Now that the penalties on FL and MI are the same as the GOP penalty....

[I spent the winter writing songs about getting better]

Read the rest of the thread now. Thanks everybody. Glad to see even people who don't like me can also see obvious hackery and name calling immaturity.

Also hilarious is how J. J. thinks Hillary can win a credentials committee appeal when the credentials committee will have a majority of Obama supporters based on how it's seated. And she couldn't even keep all her people on the RBC committee together to support full seating. And that even the Hillary campaign has basically admitted to accepting the Florida decision at least, thus rendering Florida a dead and decided issue.

[J. J.]

Will the Republican/Hillary hacks quit their whining considering they didn't make a peep about what their own party did?

No, it should have been handled months ago and Hillary gets to challenge at the credentials committee.  Of the there was Obama hack, Li'l Zack who kept on saying the wouldn't be seated and we should ignore it.  LOL.

This is an I-told-you-so.

The credentials committee will be controlled by Obama based on how it's seated. Hillary's not winning anything there.

You were the one talking about Obama having to grab as many of the uncomitted as possible. LOL. It's a moot issue. I said the 69-59 would be enacted if Michigan was seated and you scoffed.

At 50% voting strength.  That could change and it might, depending on the circumstances.  Further, they can choose to seat the delegates using there own method.

There's a problem for Obama getting the nomination...but thus no problem for Hillary getting the nomination?

No, there is a problem of not ending the process.  If the super delegates sit there and do nothing, there will be a problem.  If Obama is sitting there, in mid June, with a lead but not the nomination, it will be damaging to him.  Look, I've posted several times that the super delegates really should move in and end this (and should have a few weeks ago).

If Hillary were sitting there in the same situation, she'd have the same problem.

2.  Poll numbers.  Hillary, nationally, has been doing better than Obama.  If she should improve to the point where she is consistently beating McCain and Obama is consistently losing, you may see a shift in the super delegates.  That means Obama losing some he already has.

Let's talk about this once a single Obama delegate declares for Hillary (something that has never happened but the reverse has frequently.) Until then it's just another meaningless "what-if".

[/quote]

The last "what if" happened yesterday (and it may not be over).  Right now, about half the Democratic electorate voted for Obama, about half voted for Clinton.  Now some of those voters that voted for Obama will shift to McCain.  Obama's numbers, after becoming the presumptive nominee, may drop because of that.  They may not rebound in time for the convention and the super delegates might be looking at nominating a candidate who is dropping (or has dropped) in the polls.

Ideally, if I were advising Obama, I'd suggest that, to be really secure, he gets another 110-120 delegates (and that is possible).  Greenpages shows that if everyone is seated at full strength, he'd need 108.  I'd use the fully seated target number of 2210.  By that count it's:

Obama:   2,102.0
Clinton:   2,013.0

In SD and MT, he'll pick up some, and get it into double digits.

[J. J.]

No, the state results.

What kind of random distinction is that?

Did you just think that western Virginia was some blip?  Why do results like that need to be statewide for them to matter?

Obama is still doing fine among the working-class in states like Wisconsin.  Obama has improved some among liberals and lost some among conservatives, but other than that, he hasn't "tanked" and you still haven't proven he has.

I've always known that there were some areas that would probably not vote for a black candidate, even if it was Jesus Christ.  I never thought it would the scope that it was in OH, TX or PA (though I knew that he'd get blasted in some districts in the latter case).

[Alcon]

I've always known that there were some areas that would probably not vote for a black candidate, even if it was Jesus Christ.  I never thought it would the scope that it was in OH, TX or PA (though I knew that he'd get blasted in some districts in the latter case).

Then you don't understand the demographics of these rural areas.  A simple sociodemographic regression model (devised by Nate "Poblano" Silver) pretty much nailed those rural counties -- yes, based on the results of places that voted well before March.

Unless you're trying to say that 90% of Buchanan County, Virginia, "would not vote for a black candidate, even if it was Jesus Christ," it's obvious that Clinton was already deeply positioned as the candidate of working-class whites.  If you didn't see it, that's fair, but you can't in turn ignore that as evidence of a lack of a collapse for Obama among working-class whites.

[J. J.]

I've always known that there were some areas that would probably not vote for a black candidate, even if it was Jesus Christ.  I never thought it would the scope that it was in OH, TX or PA (though I knew that he'd get blasted in some districts in the latter case).

