November 6, 2012

Dave Weigel's exit poll analysis guide

Anybody see any 2012 exit poll demographics yet? CNN usually has the best website, but I haven't seen it up.

Dave Weigel writes in Slate:
COLUMBUS, Ohio — As I head to Republican victory headquarters, here's the starter kit. 
National exit polls from 2008. 
National exit polls from 2004. 
74 percent: The white proportion of the 2008 electorate. If it's higher this time, it's good for Romney. Sorry, folks, it just is. 
31 percent: The proportion of the Hispanic vote that went to John McCain. If it's stuck there, or lower, for Romney, it's terrible news in Nevada, Colorado, and Virginia. 
+7: The raw Democratic advantage in the 2008 election. In 2004 and 2010, there was no Democratic advantage—it was tied. If there's a Democratic advantage of 3 points or more, even considering all the moves voters have made and the regional party differences, it indicates an Obama comeback. 
- 46 percent: The proportion that Barack Obama won of the white vote in Ohio last time. He can afford to drop to 43 percent, and still win, if nonwhite turnout is constant. 
- 74 percent: The proportion of white "evangelical/born again" voters won by the McCain-Palin ticket, when they made up 26 percent of the total vote. These voters have never been asked to support a Mormon before. Mormon-baiting turned out to be the dog that didn't bark this year, but we have no great handle, yet, on this aspect of the vote, which will be crucial in the Rust Belt and Virginia.


Anonymous said...

Awful bad night for Steve Sailer and his most cherished theory of American politics. The Hispanic vote does matter! Oh well, at least he can continue to write passive-aggressive takedowns of Obama for four more years.

Anonymous said...

I could understand the Hispanic vote being a game changer in a state like NV or CO, but VA?

The Hispanic community in VA is much younger, less well established than the other two states, i.e. less citizen-y, and less likely to vote. It also much smaller. In VA it is the rural white and suburban upper middle-class Romneyites vs. SWPLs and nice ladies in NOVA and the black vote. The large number of gov employees could play in R or O's favor. Lots of those gov employees are Defense, and are mostly white transplants from red states. And let's not forget the contractors. The black non-military fed gov employees tend to live in DC or MD. It would be interesting to see a breakdown, by agency, of residence in VA or MD for fed employees.

Fed employees in VA will be voting in their perceived interests, I suspect, whether they go for Obama or Romney. Federal employment doesn't automatically equally Leftist, despite the stereotype. Many of them are by nature Centrists and Establishmentarians. VA going purple in recent years says as much about the GOP's recent idiocy as it does demographic shifts.

Whiskey said...

What's missing is women. Sandra Fluke was no fluke. Obama pushed her because that's most White women under 50 these days. They love Obama. What matters is posturing and approval of the media, not actual results.

Matthew said...

"Mormon-baiting turned out to be the dog that didn't bark this year..."

There were several YouTube videos posted by a user named newnamenoah, at least one of which received well over a million hits, and were circulated in political blogs. Newnamenoah's express purpose was to make Mormon's look evil and weird, and help defeat Romney. The account has been taken down, but the videos live on as posted by others.

Mormon bashing was very much happening this election. Whether it made the difference is another matter.

Matthew said...

"I could understand the Hispanic vote being a game changer in a state like NV or CO, but VA?"

Northern VA and the D.C. suburbs are heavily minority, especially black and Asian. My cousin's kid had a shirt with all the names of of her classmates on it (8th grade graduation, I think) and probably half the names were Asian. They are both non-white AND dependent on government largesse, via jobs, contracting, and welfare.

Anonymous said...

Tell me again Steve, how the Latino vote doesn't matter?

Anonymous said...


That's correct. Fairfax, for example, is extremely diverse. I didn't add the diversity element in my description of VA as a contest between a NOVA coalition vs. rural VA-Romneyite-Evangelical coalition.

While the melting pot in NOVA, combined with Romney's guns vs. butter spending rhetoric, probably means something for improving Obama's chances, that still doesn't mean NOVA is a stronghold of Hispanic voters. Lots of those people aren't Hispanic: Arabs, South Asians, East Asians. VA is nothing like the Far West or Southwest where the [multi-generational and very large] Hispanic vote is a big deal.

Also, how many of the parents of those students can vote? The kids will certainly vote, that's the Dem strategy, but it still isn't in play so much. Moreover, I suspect, the minority adults who can vote are younger on average than their white GOP opponents. Younger people don't vote as consistently as older people do.

Cail Corishev said...

Jeffrey, Steve has never said the Latino vote doesn't matter. He's pointed out these facts:

1) There are way more white votes than Latino ones.

2) Whites, especially married whites, already lean Republican, so you don't have to buy them off or trick them. All you have to do is get them to the polls.

2) All the evidence shows that there is a cap on the percentage of the Latino vote that can be wooed away from the Democrats. Even Jeb Bush, honorary Mexican and Hispanderer extraordinaire, barely got more than half of them. A white candidate with a white wife who is not willing to outbid the Democrats by declaring open borders and free TVs and cars for illegals can't expect to break 40%.

Given these facts, the best thing a Republican candidate can hope for is that Latinos stay home. Anything he does to encourage them to vote is giving his opponent three votes for every two he gains.

Also, since there are so many more white voters, if he can sway a couple percent of them, it doesn't matter a whit whether he gets 30% of Latinos or 35%, just like it doesn't matter whether he gets 5% of blacks or 10%. It's much easier to get a couple percent more of a large group than 20% more of a small one, especially when the small one is dead set against you.