FOR IMMEDIATE RELEASE October 19, 2010
PUBLIC POLICY POLLING: Raleigh, N.C. – The race for Senate in Pennsylvania has seen dramatic movement over the last two months and Joe Sestak has now taken the slightest lead over Pat Toomey, 46-45. On PPP’s previous survey of the state, in mid-August, Toomey led by a 45-36 margin.
Three things have happened in the last 2 months to move this race back to toss up status: Sestak has made up a large deficit with independents, he’s consolidated his base, and Democratic interest in the election has picked up with election day moving closer.
In that August survey Toomey held a commanding lead with independents, 50-23. Sestak has picked up the support of virtually all of the undecided independents and has pulled into a virtual tie with Toomey with that voting group, trailing now only 49-48.
Also on that August survey there was a tremendous party unity gap with Toomey winning over 74% of Republicans while Sestak was getting just 64% of Democrats. Although Toomey is still benefiting from a more unified party base the disparity is not nearly as large as it was at that time- Toomey is pulling 82% of Republican voters while Sestak gets 77% of Democrats.
Finally, as PPP is seeing across the country, the closer to the election we get the more Democratic voters are tuning in. In August those planning to vote in Pennsylvania this fall reported having voted for John McCain by a 1 point margin in 2008. Now the likely voter poll for this year supported Barack Obama by 4 points. That still suggests a large drop in Democratic turnout from 2008, but it may not be quite as dramatic as it appeared over the summer.
“This has become a toss up race,” said Dean Debnam, President of Public Policy Polling. “No one would have expected that a month ago but just as he did in the primary Joe Sestak is closing strong. Whether he continues that momentum right on through election day or Pat Toomey can turn things back in his favor will be the ultimate determiner of who wins this race.”
PPP surveyed 718 likely Pennsylvania voters on October 17th and 18th. The survey’s margin of error is +/-3.7%. Other factors, such as refusal to be interviewed and weighting, may introduce additional error that is more difficult to quantify.
I am no expert but I heard a very interesting interpretation about this poll on a radio show yesterday. Apparently this pool “assumes” that there will be a higher Democratic turnout for these mid-term elections that there was when Obama was elected 2 years ago. I take just about any wager against that happening.