Final March 4th Polling Outlook

by: Chris Bowers

Tue Mar 04, 2008 at 12:09


Here are all of the polling averages for tonight's five contests in a single table (source for polls):

March 4th Polling, At A Glance
State Date Polls Obama Clinton P. Delegates
Ohio Mar 04 8 42.8% 51.0% 141
Rhode Island Mar 04 4 38.8% 49.0% 21
Texas C Mar 04 0 -- -- 67
Texas P Mar 04 7 45.9% 48.1% 126
Vermont Mar 04 3 56.7% 35.3% 15

These polls lead to the following delegate projections:

Delegate Projections, Based On Current Polling
State Date Obama Clinton P. Delegates
P. Delegates Jun 07 1,194.5 1,033.5 2,272
Ohio Mar 04 64 77 141
Rhode Island Mar 04 9 12 21
Texas P Mar 04 62 64 126
Texas C Mar 04 38 29 67
Vermont Mar 04 9 6 15
Sub-Total Mar 04 182 188 370
Grand Total Jun 07 1,376.5 1,221.5 2,642

I calculated the Texas caucus results based on the average delegate results from Nevada and Maine, which since Iowa have been the two most favorable caucuses for Clinton (New Mexico was CINO, Caucus In Name Only). Overall, that comes out to a Clinton advantage of 188-182. Could March 4th be the first day that Clinton has won a victory in delegates? Obama has been undefeated on that front the entire campaign, which is one of the reasons why he is still heavily favored in delegate math.

More details on these numbers in the extended entry.  

Chris Bowers :: Final March 4th Polling Outlook
Seven Texas polls were conducted entirely in March, and generally speaking they showed similar results. Here is the average of those six polls:

Texas, 3/1-3/3, Seven Poll Average
Clinton: 48.1%
Obama: 45.9%

There is clearly momentum for Clinton in Texas, as she has turned a deficit of 1-2% into a 2-3% lead over the past week. Obama can still win as long as he dominated early voting, as the various polls conflict as to who leads in that category. If Obama enters today with a 10% lead, as the Public Strategies poll indicates, then he will still go on to win the state. If Clinton enters today with a 3-4% lead, as Survey USA indicates, then she will probably go on to a narrow victory in the primary portion of the Texas nominating contest. Turnout models for primaries are also difficult to predict, with Public Strategies noting:

We've updated our racial turnout analysis from last night and again find that if African-American turnout equals or surpasses Hispanic turnout, Obama will win.

New Democratic primary turnout patterns have made polling primaries difficult this year, as women, young voters, affluent voters, African-Americans and Latinos have turned out at higher levels than 2004. In most states so far, the net result of this unexpected turnout has favored Obama, but there have also been cases, most notably in California, Massachusetts, and New Hampshire, where is has favored Clinton.

While all of the above caveats also apply to Ohio (and indeed, to every state), the final Ohio polling picture presents a much clearer outcome:

Ohio, Eight Poll Average, only including March polls
Clinton: 51.0%
Obama: 42.8%

Clinton has halted Obama's momentum in Ohio, and even increased what was once a 6-7% lead out to 8.2%. Overall, neither state shows a particularly large amount of momentum for Clinton (4% at the most), but she got it at the right time in order to maintain her large, early leads in the state. Obama was coming from way back in both states, and any setbacks like those over the past three days would always prove difficult to overcome.

Tags: , , , , , , , (All Tags)
Print Friendly View Send As Email
Final March 4th Polling Outlook | 46 comments
All this and Clinton nets 6 delegates (0.00 / 0)
That's what's giving Democrats fits. All the attacks from Clinton, the McCain boosting, the darkened picture of Obama, etc etc. Hardball politics, for sure. But it's all to net 6 delegates.
Exit polls (0.00 / 0)
Chris,

I posted the table you asked for here.  In case you missed it, I didn't want you to think I'd forgotten.

'Course, the nubmers will be incomplete by tomorrow!

As an extension of your work on late deciders (4.00 / 3)
The table below contains information from exit polls on when people decided, and how the split.

The exit polls show a consistent pattern: Obama wins by large margins among those who decide in the last week.  Among those who decide on the day of the primary, Clinton tends to at least split, and more often win.  

There is a myth circulating about Obama outperforming polling.  This analysis is based on pollster.com averages.  The problem is pollster.com includes polls in their average from a week out.  If you limit it to just polls taken in the last 48 hours it is much harder to conclude that Obama outperforms his polling on a systematic basis.  


[ Parent ]
I'm using Chris' research (0.00 / 0)
Having trouble finding it at the moment but Chris put together the original research that I've been basing my ideas on.
[ Parent ]
I have an excel (0.00 / 0)
file with all of the exit polling, plus the pre-polling data if you are interested.

I would actually like to get about 5 people together to start a project of compiling data for research.  My goal would be to get a db together with all exit poll data from 92 onwards, and pre-election polling since 1980.

I have a good deal of polling history data, but it would be great to share the work....

[ Parent ]
Nice (0.00 / 0)
Yeah, I left that out of the table because all the elections weren't held on the same day.  Your table solves that problem nicely.

I disagree with you on one point though: both candidates should outperform the polls on a systematic basis because the undecideds vote for one or the other (or stay home).  But they don't - here's data just from SUSA and PPP (they give cross tabs, that's why) and just the most recent polls.  I think most of them are within 48 hours.  Obama outperforms the polls but so does Clinton.  Take a look:

Free Image Hosting at www.ImageShack.usFree Image Hosting at www.ImageShack.us
Click to enlarge.  Obama on left, Clinton on right.

The key so far has been the amount of the electorate that is African-American, because these same polls have tended to both underestimate the Black turnout and overestimate the Black support for Clinton (see here).

So when we see the people discussing the 'myth' of Obama outperforming polls, if you come up with an average value for Clinton and Obama, yeah, it'll be Clinton +1, Obama +7.  But that is very very misleading.  The averages are skewed by the results in states with very high proportions of Black voters.

Looking forward to tonight, past history suggests both Clinton and Obama should do a few points better than their polls, unless African-Americans in Texas overwhelm the polls and drive up turnout to 40% of the electorate.  But the other aspect is the Hispanic vote, which could also be very enthusiastic, and is not accounted for in the graphs above.

Anyway, based on previous performance, I think we can expect Clinton to get maybe 54-56% in Ohio and 49-51% in Texas, if we assume similar conditions prevail in this round of primaries.

We'll see.

[ Parent ]
Just going on record (0.00 / 0)
again to say I think you're (likely) making a big mistake by dismissing the possibility that Obama will continue outperforming his polls. This wouldn't of course preclude the recent movement toward Clinton, which seems clear. It just moves the parameters.

I know of course that we don't much of anything about the recent underpolling of Obama so of course we can't with any certainty assume it will continue. But the converse is also true: there is just as little reason to assume with certainty that it won't continue.

It seems to me that the underpolling happened. We know that much. The best explanation for this in my opinion is the pollsters missed Obama's edge in their turnout assumptions. I have seen zero evidence anywhere that this has been rectified, although it could well have. Until I'm provided with this evidence, I will continue to assume that they're using the same rough assumptions that they always have because that is what they do.

And in 6.5 hours we'll begin to know who is right...

I agree (0.00 / 0)
especially considering that turnout models are ALL using way too low of a percentage of black turnout in TX.  All are around 20%, which is the same as 4 years ago.  I just don't buy it.  

The Hispanic turnout, I have no idea.  I do think South Texas has notoriously been a low turnout area historically, but we will see.

But considering how close the polls were in Wisconsin immediately before that primary, and immediately following the "plagiarism" attacks from the Clinton camp, only to see Obama outperform by at least 10% or so...really makes me question what will happen tonight.

[ Parent ]
Latino turnout (0.00 / 0)
matters for the popular vote, but not the delegate vote, because the South Texas precincts have historically lower turnout and so fewer delegates.  The black voters are in the cities that have the greater number of delegates based on past voting patterns.  Also in the CDs with odd numbers of delegates, which are easier to take an edge in.  South texas needs 63% margins for Hillary to avoid an even split.

John McCain--He's not who you think he is.
[ Parent ]
No (4.00 / 1)
South Texas historically has the highest turnout rates in the state for the primary. Austin is the only other area with good turnout. In the general election, everybody turns out and South Texas doesn't have the same disproportionate influence.

Latinos outside South Texas turn out in lower numbers.

Also, despite what the exit poll said, African-Americans were only about 12% of the 2004 Dem primary.

Statewide, Latinos are a third of the population while African-Americans are only 11%. Both groups are registered at lower rates than whites. Latinos are more likely to vote GOP in the fall than African-Americans, but 80% of Latinos vote in the Dem Primary. They were 34% of the 2004 vote.

Those turnout numbers are based on the actual voter file.

My polling kicked African-Americans up to 22%, but that recognizes increased turnout all over, including heavy crossover from white Bush voters since the GOP race isn't competitive.

