Based on the weighted average of five late October polls (Table 1), the odds that Jon Tester will win a second term in the U.S. Senate are three to two. Four of the polls report Tester leads by one to two points, while the other — the Mason-Dixon poll commissioned by the Lee newspapers — puts Tester four points behind Denny Rehberg. Tester’s mean lead is 0.6 percent lead with a margin of error of 1.6 percent. That translates into a ballot lead probability of 64 percent, or 3:2 odds.
This is not a statistical tie — Tester has a small lead — but it is a functional tie. Whether Tester is re-elected, or is replaced by a man whose presence in the senate could make Mitch McConnell majority leader, now depends on which campaign does a better job of getting its supporters to the polls tomorrow.
Below, three graphs. The first displays the polls following the conventions. The second displays just Tester’s lead over Rehberg. The third shows registration as a function of the voting eligible population. Although the absolute number of registered voters is at an all time high, so is the population. On a percentage basis, registration as a function of VEP is slightly down from 2008.
| Poll | End Date | Sample Size | MOE | Rehberg | Tester | Cox | Undecided | Tester Lead Probability | Tester Lead |
|---|---|---|---|---|---|---|---|---|---|
| Pharos Research | 10/28/2012 | 799 | 3.5 | 46.5 | 48.2 | 68.0 | 1.7 | ||
| Rasmussen | 10/29/2012 | 500 | 4.5 | 48.0 | 49.0 | 59.0 | 1.0 | ||
| Garin-Hart-Yang | 10/29/2012 | 807 | 3.5 | 43.0 | 44.0 | 6.0 | 7.0 | 61.0 | 1.0 |
| Mason-Dixon | 10/31/2012 | 625 | 3.9 | 49.0 | 45.0 | 1.0 | 5.0 | 16.0 | -4.0 |
| PPP | 11/3/2012 | 836 | 3.4 | 46.0 | 48.0 | 4.0 | 2.0 | 72.0 | 2.0 |
| Mean | 11/3/2012 | 3,567 | 1.6 | 46.2 | 46.8 | 3.7 | 4.7 | 64.0 | 0.6 |
