The Delegate Fight: 2012

[Beet]

If the straw polls don't matter, why do candidates compete for them? They should just focus on getting their delegates elected at each step of the process, instead of courting voters. I believe that was how Barry Goldwater captured the 1964 Republican nomination.

[Mehmentum]

If the straw polls don't matter, why do candidates compete for them? They should just focus on getting their delegates elected at each step of the process, instead of courting voters. I believe that was how Barry Goldwater captured the 1964 Republican nomination.

Didn't Obama focus on getting delegates elected as well?

[tpfkaw]

If the straw polls don't matter, why do candidates compete for them? They should just focus on getting their delegates elected at each step of the process, instead of courting voters. I believe that was how Barry Goldwater captured the 1964 Republican nomination.

That was pre-reform, though.  He actually had amassed a majority of delegates before the first primary was even held, as it turns out (though they were all unpledged).

[California8429]

Example districts in Colorado:


This didn't happen everywhere, but Paul probably won Minnesota and is a lot stronger in Colorado than the straw poll shows.  Caucus states are great for Paul, primaries are black and white straight popular voting (MO didn't count - they have a caucus that actually matters in a month).

Most precincts get 1-2 state delegates for anyone wondering about "100%". Romney got about half the vote but 100% of the delegates at my caucus.

[Хahar 🤔]

If the straw polls don't matter, why do candidates compete for them? They should just focus on getting their delegates elected at each step of the process, instead of courting voters. I believe that was how Barry Goldwater captured the 1964 Republican nomination.

Because delegates are completely meaningless except in the exceedingly unlikely situation that no candidate has a majority. It's all about perception; the candidate who is perceived to have the best chance of winning will win. Straw polls are part of said perception.

[Erc]

Example districts in Colorado:
(Large image)

This didn't happen everywhere, but Paul probably won Minnesota and is a lot stronger in Colorado than the straw poll shows.  Caucus states are great for Paul, primaries are black and white straight popular voting (MO didn't count - they have a caucus that actually matters in a month).

Most precincts get 1-2 state delegates for anyone wondering about "100%". Romney got about half the vote but 100% of the delegates at my caucus.

While it's certainly probable that Paul got more delegates to the next level of conventions than his proportion in the straw poll indicates, you'll excuse me if I don't exactly trust the Paul camp's self-reported numbers.

If I can get reliable indications of what actually occurs at the next round of conventions (both in CO/MN and in IA), I will update my numbers accordingly.

[Erc]

Very, Very Preliminary Delegate Estimates:

Colorado: 
Santorum - 12
Romney - 11
Gingrich - 6
Paul - 4

Minnesota
Santorum - 18
Paul - 10
Romney - 8
Gingrich - 1

This is pure guesswork at the moment; I'll have a chance to do a more thorough analysis by CD tomorrow for something slightly closer to the truth.  For reasons discussed above, I am probably well underestimating Santorum's performance in Minnesota.

If the race is still being actively contested I don't see how Romney gets more than 1 or 2 delegates from Minnesota, In fact I bet he gets shut out.  My very very rough guess would be Santorum 20, Paul 17

After a second look at these numbers, these do indeed seem far too generous to the third (and fourth) place finishers in both cases.  Whether Romney will be completely shut out will, in the end, depend exactly on how the final slates of delegates are nominated at the convention.

[Lief 🗽]

The conventions could turn out to be really fun. You could see weird things like Paul and Santorum delegates aligning to shut out Mitt Romney (or Newt + Rick in CO).

[Beet]

If the straw polls don't matter, why do candidates compete for them? They should just focus on getting their delegates elected at each step of the process, instead of courting voters. I believe that was how Barry Goldwater captured the 1964 Republican nomination.

Because delegates are completely meaningless except in the exceedingly unlikely situation that no candidate has a majority. It's all about perception; the candidate who is perceived to have the best chance of winning will win. Straw polls are part of said perception.

Except that's exactly how it's shaping up so far as I predicted it might just before Iowa.

"Perception" is an amorphous thing; it can shatter in one day, as Mitt Romney found out. Not something to hang one's hopes on by any means.

I think there's somewhat of a myth out there that primaries are a form of democracy, when that's not necessarily true; especially for the GOP. It is really a battle for control of a private organization, the person who gets the most votes or is the most popular doesn't necessarily win.

[Erc]

The conventions could turn out to be really fun. You could see weird things like Paul and Santorum delegates aligning to shut out Mitt Romney (or Newt + Rick in CO).

Oh, they're going to be real fun.  All sorts of lovely tactical voting.

These projections try to stay as agnostic about tactical voting as possible, and assume that, in the lack of a majority for one candidate, delegates are assigned proportionally (or to the top 2/3 finishers in the case of small numbers being elected, as in Iowa or any CD).

[Erc]

More sensible Minnesota projection:

For the 13 At-Large delegates:
Santorum - 6
Paul - 4
Romney - 2
Gingrich - 1

In the 8 congressional districts, Santorum makes a clean sweep in CD's 7 and 8, gets 2 (to Paul's 1) in CD's 1 and 6, while Santorum, Paul, and Romney get 1 a piece in CD's 2,3,4, and 5.  

