Edwards As VP-An Indication of Progressive Strength

by: Paul Rosenberg

Sat Jun 28, 2008 at 14:15



Obama's FISA betrayal this week puts a different spin on the strength that Edwards has shown as a VP candidate--it contrasts sharply with Edwards' performance as a progressive.  Last week, I wrote about Edwards' strength as a case of "brand recognition" rather than simply "name recognition."   This week provides more evidence of this, contrasting him with local politicians in three states--Ohio, New Mexico and Missouri--and taking another look back at Iowa.  Now that we have data from two sets of candidates in nine states, Iowa provides evidence that Edwards did significantly better there than his already strong showings elsewhere--a strong indication that (a) the name recognition argument is oversold, and (b) further exposure to Edwards' brand has the potential to increase his already considerable strength even further.

However, I don't just want to dwell on the fact of Edwards' brand, I also want to stress the nature of that brand, which is that of a progressive populist.  

This is often missed by those who look at his support and see a lot of self-described "conservative Democrats."  But, of course, this is partly because of how people have come to identify as such.  In fact, many so-called "conservative Democrats" are more economically liberal than so-called "liberal Democrats".  This is part of the larger overall problem of trying to characterize the electorate in terms of how people describe themselves, as opposed to what they say they want, believe and care about.

I'll be writing a more general diary about that later today, but suffice it to say that as far back as 1967 it was established that a majority of self-described conservatives are either liberal or moderate on social spending issues--the heart of New Deal liberalism.  This is one of the most firmly established and significant facts in public opinion research, and yet it is routinely ignored by the punditalkcrazy, because it would totally contradict their cherished narrative of a center-right nation.

For now, however, I want to catch up on the 5 new VP polls that Survey USA released this week--Oregon, Washington, New Mexico, Missouri and Ohio--do a little comparative analysis to further consolidate the care for Edwards' unmatched strength, and extend the argument that this strength is evidence of an economic liberalism--aka "progressive populism"--that is deeply disturbing to the establishment.

This is, of course, directly in line with David Sirota's argument in The Uprising.  And it also suggests that Obama is not likely to pick Edwards, despite the "pragmatist" narrative about him, which, of course, also took a hit from his FISA stance.
Paul Rosenberg :: Edwards As VP-An Indication of Progressive Strength
Four States In Quick Review

First off, here are redone versions of four tables from states SUSA polled the previous week.  These are reodered to show the strength of candidates relative to the straight ticket.  As can be seen, Edwards dominates across the boards.

Wisconsin

Edwards is the only candidate in either round who helps Obama:  He moves it from "lean Obama" to "solid Obama" against everyone but Lieberman, who is far too weak elsewhere to be a good fit for McCain:

California

California is far more friendly to Democrats, so even though Edwards was the only one who made California a "solid" Obama state in the first round, Obama does it himself in the second round, and several candidates help marginally.

Iowa

AS mentioned in the introduction--and will be seen in our comparisons below--Edwards' strength in Iowa goes well beyond his strength elsewhere.  It is a clear indication of this strength as an active candidate, who has had a significant presence in the state.  More on this below.

Minnesota

Again, Edwards alone pushes this state from "lean Obama" into "Solid Obama" territory, except against Governor Pawlenty, whom he still bests by improving Obama's lead by 2.  Given that Obama has weakened since then, that's all the more reason to take Edwards' strength here seriously.

Five New States

With the new states, we also include the comparison charts, showing the strength of VP candidates against all their matchups combined.

Missouri

Here we have a direct tests of Edwards vs. McCaskill, one of several names floated as possible VPs, and we find that Edwards significantly out-performs her.  Again, he is the only one to show gains against all comers, and he even moves it into "solid Obama" territory against one candidate--Tim Pawlenty.  McCaskill faced similarly weak candidates, but could only gain 3 points against one--and that with Obama's baseline down nine points from where it was when Edwards added another nine against Pawlenty.  There is simply no comparison between the strengths of these two candidates.



