Remapping SUSA

Nick Beaudrot redoes the Survey USA state-by-state polling maps to offer up some shading proportional to the level of support the candidates are drawing. It makes for a somewhat more nuanced picture of the landscape though, as before, I caution you that if you aggregate 100 separate polls, some of them are going to be those outliers (Obama-McCain tied in Texas?) they warned you about when they described the margin of error:

Nick also proffers the following Senate analysis: "in the states with the ten most competitive Senate races, Obama does better than Clinton in eight of them; only Kentucky and Louisiana are better for Clinton." I don't know how well this polling will hold up, but I think this is a crucial question. The issue is not so much coattails as it is anti-coattails. If this starts to shape up as a good year for Democrats, "let's not give Obama/Clinton a blank check" will be a persuasive argument to some voters in states where Obama or Clinton is unpopular. You want to minimize the number of states like that.

Comments

The Nick-maps is the best maps.

Posted by Neil the Ethical Werewolf | March 7, 2008 1:58 PM

The best thing that can happen for the Democrats, since neither Senator Clinton or Senator Obama can win the election because the country is not ready for a woman or black president would be the following scenario.

The Florida and Michigan are not seated at the convention and the super delegates split more or less evenly between the two candidates. After a few ballots in which neither candidate gets the necessary number of votes, the convention turns to a compromise candidate who then selects Senator Obama as his running mate (he first offers the job to Senator Clinton but she declines, stating that she prefers to remain in the Senate and become majority leader). Albert Gore Jr. anyone?

Posted by SLC | March 7, 2008 2:04 PM

The best maps? I can't tell the difference between New York/Mass./etc. and Arizona and the West on them. Oh well.

Posted by red-green colorblind | March 7, 2008 2:15 PM

Sorry to go OT, but does anyone else get really creeped out by the "We're a great big family" thing over at DailyKos? Every day there's a new "diary" about "We're better than this!" or something.

Posted by Monster | March 7, 2008 2:16 PM

Oh, my God, I can't believe that I agree with SLC. Brr, that is not a happy feeling. :(

But SLC, I realize that at one point he was a bit of a Likudnick, but he has backed down considerably from that position. Are you sure he hates Palestinians enough to satisfy you?

Posted by LarryM | March 7, 2008 2:17 PM

not to mention having both spectra end on (what is basically) black is pretty confusing as well

Posted by john b | March 7, 2008 2:17 PM

Hillary's NAFTA-gate

The whole uproar about Obama/Nafta/Canada wasn’t true. Not only that, but it was actually the CLINTON campaign that was in contact with the Canadians and talking out of both sides of their ass. And then they try and turn the whole thing around on Obama!

http://firstread.msnbc.msn.com/archive/2008/03/06/738264.aspx

And I know there is a denial from the Clinton people at the bottom, but I frankly don’t believe them. Nest of snakes!

Posted by Adrienne D'Addrio | March 7, 2008 2:18 PM

SLC,
That is a very bizarre post. I can't even tell where you are coming from. If it is supposed to be facetious, it doesn't come off very well.

Posted by Stacy | March 7, 2008 2:18 PM

Hillary's NAFTA-gate

The whole uproar about Obama/Nafta/Canada wasn’t true. Not only that, but it was actually the CLINTON campaign that was in contact with the Canadians and talking out of both sides of their ass. And then they try and turn the whole thing around on Obama!

http://firstread.msnbc.msn.com/archive/2008/03/06/738264.aspx

And I know there is a denial from the Clinton people at the bottom, but I frankly don’t believe them. Nest of snakes!

Posted by Adrienne D'Addrio | March 7, 2008 2:20 PM

Hillary's NAFTA-gate

The whole uproar about Obama/Nafta/Canada wasn’t true. Not only that, but it was actually the CLINTON campaign that was in contact with the Canadians and talking out of both sides of their ass. And then they try and turn the whole thing around on Obama!

http://firstread.msnbc.msn.com/archive/2008/03/06/738264.aspx

And I know there is a denial from the Clinton people at the bottom, but I frankly don’t believe them. Nest of snakes!

