Rick Santorum’s Polling in Pennsylvania

santorum-in-paThe Pennsylvania Primary is on April 24th and the release of a new poll by Franklin & Marshall has the chattering class saying Rick Santorum’s lead has vanished in his own state!  He’s losing steamIt disappeared!  The poll which was taken on 3/20 through 3/25 has Santorum with just a two point lead over Mitt Romney – 30% to 28%.  Ron Paul polls third with 9 points and Gingrich is last with 6 points.  There are 24% of voters who are undecided.

The previous Franklin & Marshall poll was done in February.  In it Santorum led the field with 45 points  to Romney’s 16 points.  Gingrich had 9 points to Paul’s 7.  So between February to March according to Franklin & Marshall Santorum’s lead was cut by 27 points.  Santorum lost 15 points, and Romney gained 12.  I find it interesting that the undecided category grew by 2%.  Gingrich lost 3 points… so I would imagine some ended up there.

Here’s what I think is happening, so let’s not get all excited over one poll.

For starters I question whether Santorum ever really lead by 29 points.  That was one poll.  The other poll taken in February which was the first to show a Santorum lead was done by Susquehanna, and he only led by 1 point.  Susquehanna did the previous poll before that in December and Gingrich was up by 17 points.  So you have two polls in the same month which has a 28 point difference in Santorum’s lead.

Hmmmm…..

A little history lesson.  Back in May of 2006 during Santorum’s race with Bob Casey to retain his Senate seat Franklin & Marshall had Casey leading Santorum by five points which was a drop from an 11 point lead from February of 2006.  Rasmussen that same month had Casey leading Santorum by 23 points.

So there’s the question of inaccuracy in the past.  There’s also the question of inaccuracy in the present.  If you were to look at the other two polls done in March, one by Quinnipiac and one by Public Policy Polling you’d see that Santorum has a 14 point and 18 point lead respectively.  Granted both were taken before Romney’s win in the Illinois Primary and Santorum’s win in the Louisiana Primary but it’s doubtful Romney’s win in Illinois would have made that much of an impact.

So what to make of this?  One I think some people who have been pushing Romney as the nominee are getting far too excited about one poll.  Secondly, Franklin & Marshall’s polling accuracy is in question.  Thirdly is it really that surprising to see a race tighten as it gets closer?  Franklin & Marshall point out that Romney leads in the east – that’s his bread & butter – the urban areas.  Those are the parts of the state which are more moderate.  Yes, moderate voters and urban voters are coming home to roost with Romney like they have in every other contest we’ve seen.  He still has a problem with conservatives and evangelicals.

Anyway, I think its far too early to say Romney has a surge going in Pennsylvania based on one poll.

If you like what you read, sign-up to get CT in your inbox!

Comments

  1. says

    Shane–What basis would you use to question the F&M poll?  Your feelings?  The poll is run by Terry Madonna who has been the leading pollster in PA for more than 20 years.  Your bio does not read like someone with training or expertise in statistical methods.  
    Neil

    • says

      My doubts are based on the disparity with the other polls, not my feelings otherwise Pennsylvanians must be really, really fickle people.  
      I don’t doubt Romney has closed the gap some… but based on other polling done I doubt Santorum ever had a 29 point advantage.

      So anyway my point is let’s not go crazy over one poll… now if another one from a different pollster comes out showing the same trend… then I stand corrected.

  2. says

    Here’s another new poll for PA from Wenzel Strategies:

    Santorum: 45.6%
    Mitt Romney: 25.9%
    Ron Paul: 10.4%
    Newt Gingrich: 8.6%

    The polling firm is Wenzel Strategies, which conducted a telephone survey of likely Republican primary election voters from March 26-27. The results include 997 respondents, at a margin of error of +/- 3.07 percentage points.

    So you were right all along to be skeptical. :)

    • says

      Interesting, the F & M poll had 505… I know sometimes if your pool is too large that can skew it too. The key is to have a size large enough you can get a representatives sampling that is in proportion to the population.

  3. Jo Bailey says

    I live in Penna. My unofficial polling says santorum loses badly. Just mention his name and everybody either laughs or gets angry. We remember him here. The most corrupt senator in the country.