2004 November 03 Wednesday
Bush Gained Over 2000 With Women And Old Folks

Check out this CNN chart comparing Bush's takes on various groupings of the population in 2004 versus 2000. Keep in mind that these figures are preliminary. As we learned early election day from preliminary exit polling that showed Kerry ahead 2 percentage points initial statistical sampling and analysis can get things pretty wrong.

I am still digesting this. But a few things leap out.

First, there has been talk in the last couple of years about the idea of "security moms". Well this election shows evidence for this. While Bush gained 2% among males to rise to 55% he gained 5% among females to rise to 49%. So the gender gap shrank 3%.

Second, by age bracket Bush's biggest gains were among oldsters. CNN doesn't have data on 18-29 year olds. Bush gained 4% with 30-44 year olds and only 2% with 30-44 year olds. But Bush gained 7% with age 60 and older. So Rove made the best decision for Bush by supporting the Medicare drug benefit. Of course all the rest of us will have to pay for that for decades to come.

Another point: He gained most among those who had less than a high school education (a full 10% to capture 49% total!) and gained 3% from those with high school degrees and with college education short of bachelors degrees. He made a 1% gain with college grads to tie Kerry at 49% each and made no gains with people who pursued post-grad education.

Also, Bush made gains in all racial groups. His biggest gain was a 9% swing with Hispanics. I find this hard to believe and am looking forward to more detailed analyses by Steve Sailer. But the gain among those with lower education might explain a big part of the Hispanic gain. Or was Bush's worker permit proposal a factor?

There appears to be a mistake in their chart about voting and church attendance. They show huge swings in favor of Bush with those who attend church regularly. How could Bush have gained 25 points to get to 58% among those who attend church weekly? That'd mean that Bush got less than half the regular church attenders in 2000. That is absurd.

The three biggest hot button issues for voters were Moral Values (22%), Economy/Jobs (20%), and Terrorism (19%). For those who thought Moral Values were most important Bush got 80% of the vote. For those who thought Economy/Jobs were most important Kerry got 80%. For those who thought Terrorism was most important Bush got 86%. The terrorist threat definitely worked in Bush's favor. Note that Iraq was in fourth place in ranking at 15% and of those who thought that issue most important Kerry got 73%. That issue was less important and also caused less of a shift than the top 3 issues.

My guess is that absent 9/11 John Kerry would have won. Kerry needed to address the terrorist threat offer a more effective critique of what was deficient about Bush's response to terrorism. But as a very liberal Democrat Kerry wasn't about to stump for better border control and immigration policies or more capabilities for federal agencies to track down terrorists with computers and databases. So Bush wins.

Share |      By Randall Parker at 2004 November 03 12:56 PM  Politics American Domestic

Brock said at November 3, 2004 2:26 PM:

"My guess is that absent 9/11 Kerry would have won."

That comment is absurd. Absent 9/11 we would be living in a whole different world. Bush's presidency would have been completely different, and Kerry might not have won the primary or even run in it.

Kerry lost because he's a defeatist. He got the defeatist vote (moulting hawks afraid of losses in Iraq) and lost the winner's vote (concerned with defeating terrorists).

Your ideas about immigration and beefing up anti-immigrant laws may be right in some respect, but you're living in a dream world if you think any American Presidential candidate could support them and win. Only the wolves howling alone in the wilderness share your views.

Derek Copold said at November 3, 2004 2:50 PM:

I tend to agree with Brock. Absent 9/11, the world would have been much different. We might not have gone into Iraq, despite the president's desires, because there would have been no pretext. Absent that, the economy would have recovered much sooner.

As for immigration, that battle will have to be won in small steps, such as the security package the House stood by.

On an ironic note, the "conservative" White House announced today that it wanted to raise the debt ceiling so it could engage in yet more deficit spending.

Randall Parker said at November 3, 2004 4:12 PM:

If 9/11 hadn't happened then national security would have been a lower priority in the minds of most Americans. Therefore economics and social issues would have been higher priorities. That would work to the benefit of the Democrats.

Bush would have still done the Medicare drug benefit deal. So he would have picked up on some of the increase in elderly voters. But he wouldn't have gained the other surge that came from terrorism.

Of course, he might not have invaded Iraq. So he might not have had that as a negative.

