what would a county map of the us look like

[Stranger in a strange land]

most of the maps are going to look similar to the actual election maps. The places where it gets interesting is in polarized states.

The most polarized states (and by that I mean greatest difference between white vote and actual vote) are, in no order,

New Mexico
Texas
Louisiana
Alabama
Mississippi
Georgia
South Carolina
North Carolina
Virginia
Maryland

I'd also add OR.

OR is polarized, but it's polarized politically, not racially. IIRC, Obama actually won the white vote in OR. It just has very liberal Democrats and very conservative Republicans.

[Snowstalker Mk. II]

I know both of these will be extremely depressing, but can someone please do Louisiana and Alabama? I'm not asking for Mississippi because I pretty much know what it will look like, but LA and AL have a bit more regional variation.

I can already tell you what it would look like. The black belt and the area just south of it are >90% Republican. The state gets a bit less Republican as you got northward. I can upload my almost complete non-black vote map as to give you an idea. It's just missing a couple rather boring states:

How is Philly Republican?

Wait...is that a GOP DETROIT?!?!

[minionofmidas]

That's a GOP Wayne County, which isn't the same thing at all. It definitely has quite a lot of Republicans in those Western suburbs. While to the North, the city line is the county line too. Detroit itself has so few Whites you could hardly draw any reasonable conclusions on their voting patterns from citywide data.
Philly though... hmmm. It does have Republican areas, of course. It has White Democrat areas too, though.

[Snowstalker Mk. II]

Philly can't possibly be Republican though. The Republican areas are the suburbs in neighboring counties, while even the whites in the city (about 40% of the population) are Democratic. Philadelphia was like 85% for Obama, and therefore it's nigh impossible for the city to be for McCain.

I also find a Republican Baltimore hard to believe as well for similar reasons.

[nclib]

I think this map overstates white Republican strength in urban areas (Philly, Baltimore, St. Louis, Cuyahoga County, Wayne County, MI, etc.) since there is probably lower minority turnout in urban areas. One would expect urban whites to be considerably more Democratic (or at the very least, equally Democratic) than the state average given by the exit poll.

[phk]

I think this map overstates white Republican strength in urban areas (Philly, Baltimore, St. Louis, Cuyahoga County, Wayne County, MI, etc.) since there is probably lower minority turnout in urban areas. One would expect urban whites to be considerably more Democratic (or at the very least, equally Democratic) than the state average given by the exit poll.

We'd have to run a regression with a dummy variable to take care of the urban/rural issues and a variable for income. They likely have huge effects on the predictability of the white vote.

More meaningful results could be derived too if we controlled for certain ethnic whites voting behavior like Germans for GOP or Scandinavians for the Democrats. Maybe a dummy variable for each certain ethnicity.

Of course actual data is hard to come by unless we actually go out and seek a random sample of white voters, but they likely will be some pro-GOP bias from 2008.

[RI]

The map certainly has its flaws because of the lack of data that is available. Only statewide data is available from exit polls. For the map, I didn't have any real way of systematically modifying the formula between counties within a state that wasn't purely subjective.

[phk]

The map certainly has its flaws because of the lack of data that is available. Only statewide data is available from exit polls. For the map, I didn't have any real way of systematically modifying the formula between counties within a state that wasn't purely subjective.

I'm sure exit polls did poll urban whites and rural whites separately right? Oh but there is that suburban business that makes the whole deal subjective.

There is also Census data that's available for regression purposes.

Variables I would probably use are

• McCain 2008 vote - M
• Percent White - W
• Median income - I
• College education - C
• Percent Married Couples - W
• Dummy variable for urban/suburban and not - D
• Unemployment in October 2008 - Google Public Data - U

log(W) = Constant + M + W + C + U + W + D
Once you have results for every county you can easily do a high confidence interval  like (>= 99.9)to see if the number of votes works out.

Another method, though extremely crude and overgeneralizing, is seeing the correlation coefficient of %McCain and %White.

I can probably try to work on a simple linear regression result of %White and %McCain in California just to see tonight.

Whats problematic though is if there's an interaction term or if a certain variable has a higher threshold than others. We could try squaring or multiplying variables.

[Atlas Has Shrugged]

It seems I cant find a blank county map of the whole US that I can fill in (yes county by county) with paintbucket, on Windows paint. Can someone post a blank one?

[phk]

It seems I cant find a blank county map of the whole US that I can fill in (yes county by county) with paintbucket, on Windows paint. Can someone post a blank one?

Why don't you just learn how to use GIS software?