New Mexico

Similarly, though not as dramatically, Edwards shows that he is stronger in New Mexico than home-state icon Bill Richardson.  Richardson is about as strong a homestate candidate as one could hope to find, and yet he doesn't help Obama as much, even against a significantly weaker GOP VP field:



Ohio

It's three in a row as Edwards out-performs Ted Strickland handily.  Strickland did not help Obama against a much weaker field, and when Obama's lead had shrunk.  OTOH, Edwards took Obama from "lead Obama" to "solid Obama."  In the more weakened environemnt of the second poll, we could expect him to move the state from "tossup" to "lean Obama"--evan against the tougher GOP VP field he faced.



Oregon

After three dramatic states in a row, Oregon is more of a ho-hum, Edwards-is-the-best-again kind of state.  But it's at least worth noting how Edwards helped Obama at a higher level against a stronger field.



Washington

Finally, in no-contest Washington, the best a Dem VP candidate could do was add two points for Obama in round 2, while in the earlier, closer poll in May, Edwards again moved the state from "lean Obama" to "Solid Obama" against 3 out of 4 candidates.



The Big Picture--Nine-State Overview

Putting all these results together, we clearly see that the second candidates help more than the first round candidates--not that surprising, given the relative weakness of the field they face.  But Edwards is the clear exception to this pattern.

(The figures were derived by averaging Obama's prefromance between the two polls, and comparing each VP's showing to Obama's showing straight-up in the same poll.  Thus, in Iowa, Edwards gets credit for increasing Obama's lead 9.5% on avegage, but only compared to his 2-poll 6.5% average lead, not compared to the 9% lead Obama actually had in the poll that Edwards was tested in.)

Edwards' strength in Iowa is clearly visible here as dramatically more than his strength elsewhere.  It's almost five full points higher than the 4.6% average bump he provides in the other eight states polled.  This is the difference in performance he shows from having campaigned extensively in Iowa last year.  It shows that strong as he may be elsewhere, there is every reason to expect him to add even more, if he were to become the nominee.


In sum, what this shows, even more decisively than earlier analyses, is that Edwards provides a solid, proven boost to Obama in all states, that his boost is stronger than that provided by even the strongest of favorite son/daughter candidates, and that he would add even more strength over time as he campaigned heavily in key states where he was needed most.

There is, quite frankly, nothing more one could want from a VP candidate than that.  As for his politics, this past week can only make us increasingly wary of those who waffle--which is to say, most of the alternatives that have been presented.  While Edwards past record is certainly muddled, the nature of his campaign last year was not--he was the one pushing Obama and Clinton to be more progressive and more explicit across a range of major issues.

There is no doubt that Russ Feingold has a stronger lifetime record in this respect, of course, and I've said repeatedly that I like Feingold very much, and wish he had run for President.  But at present, we simply have no idea how much Feingold would help Obama, and thus how much clout he would have even if Obama were to put him on the ticket--a rather remote possibility at the moment, it would seem.

Thus, it is clearer than ever that an Edwards VP candidacy would be great for the Democratic Party ticket--and that we're unlikely to see it, precisely because it would be so good.  They don't want to owe progresives anything--and with Edwards on the ticket, they obviously would.
Tags: , , , , (All Tags)
Print Friendly View Send As Email
Edwards As VP-An Indication of Progressive Strength | 51 comments
But.. (0.00 / 0)
but you have no personal investment as picking Edwards as VP. No, none at all.  
And you have none at all in not having Edwards (4.00 / 3)
Just so others who happen by this diary -- your claimed as of last week was that you have no vested interest either way. but, every diary on the subject finds you posting disparaging remarks. You are also quite clearly someone not very interest in progressives despite what you say.
[ Parent ]
No, I DON'T Have A Personal Investment (4.00 / 2)
For one thing, I have no idea if it would be a good thing for progressives for Edwards to accept.  He might well be totally frustrated in such a role.

My purposes here is to lay out the logic to the best of my abilities--and to clearly indicate the foundations of that logic.  What people do with that is up to them.

Believe me, if I was convinced that because of all the above then Edwards should be the VP, and that nothing else could advance the progressive agenda as well as that, then I would say so, flat out.

But I don't.  What I believe, right now, is that understanding is what best advances the progressive agenda.