Popularity of immigration reform: Proposition 200 just won in Arizona. That proposition, aimed squarely against illegal immigration, even got 47% of the Hispanic vote in Arizona. Tom Tancredo just won reelection with 60% of the vote. I've posted on national polls about immigration reduction. The elites are opposed. The masses are in favor.

crush41 said at November 3, 2004 4:36 PM:

Hispanics tend to be very socially conservative. As noted, "values" was the Bush trump card. Kerry and the media did a great job highlighting job woes, but not enough to convince people that Bush had anything, distinguishable from Kerry, to do with such struggles.

Arizona and Southern Cal may be reacting to the border problem because the crime rate is out of control in cities like LA, Tucson, and Phoenix. But elsewhere, it is still a long time coming.

Kris Kobach (R), 3rd district of Kansas running for the House seat, lost last night on a campaign centered around drastic border-control changes. Specifically, advocating the utilization of the border patrol in militarizing both borders, thereby stopping the flow of illegals and their baggage. Much of the work performed by illegals could be picked up at similar cost by prison systems that would operate on a token economy (ie physical labor to exchange for shorter sentences, prison provisions, etc.)

Despite being the more charismatic, healthier, more attractive candidate, and winning all three debates against Dennis Moore (D), Kobach lost by 8% in a Republican-leaning district. The country is far from seriously addressing this issue. Unfortunately, it will take a catastrophic attack that is traced back to the border to make it happen.

crush41 said at November 3, 2004 4:41 PM:

I forgot to mention why he lost: Moore unleashed a series of ads accusing Kobach of being a white supremacist, and an anti-immigartion radical conservative. Until those ads came out, as I recall, Kobach was within 2% or 3%, but after the white supremacist attack came out, he never got closer than 10% in the polls.

Randall Parker said at November 3, 2004 4:57 PM:

Crush41, The assertion that Hispanics are socially conservative is false. They have twice the illegitimacy rate of whites. They have higher crime rates, higher rates of abuse of children, and other pathologies. Those are not characteristics of conservatives.

lindenen said at November 3, 2004 6:32 PM:

But they are characteristics of the poor. As hispanics rise economically, do they become more socially conservative? And just what exactly are their attitudes toward gay marriage etc? I doubt they are socially liberal despite the social disfunction.

crush41 said at November 3, 2004 6:34 PM:

No doubt you are more erudite than I am, but the research I've read shows Hispanics are quite socially conservative. The rate among Hispanics who oppose gay marriage is higher than among whites. The same for abortion; especially first-generation immigrants who oppose abortion six to one. Both these issues have a prominent place among voters who considered "values" as the strongest determinant of how they cast their ballot.

Higher crime rates are more strongly correlated to lower education levels and lower socioeconomic status than ideology.

http://www.rockymounttelegram.com/news/content/shared/news/stories/0323_gaymarriahge.html (gay marriage)
http://www.catholicexchange.com/vm/index.asp?vm_id=26&art_id=16878 (abortion)

John S Bolton said at November 3, 2004 6:46 PM:

Those exit polls are very far off on the latino participation; it would be only around half what they said. Bush more likely had about 95% non-minority voters. If it was that way, it means that it was dozens of times harder for him to get a minority vote, than a majority one. They must have over-sampled minorities, in order to get numbers for each group, and failed to compensate for the over-sampling in sufficient degree.

birch barlow said at November 3, 2004 6:58 PM:

I actually think that Bush's gains amongst Hispanics and the 60+ population may have a lot to do with moral issues. I think the moral issues may have gotten socially conservative people of all races out (especially since some of Bush's biggest gains were amongst those 60+ who I think tend to be socially conservative compared to younger people). According the exit polls, Bush made some gains with *every* racial group.

Getting those Hispanics who *are* socially conservative to turn out may in part explain why I-200 did *relatively* well with Hispanics.

I tend to doubt Bush's pandering worked. His prescription drug plan is enormously unpopular with seniors. His guest worker plan is enormously unpopular with all racial groups, including Hispanics, many if not most of whom do not want more immigration. Those Hispanics who *do* support open (or at least looser) borders I think tended to greatly prefer Kerry's immediate amnesty to Bush's more convoluted amnesty/guest worker plan, IIRC. Also, IIRC, Hispanics don't tend to much care about immigration either way compared to issues like the economy, education, health care, Iraq, terrorism, etc--i.e. the issues that voters in general listed as their top concerns.

I would also add that given the sample size for Hispanics in the exit polls, as well as possible regional and other biases, that Bush's gains amongst Hispanics (compared to other groups) may be largely or even entirely ficticious.

John S Bolton said at November 3, 2004 8:09 PM:

The other polls, from before the election, which were accurate in showing Bush 2-4 points ahead, showed latino support at 30%, the same as 2000, or a little lower. That would fit with Bush's vote being 1% Asian, 2% Latino, 1% black, and as much as 96% non-hispanic white. if you look at the results in the 'red states', with such broad regions going 2-to-1 and 60-40 republican; something must have happened there. Whatever it was, some of it spilled over into Florida and Ohio.

Randall Parker said at November 3, 2004 8:15 PM:

Folks, the reason I doubt the Hispanic polling numbers is that a few weeks ago Bush was polling at 30% among Hispanics and Kerry was polling at 59%. How can that group swing 14% toward Bush in such a short period of time?

Birch, As for first generation versus later generation Hispanics: It would be very interesting to know what the voting patterns are of Hispanics as a function of generation since arriving in country. By your theory if first generation Hispanics are more socially conservative then they should be more Republican too. Then later generations should drift leftward politically.

Keep one thing in mind about Hispanics across generations: scholastic achievement doesn't improve that much.

John S Bolton said at November 4, 2004 1:00 AM:

Latino median household income per person is less than half that of the majority. For a population with that sort of countrywide economic profile, a Republican vote of over 30%, when the total electorate was close to 50/50, is a political impossibility. DC voted 10-1 for Kerry and plenty of latinos live there. On the county results map for the US, the Hispanic-majority counties are easily outlined in Kerry blue, just as the delta and the black belt are.

Derek Copold said at November 4, 2004 7:23 AM:


Here's a useful piece on Hispanic family values:

It's written by Ed Rubenstein, who used to crunch numbers for National Review.

I'll reproduce the important points here:

"* Abortion: Hispanic women are two and a half times more likely to have abortions than white women (33 per 1,000 annually vs. 13) and nearly-two thirds as likely as black women (49). [Source: Guttmacher Institute]

"* Teenage pregnancy: Hispanics are high (about 94 per 1,000 vs. 32 for whites) and relatively worsening – they’ve now surpassed blacks (83 per 1,000). [Source: Centers For Disease Control, pdf file]

"* Dependency: Nearly one-third (30.6%) of Hispanics receive means-tested benefits, compared to less than a tenth (9.2%) of non-Hispanic whites and just over a third (35.0%) of blacks. [Source: U.S. Census, pdf file]

"* Criminality: just over one in every hundred adult male Hispanics (1.2%) was imprisoned in 2001 – almost a third of the non-Hispanic black rate (3.5%) and well over twice the non-Hispanic white rate (0.5%). [Source: Sourcebook of Criminal Justice Statistics, 2001, pdf file]

"* Risky behavior: The Hispanic death rate from HIV disease is 2.5-times that of whites (7 per 100,000 vs. 2.8), and about one-third of the black rate (24). [National Center For Health Statistics, pdf file]"

Sal said at November 4, 2004 10:52 AM:

A good result of this election is that in the future there will be fewer dreary television messages and specials on tv urging people to vote. It is wonderful to watch the coastal elites get slapped by the masses.

Derek Copold said at November 4, 2004 2:16 PM:

Sal makes a good point. I remember in 1992, after Clinton was elected, the media elite were creaming themselves with joy about "high turnout" (nevermind that 60% of that high turnout voted to the right of the most rightist Democrat elected--ie, for Bush I or Perot). Today, we don't see to many of those encomiums to high voter participation.

crush41 said at November 5, 2004 8:47 AM:

Derek: There is a separation between the actions of the economically disadvantaged Hispanics and what Hispanics, on the whole, hold as their values. Obviously voting is a personal decision determined by the individual. Everything I have seen shows that Hispanics are more pro-life, and more opposed to gay marriage, (what they profess are their values--not necessarily the actions of Hispanics as a whole) than whites or blacks. Clearly, the worse off one is financially, the higher the rate of occurences listed above becomes.

