A meme I've seen emerge from certain prominent pundits over the last several days is that if Hillary Clinton stays in the Democratic primary and fights through the remaining 10 contests (whew), she will likely pull ahead of Barack Obama in the popular vote. 

At that point, according to neutral observers like Bill Clinton, the NY senator will have a strong moral case to make to the remaining undecided superdelegates that she should be given the Democratic nomination, regardless of Barack Obama's pledged delegate lead. 

As both a small "d" and Big "D" Democrat, I would certainly agree with her.  It is an absolute must that Barack Obama retain his popular vote lead, in order to have a legitimate claim to the Democratic nomination.

That said, why don't we take a look at the popular vote numbers, and see how they stack up against the notion, most prominently floated by famed conservative pundit and political statistician Michael Barone, that Hillary is poised to claim a popular vote lead by the time we reach Denver?  

Like so many of the stats in this race, there is no one single hard popular vote number.  However, all of the variations show Obama with a current lead.  As always, we look to RealClearPolitics' numbers crunching to light the way.  I'll order them from the most generous margin to the least for Senator Obama:

+827, 308
Includes all of the primaries and caucuses that have participated, and excludes the illegitimate, DNC-discredited contests in Florida and Michigan.

+717, 086
Includes all primaries that have participated, all caucuses but Iowa, Nevada, Maine, and Washington, and again excludes Florida and Michigan.  The reason an estimate exists that excludes IA, NV, ME, and WA?  These caucuses did not report official popular vote numbers—just which candidate won their states' various caucuses.  Popular vote estimates for each of these states have been gleaned by comparing turnout estimates with the percentage of delegates won at these caucuses.

+532,536
Includes all legitimate primaries and reported caucuses, plus the vote from the unreported caucuses, plus Florida.  There is at least a glimmer of an argument to include Florida, as at least both Clinton and Obama were on the ballot.  Like I said: just a glimmer.

+422,314
Includes Florida and all the legitimate primaries, and excludes the unreported caucuses.

+204,227
Includes Florida and Michigan, as well as the unreported caucuses.  Being that Sen. Obama was not even on the Michigan ballot, this estimate merits not even a glimmer of legitimacy.

+94,005
The ultimate pro-Clinton estimate, this includes Florida and Michigan, while excluding the unreported caucuses.  According to TL, it's complete bullshit.

With these varying tallies laid out, let's look at the path Michael Barone carved for Hillary Clinton to win the popular vote over the next three months.  After all is said and done after the remaining ten contests, he sees Clinton ahead in the popular vote by 106,186.  How does she do it?

I have her carrying Pennsylvania by 20 percent—a 60 percent to 40 percent margin of the two-candidate (Clinton and Obama) vote.  That's better than she did in Ohio, where she won 55 percent of the two-candidate vote. 

OK, this is a possible scenario, albeit an overwhelmingly optimistic one for Sen. Clinton in Pennsylvania.  It envisions no movement whatsoever over the next three weeks for Obama, despite his recent endorsement by PA Sen. Bob Casey, his outspending Sen. Clinton 5-to-1 in the state's media markets for the next 20 days, and the potential effect the "Bosnian sniper scandal" will continue to have on the Clinton campaign. 

In fact, in spite of the Rev. Wright scandal, the latest PA Rasmussen poll has Obama closing a 10-point deficit there to 5, and the latest Survey USA has him climbing from 19 points down to 12, within the last few days.

Those results have also influenced my projections of even bigger percentage margins for Clinton in Indiana, West Virginia, and Kentucky (20%, 40%, and 30%, respectively).

Now these are simply lazy projections by Barone.  It is extremely risky to assume that just because one state is demographically similar to another, and neighboring, the vote margins will be the same or better.  It is even worse to do as Barone has done with these three, which is to in effect say, "I like the really optimistic odds I gave Hillary in PA.  I like them so much, I'm going to double-down on them in neighboring West Virginia!" 

Really?  Hillary's going to win by 40% in West Virginia?  Even with neighboring Virginia next door, which Barack won by 25 points?  Even with a sustained, well-funded media campaign like the one he's running in PA—which Clinton simply does not have the cash to compete with?

I projected a 10 percent margin for Obama in North Carolina; the Realclearpolitics.com average of recent polls has him ahead 57 percent to 43 percent in the two-candidate vote. 

I think this is a fair enough, if rather conservative prediction for Obama's looming win in N.C.  However, I find it interesting that Barone has no problem using his "neighboring state" methodology when it suits Clinton, but not when it might suit Obama.  I mean, didn't Obama win South Carolina by 28 points?  He repeats this bias again with Oregon.

I have Clinton losing also by 10 percent in Oregon.  That’s roughly comparable to her showing in the nonbinding February 19 primary in next-door Washington, where she got 47 percent of the two-candidate vote. 

 
Awfully odd of Barone to predict Oregon's results by using neighboring Washington's non-binding party primary, when he could have gotten a much more accurate reflection with Washington State's binding caucuses.  In those, Obama score a 36-point victory over Clinton.  Not 10 percent.  36 percent.  Oh, did I mention they were binding?

I have Clinton winning Montana and South Dakota by 20 percent margins, when the conventional wisdom seems to be that these states lean to Obama.  It’s true that Obama did very well in caucuses in Minnesota, North Dakota, Nebraska, Idaho, and Wyoming.  But my hunch is that the wider primary electorate will go the other way.  (italics mine)

I think you get the idea here.  In Barone's defense, he did say at the outset that he was deliberately giving Clinton a more optimistic edge in these contests.  But with many of them, he didn't just lean towards her—he completely reversed what the polls, conventional wisdom, and the actual voting trends thus far have shown.  Barone posted an unintentionally humorous defense of his predictions today, stating the following:

I was making projections that were optimistic from Hillary Clinton's point of view, and I found, to my surprise, that the results showed Clinton ahead of Obama in popular votes. (italics mine)

Do we need any further evidence that we, as a species, are doomed?

All kidding aside, I really don't see how Hillary pulls ahead of Obama in the popular vote.  With Barone's ridiculously inflated, even utterly impossible Clinton victories, she still only nets just a 100,000+ popular vote margin.  That means even in Fantasy Land, she barely bests him.  And we're not in Fantasy Land.

I guess I can forgive conservative Michael Barone for trumpeting scenarios where his wildest dreams of Democratic civil war come true.  It's quite another thing when heroic progressive writers like E.J. Dionne are posting Barone's "analyses" as fact.  From Dionne's Washington Post column today:

For now, Clinton has a strong argument for continuing.  Obama leads in delegates, but that advantage is not overwhelming. Clinton still has a chance — a pretty good one, according to an analysis posted by Michael Barone on U.S. News & World Report's Web site — of emerging from the primaries with a lead in the popular vote, though it seems impossible for her to overtake Obama in the delegate count.

E.J., please be sure to read Today's Lies.  This is one of them.