One Last Thing: Voting lines lead right to superdelegates

Let's begin with one paramount fact: The superdelegates will determine the Democratic nomination. Neither candidate will capture enough pledged delegates to win without them. The contest between Barack Obama and Hillary Rodham Clinton will not be settled by purely democratic means.

The common misconception is that superdelegates are basically like jurors: They listen to each side's case and then apply a specific set of instructions and a strict legal framework to make their evaluation. But superdelegates don't work that way. They're more like the College of Cardinals: They decide by feeling their way through moral and political, not legal, claims. And as in electing a pope, there is no right or wrong. At some point the superdelegates will send up the white smoke, and their decision will be, by definition, legitimate.

So instead of debating the "legitimacy" of their decision, it's more useful to ask how they'll do the deciding.

I have long believed that at the end of the day, the superdelegates will be swayed more by the popular vote count than by the pledged delegate count. What complicates matters is that there will be at least three different versions of the popular vote: total Democratic National Committee-sanctioned votes, and then votes including Florida and/or Michigan.

After Tuesday's primary, Clinton leads Obama in the most inclusive category, which includes the Florida and Michigan results. She has now received more primary votes than anyone who's ever run for president. If, by the end of the primary season, Obama leads in this count, then he will almost certainly be the nominee. It is very possible, however, that Clinton will retain this lead and also capture the lead in the popular vote count in Florida, but excluding Michigan. And there is a smaller, but real, chance that she will lead in the popular vote minus both Florida and Michigan.

Which leads us to the next problem - Puerto Rico.

If Clinton is to overtake Obama in the popular vote that excludes Michigan and Florida, she will need a big victory in Puerto Rico's primary. You didn't know Puerto Rico had one? Well, it didn't - until last month, when it swapped its caucus process for a primary election.

No one knows what to expect in that contest on June 1. Puerto Ricans are quite participatory. In their 2004 government elections, 52 percent of them voted, which translates to two million voters.

But let's say, just for the sake of argument, that Clinton were to win such an overwhelming victory that the Puerto Rican vote became the margin that put her ahead of Obama, even without Florida and Michigan. What would the superdelegates make of that? Remember, Puerto Ricans don't have a say in the general election; they only get to vote in the primary.

If it looks as though I'm stacking the deck against Obama, it's because I am, but only to illustrate a point: Obama is in a strong position and may well be the nominee. But he needs to win at least one of the popular vote tabulations. And to the extent that he is forced to make the case that the votes of various groups shouldn't count - Michiganders, Floridians, Puerto Ricans - then his moral claim will be weakened.

But who knows? Maybe Obama will win the total popular vote outright, and all of this will be academic.

One last lesson from the Pennsylvania result: The most interesting aspect of this race is how little events have mattered.

Elections are usually fluid processes greatly influenced by external events. Yet nearly every primary result in this cycle has conformed to a demographic model that evidenced itself shortly after New Hampshire: Obama wins blacks, the young, the wealthy, educated whites, and racially segregated whites. Clinton wins Hispanics, women, Jews, union voters, seniors, and racially integrated blue-collar whites.

Those constituencies have been so constant that if you had used them as rules to predict outcomes back in January, you could have picked nearly every winner so far to within a few points.

That's astonishing. It means we have a race frozen in amber. Which is why it's reasonable to suspect that no matter how much money Obama spends or how many scandals each candidate suffers in the coming weeks, Indiana, Kentucky, West Virginia, and Puerto Rico will go for Clinton while Obama takes North Carolina, Oregon, Montana, and South Dakota.

And then the superdelegates will have their conclave.