Matt Yglesias notes an expectations imbalance:
Let me just note that simply because Hillary Clinton is hopelessly far behind in North Carolina doesn't mean that Indiana is the only May 6 primary that matters. Insofar as the remaining contests matter at all, they all matter. The Clinton campaign did a good job of making the contests Obama won between March 4 and Pennsylvania go down the memory hole, so that I heard TV people talking about Clinton having a streak (her actual streak as of today is one win!) but she doesn't get to arbitrarily decide which states matter.
It's a particular irony that when Obama has a double-digit lead in a state a couple weeks out (say, North Carolina or Maryland or Washington state) it's seen as an Obama state, but when Hillary Clinton has a double-digit lead in a state a couple weeks out (Texas or Ohio or Pennsylvania) it's seen as a "battleground" state, because everyone assumes (so far correctly) that Obama will be able to cut dramatically into her margins and perhaps put the state in play.
On a larger point, it is remarkable the extent to which the media narrative of the race has followed the same contours as the Clinton campaign's spin over the last couple months. After Super Tuesday, it was "on to Texas and Ohio"; after Texas and Ohio, it was "on to Pennsylvania." According to CNN's numbers, in March--when Clinton allegedly began her big comeback--Obama actually beat her overall by one pledged delegate.
In general, the media has given far greater coverage to Clinton's small-to-medium-sized victories in big states than it has to Obama's big victories in small-to-medium-sized states. (Obama beat Clinton by 370,000 votes in Georgia and netted 35 delegates; Clinton beat Obama by 230,000 votes in Ohio and netted 9 delegates. Which was more important?) And there is of course the media's downplaying of caucus results. (How many news-watching Americans are aware that Obama won Texas by 5 delegates, the only metric that technically matters in the quest for the nomination?)
To be clear, I don't think the news media is just following the Clinton campaign's lead (though there is an element of this). The campaign has been very shrewd about tapping into some of the press's pre-existing tendencies. Reporters give give heavier coverage to big states for all kinds of reasons: because they're conditioned by the winner-take-all general election to think a narrow win in a big state still matters enormously; because there are too many primaries for the press to devote heavy coverage to all of them and simplifying the season down to a handful of "big" races is convenient; because reporters, like most American elites, probably tend to discount the importance of smaller states (especially those outside the Boston-DC corridor) a bit to begin with. Reporters also tend to downplay (post-Iowa) caucuses somewhat because they see them (correctly, to my mind) as abstruse, archaic, and not very democratic. And, as others have pointed out, the press is extremely receptive to the idea that the small-town working class is somehow more "authentically American" than they and their elite peers are.
None of these tendencies is particularly nefarious, but together they've helped paint a profoundly inaccurate portrait of just where the Democratic race stands. It's little wonder that 43 percent of Pennsylvania voters thought Hillary Clinton will be the eventual nominee.
--Christopher Orr