that were not already directly proportional. Less generously, this does not count, like the previous version of this exercise, caucus states with unbound delegates (see Iowa, Colorado, Maine, Minnesota, etc.) nor automatic delegates who have yet to endorse.

Where does that leave Santorum? 1075 delegates.

But hold on. What if we add another layer to this by accounting for the thresholds for receiving delegates in the various states (typically 15% or 20%)? This would have the impact of reallocating delegates of those under the threshold in proportional environments to those candidates over the threshold. That would mean more delegates. If we set the number of candidates over the threshold to its lowest value -- 2 candidates in 20% threshold states and 3 in 15% threshold states1 -- that maximizes the number of reallocated delegates.

Where does that leave Santorum? Again, this is assuming winner-take-all rules have been triggered in all the conditional states. It assumes that the likely bare minimum of candidates has crossed the thresholds to receive reallocated delegates. This is very generous.

1162 delegates. That's cutting it awfully close.

Surely the automatic delegates or the unbound caucus delegates would keep Santorum over 1144. Yeah, they could potentially serve as kingmaker until you remember that we just very unrealistically gave Santorum winner-take-all allocation where is was conditionally possible. We gave him a consistent 50% of the vote -- over 15% better than he has performed during his best stretch. Also, Santorum -- given the polls we have access to for today's races -- is very unlikely to reach that level of support across all of the Super Tuesday primaries and caucuses. That means that after today -- a day with over 400 delegates at stake -- Santorum will not be able to get to 1144.

...and neither will Newt Gingrich.

--
Well just a darn minute there, FHQ. You're cooking the books, right? What if you put Mitt Romney in the same model(s) under the same circumstances? Ah, I'm glad you asked.
  • In the first model where Romney would be at 50% support statewide and in each congressional district, the former Massachusetts governor would net 1254 delegates. 
  • In the second model that accounts for a likely bare minimum of candidates over the threshold, Romney would surpass 1300 delegates at 1341. 
Even if we simulate a scenario where Romney continues to only win half of the congressional districts, he still gets to 1152 delegates in the second more realistic model and .2

--
The bottom line here is that Romney has enough of a delegate advantage right now and especially coming out of today's contests that it is very unlikely that anyone will catch him, much less catch him and get to 1144. The latter seems particularly far-fetched given the above scenarios. And that is a problem in this race. Well, a problem for Gingrich and Santorum anyway. If all either of them can take to voters is an argument that all they can do is prevent Romney from getting to 1144, then neither has a winning strategy. That sort of strategy has a half life; one that will grow less effective as, in this case, Romney approaches 1144. Complicating this scenario even further for Gingrich and Santorum is the fact that if neither can get to 1144 or even close to it, neither is all that likely to be the candidate to emerge as the nominee at any -- unlikely though it may be -- contested convention.

These contests today may not be decisive in terms of settling the nomination, but they very much represent a mental hurdle in this race. That Santorum and Gingrich cannot get to 1144 without vastly over-performing in the remaining contests (relative to how well they have done in the contests thus far) ushers in a new phase in the race.

But how long will the "keep Romney from 1144 plan" last? With southern contests scattered throughout the rest of March, Gingrich and Santorum will have legitimate chances at wins. However, that means Illinois on March 20 and the bulk of April end up being rather tough terrain. Wins on Romney's turf become imperative to stay alive at that point for Gingrich and Santorum. By that point, though, Romney will still hold the delegate advantage and favorable contests in front of him. That is not a good combination for anyone hoping to catch him in the delegate count.

...or even keep him under 1144.

--
1 Remember that one candidate is already at the 50% level and it has been rare to see more than two candidates over 20% or 3 over 15% with the top candidate approaching 50%.

2 In states with an odd number of congressional districts, the delegate total was rounded down to the nearest whole district. A five district state would have Romney winning only two districts. This does not apply in states where there is an attempt to allocate congressional district delegates proportionally. In those states, Romney is given the partial total across all congressional districts. Look, if we are going to be generous to Santorum/Gingrich then it is equally as helpful in this exercise to be stingy with Romney. We want to poke holes in his ability to get to 1144. If we poke enough, Romney can be pulled under 1144, but it becomes more and more complicated and less and less realistic.

--
NOTE: The delegate scenarios above were constructed as part of a request from ABC News. 

Recent Posts:
The (Delegate) Keys to Super Tuesday

Race to 1144: Washington Caucuses

Fantasy Delegates


Are you following FHQ on TwitterGoogle+ and Facebook? Click on the links to join in.

15 comments:

Anonymous said...

Did this model take into account what happens if Gingrich or Santorum successfully contest the illegal winner-take-all formats in Florida and Arizona?

Anonymous said...

The RNC set it so Romney would be the nominee. The frontloaded the election with Romney friendly states and backloaded it with winner take all states like California and New Jersey and New York and in the middle they made it propertional so he would get something in the south. Yeah it's a racket. But guess what? Ron freaking Paul has been playing the game set up for Romney to win if he gets a chance. He is focusing on caucus states because he can get his people to take a majority of delegate spots even if doesn't win. It was set up for Romney and his organization, but Ron Paul as an equally formidable organization. Georgia has 76 delagates and every expect Newt to win it, but what if I tell you that even though Ron Paul is polling 5% in GA, his delagate take in GA could be half. Yep! He is working the states and other primary to milk the system set up for Romney. If he doesn't catch fire soon, it will not matter though unless this thing goes to the convention. If this goes to the convention and no one gets 1144 by first ballot, every eye should be on Ron Paul. He has a greater chance of becoming the nominee if the delegates become free agent than anyone else there.

Anonymous said...

Analysis of past performance of both Santorum and Gingrich I think is not valid yet numerous pundits use their polling numbers in primaries to make Romney appear inevitable. There is a very obvious Not-Romney majority voter block out there that has been around since the primaries first kicked off. All of the other candidates with the exception of Ron Paul have been effectively splitting the conservative vote. As the field winnowed down, that voter block split between Newt and Rick. Romney's success to date has been due largely to SuperPAC's keeping either Newt or Rick in the game. Everyone thinks Romney's inevitable. But if either Santorum or Gingrich were to drop out after today, I would wager that the other will benefit handsomely from it, would experience a step change increase in support and cause real problems for Romney moving forward. Romney needs both Newt and Rick to stay in for the next few months to win.

Anonymous said...

"Analysis of past performance of both Santorum and Gingrich I think is not valid yet numerous pundits use their polling numbers in primaries to make Romney appear inevitable. There is a very obvious Not-Romney majority voter block out there that has been around since the primaries first kicked off. All of the other candidates with the exception of Ron Paul have been effectively splitting the conservative vote. As the field winnowed down, that voter block split between Newt and Rick. Romney's success to date has been due largely to SuperPAC's keeping either Newt or Rick in the game. Everyone thinks Romney's inevitable. But if either Santorum or Gingrich were to drop out after today, I would wager that the other will benefit handsomely from it, would experience a step change increase in support and cause real problems for Romney moving forward. Romney needs both Newt and Rick to stay in for the next few months to win."

Spot on analysis...

Anonymous said...

Please tell me how the guy who has 50% of the GOP allotted delegates thus far somehow hasn't gotten any of the "conservative" vote?

Perhaps if guys like Rick and Newt wouldn't support more government programs to counter ballance liberal government programs, us "non-conservative" Republicans might be persuaded to believe that Gingrich and Santorum are actually conservative.

Anonymous said...

That may be true, but that's not what you have to show. You have to show that *Romney* can get to 1144.

Anonymous said...

"Romney would be at 50% support statewide and in each congressional district, the former Massachusetts governor would net 1254 delegates"

Dream on! Your model is a fantasy. In a two person race, Romney loses in a landslide. In a three person race, it goes to the convention.

Anonymous said...

What if Gingrich drops out in the end and his delegates support Santorum?

Anonymous said...

"Well just a darn minute there, FHQ. Your cooking the books, right?"

Your <> you're.

Anonymous said...

The election was frontloaded with Romney friendly states? You mean Iowa, South Carolina, Tennessee, Georgia, Oklahoma, and Missouri were all Romney-friendly states? Are you serious? Spare me the B.S. In addition, you can't assume that all of Gingrich's voters and all of Santorum voters are anti-Romney. That is a huge mistake. Looking at the trends of candidates throughout this primary, there is at best about a 20-25% anti-Romney group. That amount will be smaller in many states like California and New York. A good number of Gingrich or Santorum voters might well choose Romney over the other opponent. Therefore, even if one of them gets out, it will still be very unlikely that the other will be able to garner the delegates that are needed.

donald morgan said...

wonderful, we get to choose between obama and obama-light(romney).

John Morris said...

Except the other two are 'not-Romney' and Newt especially is in a get Romney frame of mind. Stop Romney from hitting the magic number and the fun begins. The real fun begins if RIck + Newt doesn't equal the magic number either and Ron Paul becomes kingmaker.

Or Mitt has to cut a VP deal with one of them, and Newt probably wouldn't go for that deal. Really wouldn't care to guess whether Rick would, will probably depend how much more bad blood we see before then.

Mike Murphy on MTP this Sunday really summed up (unintentionally) why we are hellbent on stopping Romney with his line about "Thats why you call them the base, you get them. They're free." No we aren't.

Anonymous said...

Ahhhaa, so your saying there's a chance!

Anonymous said...

Isn't it funny once the statistical facts are laid out that1) the contest was "fixed" and 2)people are convinced that a cat is a dog, regardless of the fact.

Steve said...

What if Santorum bows out and endorses Gingrich, or Gingrich bows out and endorses Santorum?