Or How I Learned to Stop Worrying

1:50AM With 98.91% of the precincts reporting the Pennsylvania secretary of state reports Clinton 54.3% to Obama 45.7%, or a margin of between 8 and 9 points. There is currently a vote difference of 193,377. Hillary Clinton, after all the sound and fury, will only net somewhere between 8 to 11 delegates coming out of Pennsylvania. Barack Obama just spent Hillary Clinton into debt and she barely got a handful of delegates.

10:38PM Philadelphia is coming in very strong for Obama, currently 65% to 35% (with over 125,000 vote margin) with 95% of the precincts reporting. The Philadelphia suburbs are yet to come in. Currently with 76% of the precincts reporting in Pennsylvania Clinton leads 54% to 46%. Her popular vote lead in Pennsylvania is currently about 140,000. That puts her on track for less than 200,000 vote margin out of Pennsylvania. It looks like, if these numbers hold (and they may narrow), Hillary Clinton has lost ground on Barack Obama with time running out. What was impossible is about to become even more so after Pennsylvania is done.

9:35PM With 20% of the precincts reporting, Clinton leads about 53% to 47%. Philadelphia is slow to come in, but enough has come in for the networks to call it for Clinton. Obama is winning Philadelphia with nearly 60%. He needed about 70% to turn the race in his favor. But 60% in Philadelphia may be enough to keep the margin of Clinton’s victory small.

8:51PM MSNBC calls it for Hillary Clinton. Now we wait for all the votes to come in to see what her margin of victory is.

8:37PM MSNBC has changed their designation from "too close to call" to "too early to call". With nearly none of the precincts reporting Clinton has a lead. The change in terminology suggests that MSNBC will eventually call it for Clinton. The real question is by how much will she win.

8:22PM There are some interesting exit polls for Obama. He is taking Philadelphia by 69%, Philadelphia suburbs by 62%, Pittsburgh by 38% and holding at 42% in the rest of the state. Those numbers are in Ed Rendell territory (Rendell beat Casey by 10 points by raking up numbers around Philadelphia). Obama is also doing significantly better amongst white men than he did in Ohio. I am still holding to my prediction of Clinton 52.8% to Obama 47.2%, but these exit polls are pleasantly surprising.

8:09PM The conversation on MSNBC is about how Clinton can exit the race gracefully. The atmospherics coming from the Clinton campaign suggest they are preparing the ground to spin a narrow victory. It’s either rope-a-dope or genuine nervousness. We shall see.

8:00PM As polls close MSNBC has it too close to call. This could be a long night.

7:26PM Polls close in about 30 minutes. Early exit polls abound. On the aggregate, unreliable, early exits Drudge has Clinton up 52% to 48%, while NRO has Obama up 52% to 47%. Both numbers probably mean practically nothing. The internal numbers from exit polls from CNN, MSNBC and others are much more interesting. Gun owners go to Clinton 58% to 42%, African Americans go to Obama 92% to 8%, first time voters go to Obama 60% to 38%, late deciders go to Clinton 58% to 42%, white males go to Clinton 55% to 45%.

Hillary Clinton’s topline number (that with gun owners) is 58% to 42%. Assuming all Pennsylvanians owned guns, Clinton would win by a 16% margin.  Given that some other demographics are going to Obama, and some substantially toward Obama, the exits suggest a single digit Clinton victory at best. The real wildcard is what percentage of the vote Obama wins in and around Philadelphia and what percentage of the total votes comes from that region. Reports suggest huge turnout in and around Philadelphia. We will find out how large in the next few hours.

Hillary Clinton’s closing argument in Pennsylvania:

Barack Obama’s closing argument in Pennsylvania:

and Bill Clinton tells you why you should vote for Barack Obama:

Pennsylvania goes to the polls tomorrow. The polls have steadied over the last week and, while he remains behind Clinton by about 6 points in the polls, there may be a slight Obama momentum going into tomorrow. My prediction is here (I am sticking to it until I am put to shame by the actual results!). But tomorrow evening we will be counting the results of the only poll that matters.

