The Reverse Bradley Effect

15 02 2008

After comparing polling data on the eve of Super Tuesday and the actual election results, two scholars from the Pew Research Center concluded that race, due to the Bradley effect and the reverse Bradley effect, can exaggerate support for a candidate in either direction depending on the make up of the electorate.

Early analysis of primary counts and polling data from the final week of the campaign indicated that pre-election polls exaggerated support for Sen. Barack Obama in two states with relatively low black populations –California and Massachusetts. But the reverse was true in Alabama and Georgia, where blacks make up a larger bloc of voters. The same phenomenon is seen in the earlier primary states of New Hampshire and South Carolina.

Obviously, there is some truth to this. Many polls did have Obama beating Hillary Clinton by several percentage points (between 7 and 13) on the eve of the New Hampshire primary, and yet lost by 3 points. New Hampshire is more than 90 percent white.

In South Carolina, Obama led in the latest polls before election by 7 -15 percentage points, but in the end beat Clinton by a 28 percent margin. South Carolina is 29 percent black, which is more than double the national average, and 68 percent white. Plus, African-American voters make up more than half of all Democratic electorate. So it’s easy to see how New Hampshire demonstrated the Bradley effect and South Carolina the reverse Bradley effect.

The Bradley effect refers to the phenomenon in which white respondents in opinion polls overstate their support for a black candidate relative to the final election results to avoid appearing prejudiced. Such outcomes have caught the attention of pollsters since Tom Bradley, a black politician, saw his “lead” in the 1982 California gubernatorial race opinion polls eviscerated on election night against his white opponent.

Pollsters and pundits have also observed the same effect in other regions and during different eras. As the phrase implies, the reverse Bradly effect refers to when support for a black candidate is understated in opinion polling data. (The understatement of support might be due to an under sampling of black respondents relative to white in opinion surveys.)

But neither the Bradley effect or its reversal explains why Obama defeated Clinton in Maine 59 to 40, a state that is more than 95 percent white even though Clinton had a sizable lead in Maine at one point.

Or why Obama beat Clinton in Minnesota (85 percent white and 4.5 percent black) by a more than 2 to 1 margin, despite trailing the former first lady in a poll fewer than 10 days before the election by 7 points.

Or why in Missouri, where polls had Clinton ahead by almost 6 points, but lost to Obama in a squeaker. Missouri is more than 82 percent white and 11.5 percent black. Missouri, Maine, and Minnesota demonstrate that even in states that are overwhelming white can often understate Obama’s support just as much as states with relatively high population like South Carolina.

This is not to say that the Bradley effect or the reverse Bradley effect does not exist or that Obama is somehow immune to either. It does, however, suggest that there is probably a more nuanced explanation for how race might influence election results.

John Derbyshire at the National Review‘s blog The Corner came somewhat close to explaining Obama pattern of success by the Illinois Senator does well:

In the places, black voters will turn out for him in droves [and in places where] people feel little racial tension, will take the candidate on his merits, and be pleased by his uplifting rhetoric.

However, in places where the proportion of black citizens is big enough to cause tension, but not big enough to swing an election, the white majority will not support Obama, and he will do badly.

The first part I think is true. The second not so much. As a reader of this blog noted in a recent email to me, “it seems clear to me that the data suggests when white voters feel on-the-ground competition between their candidate and the candidate of choice for minorities, white voters will choose the white candidate.” This of course speaks to the very local nature of racial tension in races involving candidates of color, particularly in contests for Congressional seats.

Its harder to make this point in state wide elections in states like Missouri, which is considered part of the upper South and the Midwest, and has enough black folk to sway an election. Hence why I take issue with the second part of Derbyshire assertion since it does not account for why Obama beat Hillary Clinton in Missouri, where the number of black folk there reflects the national average, helped put him over the top against Madame Inevitable. According to Derbyshire’s logic, Obama should have lost badly there. So why didn’t he?

I am not gloating about the experts being wrong so much as I am simply scratching my head at all of this trying to make sense of it all, just like they are.

Note: In a previous post I noted the non-existence of term describing a black candidate that significantly exceeds expectations in the polls on the eve of an election and win by a large margin. I obviously now stand corrected.

Note II: The reverse Bradley Effect actually refers to when black respondents understate their support for a black candidate so as not to sound as if they are supporting that candidate out of racial solidarity.




Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: