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.
I feel like smacking those Dems in the head that are still undecided. WTF is their problem? If they can’t figure out how different the views of these two candidates are they might as well stay home and let those vote that had been following the campaigns.
Mash, please tell me you’ve signed up for An Obama Minute later today. If you read this message in time, you can go to my blog and click on the link to register and make your contribution of any amount, even if you have to do it early like I did.
Cali, done. I found out about it late so I wasnt able to post on it. But I did my part at 1pm to add to his campaign war chest.
I think the trend has been up for Hillary since the debate. Secondly, in Texas and Ohio, she did better than the final polling. Thirdly, there’s been a backlash on Obama because he oversaturated the public space with his advertising. He should have had more giant ralleys and less invasion of the media. So I think Clinton will be in the low double digits, something like Clinton 55 and Obama 44. I wouldn’t bet on the Philadelphia area either. The town’s African-American’s mayor’s support for Hillary breaks down the automatic vote of African Americans for Obama.