[ Parent ]
Depressing (4.00 / 1)
Guess the kitchen sink strategy of dragging Obama through mud and claiming that even McCain is better than him is working.

Can only hope that he's able to outperform these polls again, though the momentum is clearly trending the wrong way.

And then with a net 6 delegates (if accurate) for Clinton, which will likely be erased in the couple elections before PA, that somehow justifies continuing the race anyway despite what it might mean for November.

Bleh... guess we'll see what happens tonight.

It seems to me (0.00 / 0)
that outperforming the polls and the recent movement toward Hillary are quite possibly two separate activities. In other words, both can be true.

Let's assume for the moment that there is something about the recent Obama wave that the polls are missing (as noted in the above post, I assume it lies in turnout assumptions). This underpolling cuts about 9 percent off of Obama's real margin (e.g., Obama predicted to win by 3, actually wins by 12). Then let's assume that this underpolling problem hasn't been fixed.

We can then take this 9 points and add it into recent polling. If a HRC average polling lead in Ohio has jumped from about 3 to about 7 then it's quite possible that the real numbers are Obama +6 to Obama +2. Strong movement of late toward Hillary but not enough to overcome Obama's ongoing wave.

[ Parent ]
Don't disagree... (0.00 / 0)
I guess we'll see, though.  I'm certainly not going to just assume that we can just add +9 to Obama's numbers to predict the outcome. =)
[ Parent ]
Failure of imagination (4.00 / 1)
Has been the hallmark of this primary cycle.

The Iowa campaigns failed to imagine the massive turnout that would take place (no one but Obama saw anything over 170K).  No one saw a Clinton comeback from down 10 36 hours out in New Hampshire.  No one saw the size of the South Carolina victory, and so on.

In Texas, there are two dimensions of this failure of imagination: (1) a failure to anticipate what may be enourmous african american turnout (see SC for another example) and (2) a failure to foresee Hispanic turnout over 31%.  

As I wrote yesterday, there is significant movement to Clinton.  We may look back and see the NAFTA meeting as a key turning point in the race.  My guess is she wins Texas by over 5 and Ohio by close to 15.  The problem is I don't see where Clinton gets the votes to win the nomination.  And so this season begins to look like 1980 to me with Clinton playing the weakened Kennedy, too far behind to win the nomination but perfectly capable of making sure Carter could not win a general election.  

Nail on the head... (0.00 / 0)
I don't think your predictions are accurate, but I worry that the result of a Clinton comeback here just means that Obama will continue to have to contend with two targets, both beating up on him and both claiming, apparently, that McCain is better than him.

Yeah... that can't be good for November if even your Democratic opponent says that McCain is better.

[ Parent ]
The NAFTA issue... (0.00 / 0)
It'll probably help boost Clinton's margin in Ohio, but I don't see it affecting Texas as much.
[ Parent ]
Do people mind foreign interference in our elections? (4.00 / 1)
It seems pretty clear now that the canadian conservatives leaked the info to hurt Obama, and probably help McCain, rather than Hillary, seeing her as the weaker candidate, as most conservatives do.

I'd think there might be some backlash here, that is if it gets reported on.

John McCain--He's not who you think he is.

[ Parent ]
Super Delegates great new idea (0.00 / 0)
I just read Jonathan Tasini's blog over at the Huffington Post with a new proposal on assigning super delegates.  He is saying that a series of policy positions should be promoted and an open left is sent to both Clinton and Obama to see if they will adopt these policy positions.  If so, the superdelegate commits to that candidate.  I think that's a super idea.

His diary is focused on trade and he specifically references Public Citizens paper on how to enact many of their domestic policies they must renegotiate GATS with the WTO.  I doubt most realize that to obtain health care, labor, environmental standards which sound like US domestic policy, the WTO GATS could supersede US legislation and policy, so this gives an added awareness of why a position to rengotiate trade is so critical.  It's not just trade, it's a host of domestic policy initiatives that are affected.  

To get the best policy positions, in the US national interest, working America's interests the idea of superdelegates using this situation to obtain better policy sounds awesome.

And let's face it, how many superdelegates are going to release an open letter with a corporate lobbyists agenda on it, in public?  

NoSlaves.com  


The Economic Populist

open letter, not your blog name, typo n/t (0.00 / 0)
[ Parent ]
Do we understand yet that policy isn't what most people vote on? (0.00 / 0)
It may appeal to wonks, but most voters focus on a Gestalt of the candidate, including things such as how inspiring the candidate is.  Slicing and dicing the electorate into niches and having policies to appeal to each niche misses the appeal of a candidate who stresses community and the common good, and who inspires us to something beyond self-interest.


