The Second Presidential Debate--Preview

The Lay of the Land. Contrary to most media and pundit opinion, the facts of the political terrain going into the second presidential debate are neither knowable nor simple. What's least known is the issue that most needs to be known--where public opinion stands four weeks from the presidential election. Many of the tracking polls (Rasmussen, Gallup, GW, Research 2000), Obama has a 7-8 point lead. But other polls (Zogby, Hotline, CBS, Democracy Corps) only have Obama up by 2-3. Given a 2-3% margin of error, the latter group of polls could be read as portraying the election as tied. Rassmussen and Gallup are generally the most respected of the polling operations, However, there's a big five to nine point polling discrepancy and there's no principle that would allow one to tell which poll is right and by how much.


That's the known discrepancy. What's unknown is the impact of the "Bradley Effect" where white voters tell pollsters they're going to vote for a black candidate in order to not appear racist and then vote for the white candidate. As a result, African-American candidates like Barack Obama can find that the polls overestimate their vote by as many as 10%. In the 2008 election season, Obama has found that some polls (Iowa, North Carolina, etc) accurately capture his vote, but has seen other states (for example, New Hampshire) where polls have over-estimated his appeal along the lines indicated by the Bradley Effect.


To what extent is the current polling over-estimating Obama's vote as a result of the Bradley Effect?


Nobody knows.


However, the other major unknown--the impact of Barack Obama's voter registration efforts and turnout drives--cuts in favor of Obama's candidacy. States like Pennsylvania and Colorado report huge increases in the number of people registered as Democrats. But the question is what percentage of those newly registered voters (and young voters) are going to show up and vote on election day. The Obama campaign believes that "new voters" are going to make a big difference and believes that it has evidence from their own canvassing to back up their convictions. But the public opinion polls don't have any reliable way to pick up that impact and thus the final impact of Obama's registration efforts has to be considered unknown at this time.

What Does This All Mean for the Debate? The conventional wisdom is that McCain is in a difficult box because of Obama's recent rise in the polls from 3% behind to 7-8% ahead. The idea is that McCain would like to "take the gloves off" but that going on the offensive runs the risk stimulating a backlash among undecided voters who don't like aggressive partisanship. The conundrum is heightened by the fact that conservative commentators and voters are eager to see McCain attack at the same time that the debate is being held as a townhall meeting with an audience of undecided voters who generally don't like partisanship.

But a lot of McCain's strategy would depend on whether he thinks he's ahead because of the Bradley Effect, closer than the polls indicate because of the Bradley Effect, or behind because he isn't considering the Bradley Effect.

McCain could quite plausibly think he's ahead if he takes the polls showing Obama with a 2-3 point lead as the most accurate and estimates that the Bradley Effect would give him another 5-6 points. That gives McCain a 2-3 point lead in which case McCain could very well take a disciplined, moderate approach to tonight's debate in which he lays out his case but avoids things like calling Obama "naive" that raised McCain's own negatives after the first debate. In this scenario, McCain would be patterning himself after Joe Biden's disciplined but forceful performance in his debate with Sarah Palin.

Who knew that Joe Biden would grow into a campaign role model?

Oops. I'm running out of time. If McCain estimates that the Bradley Effect makes the race closer than the polls indicate but still leaves him behind, I think he would hue to the traditional Republican line of attacking Obama as a liberal, claiming that Obama is out of touch with the "real America," and perhaps using Obama's associations with William Ayres and Jeremiah Wright to emphasize Obama's cultural marginality.

Such a strategy could have the effect of making a traditionally effective Republican appeal to swing voters while also reminding undecided voters that Obama is black and reinforcing the Bradley Effect.

That's what seems most likely to me.

If McCain assumes that he's really 7-8 points behind, he's in just as much a box as the media says. In that case, McCain's best approach would be to unveil some sort of big policy surprise, push Obama really hard on liberal vs conservative oppositions, get in a lot of zingers about Obama's "questionable associations," and hope that Obama falls apart.

The Obama campaign believes Obama is well ahead and thinks Obama could win in a landslide. But I wonder if the Obama strategy will be for Obama to push hard against McCain's association with the Bush administration and lobbyists as a way to show how tough, determined,. and presidential Obama is. That would allow them to make their argument to undecided voters in case the Bradley Effect is indeed making the election closer than it appears.

Given the uncertainties over the state of the election, the McCain and Obama campaigns could reasonably adopt any number of strategies. It will be interesting to see what they actually do.