A Nationwide Swing
The 2008 Presidential map looks quite a bit different from the 2004 map. But the binary red/blue coding can be misleading. Andrew Gelman’s graph shows that the basic “map” is pretty similar:
Instead of anything getting redrawn, you saw a more-or-less consistent, across-the-board increase in the popularity of the Democratic nominee. Very red states like Utah and Idaho got bluer but stayed red, very blue states like Vermont got even bluer, and in a few states the general “blueing” pushed them over the line. But the trend was national, not really particular to Virginia and Colorado. The states that were immune to Obama’s charms were Arizona, Alaska, and Massachusetts (all explicable in terms of home state effects) and then a set of states (Louisiana, Arkansas, Mississippi, Kentucky, Tennessee, Oklahoma) in which racial dynamics seem to diverge from the national pattern.
40 Responses to “A Nationwide Swing”
Register
RSSLogin
Topic Cloud
ArchivesBlog RollAbout Matt YglesiasContact Me
Donate
November 5th, 2008 at 1:23 pm
Some of the states where Obama did exceptionally well should also factor in the campaign’s investment, I would think. Such as IN, GA, NV, NC.
I forget what the list of 13 or so states is that the campaign targeted from the beginning. Didn’t they win every single one of them? With the possible exception of Missouri.
I just wish now that they had put more effort into helping Franken in MN.
November 5th, 2008 at 1:28 pm
Looks like Obama’s original home state was just a bit more smitten than his current home state, not that that was a huge shock.
November 5th, 2008 at 1:29 pm
You can still hope the Abba will play the inaugural ball.
In other news it seems there won’t be a sequel to Rising Sun after all.
November 5th, 2008 at 1:38 pm
Hmmm … seems to me that there is a potential geographic pattern to some of Obama’s best states by increased support–I’m looking at Utah, Idaho, Nebraska, Kansas, North Dakota, Texas, South Dakota, Montana, Colorado, Nevada, and New Mexico. That’s not an exclusive list of his better states by this measure, but most of the rest are home state candidates (Hawaii, Delaware, Illinois, and perhaps Indiana) plus some states where AA turnout was a likely factor (VA, NC, GA, and SC).
November 5th, 2008 at 1:47 pm
Louisiana has to be chalked up to demographic changes, not electoral changes.
Arkansas, well…. let’s just let that one go, shall we?
November 5th, 2008 at 1:51 pm
The states that were immune to Obama’s charms…
Matt you’re over-reading this chart. With the exception of Hawaii this is about as tight a line as you’re ever likely to see in statistical analysis – especially when you consider that neither of the candidates in this election were on the ballot in 2004.
November 5th, 2008 at 1:54 pm
Well, after all the drama we hear about DC not getting votes, we see a resident exclude it from his graph.
November 5th, 2008 at 1:54 pm
An interesting question that nobody’s addressed yet is where does this put Republican Senators who are up for reelection in 2010.
My Republican Senator, Arlen Spector, seems to have his balls in a noose. In his 2004 election, poor Arlen got the living shit slapped out of him in the Republican primary. He prevailed but was staggering. He managed to fight through the General Election as well.
But now he’s facing a Republican primary that will probably be even more rabid — and a General Election in a state that’s gotten more Democratic. Plus there’s the cancer issue.
November 5th, 2008 at 1:57 pm
Forgot to mention — Arlen faced a hard fight in the last Republican primary because he’s a moderate — the Republican right wing doesn’t like moderates.
November 5th, 2008 at 2:04 pm
BTW, probably most people here will have seen it, but for the record, the NYT has a terrific set of graphics with very rich data on this topic- who became more blue, who more red, etc. Well worth stepping through the whole thing.
November 5th, 2008 at 2:08 pm
Matt you’re over-reading this chart. With the exception of Hawaii this is about as tight a line as you’re ever likely to see in statistical analysis – especially when you consider that neither of the candidates in this election were on the ballot in 2004.
Actually, I’d say that the tightness of the fit makes outliers a bit more striking. I’m sitting here in Nashville, a blue island in a sea of red, and in a state where indications are that Obama actually may have hurt down-ticket Democrats [The General Assembly has just passed into Republican hands for the first time since Reconstruction]. I’ve posted also over on Gelman’s site, but let me restate here that I think Obama may have redrawn the map of southern politics here by detaching what could be called the Central South from the coastal South [VA, NC, and FL, but even SC and GA]. I dont buy Matt’s easy assumption that we’re just more racist here, but I do suspect that cultural politics has had more salience here than in other parts of the country; prejudice against Obama as secret Muslim/closet commie/terrorist seems to have been more rampant in TN outside the two major cities.
November 5th, 2008 at 2:18 pm
Re David’s comment “prejudice against Obama as secret Muslim/closet commie/terrorist seems to have been more rampant in TN outside the two major cities.”
———
Yeah — the increase in Republican votes in Appalachia — from western PA down through West Virginian, Eastern Kentucky , Tennessee, and Arkansas was striking. Texas looks liberal by comparison.
Which is hilarious –because the last place any self-respecting terrorist would strike is those places. I doubt Bin Laden could even find Appalachia –and if he did, he would note the uncanny resemblances to Afghanistan.
Hell, FEMA even indicated those red areas would escape Russian strikes in a full scale, no holds barred nuclear war.
November 5th, 2008 at 2:18 pm
I was just about to point out the same NYT charts as luis, which helps explain what I picked up on above. Basically, I want to say that Obama seems to have disproportionately increased his support in Bush areas, except those Bush areas that fall into the NYT-identified band. That would explain why he seemed to increase his support the most in the states I previously identified (states with lots of Bush areas, but not in the band).
November 5th, 2008 at 2:31 pm
By the way, just a small point, but things were actually not that bad for Obama in most of central and western PA, and things only got grim down at the southwest end of West Virginia through parts of Kentucky, Tennessee, and into northern Alabama. This is relevant because Appalachia really has two culturally distinct halves, a northeast part centered on Pittsburgh stretching up into New York and down into West Virginia and taking in a bit of Ohio and Maryland, and a southwest part which is essentially just what I described above.
So Obama actually underperformed less in northeast Appalachia than he did in southwest Appalachia. Among other things, I think this does suggest a partially cultural, as opposed to purely racial, explanation for his relative lack of appeal in this band.
November 5th, 2008 at 3:02 pm
Actually, I’d say that the tightness of the fit makes outliers a bit more striking.
I disagree. And in any case, Matt is still misreading his chart. Andrew Gelman has a much more informative map of the states just 2 figures down from the one that Matt has shown here. From that it’s clearer that Arkansas, Louisiana, Oklahoma, Tennessee, West Virginia, Massachusetts and Arizona performed relatively worse for Obama. Indiana and Hawaii favored Obama.
Mississippi and Kentucky didn’t do any worse than New York and Minnesota.
Matt dismisses Arizona, Alaska and Massachusetts as “house effects” – so where’s the house effect in Texas and Illinois?
Matt makes it worse by blaming a divergent racial dynamic without any basis other than his gut instinct.
November 5th, 2008 at 3:29 pm
Hawaii’s extreme outlier status is probably partly due to the excitement of having a local boy (to use the local vernacular) on the ticket, but I would not discount the importance of Hawaii’s unique status as a truly post racial state (or at least one way, way, outside the black-white divide — people can be very race conscious here: they want to know are you half-chinese and 1/4 hawaiian and 1/4 haole, or was it 1/4 filipino, because that informs them of cultural connections and possible assumption). The fact that Obama is black (or maybe a kind of Hapa), would not be a mark against him. Actually, I’m surprised and gratified that a Punahou graduate could elicit that level of support here! (You’d have to know that Punahou has a (somewhat justified) snobby (and Haole) reputation that can turn people off).
November 5th, 2008 at 3:44 pm
Yeah, the HI old boy network had been sewn up by Hillary ‘way back when, and tried to play the Punahou snob card. Didn’t play well.
In the run up to the general, Gov. Lingle tried to play the local/transplant card, making it sound like Obama was like those military brats that are in and out in two years… pretty damn funny coming from a Jewish girl who didn’t move from the mainland until her twenties.
November 5th, 2008 at 3:49 pm
No, I actually agree with Matt entirely here. Look at the exit polling in the Deep South. 88% of whites in some states voted for McCain. And other numbers are very funny. For instance, Obama narrowly won the “not worried about economic conditions” vote in Alabama, but lost the worried vote decisively. Self-identified white Democrats broke for McCain 51/47. Do you think it was his maverick appeal?
November 5th, 2008 at 4:19 pm
True tranformative elections can only be known in retrospect. No one would consider 1980 or 1932 trandformative if the incumbent president had been bounced four years later. But instead FDR in ‘36 and Reagan in ‘84 both won by landslides, thereby sealing the transformation. By contrast Nixon’s 1972 massive victory was nothing but a flash in the pan, and we know what’s becom of GOP hopes of permanent dominance after 2004.
We shall have to wait four years before calling Obama’s true effects on our politics.
November 5th, 2008 at 4:28 pm
Matt dismisses Arizona, Alaska and Massachusetts as “house effects” – so where’s the house effect in Texas and Illinois?
I don’t have time to dig for his supporting evidence, but Nate Silver came up with and defended a theory that home-state advantage increases as the national profile of the state in question decreases. He specifically cited Delaware and Alaska, for example, as states in which Democrats would dramatically over- and under-perform, respectively, when compared with the national sentiment, since both states are so used to being ignored that even a veep nod would be the biggest thing since sliced bread. And that does seem to have happened here. Illinois, on the other hand, doesn’t seem to be far off the 2008 main sequence despite being Obama’s home state, because Illinois is used to being important.
November 5th, 2008 at 7:38 pm
dbt mentions it, but to say it bluntly: the shift in demographics due to Katrina has made LA a fair bit more red. There are many democratic voters displaced who have not moved back and may never return. All assumptions about how LA “should” vote based on its more moderate voting history compared to other southern states should be reexamined.
November 5th, 2008 at 7:41 pm
What the heck is the house effect for MA? Romney?
November 5th, 2008 at 7:58 pm
I would guess that TX “underperformed” its “house effect” not because there wasn’t one, but that it was balanced out by the same racial dynamics that weigh down the vote of OK, LA and AR Matt mentions. All of those states border TX and one can assume that there is a similar effect among some of its population.
As far as IL goes, in my reading of the graph there is a noticeable house effect, although more muted than that of DE for instance. It appears less pronounced because of the noticeable bounce NM and NV received (which are nearby on the graph), which I would argue are because of the fast changing demographics of these two small states.
November 5th, 2008 at 8:00 pm
Mo,
I think the theory is that MA would underperform this year compared to 2004 because Kerry was on the ballot then but not now. Fewer excited D voters (actually just flat while every other state was more democratic).
November 5th, 2008 at 10:37 pm
This looks like normal statistical variation to me. (I.e., the scatter of the states looks fairly random and uniform.)
The only state there that I would regard as a statistical outlier would be Hawaii, and that is definitely a home state effect. Delaware is also high but not by much, and Illinois is lost in the noise.
January 13th, 2009 at 10:02 am
laptop battery
laptop batteries
February 8th, 2009 at 9:03 pm
laptop battery
February 13th, 2009 at 9:46 pm
Of course, what a great site and informative posts, I will add backlink – bookmark this site? Regards, Reader.
February 14th, 2009 at 3:31 am
I keep listening to the news speak about getting free online grant applications so I have been looking around for the best site to get one.
March 2nd, 2009 at 5:29 am
levitraVery interesting site. Hope it will always be alive!
March 11th, 2009 at 4:52 am
Great site. Good info
March 17th, 2009 at 2:31 am
Excellent site. It was pleasant to me.
tramadol
March 22nd, 2009 at 6:24 am
tramadol
It is the coolest site,keep so!
April 9th, 2009 at 6:49 am
I want to say – thank you for this! viagra