Chris Matthews was on the news this morning pretty much blaming race for Hillary’s surprise victory last night. This is understandable for two reasons – one, he hates Hillary; and two, if they can’t trust the polls that they soe dearly depend on, political television pundits will begin to show that they don’t have any greater insight into the electorate than anyone else. (This is obvious to many, but there is always the nagging fell that “Hey, they’re on TV, they must be smarter than I am.)
The phenomenon of black candidates underperforming at the polls compared to the pre-election polls is alternatively known as the Bradley effect (Tom Bradley when he ran for Governor of California in the 80’s) or the Dikins effect (for David Dinkins when he ran for mayor of NYC in 89) or the Wilder effect (when running for Gov of Virginia in 1990) or the Gannt effect (when he ran for the US Senate against Jesse Helms in the 80’s.)
But I think that there needs to be more in-depth analysis of a possible Bradley/Wilder/Dinkins/Gannt Effect (does that satisfy everyone – lets call it BWiDGE for short) in NH before it becomes the CW.
First of all, someone who understands this stuff needs to compare the way that independents were factored into the pre-election polls. An independent is more likely to have decided at the last minute that a vote for McCain was more important than a vote for Obama, maybe because Obama was favored to win, or maybe because Hillary beating Obama was more palatable than Romney beating McCain. Looking at turnout models in these pre-election polls versus actual turnout will be crucial to determining if race truly played a roll in the Hillary surprise. My guess is that the actual turnout will more closely mirror that that was used in the Suffolk University poll that had Obama’s lead within the margin of error than the polls that had Obama leading by double digits.
Importantly, both Democrats have a vested interest in proving that NH was not an example of the BWiDGE. For Hillary, it would cheapen her historic victory. (She was the first woman to win a Presidential Primary.)
For Obama if the BWiDGE become the CW, it will become a self-fulfilling prophesy. If Democratic voters think that Obama will always under perform, it makes him less electable. Democrats right now think that Hillary and Obama are pretty much equally electable (both have major pluses and major minuses on the national stage). But if Democrats think that BWiDGE is a factor in Democratic primaries, they will assume that it will play an even larger role in a national election and not want to risk Obama losing just because he is black.
BWiDGE may be the source of Hillary’s victory last night, but a big part of me thinks that it is more likely that this is an easy excuse for the political class to use, instead of demonstrating once again that the Emperor’s new clothes aren’t exactly what they would tell you they are.
UPDATE - Here
is one possible theory: Dodd and Biden supporters weren't really paying that close attention.