If you were a bookie... Polls and bets on the 2004 elections

Hmmm. Does this mean Puff Daddy is going to kill approximately 80% of America's youth?

From this, one might extrapolate it was Democrats For Bush that made the difference.

No, Lash.

The 17% means that both in 2000 and now, those aged 18-30 made up 17% of the overall electorate.

As I already said, this means that their turnout must have increased in the same pace as that of the overall electorate.

If you check Dys' post, you'll see that:



Instead of 4 out of 10 youngsters, 5 out of 10 voted. Not overwhelmingly impressive still, but nevertheless - upping the proportion by a quarter aint bad. Puff Daddy can be proud.

Thanks again to Timber for the links. (And agreed on 'Democrats for Bush' having made the difference - although you'll note below that there were actually fewer Democrats-for-Bush than in 2000.)

Lemme make a list, cause thats what I do.

Kerry's results compared to Gore's, among:

Men: +2%
Women: -3%

Working women: -7%

Married: -2%
Not married: +1%

White Men: +1%
White Women: -4%

Whites: -1%
Afro-Americans: -2%
Latinos: -9%
Asians: +1%

(If you're wondering how he can lose among three out of four of these groups and still get the same total overall percentage, it's because Afro-Americans and Latinos, who still are in majority Democratic voters, turned out in higher proportions vis-a-vis Whites, in majority Republican, than they did in 2000.

Ergo, instead of 90% of blacks, Kerry got 88% - but more of them turned out.)

Under $15k: +6%
$15-30k: +3%
$30-50k: +1%
$50-75k: -3%
$75-100k: no ch.
Over $100k: -2%

No high school: -9%
High school grad: -1%
Some college: +1%
College grad: +1%
Postgrad: +3%

Large city: -11%
Small city: -8%
Suburbs: 0%
Small town: +10%
Rural: +3%

(Surprise!! Compared to Gore, it's the cities Kerry did badly in, not the small towns!)

Northeast: no ch.
Midwest: no ch.
South: -1%
West: +2%

("West" here is not just the Pacific Coast. Go check the interactive results map on NYTimes.com. It will show you a county-by-county map of where Kerry won compared to Gore's results and where he lost (and the same for Bush compared to his previous result, of course). No surprise is that Kerry lost in the South and borderstates. Surprise is that he won plenty of extra votes in the mountain states - Montana and such. Unfortunately, it will take many more gains like those to ever make these states competitive ...)

Democrats: +3%
Independents: +4%
Republicans: -2%

(How can Kerry have won among both Dems and Indys and still not have netted a higher percentage overall? Because Dems constituted a smaller - and Reps a larger - share of the overall electorate this time.)

Liberals +5%
Moderates +2%
Conservatives -2%

(See explanation above. Kerry did better among Liberals and Moderates than Gore had done, but the effect was outdone by the increase of conservatives turning out - or voters defining themselves as conservative.)

Protestant: -2%
Catholic: -3%
Jewish: -5%
Other: +12%
None: +6%

(Kerry's gains among "other" religions (Muslims?) come equally from former Bush and Nader voters; his gains from the a-religious are mostly picked up from Nader. Nader pulled 7% of both these constituencies in 2000).

Union members: -1%

Gun owner in household: no change

Gay, Lesbian or Bisexual: +7%

(77% of gays etc. voted Kerry.)

I think thats about it for the demographics in these exit polls. Thanks for your patience (I find these fascinating! ;-)).

Thing that strikes me most is how little actually changed. Plus ca change ... really. Kerry roughly mobilised the same blocks as Gore.

Conspicuous differences are the narrowing of the gender gap, Kerry's spectacular losses among those without high school and among Latinos, his inroads among low-wage earners (an odd combination with his losses among those with low education) and the drastic narrowing, contradictory to all the post-election rhetorics (in which I contributed here), of the urban-rural divide.

