Let's continue the habit of prediction folly! Tomorrow, Pennsylvania Democrats cast their votes (in record numbers, I suspect) to determine who will earn the majority of their 188 delegates.
Hillary Clinton is counting on a wide-margin, 8-11% spread in order to give her campaign hope of winning the popular vote tally and shortening Barack Obama's lead among awarded delegates.
The polling provides little evidence of what will really happen. Some polls show an Obama surge, others show Clinton holding a steady 5-8%. Then you have the twin issues of undecided voters, who by demographic profile should swing to Clinton; and the other issue of under-polled new (and younger) voters who by demographic profile should swing to Obama.
In national polls, Obama has opened up a nearly 20-point lead. He needs a strong finish in Pennsylvania and then a quick victory in North Carolina to finish off Hillary Clinton. To date, whenever he's been close to wrapping up the nomination he fumbles and fails to close the deal. The Keystone contest could clinch it for him or prolong the campaign for another six weeks.
As with most of this campaign 2008, winning hasn't always been about the numbers directly. It's more often about defying expectations; doing better or worse than expected.
My guess...err, my prediction, is that Obama will do better than expected. While in states, like Pennsylvania, Democrats can expect Obama to under-perform polling trends, in tomorrow's contest, I think the polling will prove to have been an unreliable read of who was going to participate.
In more specific terms, I think Hillary Clinton will win the Keystone state but only by a margin of 5% or less making her win somewhat more hollow than it would have been otherwise.


