I get the sense around here that there is a lot of moping over New York, so I want to show you some data to suggest that Pennsylvania may not be the same. We have an exchange student visiting this week, so I'm a bit pressed for time (and can't phone bank unfortunately) but I'll try to be coherent as well as quick.

First, let's look at the polling data. The time series from Real Clear Politics shows that Bernie is down by 52 to 37, but it also shows the usual Bernie Magic Carpet Rise. For me, the most interesting point in this graph is what isn't there: The 11% who are missing. These are people who know who Clinton is and aren't willing to support her yet. That is most of the gap right there, so let's get canvassing!

The other big difference from New York can be seen in the supafresh FEC contributions data released Thursday. This is all reported contributions through March 31st to all presidential candidates. As far as I can tell, this includes required reporting of donations over $250 plus some optional reporting. Here is the ratio of Sanders contributors to Clinton contributors:

This is a positive ratio! I have been using this model to predict likely voters with some success (on the theory that people who put money in are likely to vote) and it tracks tolerably well in some places. It was quite accurate for Washington where I live, but much less so in the South. The New York data shows a narrower margin (10%) than actually happened, but the prediction as of the last dump (up to February 29th) was about 20%. Also remember that the delegate margin was more like 12%. So in this part of the country it seems to be in the right ball park -- which suggests that we can pull this one off.