Then you don't understand the demographics of these rural areas.  A simple sociodemographic regression model (devised by Nate "Poblano" Silver) pretty much nailed those rural counties -- yes, based on the results of places that voted well before March.

Unless you're trying to say that 90% of Buchanan County, Virginia, "would not vote for a black candidate, even if it was Jesus Christ," it's obvious that Clinton was already deeply positioned as the candidate of working-class whites.  If you didn't see it, that's fair, but you can't in turn ignore that as evidence of a lack of a collapse for Obama among working-class whites.

I didn't see it because I looking at a larger scale, Buchanan County is more of a "rural" as opposed to "working class" county, from what I could tell from Wiki.  "Working class" is more urban or suburban.

[Alcon]

I think that defining the "working class" as urban and suburban, and then pointing to his loss in Ohio/Pa. is misplaced, but OK.  How about you tell me some counties where Obama would have "tanked" since his showings in February, then.  That would be a good starting point for this.

[I spent the winter writing songs about getting better]

Will the Republican/Hillary hacks quit their whining considering they didn't make a peep about what their own party did?

No, it should have been handled months ago and Hillary gets to challenge at the credentials committee.  Of the there was Obama hack, Li'l Zack who kept on saying the wouldn't be seated and we should ignore it.  LOL.

This is an I-told-you-so.

The credentials committee will be controlled by Obama based on how it's seated. Hillary's not winning anything there.

You were the one talking about Obama having to grab as many of the uncomitted as possible. LOL. It's a moot issue. I said the 69-59 would be enacted if Michigan was seated and you scoffed.

At 50% voting strength.

Which is exactly what the DNC rules state as the MINIMUM penalty.

That could change and it might, depending on the circumstances.  Further, they can choose to seat the delegates using there own method.

Yet you unsurprisingly completely ignore my point about how the makeup of the credentials committee is determined. There is simply no way it can be less favorable to Hillary than the RBC. Then there's the fact the Hillary camp has even tacitly accepted the Florida decision. Florida is done.

No, there is a problem of not ending the process.  If the super delegates sit there and do nothing, there will be a problem.  If Obama is sitting there, in mid June, with a lead but not the nomination, it will be damaging to him.  Look, I've posted several times that the super delegates really should move in and end this (and should have a few weeks ago).

If Hillary were sitting there in the same situation, she'd have the same problem.

Obama only needs about 30 more superdelegates. He'll pick up around 7 this weekend and another good chunk the next through add-ons. Even just looking at things this week that equals 7 + 5 declared "Pelosi club" members. That means he needs 18 more superdelegates, basically what he took in the few days following May 6.

If Obama can't pick up 18 superdelegates I'll agree there's a problem, but there's no much reason to believe that's not going to happen. The point though really is Obama doesn't need a gigantic block of superdelegates to end this before the end of this week.

2.  Poll numbers.  Hillary, nationally, has been doing better than Obama.  If she should improve to the point where she is consistently beating McCain and Obama is consistently losing, you may see a shift in the super delegates.  That means Obama losing some he already has.

Let's talk about this once a single Obama delegate declares for Hillary (something that has never happened but the reverse has frequently.) Until then it's just another meaningless "what-if".

[/quote]

The last "what if" happened yesterday (and it may not be over).  Right now, about half the Democratic electorate voted for Obama, about half voted for Clinton.  Now some of those voters that voted for Obama will shift to McCain.  Obama's numbers, after becoming the presumptive nominee, may drop because of that.  They may not rebound in time for the convention and the super delegates might be looking at nominating a candidate who is dropping (or has dropped) in the polls.

Ideally, if I were advising Obama, I'd suggest that, to be really secure, he gets another 110-120 delegates (and that is possible).  Greenpages shows that if everyone is seated at full strength, he'd need 108.  I'd use the fully seated target number of 2210.  By that count it's:

Obama:   2,102.0
Clinton:   2,013.0

In SD and MT, he'll pick up some, and get it into double digits.
[/quote]

Once again you completely ignore the point that Hillary just had her best chance at FL and MI she'll ever get. Obama will control the credentials committee. This is all really moot. The decision on those is basically over, and on Florida, it basically IS unarguably over since the Hillary camp has basically accepted the result. Please note that even Hillary's supporters on the RBC voted in favor of the half seating of Florida and many crossovered to the Michigan plan.

And BTW, why is it that even people who normally don't get along with me aren't taking your side here?

[J. J.]

