Pennsylvania polls remain all over the place

No matter how often polls have been discredited throughout this cycle, it seems to me that we could usually rely on a certain consistency. In Ohio and Texas, Obama was clearly gaining in all polls before surveys released in the final 3 days showed the start of a Clinton comeback; on Super Tuesday, the trend was clearly towards Obama. Even in New Hampshire, where the final polls were so awfully off, I have always believed that the polls cannot be blamed but it is the media who didn't know how to read them given the 5 day window of NH campaigning and the last minute shift towards Clinton.

In Pennsylvania, however, polls show different trends, different momentum and different leaders. After SUSA and ARG, and even Strategic Vision and Rasmussen, showed Clinton solidly ahead yesterday, two new polls paint a picture:

• PPP finds Obama ahead, 45% to 42%. Last week, Clinton was ahead by the same margin. The poll was conducted Monday and Tuesday.
• Two weeks ago, PPP has already released a poll showing Obama ahead -- but no other institute has ever shown him leading.

• Franklin & Marshall University are more reassuring towards Clinton, who remains ahead 46% to 40% (49-42 with leaners).

So what is the effect of bittergate? There is little evidence of any effect at all, with different polls going in different directions. And it cannot be ruled out that the controversy is backfiring on Hillary Clinton. If some Pennsylvania voters had concluded that Obama had basically won the nomination but still wanted to cast a ballot for Clinton, her willingness to air an attack ad on this issue and to criticize Obama as harshly as she has could make some people think that it is too dangerous to prolong the primary. On the other hand, Clinton's ad just started running on Monday and the media is keeping the story alive, so there is only so long a politician can not take a hit over something that is repeated so often.

Thankfully, North Carolina polls are a bit easier to navigate. Two new polls today confirm the conventional wisdom that this state is favoring Obama:

• An ARG poll shows him leading 52% to 41%, which is a two point improvement for Hillary since last week. Clinton narrowly trails among women (50-44), crushes Barack among whites (57-33).
  

• Meanwhile, Insider Advantage finds Obama ahead 51% to 36%. The pollster adds that there has been little movement since the end of March, suggesting bittergate has not caught on the imagination of North Carolinians.
  
• He adds, as a warning to Clinton, "The makeup of the white Democratic vote in North Carolina is heavily populated with voters from ‘Research Triangle’ universities-based region." That is not the type of electorate Clinton ever hopes for...

These polls give us an idea of how difficult the road ahead is for Clinton. She needs to post strong numbers in PA and she then immediately needs to get equally solid results in North Carolina and Indiana. A double-digit lead in the second largest state left to vote (after PA) will simply not cut it. The problem for Hillary is that the NC electorate is in some ways similar to Virginia's, with a significant proportion of the white-collar whites (though not as much as VA, allowing her a route to stay competitive).

Finally, the last poll of this morning comes to us courtesy of the Washington Post, which finds that Obama is now ahead of Clinton 51% to 41% nationally; against McCain, Clinton trails 48% to 45% while Obama is ahead 49% to 44%. Could it be that national numbers are more important than ever since, for the first time, they have an audience (superdelegates)?