The more we understand what's going on, what the possibilities are, what the alternative entail, etc., the better able we will be to act effectively together.

And, of course, so long as I can use this approach to shed some light on progressive possibilities, I'm likely to continue doing so.

"Senate passes expanded GI bill despite Bush, McCain opposition"

[ Parent ]
Arrrrggggg (0.00 / 0)
Edwards is my favorite for VP. My comment is about something else.

Obama's FISA betrayal this week...

I'm extremely disappointed in Obama's cave in on FISA. I'm angry about it, but is every single ^%$#*@&^% post on every single ^%$#*@&^% progressive blog going to start off with

"Now that Obama has stabbed us in the back, here's my thoughts on global warming..."

or

"Today Senator Obama (who betrayed us like Brutus, remember? I told you so. Nya nya nya nya nya!) released his statement regarding Bush's policy on..."

I Don't Know... (0.00 / 0)
[ Parent ]
He won't pick Edwards or any progressive (4.00 / 3)
a) There is no political incentive to do so. Obama's support is mostly of the blind variety, that we are "tricking" the American public variety,  or the Mike Lux (I am smarter than you about politics) variety. There are other variants- none of them about accountability.

b) There is no incentive in the media

c) Most of the A list bloggers are centrist by another name.  Most, let's face it, are pretty much of the same background as the political insiders they claim are the problem or they want to be.

d) The power center in the party will be dominated by the conservatives because  many confuse Democrat with progressive. Rather than understanding the various axis at play (ie, conservative policies versus liberal politics)- they think there is only one. As I mentioned else where- this allows for the conservatives to hijack the party because the progressives are wasty resources on both conservatives and liberals. This is especially true as the GOP loses power. These people aren't just going to go away. THey are going to shift their power to a Corporate Democrat. So long as they have their thumb on the portion of the party that's in charge of the rest of us, then they don't have to worry about outcomes being different since as FISA shows- it won't be different. In fact, it can be worse- because no matter how extreme it can be labeled as "liberal." A little part of the chess came that's centrism.

e) The party mind set is still dominated by reactions to Reaganism. Most people are still dominated by that history.

f) The American people are asleep at the wheel, and reeling so much that they don't know how to deal with this. They are also deeply cynical about all of this. Jesus could come back and walk on water and they wouldn't believe it.

There are others, but this should give you a start.

Wes Clark (4.00 / 1)
Given how cowed he now seems to be to national security issues, I'm starting to expect that he will pick Wes Clark.
 Which wouldn't be such a bad thing for progressives, as Matt has argued.

John McCain is against children's health insurance, says CNN.
[ Parent ]
I expected him to choose someone who will not (0.00 / 0)
frighten the corporate interests since that seems to be the primary motivating force.
[ Parent ]
Clark would be OK (0.00 / 0)
Edwards is a better because I think he might help more in November but Clark is still a solid progressive, so that's a good thing.  
[ Parent ]
I would be fine with Clark, but I also think that he won't choose (0.00 / 0)
CLark for the same reason he won't choose Edwards. I am pretty sure at this point that if this is politics as usual that means centrism which is interested in corporate Democrats, not those who expouse returning fairness to the system.
[ Parent ]
But he is looking at Edwards & Clark (0.00 / 0)
And I think he's serious about both.  The "leak" by the CBC regarding Edwards and Nunn was about testing the response.  And Edwards backtracked from his definitive no to a "I'll consider it" for a reason. I don't know what Obama will do but I think he's got an open mind about this.  
[ Parent ]
We shall see (4.00 / 1)
I am as Rosenberg will tell you very cynical and skeptical about politicians.  
[ Parent ]
Uh (4.00 / 1)
Obama's FISA betrayal this week puts a different spin on the strength that Edwards has shown as a VP candidate--it contrasts sharply with Edwards' performance as a progressive.

Now when you say "performance", do you mean performance, or rhetoric?

Come to think of it-- what do you mean at all? The worst we can say about Obama on his FISA issue is that he was with us during the primaries, but now that the issue is truly coming to a head during the general election he is nowhere to be found; since Obama opposes amnesty and will be voting against it whenever the Senators get their figleaf vote, his "betrayal" on amnesty consists only of not actively doing anything to stop it. Which, given that the active pressure of the nominee would be almost certainly enough to determine this bill's fate, is pretty damn bad.

