MyDD :: The math: Obama has the better GE viability.

The math: Obama has the better GE viability.

by TCQuad, Thu May 29, 2008 at 08:17:30 PM EST

I, the great Quadrino, shall now look into the crystal ball and... OK, I'm only marginally adequate with basic writing and really bad at sarcastically dramatic writing. But I'd like to talk about the current polling with regards to the general election.

As I mentioned previously, Kerry had a substantial electoral college lead, almost identical to Hillary's current lead, on this date four years ago. Arguments have been made that Clinton's lead is significant. But substantial movement from voters led to that-which-shall-not-be-named in 2004. What would happen if the same movement occurred today? Is Hillary Swift-Boat safe? What about Obama?

The math is based on three assumptions:

The lead Kerry had on Bush moved based on a number of factors, however the likelihood of a voter switching their vote (the tenuous Democrats or Republicans) is constant. Assuming a similar campaign with a similar attacks, the movement would be similar.
There haven't been any significant demographic movements since the last election that would affect the first premise (that is, all the switchable voters didn't move from Texas to Oklahoma).
To allow for the different focus, campaign strategies and associated error, any differences less than three percentage points is considered to be within the margin of error (MoE) and up for grabs.

To justify the assumptions, (1) is the assumption that the math is based on, with the second part based on the premise that every candidate has potential lines of attack open against them, no matter how much they've been vetted or how electable we think they are (see: Swift Boats and a candidate that's been in the public eye for 20 years). (2) is arguable, especially the effect of Katrina, but it does not affect the math except to move Texas to MoE for Obama. Alabama, Mississippi and Louisiana stay deep red, Florida moves to the MoE for Hillary and stays red for Obama. (3) is to allow for the fact that all the data is in integers (no decimals) and for GOTV efforts in newly contested states that could make differences.

If we take the movement seen in the Kerry election and apply it to this election, what is the net result? Here's the spreadsheet output (and, for those who are allergic to spreadsheets, a very brief summary follows immediately thereafter):

The first three columns are the Kerry polling numbers (negative indicates Bush leads; positive indicates Kerry leads), the movement in the polls (negative indicate Bush gains, positive indicates Kerry gains) and the expected result. That's followed by Clinton's current and theoretical numbers as well as Obama's. All data was pulled from electoral-vote.com's archive for May 29th for all candidates.

For instance, Georgia polled Kerry -3 in May, but ended up -18. That's a -15 point difference for Kerry. Both 2008 candidates trailed currently, so with extrapolating, they really, really trail McCain in Georgia. If we apply this method to all states and count electoral votes as mentioned, you can project what the current leads would manifest as in the general election.

The summary: McCain leads Clinton 221-202 (115 in play) while Obama leads McCain 250-235 (53 undecided).

Clinton's lead is significantly more susceptible to Kerry-style Swift Boat ads or those states that she trails in are less likely to break Democratic. As such, Obama's currently the more insulated, safe candidate in the general election.

Personally, I think using polls this far out is ridiculous and not particularly useful. However, for those who would like to give weight to these numbers, remember to take into account previous trends between now and the election.

It's completely up to the individual how to use polls this far out. Either they're completely meaningless or they indicate vulnerability for Clinton. It's completely your call, as the independent reader, to decide which.

[editor's note, by TCQuad] D.C. and Delaware were alphabetically inverted at some point between now and 2004, so the numbers (and six electoral votes) between them were slightly different. It's been fixed.

Tags: Math, hocus pocus, statistics, electability, generous definition of the word proof (all tags)

Permalink :: 22 Comments :: Post a Comment

Is this snark? (none / 0)

Check the tags... Of course it's snark! I mean, I think. I don't know the exact definition...

But it was also fun!

by TCQuad on Thu May 29, 2008 at 08:19:07 PM EST

Re: Proof: Obama has the better GE math. (none / 0)

what do his GE coffers look like now? if he has lots of cash keeping up the state by state appearances will make a huge difference, but damn it will be a looooong 5 months

by zerosumgame on Thu May 29, 2008 at 08:19:30 PM EST

Obama has a much better chance at winning (2.00 / 1)

because he won the Democratic primary. Actually being the presumptive nominee and taking fire from the right while simultaneously getting hammered from the sour grapes in the party is likely to depress his numbers. But his odds are about 99.9% better than Hillary of winning the presidency because he's the Democratic nominee. That could come in handy if the Clinton folks would ever decide to act like Democrats.

by Sun Dog on Thu May 29, 2008 at 08:24:59 PM EST

Re: Obama has a much better chance at winning (none / 0)

Sour grape juice goes down smooth with Kool-Aide.

Mooseburgers? Careful Sarah. Moose bite back!

by spacemanspiff on Thu May 29, 2008 at 08:28:16 PM EST
[ Parent ]

Re: Obama has a much better chance at winning (none / 0)

Play nice! This diary was mostly playful, since I've seen this argument about electability being thrown around a lot and it's kind of silly, in my opinion.

But no need to resort to name-calling. When (or if, for the hard-liners) Hillary concedes, we will work together. Using phrases like "sour grapes" simply makes them less willing to come over before then and less enthusiastic to work with us when they do.

by TCQuad on Thu May 29, 2008 at 08:33:58 PM EST
[ Parent ]

calling names? (none / 0)

Do you mean sour grapes? Is being called a grape so painful? I like grapes.

Seriously though, why so touchy? Anyone serious about the issues isn't going to support John McCain because they're ticked off about the silly back and forth online.

And the fact is, people are currently working against the Democratic nominee because they refuse to believe he's the Democratic nominee. Or they got so used to trying to sell the idea that Obama is unelectable that now they're determined to help make it so.

At some point they're really going to have to knock it off and kissing butts isn't going to make that happen.

Probably Hillary will eventually say that it's time to rally around Obama and then that group she has managed to keep in her corner no matter what will settle down and go back to being Democrats first instead of Hillary supporers first.

Once the fight is really on with McCain, it's not going to take decent people long to figure out which side of that fight they're on and most of the Clinton folks are, of course, decent folks.

by Sun Dog on Thu May 29, 2008 at 08:49:22 PM EST
[ Parent ]

Re: Proof: Obama has the better GE math. (none / 0)

For him to either be tied or trailing McCain while Democrats are winning landslide margins in the Senate & House shows how weak Obama is.

Obama's only prayer here is if the country's economy gets so bad as we get to November.

Otherwise, we are in for a rough November.

The electoral map for Obama is worse than even John Kerry. Obama has no room for error.

He somehow losses either PA or MI ( or both), we are looking at landslide electoral defeats.

by libdemusa on Thu May 29, 2008 at 08:32:56 PM EST

Re: Proof: Obama has the better GE math. (none / 0)

Obama is beating McCain according to every metric I've seen recently, both here on MyDD and on electoral-vote. But, as I've mentioned here and previously, polls this early aren't indicative of anything.

The fact he's in the middle of a primary (and people are resistant to back him over Hillary at the moment) further confuses everything. I'd expect him to get a decent bump once the mark is reached.

by TCQuad on Thu May 29, 2008 at 08:37:33 PM EST
[ Parent ]

Re: Proof: Obama has the better GE math. (none / 0)

McCain is strong right now, and Obama is much weaker than Kerry was at this point. That will change. The economy will likely worsen, gas prices will not go down, and as McCain starts hitting the news more, people will realize exactly how much he IS like Bush. A poll just today showed that people think McCain represents change. They will soon learn they are wrong. Obama's making him run on the strongly conservative platforms he started on, and it will cost him the election.

"Tell me about your work ethic." "Well, I don't think ethnics do no work. I mean, that's they problem, really." "Overt racial prejudice. Impressive."

by vcalzone on Thu May 29, 2008 at 08:53:21 PM EST
[ Parent ]

Re: Proof: Obama has the better GE math. (none / 0)

or that could be because Obama hasn't really had a chance to slam McCain, as he's still dealing with putting Hillary's dead-end campaign to bed while being hit from all sides.

