Louisiana primary chat

[J. J.]

Lower Evangelical turnout could be bad news for Santorum.  Mittens carried the Tea Party vote in IL.  This might be closer than we thought.

No, it's just the nature of Louisiana demographics.  In the 2008 exit polls, only about half of McCain voters in Louisiana identified as evangelicals.  That's actually closer to Illinois (36%) than it is to Alabama (71%) or Mississippi (77%).

I think we discussed the different demographics.  That might end up making this race closer.

[Alcon]

Lower Evangelical turnout could be bad news for Santorum.  Mittens carried the Tea Party vote in IL.  This might be closer than we thought.

No, it's just the nature of Louisiana demographics.  In the 2008 exit polls, only about half of McCain voters in Louisiana identified as evangelicals.  That's actually closer to Illinois (36%) than it is to Alabama (71%) or Mississippi (77%).

I think we discussed the different demographics.  That might end up making this race closer.

...?

[J. J.]

By the way, way to ignore how the amount of people saying they're "very conservative" might be record breaking. Not good for Mitt.

No, I'm saying that Santorum's strongest support is from Evangelicals.  There are not as many as in other southern states, so that was a weakness Santorum.

Romney has done reasonably well with "very conservative" voters, at least better than with Evangelicals.  Nobody is talking about a Romney upset, but possibly a closer race than we thought.

[Alcon]

It's not like the Louisiana polling has been assuming a vastly more evangelical electorate or something.

[RI]

Also, Romney may have won the Tea Party vote in Illinois, but he sure won't be in Louisiana.

[Torie]

Yes, the Cajun aspect may not a be a demographic that is picked up very well by the Sean Trende model perhaps, and it will be interesting to see how much they vote like Evangelicals versus white Catholics that are in some concentration elsewhere. The best guess I would think is somewhere in-between. The Trende model would give Mittens about 33%-34% of the vote in LA, and Mittens if the model is right won't get much higher than that with Newt pulling 15% of the vote or so, and if Cajuns are labeled "non-Evangelical," it might be a tad lower, although I suspect not much, since I would guess their educational level is rather low and that would be picked up in the model.

Moreover, outside the urban zones, the Cajun vote in the rural areas is just not that large of the share of the pie anyway, and in LA the Cajun area urban zones go their own way. LA is a very class oriented state under the surface. Huey Long is not totally dead.

Anyway, this is what I was musing about when I came up with my prediction guess, when I ended up at 32% for Mittens.

[Keystone Phil]

By the way, way to ignore how the amount of people saying they're "very conservative" might be record breaking. Not good for Mitt.

No, I'm saying that Santorum's strongest support is from Evangelicals.  There are not as many as in other southern states, so that was a weakness Santorum.

And as I said, he makes that up with more religious Catholics in a deeply conservative state.

[All Along The Watchtower]

Question: are the evangelicals in Louisiana more middle-class suburban or more rural poor in demographic? Or mixed?

[ajb]

By the way, way to ignore how the amount of people saying they're "very conservative" might be record breaking. Not good for Mitt.

No, I'm saying that Santorum's strongest support is from Evangelicals.  There are not as many as in other southern states, so that was a weakness Santorum.

Romney has done reasonably well with "very conservative" voters, at least better than with Evangelicals.  Nobody is talking about a Romney upset, but possibly a closer race than we thought.

Umm, no. In the Midwest, Santorum's run about the same with both groups. In the South, he's done much better with "very conservative" voters than with evangelicals.

Maybe you were thinking of "severely conservative" voters, J.J.?

Alabama Exit Poll:
"Very conservative": Santorum 41 Romney 18
Evangelical: Santorum 35 Romney 27

Mississippi Exit Poll:
Very Conservative: Santorum 39 Romney 22
Evangelical: Santorum 35 Romney 29

Illinois Exit Poll:
Very Conservative: Santorum 48 Romney 37
Evangelical: Santorum 46 Romney 39

Ohio Exit Poll:
Very Conservative: Santorum 48 Romney 30
Evangelical: Santorum 47 Romney 30

[RI]

Question: are the evangelicals in Louisiana more middle-class suburban or more rural poor in demographic? Or mixed?

Most of Louisiana's evangelicals are in the more rural northern end of Louisiana. The southern part is surprisingly Catholic, though a good chunk of those are Cajuns in the southwest.

[Lief 🗽]

LOL, I like how J.J. ignores the numerous demographic numbers in that post that are good news for Santorum and focuses on the one that would be good news for Romney if you know absolutely nothing about Louisiana.

[Torie]

Question: are the evangelicals in Louisiana more middle-class suburban or more rural poor in demographic? Or mixed?

In the north Baptist zone, it's like Alabama - "everyone" is Evangelical. In the south, middle class urban area voters don't vote like Evangelicals. The Cajuns outside N.O. and Baton Rouge do their own thing.

[J. J.]

Yes, the Cajun aspect may not a be a demographic that is picked up very well by the Sean Trende model perhaps, and it will be interesting to see how much they vote like Evangelicals versus white Catholics that are in some concentration elsewhere. The best guess I would think is somewhere in-between. The Trende model would give Mittens about 33%-34% of the vote in LA, and Mittens if the model is right won't get much higher than that with Newt pulling 15% of the vote or so, and if Cajuns are labeled "non-Evangelical," it might be a tad lower, although I suspect not much, since I would guess their educational level is rather low and that would be picked up in the model.

Moreover, outside the urban zones, the Cajun vote in the rural areas is just not that large of the share of the pie anyway, and in LA the Cajun area urban zones go their own way. LA is a very class oriented state under the surface.

Anyway, this is what I was musing about when I came up with my prediction guess, when I ended up at 32% for Mittens.

I gave it 30% Mitten, 45% Rick, in my prediction.  I'm still expecting double digits, but it might be lower.

[J. J.]

LOL, I like how J.J. ignores the numerous demographic numbers in that post that are good news for Santorum and focuses on the one that would be good news for Romney if you know absolutely nothing about Louisiana.

We've been talking about the demographic difference for a while.  This one might be closer than we thought.

[Alcon]

J. J., are you arguing anything besides "Louisiana might be closer because the demographics are different than the rest of the South"?  I'm fine with that argument, with two caveats:

1. There's no reason to believe that the polls were expecting the wrong demographics.

2. The exit polls show demographics that are just fine for Santorum, relative to expectations.  The exit polls give us no reason to believe the polls had the wrong demographics, and the polls would have to be wrong for a reason other than predicting the wrong sort of demographic turnout.

So, if you're willing to make those two concessions, we're on the same page.  Of course, this same argument could apply to most every state with wonky demographics in most every election, and doesn't appear any likelier than it did before we saw the exit polls.

It could be closer than we thought, sure...but is there reason to believe, from the exit polls, it's any likelier to be?  Nope...

[Torie]

The Evangelical share of the pie in the exit poll just about matches the Sean Trende figures. The electorate that turned out is no big surprise at all.

[J. J.]

J. J., are you arguing anything besides "Louisiana might be closer because the demographics are different than the rest of the South"?  I'm fine with that argument, with two caveats:

1. There's no reason to believe that the polls were expecting the wrong demographics.

2. The exit polls show demographics that are just fine for Santorum, relative to expectations.  The exit polls give us no reason to believe the polls had the wrong demographics, and the polls would have to be wrong for a reason other than predicting the wrong sort of demographic turnout.

In the second case, I think you are making an assumption that LA Catholics behave more like Evangelicals.  I am not sure that is correct.

[Mr. Morden]

More exits:

http://www.cbsnews.com/8301-503544_162-57403966-503544/early-exit-polls-most-louisiana-gop-voters-say-etch-a-sketch-controversy-not-a-factor/

You must be logged in to read this quote.

[Thomas D]

Jeff Greenfield says on Twitter that Santorum won men and women by 15.

[argentarius]

The only real question is if Mitt Romney breaks 25%. And yes, after his overpolling in AL/MS, it's a very legitimate question.

[Torie]

I read that last exit poll as Mittens gets no less than 30% of the vote, and no more than 40%.  Mittens wins big with the winner and right experience group, and loses even bigger with the character and strong conservative group.

[Meeker]

I'm having a difficult time seeing how Romney doesn't break 25%. Santorum would have to win by quite a bit.

[Torie]

Jeff Greenfield says on Twitter that Santorum won men and women by 15.

If that is right, JJ's prediction looks pretty good at 45-30, with Newt getting 15% and others 10% or something, mix and match.

[Alcon]

J. J., are you arguing anything besides "Louisiana might be closer because the demographics are different than the rest of the South"?  I'm fine with that argument, with two caveats:

1. There's no reason to believe that the polls were expecting the wrong demographics.

2. The exit polls show demographics that are just fine for Santorum, relative to expectations.  The exit polls give us no reason to believe the polls had the wrong demographics, and the polls would have to be wrong for a reason other than predicting the wrong sort of demographic turnout.

In the second case, I think you are making an assumption that LA Catholics behave more like Evangelicals.  I am not sure that is correct.

No, I'm not making that assumption.  I explicitly said that "polls would have to be wrong for a reason other than predicting the wrong sort of demographic turnout."  That would include Romney performing better among Catholics, which you seem to be suggesting is more probable than assumed, based on Catholics in other states.  There is nothing to suggest the electorate is atypical, so there is nothing to suggest that the polls have systemically messed up, or sampled the wrong sort of Catholic, or something.  Is it possible?  Yes.  Do we have any indications that this is happening?  No.  Did the exit polls give us any reason to assume it's true?  No.

[Democratic Hawk]

I read that last exit poll as Mittens gets no less than 30% of the vote, and no more than 40%.  Mittens wins big with the winner and right experience group, and loses even bigger with the character and strong conservative group.

Which, of course, may destroy him in the end given that his favorables among the wider general electorate suck