This yields a total of:

Santorum - 20
Paul - 10
Romney - 6
Gingrich - 1

These are very much subject to change!  Santorum is very close to getting a majority in CD 1 (and probably will), and is reasonably close in CD 6.  Furthermore, I'm doing these apportionments on the old CD lines, whereas the BPOU conventions are held after the final maps come down.  Not to mention tactical voting considerations, etc.

[Erc]

More sensible Colorado projection:

At-Large:
Santorum - 5
Romney - 4
Gingrich - 2
Paul - 1

Congressional Districts:
Santorum wins CD's 2, 3, 4, 5, and 7 with 2 delegates to 1 for Romney.  Romney wins CD 6 by 2 delegates to 1 for Santorum, and Romney, Santorum, and Paul each get a delegate from CD 1.

Totals:
Santorum - 17
Romney - 12
Gingrich - 2
Paul - 2

Note that this was done by just eyeballing the congressional districts---effectively, I'm just guessing that Santorum won CD 7 (in fact, I would bet he didn't).  Plus, the other usual caveats.

[they don't love you like i love you]

With the way conventions work in Minnesota, I wouldn't be surprised if Romney gets shut out at the CDs, his people will probably just focus all their resources on electing delegates to state since that's the big focused on thing. The CD delegate elections in my experience have always been a bit of an afterthought, though granted in 2010 it was really just a boring formality (Even in 2008 it was somewhat ignored despite the election of quite a few national delegates, but our straw poll was binding so the delegate numbers were already set in stone.) Paul probably won MN-5 actually.

[Erc]

Paul's Path Forward

Let's take the Paulites at their word and suppose that they are doing an extremely good job at electing delegates in caucuses---so much so that they actually get majorities at the state convention and elect a slate of Paul-supporting delegates to Tampa.  This is rather far-fetched, for a number of reasons (it's harder to pull this off in high-population precincts, it's unlikely to happen if you simply don't have that many supporters, and it's completely possible this entire narrative is BS), but let's run with it.

This is the sort of thing that only works in Iowa-style caucuses---the ones with multiple rounds of conventions, where the delegates to Tampa are picked months from now at later conventions.  In primaries, Paul doesn't generally have much of a chance (his 23% showing in NH is most likely his peak for the season)---and while he may have a slight turnout advantage in other sorts of caucuses, it's clear he's not going to walk away with those either.

So what's the best Paul can possibly hope for?  That is, what are the Iowa-style caucuses where Paul has a shot of abusing the system and walking away with a lot more delegates than one would expect?  [Ignoring territories]

In fact, there are only 7, and 4 will have occurred by the end of this week (IA/CO/MN/ME).  This only leaves Washington (March 3), Missouri (March 17), and Nebraska (early June) as states where Paul has a chance to abuse the system.

There are plenty of other caucus/convention states, mind you...but they're more like Nevada, in which the straw poll vote actually matters (NV/AK/ID/ND/KS/HI), or are completely smoke-filled rooms with no pretense of democracy and no chance of Paul infiltration (WY/MT).

Other than the 7 Iowa-style states, Paul may pick up a delegate here and there in states with true proportionality and no cutoff, but these are few and far between.

[Minnesota Mike]

Paul won 49% in the city of Minneapolis so he almost certainly won the 5th CD.

http://www.twincities.com/ci_19920382

FWIW in 2008 the Paulites managed to get 6 of the 12(?) delegate spots from the 4,th, 5th and 6th districts (I think it should be 6 of 9 but the article says 6 of 12).

http://www.minnpost.com/stories/2008/05/05/1603/beyond_mccain_ron_pauls_supporters_hope_to_reshape_the_gop

[Erc]

If Paul got (nearly) a majority in CD 5, than means he must have done terribly in CD 3 (i.e. the rest of Hennepin county).  I'm going to start digging through the precinct results and see what I can turn up.

[Erc]

Minnesota Results by (Current) Congressional District:

CD	Gingrich	Paul	Romney	Santorum	Write-In	Total
CD1	642	1601	885	2963	14	6105
	10.5%	26.2%	14.5%	48.5%	0.2%
CD 2	894	2146	1547	3402	26	8015
	11.2%	26.8%	19.3%	42.4%	0.3%
CD 3	799	1804	1865	2919	6	7393
	10.8%	24.4%	25.2%	39.5%	0.1%
CD 4	348	1216	711	1307	26	3608
	9.6%	33.7%	19.7%	36.2%	0.7%
CD 5	354	1525	745	1246	21	3891
	9.1%	39.2%	19.1%	32.0%	0.5%
CD 6	833	2361	1043	3603	11	7851
	10.6%	30.1%	13.3%	45.9%	0.1%
CD 7	666	1245	652	3164	27	5754
	11.6%	21.6%	11.3%	55.0%	0.5%
CD 8	740	1354	783	3372	12	6261
	11.8%	21.6%	12.5%	53.9%	0.2%
Total	5276	13252	8231	21976	143	48878
	10.8%	27.1%	16.8%	45.0%	0.3%

This does not change any of the analysis in the previous post; delegate counts remain the same.  Paul did indeed win CD 5, but Romney and Santorum have enough support to deprive him of making a clean sweep.

