Edwards deputy campaign manager Jonathan Prince makes a few points in response to my previous item. They're as follows (I'm paraphrasing):
1.) There's no historical example of someone getting killed in between wrapping up the nomination and the convention, at least not under analogous circumstances. True, Dole took a big hit from Clinton in 1996. But Clinton was an incumbent president with no opponent, and in any case Clinton himself (like Edwards) was within the public financing system.
2.) Earned media ends up mattering a lot more than paid media, and, as the nominee, you have nearly unlimited access to earned media. The real question is how you use your earned media. John Kerry, for example, was flush with cash in 2004. But he didn't respond aggressively when he was swift-boated, and that cost him the race. (Prince also points out that the swift-boat ad was a mere $1.5 million buy--hardly an example of drowning your opponent in money.)
3.) If Edwards ends up winning the nomination, he will by definition have demonstrated that it's possible to beat opponents who've massively outspent you with a compelling message.
4.) The pre-convention mismatch is something you hear about a lot from Clinton and Obama moneymen, who are basically flattering themselves by overstating their own importance as a factor in the race.
I think these are all fair points. I'd say three things in response:
1.) Clinton/Dole is a somewhat disconcerting precedent, if not so damning, even though the parallels aren't perfect. My recollection from reading former Clinton strategist Doug Schoen's memoir is that Clinton killed Dole with a lot of below-the-radar ad-buys in key swing states. By the time Dole got up and running after the convention, Clinton had more or less defined him in these places. So I agree that the national media narrative is influenced more by earned media than paid advertising. But 1996 suggests it's not just the national media narrative that matters; state-level narratives matter, too. And a big advantage in money would allow the Republican nominee to define the race in a lot of key states.
2.) It's true that, if Edwards wins the nomination, he will have proven that message can triumph over money. But voters, donors, and the media won't know this soon enough for it to matter. After all, at the time they're casting their votes, cutting their checks, and writing their stories, Edwards will not yet have beaten Clinton and Obama. The proposition will still be unproven at that point.
3.) Even if it's possible for the more compelling messenger to win despite being badly outspent, you'd still rather have that extra hundred or two hundred million between March and the convention. It's also possible for a well-coached team of unathletic no-names to beat a traditional basketball powerhouse like UCLA, but I'd still take UCLA given the choice.
Bottom line: I'm open to the idea that Edwards won't be sunk if he gets the nomination and can't wriggle out of the federal spending limits. Maybe the biggest problem he faces is the perception that it's going to be tricky, which makes his case tough to make in the primaries. (And which is obviously why we're having this exchange.)