« News Flash! First Major Public Poll Has Bush Under 50! | Main | Will a Recovering Economy Save Bush? »

Getcher Fresh Exit Polls Right Here!

Well, not exactly fresh, but the VNS consortium has now released a public use file of the national (though not state) data from the 2002 election. There’s a lot of interesting stuff here and DR’s crack research team will release results from these data as their analysis of the public use file proceeds.

Of course, we’ll do this analysis carefully and make sure we get it right. Others may perhaps be less careful. An early example of this is a column by pollster David Winston in Roll Call that claimed, among other things, that the VNS data show that it is a myth that "Republicans can’t attract minority voters in significant numbers".

Well, not really. In fact, the VNS 2002 data are completely consistent with that so-called myth. Republicans are still having huge difficulties attracting minority voters and the 2002 election was not an exception. Where the GOP did do exceptionally well was among white voters, where they received 60 percent of the white vote. That’s up from 57 percent in 1998, the last off-year election and the best point of comparison, and also from 2000, where they received 56 percent of the white vote.

Winton claims, however, that the GOP had a breakthrough year among Hispanics. He cites as evidence a drop in Hispanic support for Congressional Democrats and rise in support for Republicans between 2000 and 2002. While Winston’s data for ‘02 are wrong and exaggerate this change, it is true that the Hispanic two party House vote was 65 percent Democratic/35 percent Republican in ‘00 and did fall modestly to 62 percent/38 percent in ‘02. However, Hispanic support for House Democrats traditionally falls at least several points from a Presidential to an off-year election, so this says little about a real trend toward the Republicans. The more pertinent comparison is to 1998, the last off-year election, where Hispanics supported Democrats by 63 percent to 37 percent. So, basically, we have a shift in off-year Democratic support from 63/37 to 62/38. If that’s a trend, DR will eat his calculator.

Well, what about the Senate races? These were the most significant races of ‘02 and perhaps a pro-GOP surge can be detected here. Nope, the Senate two party vote among Hispanics was 67 percent Democratic/33 percent Republican. Governors, then? Not here, either–Democratic support among Hispanics was a healthy 65 percent to 35 percent.

What about other minorities? Not much luck here either for the GOP. In fact, blacks and asians both appear to have increased their support for the Democrats. The two party black vote for the House went from 89 percent Democrat/11 percent Republican in both 1998 and 2000 to a 91 percent/9 percent split in 2002. And Asians increased their support dramatically for House Democrats going from 56 percent Democratic/44 percent Republican in 1998 to 60 percent/40 percent in 2000 to 66 percent/34 percent in 2002!

Much more "progress" like this among minority voters and the GOP–aka "the white people’s party"–will have a very limited future indeed.


This is tangentially related to the topic but I wanted to say it.
On the far left there is a great concern about the integrity of the new touch-screen voting machines made by Diebold. While I am usually unsympathetic to paranoid-style politics, I think there is an issue opportunity here for the Democrats. The basic facts are that Diebold is the major touch-screen machine manufacturer and the company executives are partisan Republicans who donate heavily to their party. Some independent computer experts maintain that the machines are made so that they can easily be tampered with and vote totals altered in a way that is undetectable later. Diebold seems to be vigorously resisting any suggested changes that would fix this problem. There are also Republican ties to other manufacturers. I think Democrats should propose legislation along the following lines:

1. Conflict-of-interest legislation that prohibits voting machine executives from supporting, donating to, or being political officeholders.

2. Legislation setting standards for rendering these machines tamper-proof.

3. Legislation stating that the design and software code of these machines must be matters of public record.

I think that legislation like this seems so obviously public-spirited and reasonable that it would be folly to oppose it, but I'll bet the Republicans would oppose it nonetheless because accepting criticism and making changes and sacrifices in response to it are just not their style. If they went along with it they would lose a little campaign money and a worry would be disposed of. If they didn't it would be another piece of evidence, on top of a growing pile, that they cannot be trusted. I think the Republicans are close to a tipping point on being trusted, and the voting-machine issue could be very helpful in that area.
The tangential relation to the topic is that people on the far left suspect that Max Cleland's loss in 2002 was fixed and that lack of exit poll data partly allowed it to happen.