“I just haven’t seen good evidence that [race] is a factor in polling these days.”
~Scott Keeter, director of survey research for the Pew Research Center
“The level of inaccuracy of the polls in the primaries was unprecedented.”
~Anthony Greenwald, psychologist and (overcited) co-inventor of the Reverse-Bradley Effect
“This distortion is interesting because poll numbers are part of the story journalists tell the public”
~ Bethany Albertson, political scientist and (overlooked) co-inventor of the Reverse-Bradley Effect
As Barak Obama continues to lead the polls, pundits and pollsters have latched onto the Bradley Effect. The Bradley Effect is the idea that (white majority) voters will tell pollsters they are going to vote for the African-American candidate (to prove that they are not racist), only to flee back to the white guy camp on election day. It named after the African-American gubernatorial candidate Tom Bradley, who enjoyed a huge lead in the polls during the 1982 governor’s election in California, only to loose on election night to the standard white GOP dude, George Deukmejian [1].
While most pollsters accept the Bradley Effect, and attribute it to voters disavowing their traditional American racism [2], others bitterly oppose the idea that Americans are racist at all. Their opposition is usually accompanied by extensive hand-wringing while chanting that classic refrain, “Things have changed since the 1960s!”
Fortunately, for the bigotry-phobic pundits, social psychologist Anthony Greenwald and political scientist Bethany Albertson have come up with a new strategy to disavow racism: the Reverse Bradley Effect.
What is the “Reverse-Bradley Effect”, you ask? It is a clever construct that racist white pundits love! It combines the bigotry of pollsters with the sexism inherent in the Privilege of Women’s Invisibility to argue that racism actually benefits African-Americans!
How, you ask? Well, during the historic Democratic primaries Obama actually outperformed the polls in 13 states. Since an African-American outperformed the polls in this one election, the Bradley Effect must have been reversed nationwide.
This irrefutably proves things have changed since the 1960s. Q.E.D.
But wait a minute, there was another unusual fact about this year’s Democratic primaries that the media is overlooking: there were actually TWO minority candidates in the primaries.
Remember that woman who was frequently standing on stage across from Barak Obama? The one that wasn’t Michelle Obama? I recently wrote a piece about her candidacy. That’s right, Hillary Rodham Clinton also ran for president this year. While the dudes were paying a lot of attention to Hillary’s gender earlier this year, they buried issue of gender as soon as she was out of the race [3].
So what would they have seen if they weren’t so busy marginalizing women? Well, it’s actually a little hard to tell, since no one ever investigated the role that sexism might have played in the Democratic primaries. Whenever sexism is brought up by, oh say, the creators of the “Reverse-Bradley effect”, the media made it a point to note it:
“The researchers attributed the inaccuracy of the polls to social influences. For instance, Greenwald said many women told pollsters they were voting for Hillary Clinton but ultimately cast their ballots for Obama.”
~Fox News (10/17/2008)
And then not speak another word about it.
Did Fox ever ask why people were not pulling the lever for Clinton? Hell no! The media simply uses the Privilege of Women’s Invisibility to write off ambitious women like Hillary Rodham Clinton without a second thought, as if the evaporation of female competition was part of the natural progression of the cosmos.
Patriarchal constants of the universe aside, what role did sexism play in the Democratic primaries? Well, I managed to connect some dots that the MSM ignored.
When it came to the purported “Reverse-Bradley Effect”, thirteen states were cited as evidence for the effect: Alaska, Delaware, Florida, Georgia, Maryland, Mississippi, Montana, North Carolina, South Carolina, South Dakota, Tennessee, Virginia, and Wisconsin. Now, can we connect these states in any way that would explain why Clinton would underperform in these states, without resorting to un-racism? Are there any signs that voters in these states might be prejudiced against having a woman as president?
Let’s first check to see if a woman has ever been elected to a major office in these states? After all, states who have seen any woman perform well in a high office are more likely to approve of one for the oval office. Conversely, states that have never elected women to high office may have more sexist cultures that inhibit the election of women in their states.
