Rearranging the Primaries

Two things are predictable in the months after a Democratic presidential election loss: sniping between the Democratic Leadership Council and the party's liberals -- even if there's very little basis for it -- and proposals to tinker with the schedule of Democratic presidential primaries to produce a more electable nominee. Steve Rosenthal of Americans Coming Together proposes that the primary schedule begin with the five states that were closest in 2004, to better prepare the candidate for the general election, an idea which happens to have the advantage of keeping Iowa and New Hampshire at the top of the heap. In contrast, Simon Rosenberg, a candidate for DNC chair, has proposed that "other regions" play a role in the early process, keeping Iowa and New Hampshire in the first group but ending their gatekeeping role.

I've only spent a total of maybe two frozen weeks in Iowa and New Hampshire during primary season, but it's enough to put me loosely in the "other regions" camp. But this is not an easy puzzle to work out. I've always been tempted to start in Missouri, arguably the state most representative of the various regional and demographic forces at play, but as with any large state, it would lose the retail politics aspect of Iowa and New Hampshire, and a television campaign would give even more of an advantage to money and the "wealth primary" that precedes the votes. I'm not sure this year's schedule was that badly designed: after Iowa and New Hampshire, Kerry didn't have the nomination sewed up, making South Carolina and the other Southern primaries as important as they've ever been. If either John Edwards or Wes Clark had consistently won those Southern primaries, he might well have claimed the nomination from Kerry.

Although it's also hard to argue that either one of them wouldn't have been eaten alive by the Bush slime machine just as Kerry was. Kerry had his weaknesses as a candidate, but it's hard to argue that, from the candidates available, Democratshad a stronger one. That's why the DLC doesn't have much of an argument (if you think Bush portrayed Kerry as inconsistent, imagine what they would have done to Joe Lieberman!), and also why the calls to rearrange the primaries aren't that relevant.

But the big fallacy of these schemes is this: there is absolutely no reason to think that doing well among Democratic primary voters in a given state is any indicator that a candidate would do well in that same state in a general election. Consider that it takes about 40,000 votes to come out of the Iowa caucuses with a decisive victory. It takes 750,000 votes to win Iowa in the general election. Those 40,000 hard-core, often unionized or college-town Dems are no indicator of a candidate's appeal to the vast number of less partisan, less organized and more conservative Iowans in the general. Or, one could start in the South: the key to winning the South Carolina primary, or most Southern primaries, will be African-American votes, whereas the key to winning even one southern state in the general election would be combining extraordinary African-American turnout with some appeal to Southern whites.

The problem with our last two Democratic nominees -- and this was even more true of Gore -- is not their inability to reach Democrats. It's their inability to get quite enough support from independents and loosely affiliated voters to overcome the solid Republican base. To solve that problem, there's only one solution in the primaries: an open primary. Yes, that would mean letting people who aren't officially Democrats participate in choosing the Democratic nominee. But we live in a time when such affiliations are looser and more flexible. Democrats and independents who had a strong positive feeling about John McCain in 2000 helped him win the New Hampshire Republican primary that year; if that hadn't happened, the same independents might have swung the Democratic primary for Senator Bradley. Without arguing that either McCain or Bradley would have been better candidates for their parties in the fall, because each had their own weaknesses, it is certainly true that either one of them would have had a broader appeal, going beyond their party's core.

There is certainly potential for mischief in an open primary, as in Gray Davis's effort in California to select his own right-wing opponent in 2002. But the potential benefits of finding a candidate who can reach beyond the party's base outweigh the potential mischief. Here's my proposal: Let's get rid of party registration altogether. If you want to participate in the Democratic party's selection process, for that moment you're a Democrat, and vice versa. For the Democratic party to reach the broadest population, it needs more than just people whose loyalty is deep enough that they declared their allegiance formally and months in advance.

As the next step, I would eliminate voter registration altogether, but that's a fight for another day.