You almost have to feel sorry for Democratic voters. The media has set a trap for them. If Barack Obama wins a state, they imply that it was because voters were too sexist to elect Hillary Rodham Clinton. If she wins a state, the talking heads say, it's because voters were too racist to elect Obama. Who the heck are voters supposed to vote for, then? Someone has to lose.
The media is so determined to find racism and sexism around every corner in this race that when they can't find it, they invent it. The reporting on Obama's loss in New Hampshire last week was one of the most stunning and widespread media hatchet jobs I've ever seen.
The story, as most of the national TV news stations and many of the country's big daily newspapers told it, was that the reason the polls inaccurately predicted Obama would win the New Hampshire primary by 10 points was the so-called "Bradley effect." That's when voters tell pollsters they are going to vote for the black candidate because they are too embarrassed to admit they wouldn't pull the lever for an African-American. Then, in the anonymity of the voting booth, they vote for the white candidate instead. The result is a double-digit difference between the vote percentage for a black candidate predicted by pollsters and the actual outcome.
Self-righteous faux journalists clearly loved the idea of the Bradley effect storyline to explain Obama's tiny 37 percent to 39 percent loss to Clinton in New Hampshire. For three days after Obama's New Hampshire loss, they chattered about Obama's "race problem" and the Bradley effect without apparently bothering to read the polls they were talking about.
"Methinks Paleface speak with forked tongue," MSNBC host Chris Matthews roared into the camera as he railed against racist poll respondents who supposedly lied to pollsters. "You hear me? Forked tongue!"
Matthews suggested that if pollsters used someone who sounded like Archie Bunker to conduct their polls, they would get more honest answers.
Matthews went on about how he was moved to tears by one of Obama's recent speeches and how sad it must have been for Obama to be led to believe he was going to win New Hampshire by racist poll respondents only to suffer a defeat. Obama should just accept the reality of a racist America and automatically subtract 10 points from his poll results, Matthews advised the campaign.
The problem, as members of the media would have realized had they actually read the polls leading up the New Hampshire primary, was that voters didn't lie about their intentions to vote for Obama, as they would have had to do to cause the Bradley effect. Six of the seven major polls on the race (see www.realclearpolitics.com) conducted in the days before the primary were dead on accurate about Obama's final vote percentage. They nailed it, predicting his final percentage at between 35 and 40 percent. Obama's 37 percent finish put him smack in the middle of the polls' three to four point margins of error.
What pollsters actually blew was Hillary's vote total. All seven polls underestimated her vote percentage by double digits, predicting she would get between 15 and 20 percent of the vote when in the end she got 39 percent.
Had the faux journalists bothered to read pollsters' explanations of what actually happened to throw their polls off, explanations prominently posted on Web sites like Gallup.com, they would have understood what actually occurred.
These pollsters were all in agreement -- there was no Bradley effect.
According to pollsters, on election day, most of the voters supporting Sen. Joe Biden, who had recently dropped out of the race, shifted their votes to Clinton. Almost all of the five percent of voters who were undecided in the final polls voted for Clinton as well. Edwards also lost a couple of percentage points to her that he had been predicted to get in final polls.
Then there were the women. In Iowa, Obama won more of their votes than Clinton did. In New Hampshire, they backed Clinton following the tearful incident that got so much attention.
And women mobbed the polls. A whopping 57 percent of those who voted in New Hampshire were women. That's not unusual in New Hampshire, which has a long tradition of voting for women and high female voter turnout.
The last-minute shifts were missed in New Hampshire, pollsters say, because no Monday polls were taken before the Tuesday primary, which is highly unusual.
In its rush to create a story line around Obama's race, the media overlooked the fact that in New Hampshire, which is 96 percent white, he got more than twice the votes the highly Caucasian candidate John Edwards did. And Obama barely lost to Clinton.
None of this will matter going forward, I'm sure. The next Obama loss will no doubt have a racial storyline tacked onto it, regardless of what voters actually do. That's a shame. From the beginning, Obama has downplayed his race and amplified his message. He obviously wants to be much more than just the black guy in the race. Too bad the media won't let him.