I've been thinking a lot about the problems with our current system of frontloaded primaries, and the more I think, the more the problem seems to be the fact that there is no disincentive for a state to have its primary as early as possible--to be relevant, you need to hold your primary before there is a clear winner, and therefore, there is every reason to hold your primary as early as possible. This, of course, will probably just magnify the already disproportionate influence held by Iowa and New Hampshire.
Obviously, the best reform is to just break the monopoly of these two states, and for the national party to force a more prolonged schedule. But it is getting increasingly clear that this simply won't happen.
But what if you simply made one reform, one that party insiders might likely not enjoy, either, but wouldn't particularly affect any states disproportionately?
How about simply making it so that a candidate, once (s)he has dropped out of the race, cannot pledge their candidates to another candidate on the first ballot? The practice is pretty undemocratic--I go to the polls and vote for Gephardt (for example), I elect a delegate, but that delegate ends up voting for Kerry, due to the dynamics of the election in a state after mine. My delegate should be forced to vote for Gephardt, regardless of what happens later.
This, however, wouldn't keep delegates from dropping off of the ballot in later states, and that is the key. This would give states an incentive to have later primaries--the first couple of states will have ballots cluttered with candidates that will eventually fall off of the map. Hence, their first ballot delegate counts will contain large numbers of delegates that will go to non-viable candidates. They will still have a great deal of influence, as they determine who has early momentum, but they will not invalidate the later states, who will only have the frontrunner and the challenger on their ballot, and thus, will have nearly all of their delegates count.
Thus, you will have a set of states that 'vet' the candidates, and a set of states who actually determine the nominee. This would also make the 'retail politics' angle of IA and NH more relevant--they are processing the large number of early candidates for later states' consideration. The only real downside that I see in this is that it would greatly increase the chances of having a brokered election. And I guess it might greatly increase the chances that a candidate drops out early, and thus make IA and NH even more important. But I do wonder what effects a reform like this would have.