John McCain--He's not who you think he is.
[ Parent ]
well (0.00 / 0)
I really don't care if people are so easily manipulated that they are not voting on policy, the reality is policy is what will set the future direction of this country.

All the more reason to put the arm of reason onto the super delegates and stop this Ronald McDonaldization of political elections.  

NoSlaves.com  


The Economic Populist

[ Parent ]
Super Delegates Don't Choose the Winner of a General Election (0.00 / 0)
That's a nice thought, but Super Delegates don't get to determine who the winner of the General Election will be, nor how the public at large will decide who to vote for.

I would think that after Bush was RE-ELECTED, it'd be obvious that people don't generally vote for policies (or, at the very least, in a close election, "policy" voters don't matter).

[ Parent ]
I'll take Ronald McDonald (0.00 / 0)
over some of our superdelegates any day of the week.
[ Parent ]
Most people don't understand policy (0.00 / 0)
the way we do here.  Many don't even see issues as we do, or see policies as possible solutions to problems.  Not because they aren't as smart, but because they are interested in other things and don't pay attention to politics until they have to.  See here
The mere fact that you're reading this article right now suggests that you not only think politics is important, but you actually like it. You read the paper and listen to political radio and talk about politics at parties. In other words, you view politics the way a lot of people view cooking or sports or opera: as a hobby. Most undecided voters, by contrast, seem to view politics the way I view laundry. While I understand that to be a functioning member of society I have to do my laundry, and I always eventually get it done, I'll never do it before every last piece of clean clothing is dirty, as I find the entire business to be a chore. A significant number of undecided voters, I think, view politics in exactly this way: as a chore, a duty, something that must be done but is altogether unpleasant, and therefore something best put off for as long as possible.

You go into an election with the electorate you have, not the electorate you want.

John McCain--He's not who you think he is.

[ Parent ]
Ohio (0.00 / 0)
Just proved you wrong.  When the situation is dire enough, wiping out entire families economic future, you can bet they start paying attention to policy.  

NoSlaves.com  


The Economic Populist
[ Parent ]
I have no problem with other politicians (0.00 / 0)
working the process to try and get candidates to adopt their policy positions in exchange for support, but that still doesn't do anything to address the problem of having superdelegates at all.  There's really very little reason (if any) for them.  If a candidate desires the support of politician X and gains it by promoting policy Y, then great - but exchanging superdelegate support for policy simply reinforces the undemocratic nature of the system.
[ Parent ]
One other thing (0.00 / 0)
your argument is predicated upon two suspect notions: first, that all superdelegates would exchange their support for progressive policies, and second, that all candidates aren't interested in accepting non-progressive policies.  My first thought is something like CAFE standards - some Democrats and labor leaders aren't interested in raising them, and may demand such, even though that's certainly the progressive policy.
[ Parent ]
And what about (0.00 / 0)
those of us who care about more than just trade? :)
[ Parent ]
uninspired and disheartened (4.00 / 1)
I'm trying to convince myself it's worth going to the polls after work.  While I was never unrealistic about my expectations from Obama, I was taken aback when word came that he wants Republicans in his White House.  Second term maybe, but right away?  Hell.  No.  Especially not for higher positions like SecDef, as he apparently suggested.  

If this is the "change" Obama speaks of, I don't see capitulation to the GOP as anything other than a continuation of the past 7 years.  Was it just a trial balloon?  If it was, would he have actually considered following through on that wish?  This is an alarming sign to me that Obama may not understand the underlying reasons behind the public's desire for "change".  What to do, what to do.....

Oh by all means stay home. (4.00 / 1)
I mean gee whiz--why bother in November either?  More wars will be so much fun, and besides, we can all watch accelerated climate change with another conservative Admin.  Really exciting, dude.

John McCain--He's not who you think he is.
[ Parent ]
good job on the measured response (4.00 / 1)
I especially liked the part where you assumed that my lack of inspiration in the primary race translates into not caring who wins the general election.  It's like you read my mind!
[ Parent ]
Is Hagel a typical GOPer on defense or is it the (0.00 / 0)
illusion of cooperation with the GOP (who are still part of our country and who can run circles around the Dems in Congress unfortunately). Seems more illusion than capitulation.  
[ Parent ]
you may be right (0.00 / 0)
And I certainly hope you are, but the fact remains that Hagel votes with the rest of his GOP bloc on torture, wiretapping, and any number of other issues.  I would be much less concerned if he had shown some willingness to break with the pack on the most serious issues.
[ Parent ]
Probably a trial balloon... (0.00 / 0)
But even so, he'd still be serving under a Democratic administration.  Obviously there's no way to know, but you would hope that that would at least keep him in line with the broader Administration goals (or risk getting fired, essentially).
[ Parent ]
Oh PLEASE..... (4.00 / 1)
If you don't think getting of your butt and voting to end the War is worth it, I suggest you attend your local Ralph Nader meeting where you can find others who repeat the "there no difference between the parties" nonsense.

