SP&R’s Post-Election Review Shows We Were First Firm to Show a Close Race for President

Posted on November 27, 2012. Filed under: General Surveys, In The News |

SP&R has a solid track record of accurate polling leading up to the recent presidential election, and a review of our polling shows we were perhaps the first firm to show the Presidential race was competitive in Pennsylvania – and long before other firms showed the race tightening in early-to-mid-October.  For instance, in our three main, independent polls conducted in September and October, SP&R showed a close race for president well within striking distance of Obama’s eventual 5-point margin.  In our poll conducted Sept. 18th-20th (on behalf of the Pittsburgh Tribune Review), Obama led Romney by 2 points, or a margin of 47-45. However, according to all other publicly-released polling in September compiled by the polling website realclearpolitics.com, Obama’s average lead in September was 9 points, with some firms showing Obama leading by as many as 12 percentage points. We said at the time in a published story on our blog that a double-digit lead for the President was simply not to be believed, and that it was extremely unlikely the President would carry the Keystone State by a margin anywhere near his 10-point margin in 2008 due to things like a polarized electorate and a struggling economy with continued high unemployment.  So, our early September poll was the first one to show the race competitive, and differed only 3 points from Obama’s eventual 5-point margin.  

Then, our October Statewide Omnibus Poll conducted Oct. 4th-6th on behalf of SP&R subscribers and other corporate sponsors, showed a similar 2-point margin for Obama, thus confirming and validating our September polling.  A second poll commissioned by our client the Republican State Committee of Pennsylvania, conducted in early October, showed Romney leading by 4 points (49-45).  Although this was one of the only publicly-released polls to show Romney winning, the fact that he was not over fifty percent in a state as Democratic as Pennsylvania meant Obama could not be counted out.  Plus, the movement towards Romney was entirely consistent with other polls at the time showing significant movement towards Romney after a strong debate performance on Wednesday, October 3rd.  This is because many of these same firms that showed Obama leading by double digits in September now showed margins for Obama only half as big as what they reported earlier.  So, if the race wasn’t really close all along, a position some continue to maintain, then why did both Obama and Romney decide to make a combined $20 Million investment in the state’s major media markets in the last two weeks of the election? Even the average lead for the President according to realclearpolitics.com fell to less than 4 points in late October- which is a lead within the margin of error for most polls and leading many to finally declare the race in Pennsylvania “a tossup”.

Finally, our last independent poll, commissioned by the Pittsburgh Tribune Review and conducted October 29-31, showed Obama and Romney tied at 47-47, with the remaining 5% undecided.   This same poll showed U.S. Senate Republican nominee Tom Smith trailing Sen. Bob Casey 46-45, with 10% undecided.  While it is true that the statistical tie for President in this last poll differed by 5 points from Obama’s eventual 52-47 margin, SP&R was perhaps the only firm to accurately peg Romney’s eventual margin (at 47%) and Tom Smith’s (at 45%), a fact seldom reported.  Furthermore, exit polling released on Election Day and conducted nationally now confirms that voters who made up their minds in the last week of the election broke overwhelmingly for Obama and presumably, other Democratic candidates down ballot.  Furthermore, since more than 8 in 10, or 82% of all undecided voters in our last PA poll were Democrats and Independents, it is entirely plausible these undecided voters ultimately broke for both Obama and Casey in their respective races – with Obama over performing by five points (getting 52%) and Casey over performing by eight points (getting 54%).  The fact that these undecided voters broke for the incumbents rather than their Republican challengers is largely viewed as an anomaly, and explains why Obama over performed not only public polling in PA but other states as well.  Plus, if our sampling were in error, this would not explain the fact that our last poll showing the Presidential race tied also showed Democratic Attorney General candidate Kathleen Kane leading her Republican opponent by 11 points (48-37), which turned out to be almost her precise margin of victory.

In the end, Obama won the Keystone State by again racking up a huge margin in Philadelphia (winning by more than 460,000 votes), and running ahead of Romney in the Suburban Philadelphia collar counties where he added an additional 112,000 votes to his column, albeit by smaller margins than 4 years ago.  Although Romney over performed margins former President George Bush received in 2004 in places such as Democratic-leaning Southwestern Pennsylvania and the conservative, rural “T”, it simply wasn’t enough to overcome Obama’s margins in Southeastern Pennsylvania.  Overall, although voter turnout was down from 2008, with approximately 5.5 Million votes cast (compared to 5.9 Million cast in 2008), turnout either held firm or slightly increased for Obama among key constituencies that propelled him to victory 4 years ago.  These included 18-29 year olds (19% of all votes cast), those of Hispanic origin (6% of all votes cast, up from 2008) and black voters (13% of all votes cast).

The final results show that Obama carried the state by only 5 points; not 8, 10 or more points as suggested from polling conducted by other firms in September.  Moreover, Pennsylvania turned out to be only the fifth closest state lost by Romney in the entire country, not far behind losing margins for Romney in other battleground states of Ohio, Florida, Virginia and Colorado.  This is why respected New York Times bloggers like Nate Silver in a November 25th article entitled PA Could Be Path Forward for GOP, recently discussed how PA – and not other states – was what he called the true “tipping point” that probably cemented Obama’s Electoral College victory.  Anyone reviewing the accuracy of all of SP&R’s presidential polling conducted during the last two months of the election will be happy to know that our polling was, and turned out to be, closer than the other polling firms were stating.


Liked it here?
Why not try sites on the blogroll...

%d bloggers like this: