Mark Silk pointed me to today's CBS poll showing that white evangelicals support John McCain over Barack Obama 58 to 24, with 15 percent undecided. As Mark points out, if those 15 percent break 50-50, Obama's 31-32% share would be a major improvement on Kerry's 21% share in 2004. A ten point gain among a bloc that makes up approximately a quarter of the electorate seems like a heck of a payoff.

A couple of caveats, of course: That 50-50 break is a big if -- there's a case to be made that the undecideds could break strongly in either direction -- and it isn't clear how white evangelicals are defined here, so we might have an apples-to-oranges issue.

However, the potential gain here could be a game-changer. It reminds me of Bush's 9-point gain among Latino voters in 2004. His campaign saw an opening with a solid Democratic constituency, reached out, made big gains, and set the precedent that the Democrats couldn't take them for granted. (As Latino evangelicals are demonstrating this year.) If the Democrats pull this off among white evangelicals, that's roughly 2.5 million votes picked up, and a couple of swing states sewn up

It bears mention that Bill Clinton won one-third of white evangelicals in 1992 and 1996, so Obama's potential success isn't exactly unprecedented. But Clinton no doubt got a boost from actually being a Southern Baptist running against classic country club Republicans. The real question, in my mind, isn't whether the Obama campaign's outreach is worthwhile. It's whether future Democratic candidates' success among white evangelicals will depend on their personas as opposed to their strategy and infrastructure. Time will tell.