Posted by Adrienne D'Addrio | March 7, 2008 2:20 PM

First of all, everyone needs to turn the brightness up; I have no problem telling the difference between the dark red and the dark blue. Sorry about those who are r-g color blind, but you guys have been screwing up my Catan games on Xbox Live, so we'll call it even.

Second, it's worth pointing out that while any one state may be off, in the aggregate things are probably correct. Obama does better in ALL of the Upper Midwest, Pacific Northwest, and South Atlantic Coast. Clinton does better in ALL of the mid-atlantic and the rest of the South.

Posted by Nicholas Beaudrot | March 7, 2008 2:23 PM

And, for the record, I don't agrre with the first part of SLC's statement - I don't think Clinton can win, but not because of her gender, and I do think that Obama can win, but he does have vulnerabilities, which are being heightened by Clinton's antics. But his race (though sadly an issue with some people) wouldn't be what keeps him from winning.

But I do think that a Gore draft at a deadlocked convention, with Obama as VP, could be a stroke of genius politically (and perhaps even give us a better president). Though of course it ain't happening.

Posted by LarryM | March 7, 2008 2:29 PM

According to the map, Texas is a swing state for Obama, but Clinton trails by 4 points. So much for her claim to be the only candidate who can win the big states.

Still, I have a hard time believing that Washington will give Obama a 14% victory and Hillary a 4% loss. Or that Florida swings in the opposite direction by about the same amount.

The fact that the map shows so many states with huge shifts like that convinces me that the polls aren't measuring real GE outcomes. I don't even think they're consistent with current voting patterns.

Posted by Jinchi | March 7, 2008 2:32 PM

Does anyone have a similar set of maps of George Bush's popularity? I think that's a better indicator of Democratic vs Republican strength in the November election.

Posted by Jinchi | March 7, 2008 2:35 PM

FWIW, the sense of things on the ground, here in TX, are such that a "tie" between Obama and McCain in this state is perfectly plausible. You seem to dismiss this notion too quickly as an "outlier," Matt.

Seriously. There are strong feelings here that Obama could put Texas in play.

Not so with Clinton.

Posted by Jason | March 7, 2008 2:38 PM

Re Stacy

Where I am coming from is very simple. I think that Al Gore is a better bet to win the election then either Senator Clinton or Senator Obama for the reasons mentioned in my comment.

Re LarryM

Mr. LarryM, like many others, seems to be laboring under the erroneous supposition that I hate the Palestinians. Nothing could be further from the truth. I am completely neutral relative to the Palestinians. I am all in favor of live and let live. Unfortunately, the Palestinians are not in favor of live and let live so the Government of Israel has no choice but to reply in kind^2. As I stated on a previous thread on this blog, none of the three candidates, McCain, Clinton, or Obama hates the Palestinians either, nor does former Vice President Gore.

Posted by SLC | March 7, 2008 2:40 PM

So what if he is? I don't see how that is relevant. Does Gore want to be President? Would he deserve to be after not going through the nomination process? I just don't see what you are getting at. Do you actually think what you suggested should or could happen?

Posted by Stacy | March 7, 2008 2:46 PM

First of all, everyone needs to turn the brightness up;

couldn't you turn the brightness up ?

this is what it looks like after opening it in the GIMP and cranking the brightness by +50.

Posted by cleek | March 7, 2008 2:47 PM

Re Obama and Texas - Texas would be in play if Obama picks an Hispanic VP candidate (is there a credible option there other than Richardson? Honest question; I haven't heard of one).

An hispanic VP candidate would also help in Florida (though not as much as one might think), and more generally shore up a potential demographic vilnerability. Thogh such a choice would have downsides as well.

Posted by LarryM | March 7, 2008 2:50 PM

Re Stacy

Ms. Stacy, the object of the exercise is to win the election. If Gore is more likely to win then either Clinton or Obama (which he is IMHO) and neither Clinton or Obama are able to garner sufficient votes at the convention to be nominated, I see nothing wrong with the convention, with the concurrence of Obama and Clinton of course, turning to Gore as the the man on horseback. This is what is called a draft and, however reluctant he might be, Al Gore would accept a draft.

Posted by SLC | March 7, 2008 2:54 PM

Good maps, as they always are.

However, what I'd really like to see is a map showing the difference between Obama and Clinton, so I don't have to keep going back and forth between the two maps. That is, if both Obama and Hillary beat McCain in a state by the same amount, that would be a 0, whereas is Obama beat McCain by 5 points more than Hillary did then that would be a +5, and if Hillary beat McCain by 5 points more than Obama did, that would be a -5 (which would obviously be colored differently than the +5).

I think that would be an easier way of seeing which candidate is stronger in which states and regions.

Posted by Al | March 7, 2008 2:56 PM

If Obama can compete in New Mexico and Colorado, why not Texas?

By the way, as I noted elsewhere, Clinton's support quickly falls off when you get West of the Mississippi, and it is not so great South of the Ohio/Mason-Dixon either. In other words, it appears she looks to most of the country like a typical Northeast Liberal--which is not particularly surprising for a Democratic Senator from New York.

Jinchi,

I really think Clinton's problems in the Pacific Northwest relative to Obama make perfect sense. Those states have young, liberal, urban enclaves plus Western-libertarian rural areas. Obama does significantly better than Clinton in both of those sorts of areas, and there is more or less nothing else in those states with which Clinton can compensate.

So, basically what you are seeing is Obama running up the score against McCain in the liberal enclaves and keeping things close enough in the Western-libertarian areas to win comfortably. In contrast, McCain is running up the score against Clinton in the Western-libertarian areas and keeping things close enough in the liberal areas to beat her.

Posted by DTM | March 7, 2008 3:01 PM

Yeah, I'm not from TX, but I don't know why everyone rejects this out of hand. Obama can turn out new voters and African-Americans like nobody's business. Huckabee got 36% on Tuesday, and with no anti-gay stuff this year, a lot of those bible thumpers may stay home. Independents may break for Obama in a big way. It's possible that hispanics, especially conservative leaning ones, might be trickier. It might depend on whether McCain's immigration flip-flops are a vulnerability or a strength.

But it doesn't seem like Obama will have a shortage of cash or organization. Just making the McCain camp worry a bit about Texas seems like it would be a pretty useful distraction, even if it doesn't end up blue.

Al- Is Nick's "Krugman Index" map what you're looking for?

Posted by jack lecou | March 7, 2008 3:05 PM

Jack lecou - yes, thanks!

Maybe I should have figured he'd have anticipated that.

Posted by Al | March 7, 2008 3:09 PM

Matthew, I am wondering about your disclaimer: "if you aggregate 100 separate polls, some of them are going to be those outliers"

While I am not saying it is wrong, I do have a couple of remarks:

1. Are those 100 separate polls, or are they in fact 50 polls with 2 questions each? That could have an impact on the error.

2. I've seen people saying that on average 1 in 20 polls is completely wrong; I'm guessing such general poll error theory refers to polls made by various pollsters using various methods, some of them with a terrible track record (e.g. ARG). But this is Survey USA, which has proven its efficiency during these elections, and also used the same methodology for all 50 states.
------

Also, for your example, I can see a close contest in Texas at this very moment, after a few weeks of intense and enthusiastic democratic campaign in Texas. That this enthusiasm will probably melt over time and Texas will get back to being safe Republican, it's another thing.

Posted by davidv | March 7, 2008 3:11 PM

I am from the Northwest, and I'll strongly second DTM's analysis there, though I do think McCain's support in the cities could fall off as the news that he ain't no liberal gets out.

Posted by jack lecou | March 7, 2008 3:11 PM

2. I've seen people saying that on average 1 in 20 polls is completely wrong;

That's exactly right if you use 95% confidence intervals, which I believe is the standard in the sciences.

Posted by Matt B | March 7, 2008 3:18 PM

Yep, the 1:20 number is coming from the 95% confidence interval, which dictates the reported margin of error. So, the expectation is that due to sampling error, 1:20 polls will find a result outside the reported MOE from the underlying number in the sampled population.

But of course these are really curved probability distributions, and there is nothing magical about the MOE. For example, there is not an even probability distribution between the two ends of the MOE range (which is why it is inaccurate to say two people within the MOE are "tied"). Similarly, there is not an even probability distribution outside the MOE, so saying 1:20 polls will find a result that is actually outside the reported MOE doesn't mean those polls are likely to be "completely wrong".

