The Flathead Valley’s Leading Independent Journal of Observation, Analysis, & Opinion. © James R. Conner.


5 April 2012

Cumulative polling results suggest Tester is in trouble with the voters

A new Rasmussen Poll reports that Republican Rep. Denny Rehberg leads incumbent Democrat Sen. Jon Tester by three points. This is consistent with other polling results over the last 18 months. Tester is in trouble with the voters, and no amount of partisan spin-doctoring can obscure the facts.

A comparison with the polls (graphs below) for the 2006 Senate race, which Tester won with a narrow plurality, brings Tester’s predicament into sharp focus.

From mid-spring 2006 through the election in November, Tester almost always led Burns by a few percent. Although the leads were slim, usually within the sampling margin of error, they were consistent. The probability that Tester was actually leading was high, as can be seen in the two graphs below. Tester won by a narrow plurality (Burn’s most likely lost votes to Libertarian Stan Jones).

2006: April to Election Day

PDF for printing


2006: Labor Day to Election Day

PDF for printing

Now, the advantages of incumbency notwithstanding, Tester has for the last 18 months trailed GOP challenger Denny Rehberg by a few points. Rehberg’s lead is small, often within the sampling margin of error, but consistent, and the probability that he’s actually leading is high.

2012 campaign: recent polls

PDF for printing

As November approaches, the frequency of polling will increase, giving us a better sense of who’s in the lead, but Tester’s trailing in the polls while leading in campaign contributions suggests that something fundamental is amiss with his campaign.

Democrats may be tempted to dismiss the polls as statistical ties, or as biased in some way. That would be a mistake. Public polling has come a long way from the Literary Digest poll of 1936 that predicted Alf Landon would clobber Franklin Roosevelt. When many polls are in close agreement, the odds are high that the results are reliable within the limits of the polling process.