But the same is true of Edwards: he was with us during the primaries, but now that we're in this post-primary malaise is doing nothing active to campaign against amnesty himself-- in fact correct me if I'm wrong but he hasn't even issued a toothless "I oppose the amnesty provisions of this bill" as Obama did. Given, since he is not a Senator and no longer has either the influence Obama has as nominee nor any particular access to the national microphone, there's absolutely no reason we would expect Edwards to be campaigning against or issuing statements on FISA at this point.

But still-- I do not see a lot of reason to take seriously the idea he takes this issue any more seriously than Obama. If Edwards was the nominee right now rather than Obama, how would Edwards act on FISA? The way that Edwards and Obama acted on FISA during the primaries, when they needed our attention and votes? Or the way that Obama is acting now, and Edwards acted before 2005?

John Edwards has a 50% lifetime rating from the ACLU, the lowest of any top-tier Democratic presidential candidate... During the 107th and 108th Congress, Edwards voted in alignment with the ACLU position 29% of the time, voted against the ACLU position 29% of the time, and was absent 42% of the time. This gave him a 50/50 rating from the ACLU, which is not particularly impressive for a Democratic candidate. On the other hand, it's worth bearing context in mind: Edwards served in the U.S. Senate from January 1999 to January 2005, during which most of the Bush administration's problematic counterterrorism bills were supported by the majority of senators of both parties. Edwards' biggest problem is that he simply wasn't present for six of the nine votes in 2004 on which the ACLU took a position. If he had been, his rating would probably be substantially higher.

What You're Doing Here (4.00 / 2)
is recapitulating the reasons that I and many other progressives on the blogs leaned toward Edwards for many of the things he said during the campaign, but remained largely uncommitted until very late in the game.  It was only after a long time that many folks became reasonably convinced that he really had changed in significant ways, and would make a better candidate and a better President.

But, rather obviously, given how late a significant chunk of the blogosphere came to support him, there is significant room to disagree.

However, the sort of analysis I'm doing here is very much dependent on broad public perceptions.  If I were passionately advocating for Edwards on the grounds of "who's the best", then that would be something quite different.  But I'm not.

"Senate passes expanded GI bill despite Bush, McCain opposition"

[ Parent ]
I agree: Some perspective is needed (0.00 / 0)
when it comes to comparing Obama to Edwards and others who are supposedly more "progressive."

What evidence is there that Edwards would have taken a different tone or approach had he prevailed and become the nominee?  Who is the last nominee of either party who did not make an attempt to "move to the center" for the general election?

[ Parent ]
Oh, Jeez (4.00 / 5)
I could point to some of Edwards's better moments on civil liberties, his opposition to the flag burning amendment, for instance--he was one of only 2 redstate senators to oppose it--but I don't think we need a rehash of the do-we-trust Edwards argument.

Leaving aside O's and E's deepest feelings-which we can only guess at--it should've been clear that Edwards was eager to run as a progressive--a populist progressive, no less--and to identify with the progressive movement. Indeed, I supported Edwards not because I thought he was a better progressive in any pure sense but because he believed that the path to the White House ran relatively left and that he believed in partisanship. While it's probably true that he would've been more likely to run to the middle on an issue like FISA, as opposed to trade and other economic issues, he staked his campaign on combativeness, on fighting both the GOP and corporate power. What's more, maybe because he's a white man with a Southern accent who strikes some people as a cultural conservative, he would've felt very little need to cover his right flank. All adds up to the right position on FISA. God fucking knows he wouldn't have hired a Wall-Mart fan to run his economic policy shop.

In short, when trying to gauge how a pol with run. don't try to look into a his heart, try to see the landscape from his perspective.

Incidentally, as someone who about 30 blog posts urging the sphere to get behind Edwards, I view this criticism of Obama with a mixture of vindication, annoyance, and sadness.

[ Parent ]
Except That (0.00 / 0)
I never trusted Obama after he aligned himself with Lieberman. So this is not really new for me.  It's more re-renewed.

But I agree 100% with your line of argument.

"Senate passes expanded GI bill despite Bush, McCain opposition"

[ Parent ]
I must say, (4.00 / 2)
seeing endless charts with, in every one of them, the highlighted line of the Obama-McCain baseline, and every single Edwards matchup below that line, and every single other matchup above that line, does have an effect. I have no idea what it is that Edwards has or how he has gained it, but it seems pretty significant. I wish they would also test some other candidates who are widely and well known outside of a state or two. I'm not sure who -- Al Gore, Hillary Clinton, John Kerry, Howard Dean?* (I'd be particularly interested in Gore's numbers, even if it's unlikely he would accept). Edwards has definitely achieved something significant; what I'd like to know is how significant, and how much hope any of the other candidates would have of matching it once they becomes known at a national level. I do think that anyone who is outperformed by Edwards in their own state is essentially out of contention: this means Rendell, Strickland, Richardson, and McCaskill.

Also, it's striking (and troubling) that in nearly every state, adding a little-known VP candidate next to both Obama and McCain almost always causes Obama to lose ground and McCain to gain it. Anyone have an idea why this might be?

Anyways, I'm still wishing for Feingold; but Obama's FISA stand pretty surely takes that off the table, if it was ever on it.

* This does raise the question: is there any way to gain wide national attention besides running for President? (Maybe leading a chamber of Congress; but neither Pelosi nor Reid seem like good choices).

A conjecture: (4.00 / 4)
Also, it's striking (and troubling) that in nearly every state, adding a little-known VP candidate next to both Obama and McCain almost always causes Obama to lose ground and McCain to gain it. Anyone have an idea why this might be?

I have a conjecture:  The Republican primary ended long ago, and the wounds have pretty much healed from it.  The Democratic primary only ended this month, and was a fiercer and bloodier battle than the Republicans had.  

Clinton supporters from the primary who are still holding out hope for an Obama/Clinton ticket are expressing their disappointment to pollsters by giving their thumbs down to other match-ups.  Edwards supporters from the primary, meanwhile, were given the opportunity to thumbs-up their preferred candidate.

[ Parent ]
That's a good theory. (4.00 / 1)
Quite plausible. Another reason why it would be to interesting see Hillary's name in the list (IIRC she was tested in a couple of states? but I can't remember what the results were).

I wonder where all those Edwards supporters who supported Obama in the primary but are defecting to the Republicans without him on the ticket would be coming from, though, so that part is not as plausible to me.

[ Parent ]
Lord, SUSA again? (4.00 / 2)
In that case, I'll repeat my objections to the way you're using this polling:  

1: SUSA is (with few exceptions) polling Edwards (who has run two national campaigns, and who has visited the states they're polling as a candidate) versus governors and senators who have not taken their message to the people of the states being polled.  This isn't just "name recognition"; it's message and personal recognition as well.  

For example, in Wisconsin, he's measured against a slew of people who've never visited Wisconsin in a political capacity, and who have never addressed Wisconsinites with their message.  Of course, Wisconsin is likely to prefer John Edwards to a bunch of non-entities.  

If SUSA had compared apples to apples, your interpretation of their results would be more credible.

2. You've discussed three of the exceptions that SUSA has polled in this diary:  Obama/Strickland in Ohio, Obama/McCaskill in Missouri, and Obama/Richardson in New Mexico.  Obama/Edwards outperforms Obama/Strickland; you and I both agree there.

You also claim that Obama/Edwards outperforms Obama/Richardson.  I dispute that claim:  

Obama/Edwards was polled in New Mexico on May 19; Obama/Richardson on June 24.  (You call these "Round 1" and "Round 2".)  In "Round 1", Obama/Edwards outpolls the generic match-up by 3-9% depending on John McCain's running mate; in "Round 2", Obama/Richardson outpolls the generic match-up by only 3-5%.

However, the partisan breakdown of the two polls were different.  In "Round 1", 52% of the sample were Democrats, and in "Round 2", only 48% of the sample were Democrats.  That 4% difference gobbles up the apparent 4% advantage that Edwards has over Richardson.  (Incidentally, so does the margin of error.)

