Pennsylvania 2010 Senate True Vote Analysis
June 18, 2011
This is an analysis of the 2010 Pennsylvania senate race in which Toomey (Rep) defeated Sestak (Dem) by 80,000 recorded votes (51-49%). The True Vote analysis indicates that Sestak had at least a 51.5% vote share and 120,000 vote margin.
The unadjusted exit poll (2664 respondents) was a 50/50 tie: Sestak had 1331; Toomey had 1333. The Final exit poll was forced to match the recorded vote: Sestak 49.1%; Toomey 50.9%.
The Final indicated that 49% of the votes recorded in 2010 were cast by returning Obama voters and 45% by returning McCain voters, a 4 percent margin. But Obama won the PA True Vote by 15%. Therefore it is logical to assume that there were many more returning Obama voters than indicated in the 2010 PA exit poll. If this is the case, then it is beyond a reasonable doubt that Sestak won the election.
Given Obama’s 57% PA True Vote and a) 60% returning Obama voter turnout, b) 70% McCain turnout, c) 2010 PA exit poll vote shares, Sestak won by 122,000 votes (51.5-48.5%).
As in the Wisconsin exit poll, vote shares were not available for returning third party (Other) voters and new (DNV) voters. However, the exit poll indicated that returning third-party 2008 voters and new voters each represented 3% (119,000) of the total 2010 vote. This is a RED FLAG! Obama won third party voters by 66-20% over McCain. It is also impossible that returning third-party voters comprised 119,000 of the 2010 vote. There were only 81,000 third-party voters in 2008 and approximately 55,000 returned to vote in 2010. Therefore, the returning third-party mix was changed to 1.4%. New (DNV) voters increased from 3% to 5.2%.
The sensitivity analysis tables display Sestak vote shares and margins for various scenarios: Obama 2008 voter turnout in 2010, Sestak shares of returning Obama voters and Toomey shares of returning McCain voters.
The Final 2010 PA Exit Poll is eerily similar to that of Wisconsin. Apparently, the exit pollsters forgot to adjust “When Decided” to match the recorded vote. Feingold and Sestak each won the “When Decided” category. Toomey needed 53% of the 27% of voters who decided in the last week. He had just 42%.
The Democrats led in Party ID by 40-37% over the Republicans. Democrats virtually always win when they have a 3% edge in Party-ID.
In the “Voted in 2008” category, Sestak led the 94% who were returning Obama and McCain voters. Vote shares for returning 3rd-party and new voters are not available.
The 2008 Presidential True Vote analysis indicates that Obama won nationally by 58-40% – a 22 million vote margin (only 9.5 million was recorded). Forcing the State and National Exit Polls to match the recorded vote is standard operating procedure.
In order to force a match in 2004 and 2008, the exit pollsters had to assume an impossible number of returning Bush voters from the previous election. The returning voter mix should reflect the previous election True Vote, not the recorded vote. The Final is forced to match the recorded vote by adjusting the returning voter mix and/or the vote shares. In 2004 adjusting the mix was not sufficient to match the recorded vote, so the exit pollsters had to increase Bush vote shares as well.