Also, it's unlikely that those responsible for the statistics laid out actually came out to the polls. Those that did vote, I maintain, are probably more socially conservative than their other minority counterparts and maybe even whites.

Derek Copold said at November 5, 2004 8:56 AM:

Crush, look at those statistics. They're not the same as white, they're far in excess in a lot measures. Add on to that the Hispanic culture's acceptance of government paternalism and you have a natural Democratic constituency. The only way a Republican can win any increase is by acting like a Democrat, which is pretty much what Bush has done for the past four years. So in the end, they may vote for the "R", but in substance, the "D" has really won.

Randall, As far as the numbers go, there's a contrary survey out now that shows only 31.4% of Hispanics voted for Bush. Here's the link:

I don't vouch for the reliability of the numbers, as I know nothing about the WCVI. However, these numbers are pretty much in line with all the pre-election polling

birch barlow said at November 5, 2004 12:52 PM:

Interesting map here:

Looks to me like the heavily Hispanic areas of Texas (such as the southern part of Texas near the Rio Grande) went pretty heavily Democratic. I'm thinking WCVI, as well as the pre-election polls, are more in line with the actual Hispanic vote than the exit poll numbers posted on CNN.

Derek Copold said at November 5, 2004 2:12 PM:

I have to agree. The total Texas vote was 61% for Bush. If the Hispanic vote was truly 59%, then Hidalgo, Starr and other border counties should have been at least light red. Instead they very from white (Cameron county on the tip of Texas, which has a large white population in its western half) to dark blue (Zavala County, which elected a county slate from the La Raza Party).

birch barlow said at November 5, 2004 4:51 PM:

Also look at Orange and San Diego counties in CA (the two southernmost coastal counties in CA, traditionally very conservative). It looks like they were close to 50-50. I don't see how that could be without the Hispanic and probably Asian vote going waaaay left. I'm guessing the Hispanic vote may have gone close to 40 points for Kerry nationwide (maybe close to 50 in the more liberal California) and the Asian vote may well have gone close to 35 points for Kerry. It's looking to me like the exit polls on the CNN website greatly overestimated minority support for Bush.

The bottom line: Bush's pandering to Hispanics and other minorities *may* have given him a boost with white soccer moms, but probably backfired with actual minorites.

John S Bolton said at November 5, 2004 10:54 PM:

It looks like what happened with the exit poll was an enormous over-sampling of minorities, especially latinos, but also blacks. Then they scrambled to reduce the excess of Kerry votes, by deleting 4 points of latino Kerry votes, while leaving the latino Bush votes untouched. Starting from 12%, deleting 4 points of latino Kerry votes, leaves an 8% latino proportion, but the Bush percentage has been falsely raised from ~30 to ~45%. If the totals were readjusted to 6% Latino and 9% black, that would fit the results of Bush ahead 3 points, instead of behind by 3 points, as the exit poll had it before they started readjusting their numbers. Also the Latino Nader percentage would fall closer to 1%, where it should be. The Asian percentage of the voters was 2%, according to the exit poll; a little under half their percentage of the population, as is to be expected of a largely foreign-born population. The latino percentage of the total vote should also be around half their percentage of the total population, or around 6%. These are the raw materials of our political history, and the exit-poll managed to bungle it dreadfully.

John S Bolton said at November 6, 2004 3:27 PM:

The exit-poll, with many millions of network money, might have got infiltrated by activists, sociology majors and others with political and ethnic axes to grind. As soon as those in charge realized that they had been used this way to the extent of pushing the poll six points toward Kerry, they tried to correct it by eliminating four points of the total. It looks as if they took these four points out of the Latino Kerry support, without reducing the Latino Bush respondents total. If they removed another two points, exclusively from the Latino Bush support, that would bring the poll close to the truth. The same process which generated the overestimate of latino voting, also would have oversampled the black vote by two or three points of the total vote. Some factor caused almost a two-fold oversampling of latino votes, and significant overestimate of the black vote.

Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Web parapundit.com
Go Read More Posts On ParaPundit
Site Traffic Info
The contents of this site are copyright ©