Alec MacGillis has an interesting post yesterday arguing that the national nature of the GOP presidential campaign this cycle hint at a "declining legitimacy" of Iowa and New Hampshire.
That may be true if the justification for the current system is primarily based on the advantages of retail politicking. It's clearly in decline, for this cycle at least (and MacGillis quite properly is cautious about extrapolating out from current trends), or at least if we assume that there won't be a late Santorum surge in Iowa and a Huntsman surge in New Hampshire. Which could, of course, happen.
But I don't believe that retail politics is the real reason to have the current system. No, what is really important is having a sequential system of primaries and caucuses rather than a national primary. And the case for sequential is certainly a lot stronger after this year's clown show during the invisible primary. At least, that's the case if you believe (as I do) that what we've really seen is that practically any candidate can get a temporary bubble.
The problem, of course, is that if that bubble happened with just the correct timing for the national primary, a party might well be stuck with a Newt Gingrich, Herman Cain, or even Donald Trump. Indeed, that's basically what happened in last year's GOP primaries in Nevada and Delaware; a candidate who couldn't hold up to serious scrutiny peaked at the right time and wound up winning a one-shot primary.
Sequential primaries are a good way to prevent that. And if you're going to go sequential, then sticking with the traditional small states at the beginning of the process seems as good a plan as any, whether or not the supposed benefits of retail politics are real.