Re: So This Is Depressing

1

Oh, fuck, I didn't follow Rosenfeld's link when I read that post.


Posted by: washerdreyer | Link to this comment | 04-14-06 2:35 PM
horizontal rule
2

The Post is a little unclear, contrasting R's are 25% more likely to vote for a D when the R senatorial candidate is black, with D's are 38% less likely to vote for a D (meaning, I'm guessing, either vote R or stay home) when the D House candidate is black. I get suspicious when stats that should be symmetrical aren't -- why aren't Senate races compared with Senate races, and why aren't switchers only compared with switchers only? I'd like to see the full study when it's available.

But there does seem to be at least some D effect, and Rosenfeld should have mentioned it.


Posted by: LizardBreath | Link to this comment | 04-14-06 2:40 PM
horizontal rule
3

But racism doesn't exist!!! Everyone knows that. After all, look at Clarence Thomas and Condoleeza Rice.


Posted by: bitchphd | Link to this comment | 04-14-06 2:41 PM
horizontal rule
4

Don't forget Oprah!


Posted by: apostropher | Link to this comment | 04-14-06 2:48 PM
horizontal rule
5

I found a link to something that sounds like it, but only the abstract is available


Posted by: washerdreyer | Link to this comment | 04-14-06 2:50 PM
horizontal rule
6

I'd want to see breakdowns by state, age, etc. I'm not too worried about us; I'm pretty sure we can backdoor Obama into the White House. And eight years (or even four), it will cease to matter.

I've long thought that the most important thing Clinton did, if we take a long view, was to construct his Cabinet that looked like America.


Posted by: SomeCallMeTim | Link to this comment | 04-14-06 2:53 PM
horizontal rule
7

I suspect they picked Senate races for Republicans instead of House races because there are even fewer races to pick from. There are currently no black Congressional Republicans, which is pretty amazing when you consider there's something like forty black Democrats in the House.


Posted by: Isle of Toads | Link to this comment | 04-14-06 2:56 PM
horizontal rule
8

LizardBreath,

To your credit, you point out (as TAPPED does not) that the effect you cite appears to go both ways. Without seeing the underlying data, however, it is not clear how much any of this tells us.

First, we don't need a study to know that race often plays an ugly part on both sides in partisan politics.

Second, it is not clear what the study really means, given what little we know. For example, according to TAPPED, Alan Keyes's senatorial bids in Maryland account for two of to Senate races analyzed. I'm a Republican and I would not vote for Keyes for a whole host of reasons having nothing to do with his race. Indeed, apparently only five Senatorial races were analyzed. How could one possibly control for other variables to say that people did not vote for their party's candidate because of race, particularly with such a small sample.

I think sampling problems probably explain why the study reports on House races for Democrats. As far as I can tell from the articles, for whatever reason, the author did not study Senate races involving white Republicans and black Democrats.

It would be interesting to see the data, but this seems to be one of those things that is really hard to study rigorously.


Posted by: Idealist | Link to this comment | 04-14-06 3:02 PM
horizontal rule
9

Idealist is right. Alan Keyes is a freak.

Here is another interesting Ebonya Washington study .


Posted by: Joe O | Link to this comment | 04-14-06 3:06 PM
horizontal rule
10

I'm a Republican and I would not vote for Keyes for a whole host of reasons having nothing to do with his race.

Mmm. The whole 'batshit insane' thing, perchance?

I don't know how hard it necessarily is to study rigorously -- if you broaden it to include state races, rather than just Federal races, the data set has to be a decent size.


Posted by: LizardBreath | Link to this comment | 04-14-06 3:08 PM
horizontal rule
11

I wouldn't vote for Keyes either, but I also wouldn't vote for Santorum or any of the batshit insane white Republicans who seem to keep getting elected--and who get covered as if their batshit insanity actually constitute a valid political viewpoint, whereas batshit insane black folks always get covered as merely being kooks (and often, as if their kookery were somehow intrinsic to their race).


Posted by: bitchphd | Link to this comment | 04-14-06 3:14 PM
horizontal rule
12

The whole 'batshit insane' thing, perchance

Well, that's a bit strong, but he is out there on some things.

I don't know how hard it necessarily is to study rigorously -- if you broaden it to include state races, rather than just Federal races, the data set has to be a decent size.

You may be right. It just seems from the articles (and they may simply have not reported it) that the study does not do this. It would be interesting to see.


Posted by: Idealist | Link to this comment | 04-14-06 3:16 PM
horizontal rule
13

The Keyes vote.

I have to say, most of the top hits don't seem out of the ordinary for a typical conservative Republican. Santorum, Coburn, etc. etc. But I've always taken Keyes' batshiticity to be something that is well known enough not to require citation.


Posted by: Matt Weiner | Link to this comment | 04-14-06 3:28 PM
horizontal rule
14

Per L.B.'s note about Giuliani, I'd be curious about the effect in mayoral races. In Philadelphia the Katz (white, Republican) vs. Street (black, Democrat) race was a whole lot closer than you'd expect in a city with a 4-to-1 Democratic voter registration advantage. Of course, that race was weird for a lot of other reasons too.

It all reminds me of the famous difference between the poll questions: Do you attend a religious service regularly? and Have you attended a religious service in the last 10 days? People tend to respond to the first question based on how they would like to think of themselves ("Of course I would vote for a person of a different race!") whereas the precision of the second question pushes people to think more literally ("oops, no, I didn't actually vote for X very-well-qualified candidate last time").