Well what do you know? Eleven of the thirteen states have never had a female governor! Of the two states that had female governors, Judy Martz of Montana had an approval rating as low as 20%! Over half of these states have never had a female senator.
And while there are only six states in the US that have never elected a woman to the House, half of them are on this list! [5]
Let’s look at another proxy for progressive/regressive attitudes towards women: contraceptive equity. If a state espouses the idea that women are brood mares, chances are they’ll oppose trailblazing women like Clinton who are themselves far from the birthing sphere. Conversely, a state that mandates access to women’s contraception alongside men’s Viagra may have an electorate that can actually comprehend women’s equality [6].
So how do the “Reverse-Bradley” states stack up? Of the thirteen states, nine do not have statutes for contraceptive coverage, and only one (Georgia) has a statute that mandates contraceptive coverage without “religious” exceptions. [7]
I’m sure there is plenty of other compelling intersections of sexism and racism in this year’s election, but we won’t see them because the media is making it a point to keep them invisible. The pollsters won’t find them either unless they get clotheslined by several hundred black women on their way to CNN headquarters with their latest post-gender, post-race analysis.
The truth is that many political polls do not use pure random sampling. Instead, they “target” polls at particular demographic groups (young people under 30, white male heads of household, etc.) and weight those groups based on their assumed proportion of the total voting population. In other words, if young women are believed to be 10% of the vote in a state, then they will only ever effect 10% of the polling result, no matter how many of them actually go out to the polls. [8] These polls will only be right when the current definition du jour of “likely voters” is correct. If the consensus on the “likely voter” definition is wrong (which it obviously was during the primary battle between Clinton and Obama) then the polls will be off no matter what. So the “Reverse-Bradley effect” is really a reverse mia culpa: the people were wrong, not the assumptions of pollsters.
Worse, I suspect that another ugly truth is lurking behind the pollster’s demographic groups: that pollsters do not regularly record the opinions of female voters.
I have asked some of my voting-age, swing-state, landline-owning female friends if they have ever been polled for an election. It turns out that none of them have ever been asked about their voting habits. Don’t get me wrong, several of their homes have been polled numerous times, but when a pollster calls, they typically say, “Can I speak to the man of the house, please?” One of my friends said that her father was not available, but that they can talk to her mother or herself. The pollster hung up. The pollsters know her father’s political party, stances on issues, even religious denomination, but they know nothing about the voting habits of the other two household members who are female and who vote.
Then again, this is purely anecdotal speculation based on talking to a few of my friends. I don’t have any hard data because no one has collected any. It would be nice if someone would conduct a poll on the matter, but I’m not holding my breath. The “Reverse-Bradley effect” already shows just how far reporters and pollsters will go to ignore anything that involves women.
“Reverse-Bradley Effect”? Give me a break!
POST-ELECTION UPDATE!
Now that the election is over and Barak Obama has officially won, let’s look back and see if there was any “Reverse-Bradley Effect” to speak of. Let’s use the exit polls as a benchmark, since they count actual voters as they leave the voting booth rather than relying on biased speculations as to which demographics will do whatnot. And since Americans are raised from birth to separate one human being from another using the constructs of gender, race, age, etc., I’m confident that exit pollsters will be able to accurately record the “basic demographic information” of the people they were talking to.
Checking the Gallup Polls for November 2nd, Obama led McCain 53% to 40% in registered voters, and 53% to 42% among their estimates of likely voters. Looking at the National Exit Polls, Obama led McCain 53% to 46%. White voters broke for McCain, and everyone else broke solidly for Obama. (I’d also like to point out that women across all racial and ethnic lines voted for Obama more often than the men.)
The exit polls unfortunately show that racism is alive and well in this country. But they also show little sign of any Bradley or Reverse-Bradley effects. The only discrepancy in poll numbers was the 4% of McCain voters that were not recorded by the election season pollsters.
I wonder what Dr. Greenwald would say about these results? As of December 2008, there is no new comments on his website.