There has been a fair amount of talk about the Bradley Effect lately. This is the idea that black candidates look like they're ahead and then lose or win by considerably less than polls would indicate. The mechanism is that racist voters are unwilling to admit to being so and will say they are "undecided" when really they have already decided to vote against the black candidate. Note: They typically say they are undecided. They aren't presumed to be saying, "Oh sure, I'm for the black guy," when they really aren't. Fivethirtyeight.com has a nice little discussion of whether and where the Bradley Effect has really been a factor against Obama in this election cycle. They come to the conclusion that it doesn't exist. But let's be pessimistic. Let's assume it does exist. What then?
The Bradley Effect tends to be lurking in the undecideds. If we had a situation where Obama was ahead 44% to 36% with 20% undecided that would be a very bad sign. It could suggest that large numbers of respondents who were afraid of being thought racist for supporting McCain were instead saying they were undecided. Shy racists typically say they are "undecided" not that they will actually support the black guy. So what do the polls say now?
Pollster.com and Real Clear Politics both have Obama at about 50%. That suggests that the undecideds are not a big factor in the national numbers. RCP has Obama at 49.9% and McCain at 42.4% with 7.7% undecided and Obama with a 7.4% lead. If you take the difference between what Obama has and 50% and divide that by the undecideds it shows that McCain would need 98.7% of the undecideds to win. If Obama gets just 1.3% of the undecideds, he’s over 50% in the popular vote. I figure that any state that shows Obama consistently at or above 50% is a cast iron lock. That includes, amazingly, Virginia! Any state with Obama having a lead that is as big or bigger than the number of undecideds is very likely to be an Obama win. Anything else requires some worry.
For example, in Ohio RCP has Obama with 48.9% and McCain with 45.5% and 5.6% undecided. I figure that if Obama gets just 19.7% of the undecideds he’ll get over 50% and win Ohio. The conventional wisdom is that undecideds will break about 50-50. The Bradley effect says that they will break overwhelmingly against the black guy. But to put Ohio in the McCain column, it would have be more than an 80-20 break in McCain’s favor. The situation looks even better for Obama in Florida. North Carolina could be dicey however. There Obama has 47.9% to McCain's 46.7% with 5.4% undecided. So Obama needs 38.8% of the undecideds to vote for him if he is going to win North Carolina. That could be a tall order in a state in the "Deep South."
But the bottom line is that even if there is a Bradley Effect at work here, Obama is very likely to win the election because his leads in states like Florida and Virginia may be insurmountable.
Wednesday, October 15, 2008
The Bradley Effect
Posted by Raised By Republicans at 9:22 AM
Subscribe to:
Post Comments (Atom)
7 comments:
I strongly suspect the Bradley effect does not exist in this election. The "Bradley effect" requires a few things to work (1) race has to be an issue in the election, to stir fears in voters' minds (2) the white candidate has to be seen somehow as an illegitimate choice to some voters who nonetheless intend to vote for that person.
This is not a local race or even a governor's race. Nor has race been (much) of an issue in this campaign. This is a massive national contest between a Democrat and a Republican, both of whom share the national stage as legitimate choices for any person. Both candidates are very well known to the public.
I doubt any person who supports or leans McCain feels obligated to say he is for Obama for any reason, least of all because Obama is black. McCain is a totally legitimate choice for all kinds of reasons not relating to race.
All that is going to happen is what happens in every election: some voters are reluctant to admit their true preferences to their work colleagues, friends, and families, if they believe these colleagues have a cohesive preference and the voter's political preference diverges from that group.
This "neighborhood effect" may be present in this election, but I suspect it is with *black* voters, not white ones, given the 95% ratings for Obama in that community.
Let's not forget either that a there are ton of new voters that are being registered by the Democratic party. And many of those are young and multiracial. They aren't going to have problems with racism.
I suspect that we are going to see a big demographic shift in the election if the young voters who are signing up now actually vote! I think they will and they will because it's historic, because they want to say they voted for the first African American president. so I say there is a new thing at play: The Obama Effect: race is a benefit.
Looking at it the other way around, the young and multiracial potential new voters could have a problem with racism: they could be voting against the white guy (quite possibly due to him representing those who have been racist towards said voters). It's still racism, even if the person being racisted* is white.
*apologies, couldn't think of an alternative, concise way of explaining that, so decided to briefly channel Dubya, before he's gone from all our minds.
Having said which, I'm just making a point - I don't actually think that this is the case, I personally think that younger people 'see' race less, due to attitudes changing throughout society. Talk to your average child, and they have no qualms about describing their friend via skin colour, or nicknames such as Bournville (a brand of dark chocolate in the UK) or Snowflake. Those last two names courtesy of seven year olds in South Wales, who were hauled into the headmaster's office to be talked to about the bullying names they were using. Turns out that they were best mates actually, and the names were terms of endearment.
Fair point Pombat.
Actually, the fivethirtyeight.com story argues that Obama is a net winner from a reverse "Bradley Effect." We could call it the under counted youth and black vote effect.
I prefer the "Obama effect". More conscise.
Pombat is right that younger people have different attitudes about race than even people of my (mid-30s) generation. I think, Pombat, that they do "see" race in one sense, that is to say they know very well who is black and who is white. But racism has been usefully defined as "the association of character traits with a set of physical characteristics." This is precisely what younger people no longer do. Indeed, the set of physical characteristics that define race is largely limited to skin color now, rather than the whole panoply of physical characteristics (curliness of hair, width of lips and nose, etc.).
Early US cases from the 19th century over runaway slaves infamously assigned the burden of proof by racial characteristics. If a person "looked" negro, he had to prove he was not a slave, but if person "looked" to be white, he did not. The case law reads like an excercise in physiognomy. Younger people especially just do not conceive of "negro features" the way we used to.
Even among African-Americans, there used to be much more of an emphasis on social gradation through skin lightness. I have heard this referred to by a black friend as the "brown paper bag test" - meaning that certain blacks were excluded by blacks from higher-class black institutions if their skin was darker than a brown paper bag. Hair-straightening is no longer important to status within the African-American community. That is largely a relic of history.
To see other examples of this, just take a look at racist cartoons from as late as the 1950s. These images of massive lips and so forth can seem comical and archaic. These images no longer inform the imaginations of people today, not even of most modern racists. If you can stomach going to the aryan nations website (as I have done), you will find lots of praise of Nazis and such, but no prominent racist imagery of this kind. There are no such caricatures of Obama (except for his ears), but there is a picture of him wearing a turban.
Post a Comment