I think that defining the "working class" as urban and suburban, and then pointing to his loss in Ohio/Pa. is misplaced, but OK.  How about you tell me some counties where Obama would have "tanked" since his showings in February, then.  That would be a good starting point for this.

One in PA that I can think of is Allegheny County (Pittsburgh), in PA.  In February, looking at the demographics, I would have said a reasonably good victory for Obama.  Montgomery County looked like a likely Obama victory them, Beaver, Lawrence, Lehigh, Luzerne, Northhampton, Berks; except for the Hillary connection, I'd expected Obama to do okay in Lackawanna.  Even Blair and Erie were a possibility for a win. I expected some bad counties, like Cambria, Bedford, possibly Westmoreland.

If you would ave told me in January that Obama would be running against Clinton in PA and outspending her three to one, but go on to lose Allegheny, Montgomery, Beaver, Lawrence, Lehigh, Northhampton Counties, I would have called you insane.

Up to the primary, I though Allegheny was a possibility but with a loss of main CD there (14th ?) but a narrow win county wide.  Some of the working class counties are actually worse than the rural counties!

[Alcon]

What prior county results were you using to make that assumption, though?  You can't just say "I expected Obama to win Allegheny" and leave it at that.

Silver's regression model essentially nailed Allegheney, by the way...so it did vote relative to previous county results, so you're going to have a hard time substantiating your position here.

[J. J.]

Will the Republican/Hillary hacks quit their whining considering they didn't make a peep about what their own party did?

No, it should have been handled months ago and Hillary gets to challenge at the credentials committee.  Of the there was Obama hack, Li'l Zack who kept on saying the wouldn't be seated and we should ignore it.  LOL.

This is an I-told-you-so.

The credentials committee will be controlled by Obama based on how it's seated. Hillary's not winning anything there.

You were the one talking about Obama having to grab as many of the uncomitted as possible. LOL. It's a moot issue. I said the 69-59 would be enacted if Michigan was seated and you scoffed.

At 50% voting strength.

Which is exactly what the DNC rules state as the MINIMUM penalty.

Which can be waived (and were for other states).

I just realized it.  The half delegates could vote to let their "other half" be seated.  The can do it at the convention.  Now THAT is tricky.

That could change and it might, depending on the circumstances.  Further, they can choose to seat the delegates using there own method.

Yet you unsurprisingly completely ignore my point about how the makeup of the credentials committee is determined. There is simply no way it can be less favorable to Hillary than the RBC. Then there's the fact the Hillary camp has even tacitly accepted the Florida decision. Florida is done.

No, there is a problem of not ending the process.  If the super delegates sit there and do nothing, there will be a problem.  If Obama is sitting there, in mid June, with a lead but not the nomination, it will be damaging to him.  Look, I've posted several times that the super delegates really should move in and end this (and should have a few weeks ago).

If Hillary were sitting there in the same situation, she'd have the same problem.

You must be logged in to read this quote.

It's more like 80 when you look at the unsanctioned states, and that assumes god wins MT and SD.  He basically has to say the , "Even if MI/FL are seated in full, I still have enough."

Once again you completely ignore the point that Hillary just had her best chance at FL and MI she'll ever get. Obama will control the credentials committee. This is all really moot. The decision on those is basically over, and on Florida, it basically IS unarguably over since the Hillary camp has basically accepted the result. Please note that even Hillary's supporters on the RBC voted in favor of the half seating of Florida and many crossovered to the Michigan plan.

It's neither her best nor her last chance, unfortunately.  She can do it, with enough delegates,  The question is, can Obama get enough to stop her.  I just had an epiphany.  It would not be pretty, but it's possible.  A parliamentarians nightmare but a coup masters delight.

You must be logged in to read this quote.

So where a number when I said Hillary would win NH.

If Omaba can't get to 2210, there is a fair chance that Hillary can pull it off.  It would not be pretty, but it wouldn't violate the rules either,   

[J. J.]

What prior county results were you using to make that assumption, though?  You can't just say "I expected Obama to win Allegheny" and leave it at that.

Silver's regression model essentially nailed Allegheney, by the way...so it did vote relative to previous county results, so you're going to have a hard time substantiating your position here.

I'm not a regression model, I'm looking at the pattern in other elections.  IIRC, Pittsburgh tended to vote the same as Phila (when the candidate wasn't from Pittsburgh).  I've been going back to the governor's races; the presidential one was too far back, 1988.

You'll have to post the link to the model; I attempt to understand it/

Now, you've asked me telly you what counties surprised me, and it was the ones listed. 

[I spent the winter writing songs about getting better]

Which can be waived (and were for other states).

I just realized it.  The half delegates could vote to let their "other half" be seated.  The can do it at the convention.  Now THAT is tricky.

Sure they could. The question is will they attempt (since the state parties consider the matter closed and have accepted the compromise, and if it'll pass. Not likely.

Your scenarios you bring up here are similar to someone saying a candidate could lose the popular vote and not win enough states to win an electoral victory, but could still win by persuading enough electors for the other candidate. Just because it's theoretically possible doesn't mean it has a likely enough chance of happening to be discussed as a serious possibility.

That could change and it might, depending on the circumstances.  Further, they can choose to seat the delegates using there own method.

Hillary has no chance at a majority of pledged delegates, even with FL and MI seated fully (which in this case moot since it'd become a catch-22 scenario) and since the credentials committee is based on the proportion of pledged delegates...well it's obvious. You're the one who needs to explain how it can change, or what type of "their own method" could be used. Without details these scenarios are nonsense, it's like me saying there's a way Obama could lose Pennsylvania, Colorado, Virginia, Ohio, Florida and New Mexico and still beat McCain, and refuse to explaiin how it happens.

So how does Hillary get a majority on the credentials committee? (And not just that, but a majority that will rule in her favor on everything, which her supporters on the RBC did not.)

Furthermore, the vote on Florida's current situation on the RBC was unanimous. So you're basically saying that Hillary can get a unanimous decision by a committee fairly favorable to her overturned by a far more unfavorable committee? LOL!

No, there is a problem of not ending the process.  If the super delegates sit there and do nothing, there will be a problem.  If Obama is sitting there, in mid June, with a lead but not the nomination, it will be damaging to him.  Look, I've posted several times that the super delegates really should move in and end this (and should have a few weeks ago).

If Hillary were sitting there in the same situation, she'd have the same problem.

It's more like 80 when you look at the unsanctioned states, and that assumes god wins MT and SD.

The Florida Democratic Party has accepted the current situation and considers the matter closed. So has the Hillary campaign tacitly by only saying they're keeping the possibility of a Michigan challenge on the table, not both states.

The Michigan Democratic Party also considers the matter closed. Hillary may not, but the dispute is only over 4 half-delegates. So that means Obama only needs 2 more in that scenario.

The issue is as far as the states involved are concerned, is closed. You're the only one who keeps thinking it'll still be a big deal and claim that Hillary can somehow magically win a majority on the credentials committee without explaining that it happens.

He basically has to say the , "Even if MI/FL are seated in full, I still have enough."

And if he doesn't, but still has a majority at the convention with FL and MI seated as is...he's still the nominee. So no, he doesn't have to say that.

It's neither her best nor her last chance, unfortunately.  She can do it, with enough delegates,  The question is, can Obama get enough to stop her.  I just had an epiphany.  It would not be pretty, but it's possible.  A parliamentarians nightmare but a coup masters delight.

Except as Obama has now clinched a majority of pledged delegates under any scenario, she can not get enough delegates to do it. So it's not possible. You need to explain how Hillary can win a majority on the credentials committee.

[J. J.]

Which can be waived (and were for other states).

I just realized it.  The half delegates could vote to let their "other half" be seated.  The can do it at the convention.  Now THAT is tricky.

Sure they could. The question is will they attempt (since the state parties consider the matter closed and have accepted the compromise, and if it'll pass. Not likely.

Not likely at the current time.  This thing is two months away and it's close.

That could change and it might, depending on the circumstances.  Further, they can choose to seat the delegates using there own method.

Hillary has no chance at a majority of pledged delegates, even with FL and MI seated fully (which in this case moot since it'd become a catch-22 scenario) and since the credentials committee is based on the proportion of pledged delegates...well it's obvious. You're the one who needs to explain how it can change, or what type of "their own method" could be used. Without details these scenarios are nonsense, it's like me saying there's a way Obama could lose Pennsylvania, Colorado, Virginia, Ohio, Florida and New Mexico and still beat McCain, and refuse to explaiin how it happens.

[/quote]

And Obama will not get a majority of the popular vote.  Neither metric is the one used to choose the nominee. 

You must be logged in to read this quote.

She doesn't in the current circumstances.  Will those be the same circumstances in two months?  I don't know.  I can see a 1976 situation.

The Florida Democratic Party has accepted the current situation and considers the matter closed. So has the Hillary campaign tacitly by only saying they're keeping the possibility of a Michigan challenge on the table, not both states.

The Michigan Democratic Party also considers the matter closed. Hillary may not, but the dispute is only over 4 half-delegates. So that means Obama only needs 2 more in that scenario.

The issue is as far as the states involved are concerned, is closed. You're the only one who keeps thinking it'll still be a big deal and claim that Hillary can somehow magically win a majority on the credentials committee without explaining that it happens.

You must be logged in to read this quote.

And if he doesn't, but still has a majority at the convention with FL and MI seated as is...he's still the nominee. So no, he doesn't have to say that.

BRTD, Obama needs to get into that position to end the process fully.  That's why I'd like to see him hit really 2210+.

Except as Obama has now clinched a majority of pledged delegates under any scenario, she can not get enough delegates to do it. So it's not possible. You need to explain how Hillary can win a majority on the credentials committee.

Again, it's not the pledged delegates any more than it is a majority of the popular vote.  It's the delegates.

As for what can happen, 1976 for the Republicans.  A rule change would have released some delegates from Ford to Reagan; it almost happened.  What was the political impetus for this?  Reagan winning in a landslide?  No.  A Ford scandal?  No.  Polls showed that Reagan would lose by slightly less against Carter than Ford would; IIRC, the polls were showing that Reagan would do about one point better against Carter, but still lose badly.  That was almost enough.

You want to tell me what the polls will look like in August?

[jimrtex]

Will the Republican/Hillary hacks quit their whining considering they didn't make a peep about what their own party did?

The GOP imposed the same penalties on Michigan, Florida, New Hampshire, South Carolina, and Wyoming.  Nobody kept McCain from campaigning in Florida or South Carolina, or Romney in Michigan and Florida.

Weren't the original Democrat sanctions supposed to be 1/2 the elected delegates, and all the superdelegates, etc. from Michigan, Florida, and New Hampshire.

[jimrtex]

... Li'l Zach the Barak Hack ...

Okay, I don't like BRTD either, but dude, you're 46.

It's not a question of liking him.  It is a question of Li'l Zach the Barak Hack's blatant, and false, spinning on this issue.

Am I allowed to ask why you remove the "C" from "Barack" when there's a "C" in "Zach"?

Maybe Zach is a hack for Ehud?

[Erc]

The real point at issue is not the final penalty the DNC put on MI/FL...at this point [J.J.'s comments aside], it wouldn't have mattered what was decided last Saturday.  Obama would still have won the nomination handily even under the worst case scenario (MI/FL fully seated, no interference with Michigan's Delegate Selection process)...it would just happen a week or two later.

The stripping of MI/FL's delegates had an impact on the race as they affected the media cycle in January, and what an impact they had.

Even forgetting Michigan, just imagine if Florida had had half-delegates as the Republicans did.  Coming three days after Obama's win in SC (or even on the same day, if SC hadn't moved up three days), Clinton's big win (perhaps closer than in real life, but still a large win nonetheless) would have shut down Obama's ever-gathering media frenzy at a critical time.  I feel that could have made a rather substantial difference on Super Tuesday, much as (though to a lesser extent) McCain's win in FL did.  (Edwards, coming in at under 15% in Florida in this scenario as well, would probably still drop out at around the same time).

[J. J.]

I am worried about a 1976 situation, and that does neither candidate (or the republic in general) any good.

[Iosif is a COTHO]

I still can't understand why Obama didn't just give her the damn delegates. He'd still have won.

[Alcon]

I'm not a regression model, I'm looking at the pattern in other elections.  IIRC, Pittsburgh tended to vote the same as Phila (when the candidate wasn't from Pittsburgh).  I've been going back to the governor's races; the presidential one was too far back, 1988.

You'll have to post the link to the model; I attempt to understand it/

Silver's site is currently down, but it's a fairly simple regression model involving ethnic identification, income, education levels and other variables.  But the contention here is that Obama crashed among the working-class.  Your comparison point cannot be other, past races in Pennsylvania; that doesn't establish a true relative index.

I want to see actual evidence of huge degradation, more than the normal process of increasing polarization throughout the primary season.  "I expected the Pittsburgh area to be better for Obama" is not one.

Now, you've asked me telly you what counties surprised me, and it was the ones listed. 

I didn't ask you what counties surprised you.  I could point to Curry County, Oregon, and tell you that Obama is surging among the working-class and the old.  That would be disingenuous.  I asked you which counties demographically and culturally similar to Southeastern Pennsylvania voted previously, where Obama's performance had since fallen.  There are none.  And the exit polls agree with me.

The working-class in Oregon was always going to be better for Obama than the working class in New Hampshire, which was always going to be better than in Pennsylvania, etc. etc. etc.

[Torie]

Alcon what are the variables in the regression analysis that "explain" why Obama does better with the white working class in Oregon than  in Pennsylvania?  I have my own little theories, but I was wondering what was contained in the set of variables used in the formula.

[Alcon]

Alcon what are the variables in the regression analysis that "explain" why Obama does better with the white working class in Oregon than  in Pennsylvania?  I have my own little theories, but I was wondering what was contained in the set of variables used in the formula.

Honestly a good question.  Silver does "near neighbors" analyses too -- actually stolen from baseball -- that identify similar congressional districts.  He's never made his methodology exactly public, and it does have errors, since not everything is quantifiable.  He did poorly with Kentucky, and after seeing his model's underestimation of Clinton in southeastern Indiana, it wasn't surprising.  But fundamentally, his prediction in the Pittsburgh area was solid.  The ethnic working-class folks voted Clinton heavily.  The white-collars voted Obama moderately.  The blacks voted Obama heavily.  His only error was under-estimating the strength of the working-class ethnic vote, and that's mostly because some religious differences and cultural differences (working-class folks in PA are a lot less WASP/"WASNR"-y than in OR; the "redneck"/"ethnic American" variable affected KY+WV) weren't quantified in the model.  They can't be, but we all know they're there.

But I think I'm getting a little off-topic.  Whether the area was demographically predictable was non-point.  What really matters is evidence of a relative collapse, which I really don't see.

I do wonder a lot how much it affects state results, whether a state has already voted or not.  When I see these race re-hash polls with Obama +13 in California, I wonder whether Oregon has trended similarly or whether people are much more apt to change their mind after they voted.  Obviously that question affects any relative comparison.

[Mr. Morden]

Btw, when did Silver actually construct his regression analysis?  Obviously, if it was derived after most of the primaries were already held, then it's a lot less impressive than if it was derived early on.  Once you already know how everything worked out, it's not necessarily that hard to fit a good model to the data.

[Alcon]

Btw, when did Silver actually construct his regression analysis?  Obviously, if it was derived after most of the primaries were already held, then it's a lot less impressive than if it was derived early on.  Once you already know how everything worked out, it's not necessarily that hard to fit a good model to the data.

The model is polymorphic.  It adjusts itself as new results come in, and he's tweaked variables.  Of course, that makes its successes less impressive as time goes on, but the intent wasn't to establish a model that could predict the outcome of the entire primary season.  It is useful in that it shows relative "expected" performance differences over time -- that is, areas where Obama and Clinton out-performed quanitifiable demographics relative to past races.

My point was that there are two conclusions to be made about the fit of Allegheny County to the prior model.  Either:

1. It was accurate, and Obama essentially performed at par with the demographics he was expected to.

2. Obama deteriorated among the working-class, but if he hadn't, the model would have underestimated his performance because of some non-quantified/non-quantifiable cultural variable we didn't see.

I don't know what Obama demographic in the Pittsburgh area Silver's model wouldn't pick up, so (2) seems doubtful to me.  In fact, being that the two greatest flaws of the model have been an underestimation of Clinton's strength among some working-class ethnic Catholics and evangelicals, and an overestimation among working-class WASPs (because Census numbers don't really reflect these cultural differences beyond the "American" value), I'd argue the opposite would be more likely to be true -- which explains why it didn't fare as well in the outlying Pittsburgh metro.

I guess you can retroactively construe any result to "fit," but I'm just saying, I'm very skeptical of Allegheny County as strong proof of an Obama collapse among the working-class.

[Torie]

Tweaking a model as more data comes in, has a data mining aspect about it (which is a no-no, and one of the worst insults that can be levied in the statistical world). And it tends to bury whatever is happening out there in the news, that might be changing voting behavior. Far better is to have a model ex ante, and see how it works.

In any event, I suspect that Clinton has increased her support over time with economically stressed white voters as an overlay to all of the cultural stuff, but I could be wrong.

[J. J.]

So basically we don't have any evidence that there was "always" the problem with working class white votes for Obama?