(I'm encouraged that even with a less-Democratic sample, the generic Obama/McCain matchup in Round 2 outperforms the matchup in Round 1; that suggests there's some definite movement in that state.)

A similar situation occurs in Missouri, except that this time, the Democratic sample size remained the same in both surveys, but there's a 5% difference in the Republican/Independent balance.  

You neglected to mention the difference in samples chosen, and you also neglected to link to the actual polls so that readers could more easily verify this difference for themselves.  

3. In May, SUSA polled Kansas, asking about the Obama/Edwards and Obama/Sebelius matchups to the same sample.  In this poll, Edwards and Sebelius each "won" two of the matchups, leaving neither with an advantage over the other.  

4. If in May, I ask a class of junior high students who their favorite musician was, and then in June, I ask the same question to a group of senior citizens, I could claim a "Frank Sinatra bump", and anyone would see how absurd this claim was.

You've closed with a table, that I infer you've generated yourself, of an "Edwards Bump".  You've calculated this bump by taking a figure generated from the Round 1 data, and subtracting it from a figure generated from the Round 2 data.  Since these two sets of data came from different populations, and SUSA did not keep the party affilation constant between rounds, your table is flawed.  Not to the same degree as the "Frank Sinatra bump", but flawed in the same way.

Asked And Answered, If You'd Bothered To READ What I Wrote (4.00 / 1)
SUSA's polling is disappointing, to say the least in its selection.  However there are enough examples of comparing Edwards to people who have campaigned in their various home states, including some who might be called institutions.  He wipes the floor with virtually all of them.

A noted exception being Sebelius, with whom he is merely tied, in a state that even they seem unlikely to turn blue.

So, he beats Richardson in New Mexico, Rendell in Pennsylvania, McCaskill in Missouri, and Strickland in Ohio.  Soundly.

You also claim that Obama/Edwards outperforms Obama/Richardson.  I dispute that claim:  

Obama/Edwards was polled in New Mexico on May 19; Obama/Richardson on June 24.  (You call these "Round 1" and "Round 2".)  In "Round 1", Obama/Edwards outpolls the generic match-up by 3-9% depending on John McCain's running mate; in "Round 2", Obama/Richardson outpolls the generic match-up by only 3-5%.

However, the partisan breakdown of the two polls were different.  In "Round 1", 52% of the sample were Democrats, and in "Round 2", only 48% of the sample were Democrats.  That 4% difference gobbles up the apparent 4% advantage that Edwards has over Richardson.

The last sentence (in bold) is so egreggiously wrong that I find it hard to believe you're a math teacher.  From a purely mathematical POV, it could only hold true if both men had 100% support among Democrats and 0% support among independents and Republicans.

Then there's the pollster's argument--that partisan ID is fluid, and generally should not be controlled for unless it is obviously out of whack, in which case, the poll itself is probably best regarded as an outlier and left at that.

3. In May, SUSA polled Kansas, asking about the Obama/Edwards and Obama/Sebelius matchups to the same sample.  In this poll, Edwards and Sebelius each "won" two of the matchups, leaving neither with an advantage over the other.
 

As I already noted above.  But Kansas is really not one of those states that Edwards has campaigned in much, had TV ads, etc.  So the fact that he's even with the most popular Democrat in the state in living memory hardly diminishes him.  In fact, the Iowa figures strongly suggest that if he had campaigned there as much as Sebelius had, he would have bested her by about 5 points.

You've closed with a table, that I infer you've generated yourself, of an "Edwards Bump".  You've calculated this bump by taking a figure generated from the Round 1 data, and subtracting it from a figure generated from the Round 2 data.

No, that's not what I've done.  Your reading and math skills are, apparently, equally flawed.



"Senate passes expanded GI bill despite Bush, McCain opposition"

[ Parent ]
Oh, lord, you're right. It's worse. (4.00 / 1)
You've closed with a table, that I infer you've generated yourself, of an "Edwards Bump".  You've calculated this bump by taking a figure generated from the Round 1 data, and subtracting it from a figure generated from the Round 2 data.

