I think you're absolutely right, Kos, that one possible scenario involves Dean and Kerry duking it out to Gephardt's benefit, and I'm sure there will be enormous pressure placed by the media, the party, and the CW to boil the primaries quickly down to two candidates. I frankly think it's more fun, however, to speculate about the many ways in which the CW may be dead wrong, and I suspect some of these scenarios are as likely as any others we've heard.
Here are some CW "rules":
1) Candidates who fail to meet fundraising expectations will bag out early.
-- Don't count on it. Lieberman does seem to have more to lose with the Q2 numbers (with Edwards right behind him), but no one besides Moseley-Braun will drop out before 2004 over money. Their egos won't let them. Lieberman will count on WMDs showing up in big numbers or a terrorist attack--something that could hamstring others and catapult him into the game. He wouldn't mind it either if Gephardt said some more lamentable things like the Executive Order comment. Everyone hopes Dean flames out, victim of his own success. They want more MTP moments, and they'll be floating damning comments with the media. They all see his loss as their gain, and they're right.
2) Poor performances in early primaries will force the race down to two candidates.
-- As with #1 above, a lot of the candidates are going to hold out hopes for an 11th-hour miracle to catapult them. A Dean flameout may become their favorite dream, but we may hear lots of creative rationales for staying in the race. Lieberman could do poorly in most of the early primaries and still declare that Super Tuesday victories in New York and California are going to get him back in the running. (Of course, if he runs out of money, that fantasy will die.) Graham may stick it out throughout the primaries, in effect becoming an issues candidate (and helping maintain his viability for VP). Kerry's and Gephardt's warchests may help them limp along longer than they should. And Dean, as you point out, may not be allowed to drop out, if his volunteers have anything to say about it. Texas, Washington, and California, to name just three states, seem to have VERY active volunteer networks in place for Dean, and I think we'd hear promises from these groups to get him back on track, if he fares poorly in many early contests.
3) Clear victories will help frame the race.
-- What if Iowa's results look something like this: Gephardt 24%, Dean 23%, Kerry 12%, Edwards 10%, Lieberman 10%, Kucinich 10%, Graham 7%, etc.? Gephardt's a "loser" because he isn't a clear winner, Kerry's a loser in comparison to Dean, and Graham's a loser because he's in the back, but everyone else can spin victories of various sorts. Then they could all get shuffled in New Hampshire. Call it golf scoring: We may have a narrow winner in each early primary, followed by a bunch of guys tied for second and third. And because there is no clear frontrunner, it'll be harder to discern (or concoct) a "message" from each primary. McCain's New Hampshire spanking of Bush was widely interpreted as a message of his strength among Independents, but Bush's solid Republican victory in South Carolina became the door slamming shut: McCain had had his moment, but the presumptive winner was back on top. That's not a viable story line in a race with no clear leader, and results from the early primaries could reasonably leave people wondering who among Dean, Gephardt, and Kerry is going to take it. (I think it's less likely that others could move themselves into consideration for frontrunner, but it's certainly possible.)
My own state of Virginia could be wide open come February 10. If the race hasn't yet narrowed, I could see a finish like the one I posited for Iowa (with the candidates shuffled around, and perhaps with Kucinich and Graham farther out of the picture). Which previous primaries would necessarily predict how Virginians would vote? We have a huge tech sector in NoVa, a massive military presence in Hampton Roads, a large minority population, and a Southern white vote. Virginia Democrats have also been effectively muted in Presidential politics for a long time, so there are going to be many of us swarming around. And results out of Virginia could only muddy the outlook further.
4) Candidates who lack key early wins will be sent packing.
-- If we get a lot of "golf scores" in early primaries, the field will remain large. The longer it stays large, the more likely we are to get more golf scores, and the more likely we are to hear about the possibility of a brokered convention. Why drop out if no one else is heading for a majority of delegates? We could see stories calling Lieberman "the second-place kid" or Edwards "Mr. Patient." Remember that Northern California poll of a few weeks ago? Dean, Kerry, and Lieberman were tied for first at 13%. February 2004 polls coming out of New York and California could predict similar results.
5) Voters will choose between moderate and liberal candidates.
-- Some will, but many won't. Gephardt is getting spun as more moderate than Dean, but many of his positions are far to the left (e.g., his health care plan). Kerry is fighting hard to avoid anyone to his left on women's issues, but he's Mr. Moderate on defense. Dean is getting called everything, and most of his supporters seem not to care what label is used. Look, most Democratic voters are scared to death of another four years of Bushco, and they will be seeking someone who can beat him. If given the choice between the guy who best matches their positions and the guy who seems most electable, which way are they going to vote?
On the other hand, crappy Q2 numbers could seriously undermine chances for Lieberman, Graham, and possibly Edwards. Fundraising and polls may make it apparent well before January that these guys have no shot. Kucinich may finish stronger than several of them, simply because he's out on the Left. If Dean, Gephardt, and Kerry are widely perceived as the only viable candidates before the first vote is cast, much of what I have written above goes out the window. And by the way, I'm not trying to play favorites here. Graham and Edwards are the two candidates who don't seem to be catching fire in early polling, and Lieberman is reportedly struggling with money.