By:
TJHalva |
Comments [0] | Category:
VFA Original | 10/21/2008 10:50:20 PM CT
The Minnesota Senate race is likely the most contentious battle this election cycle, and will likely end up being the most heavily financed when all is said in done. The race features Incumbent Republican Norm Coleman, Democratic challenger Al Franken and Independent Party of Minnesota candidate Fmr. Sen. Dean Barkley. Coleman and Franken remain the front runners, but Barkley has consistently polled in the high teens; while down, Barkley is not out.
A new round of Minnesota polling was concluded on October 18th from two local media pollsters based in Minneapolis; the Star Tribune Newspaper, and KSTP-TV (the ABC affiliate). Each pollster performed Presidential and Senate polls sixteen days prior to their most recent release. I've included the results of their new polls, their old polls, and our projections for both time frames below:
Star Tribune [10/17] SurveyUSA [10/18]
Dem Rep Dem-Rep Dem Rep Dem-Rep
President 55 37 18 46 47 -1
Senate 43 34 9 33 43 -10
Pres-Sen 12 3 9 13 4 9
Star Tribune [10/1] SurveyUSA [10/2]
Dem Rep Dem-Rep Dem Rep Dem-Rep
President 52 41 11 50 44 6
Senate 39 36 3 39 41 -2
Pres-Sen 13 5 8 11 3 8
VFA Projection [10/2] VFA Projection [10/18]
Dem Rep Dem-Rep Dem Rep Dem-Rep
President 52.93 38.97 13.96 49.98 43.71 6.27
Senate 41.61 36.64 4.97 38.61 40.15 -1.54
Pres-Sen 11.32 2.33 8.99 11.37 3.56 7.81
The bolded row at the bottom of each pollster's result represents the correlation between that pollster's Presidential and Senate race findings. If you look at all six results, the bolded numbers for each column reasonably agree with the other five. This result shows that within the meta sample of all four polls and our projections the correlation remains consistent. Using this consistency and our projections on October 18th, I have derived a set of simple formulas that correlate Senate and Presidential performance:
An Obama victory in Minnesota with just 50.97% of the vote correlates to a razor thin victory by Al Franken using the above formulas (see note at bottom). Given that the Minnesota Senate race is a three way ordeal and the fact that our projection neglects to allocate independents or uncommitted voters I'll dive deeper into the meta sample.
If we assume that Barkley receives eighteen percent of the vote, as he has in the last four polls, the independents cannot save Norm. As it stands, under this assumption, Barkley is receiving roughly 63% (11.37 / 18) of his backing from Obama supporters, just 20% (3.56 / 18) from McCain supporters and 17% from Independents. Again under the eighteen percent assumption, just 3.24% of population is left undecided. If we just arbitrarily allocate this demographic to Norm Coleman, Obama must increase his popular vote percentage by 4.19% (3.24 / .7725) to 55.17% in order to guarantee a correlated victory for Franken.
Although the current polling doesn't show this kind of lead for Obama; if I were Norm Coleman I would be looking to fundamentally alter the "coattail coefficients" to better situate myself for reelection. Looking at the MN Senate graph below it appears that the scorched earth strategy doesn't seem like an option. Coleman needs a game changer, but with just 14 days to go his options seem limited.
Although Norm has the lead in the interim, his support is deteriorating at a much greater rate than his opposition.
Note: The threshold of victory for a Senate race is found by averaging the projection of the two leading candidates and dividing by two. In the case of Minnesota, a popular vote greater than 39.98% should result in a Senate election victory.
0
Response(s) to
Minnesota Meta-Polling
Leave a Reply:
Name: (Defaults to Anonymous)
Type the characters you see in the image below: