Department of Correlations
Greg Mankiw observes:
Obama leads in 18 out of the 19 states with the largest recent declines in home prices, whereas McCain leads in 13 out of the 14 states with the largest recent increases in home prices.
Marc Ambinder says “this statistic explains the election.” Obviously that was a bit tongue-in-cheek, but I actually think this statistic is pretty useless. For example, the data shows a large decline in home prices in California and also a sizable Obama lead but we hardly need to look to this data point in order to explain why a Democrat would be doing well in California. You need to look at the minority of states where presidential elections are actually close.
Here you find Nevada and Florida as states that John Kerry lost but where Obama has the lead and where you’ve also seen large declines in home values. But home prices are up in Iowa and Indiana, both states where Obama is doing much better than Kerry did.
The economic situation definitely has an impact on voting behavior, but this particular statistic doesn’t seem to explain very much. It might be interesting, however, to bore down into, say, the Florida data and see if Obama’s increases over Kerry’s performance are disproportionately coming from the parts of the state where the price declines are largest.
22 Responses to “Department of Correlations”
Register
RSSLogin
Topic Cloud
ArchivesBlog RollAbout Matt YglesiasContact Me
Donate
November 2nd, 2008 at 3:58 pm
looks to me Mankiw, uncharacteristically, confuses cause and effect. largest home price rises occurred naturally in areas that lean democratic. shifting demographics both increase the demand for housing and the support for democrats. I’m not surprised house prices are falling in California. I’d be surprised if they were falling in Idaho.
November 2nd, 2008 at 3:59 pm
It’s true, once you look at the actual states it doesn’t appear to explain much. Although, maybe Florida – they were down 12% in home prices. But I think that was destined to be an Obama state anyway. In a 7-point election, a slightly Republican-leaning state is going to go Democratic by a couple points.
November 2nd, 2008 at 4:20 pm
What liberal venom said. There is a reason the housing bubble occurred in California but not in say, Wyoming.
November 2nd, 2008 at 4:23 pm
The data SHOW. Ugh.
November 2nd, 2008 at 4:34 pm
This once again explains why we should ignore Marc Ambinder.
November 2nd, 2008 at 7:19 pm
You argument is that there should be a multiple regression with Kerry – Bush as well as house price changes on the right hand side. Of course you are right.
Mankiw knows what’s wrong with arguing from a simple correlation. he knows that the country is divided into flatland (where the cost of a house is roughly the cost of building one) and tne zoned zone (where the bubble expanded and burst). Flatland is not identical to Red America but they are similar : wide open spaces (not between the ears) are correlated with Republican voting and maegasprawling flatland is found where people hate regulation so much that they don’t even have much perverse zoning regulation.
November 2nd, 2008 at 7:30 pm
As I’m sure Steve Sailer will soon be telling us, home price declines map to ethnic composition.
November 2nd, 2008 at 8:01 pm
Mankiw, of all people, should know that correlation does not equal causation.
My god, they’re all just going to revert to their natural wingnut state again, aren’t they? This brief moment of sanity from the “sane” right was really just a Tiffin phantasm, gone after the last of the morning’s scone has trickled through the small intestine.
November 2nd, 2008 at 8:55 pm
Obviously commenters are not clicking through to read Mankiw’s post. Mankiw says NOTHING AT ALL about which is the cause and which is the effect. Mankiw simply notes the correlation.
It’s Ambinder (and Matthew) who are making the argument about causation. Just another example of why Marc Ambinder is an utter moron.
November 2nd, 2008 at 9:30 pm
Thanks
November 3rd, 2008 at 12:22 am
Just to confirm everyone’s suspicion, Mark Ambinder was wrong, the variable was insignificant
March 1st, 2009 at 5:08 am
viagra
It is the coolest site,keep so!
March 11th, 2009 at 4:19 am
I want to say – thank you for this!
March 14th, 2009 at 5:01 am
I want to say – thank you for this!
xanax
March 22nd, 2009 at 10:20 am
buy viagra online
Very interesting site. Hope it will always be alive!
April 3rd, 2009 at 4:01 am
Incredible site!
cheap brand pfizer viagra
April 9th, 2009 at 4:58 am
I want to say – thank you for this! viagra