No, that's not what I've done.

Oh, lord, you're right.  What you've done is even worse.  For the benefit of other readers, let me explain your calculations, since you haven't fully done so:  

Column 1 is the state.  

Column 2 is labeled "Base".  The "Base" appears to be generated by taking the two running-mate-less matchups, one in Round 1, and one in Round 2, and averaging the results together.

For example, in Wisconsin, you took the figures of 6% and 9% for the two matchups, and averaged them to get a figure of +7.5 for Barack Obama.  This figure uses "Round 2" data.

In doing so:

(a) You have averaged figures obtained from two different surveys with different party breakdowns.  
(b) You have averaged a "registered voter" survey with a "likely voter" survey.  
(c) You have weighted each round 50/50, even though the number of people surveyed differed in each round.

(Washington state appears to be an exception to this rule; but could be explained if you inadvertently averaged 16 and 15 instead of 6 and 15.)

Column 7 is labeled "Edwards Bump"

This figure is taken exclusively from Round 1 data, by averaging the four "v. Head-to-Head" margins involving John Edwards.  

Still using Wisconsin as an example, you averaged (2 + 4 + 6 + 9)/4, the last four entries in the last column of the Round 1 table for that state.  The average, 5.25%, you rounded to 5.3 in the table.  

Column 6 is labeled "w/Edwards", and is the sum of the entries in Columns 2 and 7.  

So you've taken data that was composited from two surveys, and added to it data that was composited from one survey to generate that nice blue line that appears at the top of your graph that makes Edwards look so attractive.  

Yeah, that's real valid.   Thanks for the opportunity to correct point 4.

As for point 2, you write:

The last sentence (in bold) is so egreggiously wrong that I find it hard to believe you're a math teacher.  From a purely mathematical POV, it could only hold true if both men had 100% support among Democrats and 0% support among independents and Republicans.

Obviously false.  

That's one way it could hold true.  A different way it could hold true is that each man has support of a large enough number of Democrats that the remaining difference could be explained by rounding.  

Here's an example:  

In question 5 of poll 14036, 539 voters were asked whether they supported Obama/Richardson over McCain/Bloomberg, and SUSA reports that 51%/44% supported Obama/Richardson.  Since SUSA rounds to the nearest whole number, we can at best conclude that somewhere between 273 and 277 respondents said they supported Obama/Richardson over McCain/Bloomberg.  

SUSA then gives a party breakdown as follows:  48% of respondents (between 257 and 261) were Democrats, and 75% of them (between 191 and 197) preferred Obama/Richardson.  The uncertainty in the numbers is due to the fact that 48%, for example, represents a number in the range 47.5 <= x < 48.5.

If we alter the party breakdown to match the breakdown in Round 1, but we keep constant the other percentages (including the figure that 75%, not 100%, of them preferred Obama/Richardson), we discover that as many as 288 of the 539 respondents would have supported Obama/Richardson, changing the 51/44 7-point advantage to a 53/42 11-point advantage.  

In other words, the 4% shift in party affiliation can gobble up the 4% difference between Edwards and Richardson, even without the additional assumption that 100% of Democrats supported both candidates.

[ Parent ]
I'm Making The Best Use of Available Data (0.00 / 0)
And believe it or not, folks all over the intertubes--some of them even perfessors, like Charles Franklin of Political Arithmetik and Pollster.com--also average disparate polls together.  Some of them even (gasp!) by different pollsters! (Katie bar the door!)

You, however, prefer pure speculation:

Clinton supporters from the primary who are still holding out hope for an Obama/Clinton ticket are expressing their disappointment to pollsters by giving their thumbs down to other match-ups.  Edwards supporters from the primary, meanwhile, were given the opportunity to thumbs-up their preferred candidate.

So, are they getting their instructions via tinfoil hats?  Tooth fillings?  Lights in the sky?  Or MySpace?  And if the answer is MySpace, then why didn't those campaigns make such sophisticated use of it during the campaign?

Inquiring minds want to know!

"Senate passes expanded GI bill despite Bush, McCain opposition"

[ Parent ]