Reliable polls taken in the last 2-3 days showed Clinton beating Obama by an average of 5.8%. But the day before the primary I predicted the spread would be 9.5%. Current reports place it at 9.3%, with 99% of precincts reporting. Why did I get it so right when the polls got it so wrong?
One reason is dumb luck. Accurate predictions are difficult with numbers as fuzzy as polling data. That said, there are other, more ominous reasons why I did better than the pollsters. The one that should be of greatest concern to the Democrats is the Bradley Effect, where white voters are reluctant to tell pollsters that they won't vote for a black candidate. While the Bradley Effect is highly controversial, the Texas and Ohio results persuaded me that it's real. So I increased Clinton's margin accordingly.
It didn't have to be this way. In the first few months of the campaign, a great many voters didn't seem to perceive Obama as "black." Pundit talk that he was somehow "postracial," possibly as a result of his multi-ethnic parentage, seemed to reflect itself in public reaction to his candidacy. There would have been a Bradley Effect in any case, but it might have been smaller than it turned out to be yesterday in Pennsylvania.
Which leads me to the second reason I called this race accurately: The Clinton campaign has been successful in ghetto-izing Obama as a primarily "black" candidate. Bill Clinton's Jesse Jackson remarks were the opening salvo of a war to do exactly that, and it has worked. Even Jon Stewart's joke when interviewing Obama on Monday, while funny -- "Do you plan to enslave the white race?" -- played into the "black candidate" angle.
From what we seen of the Clinton psychology, and Bill Clinton's in particular, it's easy to understand why they did it. Obama was getting heavy support from black voters, yet presenting himself as "postracial" to the white electorate. That's having your cake and eating it, too. To people as resentment-driven as Bill and Hillary Clinton, the temptation to go after him on race must have been irresistable. (Thus we get Bill saying he was the race card victim on Monday - adding "I don't think I should have to take any shit about this" - and then denying it on Tuesday, despite the audio tape. What he was really saying was "I should be allowed to get away with this shit.")
So the difference between 5.8% and 9.3% boils down to what I said when Geraldine Ferraro made her comments last month: The Ferraro strategy was deliberate, it was coordinated, and it will work. Well, it has. And it's resulted in lasting damage to Obama -- damage that the GOP was in no position to inflict for itself. Why? Because, as Democrats, the Clintons and their surrogates have been able to make racial arguments that would have been considered unacceptable coming from Republicans. Assuming that Obama will still be the nominee, the Clintons have done McCain's dirty work.
Don't get me wrong: There would have been a Bradley Effect anyway, though I don't think it would have been this big. And other factors hurt Obama, including ABC's attacks during the debate and the tenacious loyalty Clinton supporters have toward their candidate. Obama underestimated the ferocity of the resistance he was going to face from both the Clintons and the media, and didn't build a strong enough firewall against some of these personal attacks. However vacuous or distracting you think issues like Jeremiah Wright and Bill Ayers are, it was inevitable that they would be used against him. His failure to adequately plan for that is hurting him now.
Despite what others say (like Slate's "Hillary Deathwatch"), I've never put Hillary's chances for winning the nomination at anything less than 30%. I still don't. There are powerful forces out to eliminate Obama from this campaign, and Clinton's determination is great. It's a mistake to underestimate either of these factors.
Here's an impression I can't back up with data: If Hillary had run a clean campaign, she would still have won PA. The margin would have been closer to 5 points than 9, but it would have left her in a position to make the electability argument she's been pushing for months. Her question -- "Why can't he close the deal?" -- is a legitimate one. Superdelegates and Obama supporters might be taking a second look at Hillary after yesterday's results, but her campaign has created too much bitterness for that.
Now it's too late: She's inflicted some serious wounds on Obama, but the way she's done it has made it all but impossible for superdelegates to accept her as an alternative. His supporters are too angry over her tactics to accept her on the basis of electability alone. Obama emerges from Pennsylvania damaged, but choosing Hillary instead would shatter the party. (Even the New York Times finds her campaign strategy "mean, vacuous, and pandering.")
Ironically, a smaller margin in Pennsylvania would have helped her more than this one did, if she had won it cleanly. Instead she's won a genuine -- but Pyrrhic -- victory, one that doesn't advance her chances for the nomination. And the damage to the party's November prospects is deep and lasting.
Final numbers: 1,258,245 Clinton, 1,042,297 Obama, total votes 2,300,542. That's 54.7% Clinton, 46.3% Obama (slight rounding).