Hagel is being floated as Sec Def for one reason: he has been vocal in his opposition to the War.  

[ Parent ]
can you read, or do you choose not to? (0.00 / 0)
How is voting in the Texas Democratic primary going to end the war?  Where did I say there is no difference between the parties?  I expressed dismay that Obama is showing warning signs that he may not really be all that progressive.  How that relates to what you wrote, I have no idea.
[ Parent ]
Hagel and Lugar (0.00 / 0)
I agree with the idea that appointing Republicans to top cabinet positions may serve to emphasis the point that Republicans like to make, that no Democrat is qualified to run the defense department.  And certainly, I would be very disappointed if Obama appointed a Republican to both State and Defense.  However, I have generally agreed with a lot of Hagel and Lugar's positions on foreign policy.  They would not have any real domestic influence, and presumably Obama could use them as a tool to get Republicans in Congress to support his foreign policy decisions.  So, for me, the test is whether Hagel and Lugar are the means to achieving a better foreign policy, or if they are the ends (if they are there only so Obama can say that he is a bipartisan president).  Personally, I think the answer leans more towards the former, so I hope that you won't take this as a cause not to vote for him, but to be fair, we simply don't know the answer right now.
[ Parent ]
great points (0.00 / 0)
Thanks for your input, that's exactly what I was hoping to get - different perspectives.  It's that uncertainty you speak of that makes me hesitant, but there's always some uncertainty with political candidates.  I guess it just rubs me the wrong way that the GOP has done so much damage to the country and yet some of them might be welcomed into his administration.  I suppose I'll have to get over it and hope for the best.
[ Parent ]
this is another prove that caucuses are anti-democratic (0.00 / 0)
Even Hillary will win TX, she will get less delegates in TX?
this is absolutely ridiculous...  
They're not perfect, and should be studied for the future (0.00 / 0)
but even less perfect are the points the Clinton campaign has made on superdelegates (which are much more undemocratic).  No one complained about how undemocratic the caucuses were before Hillary was losing them.  
[ Parent ]
well: nobody had a reason i guess (0.00 / 0)
now we have plenty of reasons and proves, so it is not a complain it is THE observation:
if you want a legitimate nominee, than remove caucuses, use only close primaries, replace delegates with direct vote (one voter - one vote, no friken delegates, super or not) and make primary as one nation-wide day. Otherwise you always will have probability of a mess and division like we have now plus GOPs and independents will dictate who is Democratic Nominee.
[ Parent ]
Should Clinton win three of four (0.00 / 0)
It will simply be up to Obama to face all the added media heat from the Chicago trial and the Clinton spin that he can't win the big states.

If he's a leader, he will shrug today off and continue to lead. If Clinton is a superior leader, she'll move past her own empty rhetoric (about the delegate counts) and start making serious inroads to move back to a leading position.

Supporters of both can gnash & lash, but that's the reality. If Obama doesn't win TX today, the campaign continues to May or June. McCain can have carte blanche taking potshots at both, while the economy craters.

And even then, one of this pair will emerge as the leader: the best leader will prove they deserve the nomination. So if you're confident that describes your man/woman, there's little point in complaining. Show some faith in your candidate to prevail.

Hillary is in leading position (0.00 / 0)
All Obama has is:
1) false momentum form anti-democratic caucuses and open primaries where GOPs and independents interfering with nomination of Democratic nomination.
2) Racially-motivated black vote
3) sexism of male voters
4) kids brainwashed by hopeshit.

There is no any substance for Obama's lead, except one speech in 2002, when he was responsible for nothing.

Hillary is only Democrat standing and she is leading among DEMOCRATIC VOTERS.

[ Parent ]
Final March 4th Polling Outlook | 46 comments





Donate to Open Left




blog advertising is good for you
blog advertising is good for you
LIBERAL INSTITUTIONS
STATE BLOGS
USER MENU

QUICK HITS
SEARCH

   

Advanced Search

Home | RSS/XML Feed | Register | Contact

  © OpenLeft 2007-2010, Powered by SoapBlox