Working women: -7% <- That's what I thought would happen with the sex-in-the-city vote.

Thanks for the clear info Nimh. You deserve syndication.

Yes I appreciate Nimh's work a whole lot even if he does think I'm a blooming idiot.

Nimh: excellent work. Syndication indeed!

The city vote: -8% for mi-sized and -10% for large. As you note, that's the oddest one. Could it be a combination of a reduction in the votes of Catholic and Jewish groups--tying into the religious vote and Bush's affirmation of friendly ties between the US and Israel?

edit: To add to the above, the Latino and African-American vote no doubt contributed. If this can also be accounted for partly by religion, might the religious vote be the answer to the city question? That seems counterintuitive, but who knows...

Thanks nimh, it helps to see it broken down into groups. I wonder what it would have looked like if someone from the middle of the country had been in the running? Kerry is soooo north east rich liberal (although he isn't liberal at all).

Is there any breakdown between Latinos in different parts of the country? I have a feeling that Cubans are much more conservative than Mexicans. Not sure about Puerto Ricans.

An interesting perspective

Graphics from the NYT Election 2004 Interactive Pages


1984, 1988

1992, 1996

2000, 2004

Yeah, nine more states than Dukakis got, but one less than Gore.

Lemme go pilfer some more of those interesting NYTimes maps, before they're taken off-line in the future.

Who won and by how much: a map by county



---

Note: the maps below do NOT indicate where Bush and Kerry got their highest vote percentage; it indicates where they gained the most, in terms of percentage points, compared to last time.

(If you have seen maps that simply show each of the two candidates' support levels by county, lemme know where so I can put these two and those two side by side).





Note how Kerry did a lot better than Gore in not just Colorado, but also the Dakotas, Montana, Idaho. Odd.

Note as well how Kerry also increased his share of the vote in much of Ohio; but not in Florida.

Question to the experts on this one: it says that Kerry lost, in terms of percentage points at least - obviously not necessarily in terms of actual votes, because of the increased turnout - in two-thirds of the counties. But his overall share of the national vote was the same as Gore's.

So does that mean that where he did win, he did so much more strongly than how he lost where he lost? Or perhaps he lost percentage points in a lot of counties with a comparatively small population and won in a few counties with a lot of votes? But that seems in contradiction with the exit polls that say he did better in small towns and rural areas, and worse in large cities, than Gore.

On that one I'm confused. But otherwise its a really interesting map.

The election maps from recent years pretty much answer Diane's question about "someone from the middle." The stripe down the center for Clinton provides a pretty sharp contrast with the recent candidates. Good maps! Also a pretty good reminder of what a truly dominating victory looks like (Reagan).

These things take time ... be patient, we're working on it

Just a nit-picking point here, and I may be wrong, but by my calculations, Kerry lost, regardless of margin or change of margin election-over-election by comparison with Gore, not 2/3 of America's counties, as nimh references, but 82% of America's counties. But then that's by my calculations.

Just looking at the map for the western states, excluding the Pacific corridor, it is interesting that most Kerry counties fall a) inside the areas containing the most liberal cities such as Santa Fe and Taos NM, pure bastions for the ultra wacko radical left wing here, and b) within the Indian reservations or lands controlled by the Bureau of Land Management and thus highly populated by government workers or people dependent on government salaries, or c) areas heavily populated by Hispanic or black Americans.

Again I wish there was some way to assign demographics to the counties won by each across the country, but you're talking about a HUGE amount of work to produce that.

Oooops ... I see now I misundertstood what nimh was gettin' at there ... Invalid nitpick .... sorry.

Not exactly 3-D, but similar idea:



From here, very interesting:

http://www-personal.umich.edu/~mejn/election/

by county:



This is what you're looking at:

Soz, that is the most helpful and effective work I've ever seen for a national election. It clears up so much that seemed terribly skewed by the so-called 'normal' maps. Many thanks!

perspective...gotta love it