It's all about McCain/Bush now...

by thereisnospoon on Thu May 29, 2008 at 09:49:57 PM EST
[ Parent ]

James Carville doesn't agree with you (none / 0)

http://tpmelectioncentral.talkingpointsm emo.com/2008/05/carville_on_obama_i_thin k_he_w.php

by jaywillie on Thu May 29, 2008 at 10:14:14 PM EST
[ Parent ]

Obama's Woes: A Tale of Three States (none / 0)

I have included the following article from Real Clear Politics in its entirety, as it concisely reflects my sentiments:
____

http://www.realclearpolitics.com/article s/2008/05/obamas_woes_a_tale_of_three_st .html

May 29, 2008
Obama's Woes: A Tale of Three States
By Richard Baehr

If you want evidence that the Democrats are taking a huge gamble by nominating Barack Obama as their Presidential candidate, you need look no further than the current state of the race in three Southern/border states.

In 1992 and 1996 Bill Clinton won Kentucky, West Virginia and Arkansas. In 2000 and 2004, George Bush won all three states. In the current Democratic Party nominating contest, Hillary Clinton won all three states by huge margins -- 30 points or more in each case. West Virginia (3%), and Kentucky (7%) have relatively small black populations. Arkansas is just over 15% African American (in the same range as Florida and Tennessee).

The three states have 19 Electoral College votes among them, almost as many as Ohio (20). In 2004, Bush won the Electoral College by 286-252. Had he lost Ohio, Kerry would have been elected. In 2008, Ohio will undoubtedly be a battleground again.

Were the Democratic nominee Hillary Clinton, the Democrats would be in very good shape even without Ohio. That is because current surveys show Hillary Clinton winning all three states by solid margins over John McCain. But John McCain trounces Barack Obama in the same three states by over 20% in each case. So with Clinton as the nominee, these states vote as they did when her husband was the nominee. When Barack Obama is the Democratic nominee, these states vote as they did when George Bush was running. The differences in the poll results are shocking. Clinton wins Arkansas and Kentucky by 14% and 9% respectively. McCain wins against Obama in the two states by 25% and 24% respectively. This means the shift from Obama to Clinton is a change of over 34% margin in one state, 38% in the other.

Roughly 40% of the voters who are for Clinton will not support Obama in these two states.

The poll results in these states suggest a few things:

(1) If Clinton were the nominee, she would have an opportunity for a broad based national victory. In addition to the three states above, Clinton appears to be the far stronger nominee in Ohio, Florida and Pennsylvania. While both Democratic candidates seem to have a good shot at winning Pennsylvania's 21 Electoral College votes (the Democrats have won the state 4 straight times), Clinton has a much better shot at winning Ohio (20) and Florida (27). Bill Clinton won Ohio twice and Florida in 1996. In other words, Hillary Clinton has the opportunity to create an Electoral College map that looks more like Bill Clinton's races in 1992 and 1996, when he won 32 and 31 states and 370 and 379 Electoral College votes, than the much closer Presidential races in 2000 and 2004.

(2) If Obama is the nominee, the 2008 race looks like it will be a squeaker again, with a few tight races in smaller states -- Nevada (5), New Mexico (5), Iowa (7), Colorado (9), New Hampshire (4) and Wisconsin (10) being decisive. There is far less margin for error with Obama than Clinton. If Obama is shut out of Florida (which seems likely) and loses Ohio, where he ran very poorly in the Democratic primary and trails in most surveys at this point, he will need to hold all the Kerry states from 2004 and pick up the needed Electoral College votes from the West and Iowa, and perhaps Virginia. (13).

But holding all the Kerry states is easier said than done -- with Michigan (18), New Hampshire and Wisconsin all at risk, with 32 Electoral College votes among them, and perhaps Pennsylvania too, another state in which Obama ran poorly despite spending well over ten million dollars on campaign ads for the primary. Some of the same voters who appear to be soundly rejecting Obama in West Virginia, Arkansas and Kentucky, also populate Pennsylvania and Ohio, though not in the same percentages.

At this point, Obama appears to be the all but certain nominee. This is despite Clinton winning the same number or perhaps slightly more total popular votes, and winning virtually all the contested primaries since February when the Reverend Wright story surfaced. Obama's wins since then have been in states with very heavy African-American voter percentages -- North Carolina and Mississippi, and in very liberal Oregon. With almost all super delegates now breaking for Obama, he could wind up with close to a 10% delegate margin, but be only even in the popular vote. This will result from a 5% elected delegate margin (built on low turnout caucus victories), and perhaps a 30% super delegate margin, if the remaining 200 or so uncommitted super delegates break as this group has since North Carolina (62-10 for Obama). A system built on proportional distribution of elected (pledged) delegates will have grossly expanded the popular vote margin to give one candidate a decisive victory among these delegates, which has been used to justify the shift to that candidate of super delegates.

Obama's caucus victories (he won every caucus except Nevada) were built on enthusiasm from activist left wing Democrats and young voters willing to participate in the several hour process. In the general election, the participation rate will be much higher and voting patterns should mirror much more closely the results in primary states, where Obama has not done nearly as well.

(3) The serious daily tracking polls, Rasmussen and Gallup, show John McCain with a 3-4 % national popular vote lead. This is a stunning result given the general weakness of the Republican Party this year, the low approval ratings for President Bush, the strong perception the county is headed in the wrong direction, the still unpopular Iraq war, the economic slowdown, gas and food prices, and what was supposed to be a consolidation among Democrats for Obama now that the media chorus has awarded him the nomination.

It is I think a reflection of the weakness of Obama as a candidate. Starting with the revelations about Reverend Wright, the Obama campaign has dropped from the semi-celestial status it enjoyed in the eyes of many. The gaffe-a-day express, the foot in mouth disease among Obama advisors, the glaringly weak posture on national security and foreign relations the candidate has put forth (and for which several times he has been forced to backtrack), all have damaged Obama's chances. Now he is a mere mortal -- except to the true believers, and they are not enough to put him over the top in a general election.

He could still win with the huge financial advantage he will undoubtedly have, but it will be close and hard fought, and he will have to be lucky to triumph. The Democrats in their "wisdom" will take a pass on what could have been a much easier road to victory with Clinton.

John McCain may not be the perfect Republican, but he may appear much safer, more experienced, and less of a risk than the untested junior Senator from Illinois for many voters.

Richard Baehr is the chief political correspondent of American Thinker.

by Liame on Thu May 29, 2008 at 08:57:54 PM EST

Re: Obama's Woes: A Tale of Three States (none / 0)

I'll be honest, the chief political correspondent of the conservative "American Thinker" (currently on their front page: True liberals must be appeasers) isn't exactly the source I'd run to for unbiased opinion. After all, this is a gentleman who, assessing a Clinton presidency when she was the front-runner claimed that it would be a close contest and now claims she would have been the easy victor while Obama is close. I'm guessing he's just attacking whoever's in the front on the Dem side.

However, since you brought it up and agree, I'd like to analyze it a bit.

I'm not sure where the polls will end up. I've said this twice today, because I think it's important to emphasize. We're too far out and there's too big of a deviation from before to later.

I think many Clinton supporters who cannot imagine not voting for Hillary will eventually end up voting for Obama, simply because McCain's position are so offensive to everything that they stand for. I don't think Clinton would create this broad-based coalition that's been talked about with 370 electoral votes. It would require taking every election that's currently +5% McCain or more favorable. No losing ground, all progress. Possible, but not something I'd bank on.

Currently, she holds NC, which didn't favor Bill, Gore or Kerry. Florida went for Bill the second time around, but not the first, and I think she's getting more leverage out of disenfranchisement-gate than we realize. Kentucky's been going progressively more red the last sixteen years. Nevada was a squeaker twice for Clinton in a good way, twice for Dems the other way. Those are the states Clinton needs to keep in her corner and that I think will trend away from her over time, even in the absence of a Swift Boat type attack. WI/MI are likely pick ups to offset FL, leaving her about 28 down from current, probably around the 290 range.

Obama, on the other hand, has a chance to build a different coalition. Maybe I'll write about this later, but you touch on it. Caucuses show us where the youth of our party is. If we motivate the youth, get them to turn out, involve them in the process and inspire them for the future, we can build a coalition not of just 50 states, but of 50 years. It's a slim chance, but a really intriguing one.

As for this year, I think he'll pick up IN (close, but the bonus of proximity to the West, the second district which went blue a couple of years ago and a general likeability he should pull it out), MI, WI (easily) for a bonus 38, making Ohio irrelevant (even though he's leading). He may get other bumps from the "hard-working white vote" when Clinton proxies start endorsing him and pulling some of the vote along their coattails.

I don't think Obama is weak as a candidate, I think he's mortal. He hasn't gotten the natural consolidation bump (as seen by the hard-liner attitudes here) and we haven't seen him able to completely turn and attack McCain. For that matter, I'd like him to turn and attack people use phrases like "gaffe-a-day". Some of these "gaffes" have been completely ridiculous, like people making a big deal over which concentration camp his uncle liberated ("oh, no, he wasn't accurate over something from before he was born..."). McCain will be painted with that brush, by everyone, when the general election campaign starts ("Sunni, Shiite, it doesn't matter to President Ackhmininidijad").

On the other side, I don't think Clinton is exceptionally strong as a candidate, she's benefitting from a very vocal and determined following that hasn't quit on her. She hasn't expanded her base since the third quarter of last year...

If she concedes and they come over to Obama's camp, that's a really powerful movement heading into the GE. Especially after the convention, when a lot of unity feelings start up, it'll be fun to watch!

by TCQuad on Thu May 29, 2008 at 09:48:50 PM EST
[ Parent ]

Re: Obama's Woes: A Tale of Three States (none / 0)

I think it is more plausible that there may be an even larger erosion of support from Obama in the swing counties of the traditional swing states, once he is the presumptive nominee and given that Clinton is not also on the ticket. It may be that, in these swing areas, Clinton's continued presence in the primaries and the continued potential of her to have a mitigating role to him, for voters that are not inclined to support Obama, has propped up his numbers somewhat and kept these same votes from looking elsewhere.

I also think that since super-Tuesday, Obama's polls have tended to show their upper limit earlier, and that late breakers will more likely favor McCain as the safe choice, as they have favored Clinton.

Fivethirtyeight ranks Ohio, Pennsylvania and Michigan as the top swing states, ≥10% for a state being pivotal, for Obama, and these are states that have favored Clinton and show certain ominous signs of the `swing counties of the traditional swing states' mentioned above. That same site also gives Obama a 38% chance of winning Virginia, a state upon which his campaign seems to have a lot riding. Colorado is the only highly probable new state that Fivethirtyeight predicts.

This information, with the backdrop that I believe he needs a mandate, a large win, to truly have a chance at change, gives me pause.

by Liame on Fri May 30, 2008 at 10:22:26 AM EST
[ Parent ]

Re: Obama's Woes: A Tale of Three States (none / 0)

I haven't heard of Fivethirtyeight before, but I'll check it out and run through the statistical paces with it when I get a chance.

I'm not sure about people supporting Obama because he's fighting. Like I said before, I don't expect many Clinton supporters to break hard right into McCain's camp. Even if you don't give Clinton supporters a natural propensity to vote Democratic and put them to to-be-determined camp, if their position is even partially motivated by the issues they'll break completely into Obama's camp. Every position Clinton trumpets is marginally different than Obama's, when compared to the difference with McCain.

As for the map, Obama's come up 11 points (to even now) in Virginia without really campaigning there. Kerry lost there because he didn't campaign, starting from double-digits down. Decent voter registration and GOTV efforts should swing it, though the portents of that may not manifest in the polls for a few months.

Ohio, Penn, Michigan are all likely Democratic independent of nominee. Ohio's got Republican corruption to play off of. Pennsylvania has a large reservoir of unregistered African American voters (RealClearPolitics put it at about 200,000). Michigan's depressed because of a primary rules battle, but is already showing greater favorability than a month ago.

It's a long, tiring process. In a couple of months, we'll have a better idea (assuming it's a one on one match-up then) of where everything could end up. The cynic in me says that no one can win a mandate-style landslide this year, though the optimist can see the paths it could happen along.

by TCQuad on Fri May 30, 2008 at 07:00:12 PM EST
[ Parent ]

Convenient choice of MoE (none / 0)

If the MoE chosen were +-4% instead of +-3%, Obama loses Pennsylvania, Montana, and Iowa to the MoE tally.

In short, BS assumptions = BS results.

by Sieglinde on Thu May 29, 2008 at 09:05:32 PM EST

Re: Convenient choice of MoE (none / 0)

Actually, the Democrats don't win with any margin of error less than 3%.

0 - Clinton tie, Obama loss
1 - Clinton loss, Obama loss
2 - Clinton loss, Obama loss
3 - Clinton loss, Obama win
4 - Clinton win, Obama loss
5 - Clinton win, Obama win

I happened to choose 3% before I did the math. Not using any wouldn't have been accurate at all, since there were were no decimal places in the original data. Since the most reputable polls' best accuracy was usually in that range, I thought it appropriate.

This, of course, wasn't meant to be an exhaustive mathematical study, but rather to make the point that polls this far out can't predict what will happen and they make a rather silly metric for electability without considering what will happen down the line.

by TCQuad on Thu May 29, 2008 at 10:00:59 PM EST
[ Parent ]

Re: Convenient choice of MoE (none / 0)

"This, of course, wasn't meant to be an exhaustive mathematical study, but rather to make the point that polls this far out can't predict what will happen and they make a rather silly metric for electability without considering what will happen down the line."

You made the opposite point, actually. Your conclusion is that Obama is stronger in November, based on your cooky math, but then your point now is that "polls this far out can't predict what will happen ...".

What're you smoking, and can you please pass it around.

by Sieglinde on Fri May 30, 2008 at 11:28:17 AM EST
[ Parent ]

Re: Convenient choice of MoE (none / 0)

No smoking in the apartment, sorry. But this line of reasoning was intended to show that polls this far out mean very little, given the fact that a lot of movement will occur before the general. The tags I assigned the diary were "Math, hocus pocus, statistics, electability, generous definition of the word proof". My first comment established this as not particularly serious, but rather a fun exercise in math.

I even explicitly stated in the diary:

Personally, I think using polls this far out is ridiculous and not particularly useful. However, for those who would like to give weight to these numbers, remember to take into account previous trends between now and the election.

You took this diary as "ZOMG, Obama's so much better!" when it was actually a perhaps-too-subtle twist on the use of the validity of May polls to determine electability.

The numbers this far out mean very, very little. If you assume that the campaign between now and November will be candy and flowers, Clinton has an electability edge. If you assume that the campaign between now and November will be exactly like 2004, Obama has an electability edge. In both cases, you're assuming things that can't be logically (but can be emotionally) justified.

by TCQuad on Fri May 30, 2008 at 06:26:23 PM EST
[ Parent ]

Re: Convenient choice of MoE (none / 0)

"The numbers this far out mean very, very little."

Like all things, it depends. A Clinton win in November will be fought in the traditional battleground states (OH, FL, PA), where there's lots of data and analysis over the past few election cycles. Notwithstanding the fact that she's still a woman, she appears to have overcome the "not-white-man" disadvantage among the electorate, and so therefore she is the Democrat's "safe" choice.

Obama, on the other hand, is banking on a vastly altered map. As Donna Brazile deftly put it, who needs those white blue collars when we have all the African Americans and the young people behind us. The list of battleground states, they like to believe, is different this year. So therefore, numbers are highly suspect when talking about Obama's electability-- or if they're to be used, it's really to Obama's disadvantage, primarily because no one has yet pulled a victory in the past by turning out the youth vote dramatically. Perhaps Obama can do it. I won't put my own money on it though.

So therefore, I don't buy your projection that "if you assume that the campaign between now and November will be exactly like 2004, Obama has an electability edge". The Obama campaign (as well as Donna Brazile and her cohorts) disagrees with you on that.

by Sieglinde on Sat May 31, 2008 at 08:41:46 AM EST
[ Parent ]

LMAO (none / 0)

Thankyou, I needed a good laugh.

by LindaSFNM on Thu May 29, 2008 at 09:15:29 PM EST

Re: LMAO (none / 0)

No problem. As you can probably tell from my tag "generous use of the word proof", this wasn't meant to be particularly serious. But I find the analysis of May polling data as an electability metric to be amusing, so I thought I'd contribute in a different (and just as valid) manner.

I don't believe a word of it, though.

by TCQuad on Thu May 29, 2008 at 10:05:02 PM EST
[ Parent ]

You are not logged in.

In order to post a comment, you must be logged in. If you have a member account, please log in to comment.

If not, you can make an account right here. It's quick and free.