What this analysis does not account for:

1) The new CD borders.  Given the timing, it's unclear whether the BPOU conventions will use the new CD borders when electing delegates to the District Conventions, or how precise they're going to be with them.

2) Rounding errors.  Precincts only elect an integral number of delegates; if this number of delegates is small, this tends to disfavor small candidates.  If supporters of a particular candidate can manage a majority in a given precinct, they can likely elect a full slate of their supporters.  This would tend to favor Santorum (and perhaps Paul in certain places) at the expense of the other candidates.

3) The number of delegates each precinct / BPOU is entitled to is not determined by the turnout, but by the vote for Tom Emmer in the 2010 Gubernatorial Race.  As the two tend to be correlated, this isn't a huge effect.

4) Straw Poll results don't necessarily have anything to do with the delegates elected to the BPOU conventions (the Stealth Paul strategy).

Of these, 2) and 4) will probably have the largest effect.  I can't do anything about 4) without tracking down every delegate elected in the state of Minnesota (or taking the Paul campaign's word as gospel, which is inadvisable).  If I can find more information about how many delegates exactly each precinct caucus elected, I may be able to account for 2) in some fashion---but more likely there won't be any changes to the count until after the BPOU conventions, if even then.

Long story short...expect Santorum to do even better than these numbers in the end, Romney to do worse, and Paul to stay about the same (with an outside chance of doing much, much better if 4) is true).

The major question is, of course, whether Santorum can gain a majority at the statewide convention; this seems very likely, in fact.

A plausible final result (Santorum gains a majority statewide and in CD's 1 and 6):

Santorum - 29
Paul - 4
Romney - 4

[Lief 🗽]

It's times like these I wish the forum had a "like" button.

[Erc]

I sadly cannot do similar analysis for Colorado at the moment, as the Colorado GOP does not appear to provide precinct-level results, making the Denver environs (Adams/Arapahoe/Jefferson counties, specifically) complete guesswork.

For amusement value...the AP currently projects Santorum to win all 37 delegates out of Minnesota.  While this may end up being closer to the truth than my projection, at the very least one would expect Paul's win in CD 5 to screw up this projection.

[Erc]

Redoing the analysis under the assumption that if a candidate has a majority of supporters in a given precinct caucus, they get to elect all the delegates in that precinct caucus.  (Basically, using the logic at the precinct level [but not the BPOU level] that I had used at the CD level to give Santorum CD's 7 and 8 ).

Santorum wins a majority statewide, and in CD's 1, 6, 7, and 8, receiving all delegates from those jurisdictions.

Santorum is just short of a majority in CD 2, and splits the delegates 2 - 1 with Paul.

In CD's 3 and 4, Santorum, Paul, and Romney each win a delegate.

In CD 5, Paul is still short of a majority, and wins 2 delegates to Santorum's 1.

Grand Total:
Santorum - 30
Paul - 5
Romney - 2

As usual, this is approximate, but it should be closer to the truth than my earlier projection.  I am updating the main page accordingly.

I could do the same thing taking the intermediate-level (BPOU) winnowing into account as well, but I think it's rather pointless until we get the new CD maps (and we figure out whether the MN GOP is using them).

I could do the same thing in Iowa at some point if I find good precinct-level results, but it likely won't make a major difference outside of Dutch Reformed country.

[Landslide Lyndon]

Great job Erc.

[they don't love you like i love you]

Where are the Minnesota precinct results? I can only get results on the SoS site of what reported, not the actual results.

[Erc]

Took me a while to find them.  Big old nasty set of files can be found here.  You need to crosscheck a few of them to make any sense of the files.

[they don't love you like i love you]

Bleh and we're usually pretty good at getting good and easy to use data out.

So Ron Paul won "only" 53.85% (14 votes lol) in my precinct. Of course Romney came in second with exactly half that, and Santorum had only 2 votes.

The Paul dominance of various college dorm precincts is pretty amusing. Here's something a little odd, Minneapolis 7-10, home to North Central University (Assembly of God), the most Republican precinct in Minneapolis (McCain actually broke a third of the vote) also gave a majority to Paul and second place to Romney, the latter being more surprising than the former. Santorum in fact received only one vote, to Romney's 7 and Paul's 11. In other words it didn't vote much different than most other Minneapolis precincts. People at NCU were really that uninterested?

[Erc]

As noted elsewhere on this forum, AJ Spiker, not Bill Schickel, has replaced Matt Strawn as State Chairman in Iowa after winning a narrow vote.  He is a noted Paul supporter.

I believe this is the first State Chairman in any state to officially endorse a candidate (most seem to stay neutral).

County Conventions in Iowa take place on March 10th...we'll see if there's anything interesting to report out of them at the time.  I would not be surprised if Paul's estimated delegate numbers go up.