Over at Cliopatria, K. C. Johnson started an interesting discussion of the current debate over “voter disenfranchisement” and the Democratic party.

This post began life as I was composing a response to a comment by Nicholas Norden there: “It's discouraging that the Obama campaign has characterized the popular vote metric as legitimate.” But my reply got so long, I decided to put it here, instead.

The choice of metric depends on what one hopes to measure.

One purpose of the Electoral College was to choose a president by a weighted vote of the states as opposed to the general populace. By the time of Andrew Jackson's presidency, the ideal election was one in which the man with the majority--or at least plurality--of the popular vote as well as a majority of the electoral votes became president.

Concerning the nomination, the convention system, with its interlaced caucuses and conventions, emerged at the beginning of the Jacksonian era as the most democratic alternative available in a time of comparatively slow transportation and communication. It seemed and probably was the best method available to measure on a state-by-state basis the popular support for candidates within each party. It worked relatively effectively but was also prone to cronyism.

In the Progressive era, with its rapidly improving communications and a growing middle class distrust of party machinations, the primary began to emerge as a more democratic alternative for delegate selection. However it was not until the late 20th century that it had become the more legitimate approach for most people in most states. The remaining caucus systems retained a certain legitimacy, but I honestly believe that they have been viewed as"quaint" by people who do not participate in them.

At least until lately.

Of course, even primary systems differ. Does one wish for a winner-take-all system that echoes the less-than-democratic electoral system? Or should a primary apportion delegates according to the popular vote, an approach that is more democratic but tends to slow the selection process? Again, what is the primary supposed to measure? The most popular candidate? The one most likely to win in the most powerful states? Something else?

I think the American people are shifting more and more toward a national voter metric--the candidate with the most votes should win. Whether nomination or election. Period.

Is that wise? Or, at least in the election process, should we still give smaller states a"leg up" by weighting them more heavily. There are good arguments for that. In fact, I tend to agree with some of them, but the cost or a minority president can be high. In the case of the Bush Administration's first term, his minority of the vote, along with the Supreme Court's clumsy intervention, exacerbated an arrogance concerning executive power already in place. This was because Bush realized, correctly, that the one thing he could never, ever do was question his own legitimacy. The minute he did so, any hope of governing effectively as a conservative would have been damaged badly.

Right now, the tendency is to look upon the Bush administration as the acme of presidential horrors. But even a more competent president would have been faced with tough choices if the electorate, based on the popular vote, had chosen someone else.

So what do you want to measure in a primary and in an election?

Postscript (7 April, 9:20 am CT): My thanks to Sean Wilentz for providing, albeit accidentally, an argument on why the Democratic primary should be a state-by-state winner-take-all system in order to better pick a successful presidential candidate.

As usual with Wilentz, when he is functioning as a politician, one has to cut through the self-righteous polemics. But, however self-serving, the point he makes is worth considering. The proportional system may be a better measure of what Democrats want than a winner-take-all system is, but is it a poorer measure of which Democrat may win?