People have been talking about the popular vote as a possible trump card for Clinton. As in, Yes, Obama is going to win the pledged delegates. But what if Clinton wins the popular vote? The implication is that a popular vote win by Clinton could convince superdelegates, who will decide the election, to swing toward her instead of Obama.
Here’s the problem with that: The popular vote isn’t as pure a number as people think. For all the biases of the Democrats’ pledged-delegate selection system—proportional allocation, caucus math, open vs. closed voting—the popular vote has its own inadequacies. Namely, it understates Obama’s success in caucus states.
Caucuses have relatively low turnout compared with primaries. "To me, the caucuses don't provide the broad base of participation that I have fought for my entire life," Clinton said even after winning Nevada. That’s why Obama’s strength in caucuses—he’s won all but two of the 15 caucus states so far—irks Clinton so much. They privilege energetic young people with free time, also known as Obama’s base.
But when you’re talking about the popular vote, the relatively low caucus turnout turns into an advantage for Clinton. Obama won Kansas, for example, with 74 percent of the vote. But only 37,000 of the state’s 401,000 registered Democrats—about 9 percent—turned out to caucus. Primary turnout, on the other hand, has been more than 30 percent in many states this year (although each state’s registration system is different, making the exact numbers difficult to measure). So had Kansas held a primary instead of a caucus, the state would have contributed more toward Obama’s popular vote tally. (For relative turnout in the 2004 caucuses and primaries, see here .)
The Clinton counterargument would be, Well, if Kansas had held a primary, Obama wouldn’t have won by as much, if at all. But that doesn’t change the fact that the popular vote does not fully reflect the results of the system agreed to by the party. You could conceivably calculate an alternative "popular vote" that extrapolates caucus results to imagine what the total tally would have been, had more people showed up to caucus. But that comes with its own dangers, since you don’t actually know how those people would have voted.
Still, the main point stands: The popular vote is tainted. If Florida and Michigan revote, there’s a chance Clinton would narrow the lead in the popular vote, and possibly even take it. (She still probably can’t catch up delegatewise.) In which case, keep in mind that even the popular vote has its flaws.