Here’s a really interesting article (albeit a few months old) from the Pew Research Center that concerns a point we’ve touched on before: Minority political candidates tend to do better in pre-election polls than in the actual elections, suggesting that voters want to sound color-blind to pollsters but in fact carry a strong racial preference into the booth. The article cites one of the elections that we wrote about in Freakonomics:
Also in 1989, Democrat David Dinkins, an African American, won victory over Republican Rudy Giuliani in the race for mayor of New York by a slight two points, despite leading by 18 points in a poll conducted by the New York Observer a week before the election.
The news from the Pew Center, however, is that things may be changing, on two dimensions:
1. According to the polls cited by the article, the percentage of Americans who say they would vote for a black president is at or near historical highs; and, more significantly:
2. Based on the 2006 midterm elections, the gap between a minority candidate’s pre-election poll numbers and actual election numbers seems to have shrunk. In other words, it seems that fewer people lie to pollsters — and, therefore, that black candidates have a better sense of where they truly stand than they did in the past.
This is all written, not surprisingly, in the context of Barack Obama’s electability. If the Pew article and supporting evidence are to be believed, then the story here is substantial: If, for instance, at the end of this very long election cycle, the polls say that Barack Obama is favored by 53% of voters, then he might actually receive 53% of the vote.
(Hat tip: Maya Drucker)
and 
. You can also watch it on 
Where to begin with this? First, polls are such crap. They’re like “statistical analysis” – a fairly crude tool used to eliminate anxiety and help people who don’t have a clue decide what to to say or do. They take on a significance of their own that then wags the dog (in research or politics.) Second, “race” is such BS. It was something created to justify discrimination and has become very antiquated over years of social integration and reproduction. As long as people keep asking whether we’re obsessed with “race” – we’re obsessed with race. Barack Obama is “black” because it serves someone’s purposes to label him that way. It’s just as idiotic to focus on how “unbiased” we are as to judge people by their skin color. It’s very presence as a subject indicates that we continue to see people as abstractions and not as individuals.
Where to begin with this? First, polls are such crap. They’re like “statistical analysis” – a fairly crude tool used to eliminate anxiety and help people who don’t have a clue decide what to to say or do. They take on a significance of their own that then wags the dog (in research or politics.) Second, “race” is such BS. It was something created to justify discrimination and has become very antiquated over years of social integration and reproduction. As long as people keep asking whether we’re obsessed with “race” – we’re obsessed with race. Barack Obama is “black” because it serves someone’s purposes to label him that way. It’s just as idiotic to focus on how “unbiased” we are as to judge people by their skin color. It’s very presence as a subject indicates that we continue to see people as abstractions and not as individuals.
As a side note, if anyone wants to read the best thing written about Barack Obama that I’ve seen, check out Larissa MacFarquhar in the May 7 issue of the New Yorker: http://www.newyorker.com/reporting/2007/05/07/070507fa_fact_macfarquhar
As a side note, if anyone wants to read the best thing written about Barack Obama that I’ve seen, check out Larissa MacFarquhar in the May 7 issue of the New Yorker: http://www.newyorker.com/reporting/2007/05/07/070507fa_fact_macfarquhar
I would wonder if it’s about lying or lack of participation.
Peopled that are polled aren’t necessarily the same ones that will actually go and vote.
I know there is some evidence in post-voting polls that lying still occurs, but I don’t believe the gap is as great.
Also, how much influence does the poll have on the voter turnout? If a week before the election a poll shows a controversial candidate leading by a 20% or greater margin, you would possibly see two types of responses.
1. Those opposed would be motivated to really crank up their efforts.
2. Those supporting would possibly place less importance on showing up, since they have such a large lead.
Of course you could also argue that those opposed might give up hope at such a wide margin and not show up.
Either way, I don’t believe that people would lie to a pollster just to look less racist. Unless of course the race of the pollster played a part in their decision.
I would wonder if it’s about lying or lack of participation.
Peopled that are polled aren’t necessarily the same ones that will actually go and vote.
I know there is some evidence in post-voting polls that lying still occurs, but I don’t believe the gap is as great.
Also, how much influence does the poll have on the voter turnout? If a week before the election a poll shows a controversial candidate leading by a 20% or greater margin, you would possibly see two types of responses.
1. Those opposed would be motivated to really crank up their efforts.
2. Those supporting would possibly place less importance on showing up, since they have such a large lead.
Of course you could also argue that those opposed might give up hope at such a wide margin and not show up.
Either way, I don’t believe that people would lie to a pollster just to look less racist. Unless of course the race of the pollster played a part in their decision.
Polls are in that same Jeopardy! category as real estate agents, variable life insurance plans, lottery tickets – “I’ll take Crap I don’t need for $1000 Alex”
Do we all need metrics to establish who we tout at the water cooler but don’t actually get out to vote for (that is unless we’re talking about something really near and dear to our stupid hearts like American Idol)?
I’m debating on writing in Sanjaya for president.
Polls are in that same Jeopardy! category as real estate agents, variable life insurance plans, lottery tickets – “I’ll take Crap I don’t need for $1000 Alex”
Do we all need metrics to establish who we tout at the water cooler but don’t actually get out to vote for (that is unless we’re talking about something really near and dear to our stupid hearts like American Idol)?
I’m debating on writing in Sanjaya for president.