That said, a lot more can go wrong with polls than sampling error, and I think people are starting to figure that out. In that sense, a lot more than 1:20 of them could be very wrong, as a result of all the other possible sources of error. In fact, they could all be wrong if there is some systematic problem with their methodology, and we have lots of evidence of systematic polling problems this year.

Posted by DTM | March 7, 2008 3:29 PM

I'll say the same thing I said yesterday about this:
Speaking as a Democrat who not only wants to see our party retake the presidency but also expand its majorities in Congress and, in general, expand its coalition and reach across the country to put greater parts of it into play -- in other words, ditch 50%+1 and offensively compete in states outside the Gore/Kerry paradigm plus one more state (be it Ohio or Florida) -- I've got to say that I find the Obama map a lot more appealing. He's winning more states and, if one looks at the percentages, is competing more strongly in a lot more states that are currently in the McCain column.

And I'll add:
While one can choose to consider or dismiss the results of these polls (I'm inclined to give them some weight considering the reliability of the pollster, which has consistently one of the best and most accurate over the past six years, but especially in this volatile election cycle), it's hard to deny that it's definitely a map that looks more favorable to Obama than Clinton. It really should put to rest some of these lame memes such that losing one bloc of voters in a primary makes a candidate less likely to win that bloc of voters in a general election, and it lends credence to the idea that Obama's demonstrated strength with independents in this primary can remain a real asset as we move into a general election cycle in which the Republican brand and its standard-bearer have taken a real battering.

Posted by jbryan | March 7, 2008 3:57 PM

I don't think Clinton has ANY senatorial coattails. I hate to say it, but Louisiana is now a solid red state, and the Bush administration's atrocious Katrina response is the reason. (I'm paranoid enough to think they did it on purpose.)

Here's the math: in 2002, Sen. Mary Landrieu (D) was elected by 42,000 votes. Since then, Louisiana has lost 219,000 people due to Katrina, of whom the overwhelming majority are Democrats (African-Americans from NO, and Cajuns from the coastal areas). I'd love to be wrong, but Landrieu will be defeated this time.

And Clinton will help us defeat Mitch McConnell in Kentucky? Unlikely. He's a very strong candidate, and was elected by a big margin last time. There is no strong Democrat to challenge him.

As far as senatorial coattails it's advantage Obama.

And Texas? The math is not in favor of the Democrats taking Texas. In 2004, Bush won Texas by 1.7 million votes. And Kerry had essentially the same votes in Texas (2.83 million) as the TOTAL number of people who voted in the Texas Democratic primaries this year (2.82 million). Probably the Dems can pick up a few more voters (especially since McCain will not be a favorite son) but 1.7 million more? Doubtful.

Posted by Anne E | March 7, 2008 3:57 PM

Anne E,

I think it is difficult to quantify the importance of McCain not being a former Governor of Texas, particularly in a state like Texas which more or less views itself as having the same relationship to the United States as Britain has to the EU.

Posted by DTM | March 7, 2008 4:04 PM

Yeah, Obama's map seems a lot better. As long as it doesn't get too crazy with McCain taking Jersey, which I don't believe is really possible you have the following: Clinton is very vulnerable outwest and even though she is more competetive in the South, there are not many extra electoral votes there that she really has a shot at. Florida/Penn/Ohio are her only advantages. Obama is still really close in those places, does a much better job securing the West, can with on the southeast coast and even puts Texas into play. Maybe his politics aren't the most progressive you could imagine. But good, clean government? After better foreign policy, that's not got to be the next biggest thing we need after the Bush era. Obama may not be perfect, but I think he'll do a better job there than Clinton. Do any Clinton supporters really disagree?

Posted by mpowell | March 7, 2008 4:29 PM

DTM

I agree that GWB's advantage in Texas is hard to quantify... so I had to try and quantify it! After doing so, I agree a 1.7 million gap is much too pessimistic, but I think either Democrat taking Texas is a very long shot.

In 1996, Bill Clinton had a lot going for him in Texas. He is a southerner, he was an incumbent presiding over a strong economy, and Bob Dole was a lackluster candidate. Most important, Dole was afflicted with the continued presence in the race of Ross Perot who got 379,000 votes in Texas. Still, Dole beat Clinton by 276,000 votes. Assuming that most of the Perot votes would have gone to Dole in a two-person race, it looks like the Republicans have a 600,000 vote structural advantage which will be hard to overcome.