It is the last weekend before the Pennsylvania primary. Most of the polling has been completed and now all that remains is for the Democrats in Pennsylvania to make their preferences known. In the last month and a half, Barack Obama has shrunk what appeared to be a 20 point Hillary Clinton advantage to about 5 points, according to the latest poll of polls at Pollster.com. According to Pollster, Hillary Clinton leads Barack Obama 47.4% to 42.3% with 10.3% undecided. Because of the large number of undecideds among voters whose demographic favors Clinton, many are saying that this could result in a double digit Clinton win in Pennsylvania. If the results indeed show a double digit Clinton win it would conform to what has been long expected in Pennsylvania. However, my sense is that the results will be much closer.

My prediction for Tuesday is that Hillary Clinton will win 52.8% to 47.2%, approximately a 6 point victory over Barack Obama.

When reading polls this late, the conventional wisdom is that late deciders break for the candidate who is ahead - Hillary Clinton in Pennsylvania. A look at the polls before the Ohio and Texas primaries seem to confirm this observation. In both cases, Hillary Clinton received a higher percentage of the vote than what the final polls indicated. It appeared that those shown as undecided in the polls went toward Clinton. However, a more likely explanation is that those undecided in a primary within a day of the primary are likely to just stay home. A closer look at the polling and the results support that conclusion and offer some clues as to what we might see in Pennsylvania.

Currently in Pennsylvania Hillary Clinton garners 47.4% to Barack Obama’s 42.3% and 10.3% show as uncommitted in the polls. The decideds amount to 89.7% of those polled. If only those who appear as decided vote on Tuesday one would expect Hillary Clinton to get 52.8% (47.4 divided by 89.7) and Obama to get 47.2%. I posit that this is the most likely outcome as the undecideds this late in the game are likely not motivated enough to get to the polls. Undecideds are not the same as late deciders that will show up in the exit polls. The late deciders are already likely factored into the polls this close to the primary. Just for comparison, if we were to give all of the "undecideds" to Clinton (as the conventional wisdom suggests) the outcome in Pennsylvania would be Clinton 57.7% and Obama 42.3%. That outcome is highly unlikely, but you can consider that to be Hillary Clinton’s top-line number.

Now, let us look at the numbers in Ohio and Texas. Hillary Clinton won Ohio 54% to 44%. The final polls in Ohio before the primary had Clinton winning 49.6% to 43.6% with 6.8% undecided. The decideds amounted to 93.2% of those polled. By our calculations Clinton would have won Ohio 53.2% to 46.8%. If the undecideds had moved into Clinton’s column, she would have won 56.4% to 43.6%. As it turns out the results are very close to our prediction that assumed the "undecideds" stayed home.

In Texas, Hillary Clinton won 51% to 47%. The final polls in Texas before the primary had Clinton winning 47.8% to 45.9% with 6.3% undecided. The decideds amounted to 93.7% of those polled. By our calculations Clinton would have won Texas 51% to 49%. If the undecideds had moved into Clinton’s column, she would have won 54.1% to 45.9%. Once again the results were very close to our prediction that assumed "undecideds" stayed home.

You will notice that in both Ohio and Texas Obama underperformed the predictions by about 2%. The 2% went to "other". A similar percentage in Pennsylvania may go to "other" giving Clinton a 52.8% to 45.2% win. That would give Clinton a 8% margin of victory.

How the final numbers ultimately turn out will depend a lot on the get-out-the-vote efforts of each campaign. But the numbers suggest a likely 6 point Clinton victory and maybe a margin as high as 8 points. With this kind of a margin, out of Pennsylvania’s 158 delegates Clinton is likely to get at best 84 delegates to Obama’s 74 delegates - a net gain of 10 delegates. In reality, because of Obama strongholds in delegate rich areas of the state Clinton’s net delegate gain is likely to be smaller. When all is said and done, Hillary Clinton will likely come out of Pennsylvania - her last best state - without winning anything close to enough delegates to cut into Barack Obama’s insurmountable delegate lead of 163. For all the sound and fury surrounding Pennsylvania, this "neck and neck" race has been lost since February. Hillary Clinton hopes to keep up the illusion of this "neck and neck" race by a double digit win in Pennsylvania. However, if the polls hold and the voters do their part the illusion will have died in Pennsylvania.