Posted by: Witt | Link to this comment | 04-14-06 3:35 PM
horizontal rule
15

I remember in 1983 after Harold Washington won the Democratic primary for mayor of Chicago, thousands of white folks who'd never voted for a Republican in their lives suddenly decided to support Bernie Epton (whose slogan was the subtle "Before It's Too Late") in the general election. More recently, the racists must have been very upset when the Republicans picked crazy Alan Keyes to run against Obama for the Senate. "Oh, shit, we have to choose between two black guys?"


Posted by: Frederick | Link to this comment | 04-14-06 4:30 PM
horizontal rule
16

Maybe it's just being a physicist (and being married to a social scientist in training), but I would want to see very careful cross-checks and sample studies before beginning to take this seriously. I'm sure there is some effect due to bias, but I would suspect that it is principally secondary, expressed in the circumstances leading to a black candidate running in a usually-white-candidate district, hence the first sentence. Ideally you'd want to run an expt where you had pairs of a priori equally attractive candidates and randomly chose whether the black or white one ran, and since you can't do that you have to control for horribly messy variables. With small samples and wacky outliers.


Posted by: rilkefan | Link to this comment | 04-14-06 10:42 PM
horizontal rule
17

#11

But if you watch guys like Santorum on TV, they're much slicker than Keyes. Santorum and his ilk know they've got to tone it down a bit to get elected. They'll get up and spin and lie like crazy to cover their real intentions. Keyes is like Pat Buchanan. He genuinely thinks there's a majority out there lurking, waiting for his brand of crazy, and then seems genuinely surprised when it doesn't materialize.


Posted by: gswift | Link to this comment | 04-15-06 3:19 AM
horizontal rule
18

Here is a link to the article:
http://www.nber.org/papers/w11915.pdf

This is the authors website:
http://www.econ.yale.edu/faculty1/washington.htm


Posted by: RA | Link to this comment | 04-15-06 12:09 PM
horizontal rule
19

On reading the article, I would say that the author acknowledges that there are a lot a variables for which she cannot control, and I think there are more that she has not mentioned in the article.

Even more, here is the author's conclusion:

There is much anecdotal evidence that Black candidates increase voter turnout. In some
stories, Black Americans come to the polls in support. In others White Americans come out in
opposition. I quantify the impact of Black candidates seeking election to the House of
Representatives and higher offices. Black voter turnout increases by 2.3 percentage points for every Black Democrat on the ballot. White voter turnout increases by 2.2 percentage points, an increase that is numerically greater given the relative numbers of Whites and Blacks in the
average district. The Whites who are propelled to the polls come from the ranks of both
Democrats and Republicans. However both White Democrats and White Republicans have a
reduced likelihood of voting for a major party candidate when that candidate is Black.


What explains this increased turnout? Candidate experience does not. Nor do voter
attitudes. Nor does increased media, party or voter attention. Perhaps a perception of an extreme
liberalness on the part of Black candidates is a factor. This is consistent with the increased turnout
in response to Black Democratic but not Black Republican candidates. To further test this validity
of this explanation, data on candidate ideology and on voters' perception thereof are needed. The
collection and analysis of such data is the subject of future work.


There likely are a lot of things going on here. Racism (both black and white) likely is a part of it, but the truth seems to be vastly more complex and nuanced. As always, (and particularly since the results for which there are robust data seem to show that the effect cuts across race and party affiliation) the "Republicans are racists" explanation, while always fun to throw out there, is awfully incomplete.


Posted by: Idealist | Link to this comment | 04-15-06 1:09 PM
horizontal rule
20

Hmmm, she's got a time and state dummy there but I don't see them discussed much (I only skimmed the paper). Have these effects been constant over time? Are they stronger in some regions then in others?
There's a whole lot of interesting stuff to be examined here.


Posted by: radek | Link to this comment | 04-15-06 4:59 PM
horizontal rule
21

There's a whole lot of interesting stuff to be examined here.

There is. Ultimately, the problem with this kind of research is that it is so hard to control for all of the variables and to come up with statistically significant correlations that say something meaningful. I think the author is trying, but it is particularly hard in this area--the reasons people vote for whom they vote for are complex and often not well thought out--plus there is a sample size problem. I suspect that there are very few elections that are the same except for the race of the candidates, if you are controlling for a lot of variables.

I worked for several years on death penalty research, one of the many goals of which was to show, in a general and statistically rigorous manner, a relationship between outcomes in capital cases and race. After almost ten years of work (and long after I left the study) I think they teased out a few correlations. They things they showed were not surprising, what was surprising was how hard it was to show the correlations in something resembling a statistically rigorous fashion.


Posted by: Idealist | Link to this comment | 04-15-06 5:29 PM
horizontal rule
22

BitchPhD,

Do you seriously think Alan Keyes was cast in bad light on account of his race? Or Cynthia McKinney? Or Al Sharpton, or Jesse Jackson?

If anything, Al and Jesse are getting more respect than their racism and profiteering would command were they not the national arbiters of blackness.


Posted by: Joe | Link to this comment | 04-16-06 2:50 PM
horizontal rule
23

The question wasn't addressed to me, but I certainly think that a white Republican expressing support for every one one of Keyes' policy preferences, and a white Democrat doing the same for McKinney's would each be perceived more positively by the population in general.


Posted by: washerdreyer | Link to this comment | 04-16-06 2:55 PM
horizontal rule
24

hott lesbian schoolgirl action


Posted by: hott lesbian schoolgirl action | Link to this comment | 04-16-06 3:11 PM
horizontal rule