One other factor- in 2004 there were 39,000 votes in Texas for the Libertarian candidate. I suspect Libertarians will mostly be open to voting for McCain, especially if HRC is the Democratic nominee.

Posted by Anne E | March 7, 2008 4:43 PM

Anne E,

I think migration may have changed Texas a bit since 1996--in fact, many of those people you previously noted as leaving LA after Katrina ended up in Texas. Also, there are all sorts of different Southerners, and I am pretty sure most people in Texas would not take kindly to the suggestion they are Southerners of the Arkansas sort. Indeed, I suspect many would prefer a comparison to Kansas rather than Arkansas.

But in any event, I agree the Western-libertarian vote in Texas could be the key, which is not such great news for the Democrats in a Clinton-McCain matchup. But Obama does pretty well among that group, and McCain's war support (and the associated fiscal impact) could be a big problem for him among those folks.

Posted by DTM | March 7, 2008 4:58 PM

man this is the second or third map of Nicks where I can't distinguish the extreme colors. Even cleek's map doesn't help much.

I am thinking he needs to choose different palletes

Posted by Litch | March 7, 2008 5:36 PM

Here's a recolored map

Somebody has to tell me what software these maps are made with.

Posted by Arnold Evans | March 7, 2008 5:53 PM

http://img99.imageshack.us/my.php?image=geshiftvw5.png

Posted by Arnold Evans | March 7, 2008 5:58 PM

My understanding is that the vast majority of Katrina evacuees in Texas are not registered to vote in Texas. The efforts that I know of (by the NAACP, and legislation sponsored by Nancy Pelosi) have been aimed at empowering Katrina evacuees to vote in Louisiana (by satillite voting stations and absentee ballot) but I believe that this has been largely unsuccessful.

I hope it doesn't come across like I'm saying "don't contest Texas". The Dems absolutely must contest Texas ferociously every year- the demographic trends will eventually flip the state purple. I just think the payoff will come in 2012 or 2016.

Posted by Anne E | March 7, 2008 5:59 PM

http://img530.imageshack.us/my.php?image=obamaclintonmccainlt7.png

Photoshop adjust levels

Posted by Arnold Evans | March 7, 2008 6:11 PM

I think that this should be Democratic year, if for no other reason, economy is doing badly, and in a way that can be convincingly blamed on Republicans (oil and lending practices, big bussiness screwing the little folks as Federal regulators look the other way). If there is anything in Iraq that can be presented as victory, it will look like a hollow victory at best.

Is it enough to elect a Democrat as a Senator from Texas? At least, it is worth to try. My dream is to get such a Democratic majority that it could impeach reactionary judges if necessary.

Posted by piotr | March 7, 2008 6:43 PM

While we are dreaming big dreams...

The biggest problem facing a Democratic president is the 60/40 rule in the Senate. Unfortunately, I don't think we can get to 61 senators this year. Matt was so right that the Dems should have responded to the nuclear option by doing their best to eliminate the fillibuster altogther.

Failing 61 senators, the next best thing is to hold "moderate" Republicans from moderate states accountable for their actions in the senate. Harry Reid needs to do a much better job of letting the voters know how "moderate" Republicans are casting votes to block anything Mitch McConnell doesn't like.

Posted by Anne E | March 7, 2008 7:06 PM

Al Gore 2008 - if this bloodbath goes on into the summer he might be our only chance to win in November.

Posted by Brian D | March 7, 2008 10:47 PM

Considering only the states polling outside the margin of error--approximately 4%--this survey shows a far clearer advantage for Obama than may be readily apparent.

In an Obama/McCain match up, Obama is currently beating McCain 241 to 153 in electoral votes from states polling outside the MoE. This means Obama would only have to win 21% of the remaining 144 electoral votes to reach the magic number of 271.

In contrast, Clinton is currently beating McCain 233 to 213 in electoral votes from states polling outside the MoE--meaning Clinton would have to win 41% of the remaining 92 electoral votes to win the presidency.

Posted by Anon | March 8, 2008 10:34 PM