| Poll | Date | Sample | MoE | Romney | Santorum | Paul | Gingrich | Spread |
|---|---|---|---|---|---|---|---|---|
| RCP Average | 2/23 - 2/26 | -- | -- | 37.8 | 35.3 | 12.3 | 10.0 | Romney +2.5 |
| PPP (D) | 2/26 - 2/26 | 421 LV | 4.8 | 39 | 37 | 13 | 9 | Romney +2 |
| Mitchell/Rosetta Stone | 2/26 - 2/26 | 858 LV | 3.3 | 35 | 37 | 8 | 9 | Santorum +2 |
| WeAskAmerica | 2/26 - 2/26 | 984 LV | 3.1 | 37 | 33 | 18 | 13 | Romney +4 |
Unlike the other two polls, which inflate Gingrich and Paul’s support (the ridiculous We Ask America Poll has Paul at 18% and Newt and 13%, thus giving Willard a slender 4 point lead), this poll more realistically puts Newt and Paul in single digits, since both have abandoned Michigan and it is in fact a two man race.
Most significantly, the trend lines are in Santorum’s favor and Mitchell Rosetta Stone is the only one of the three polls taken last night to have polled on Thursday night so they can examine these results and compare them to those they got on Thursday to show the trend. 858 voters is a large sample with a better than average indicia of reliability.
As I said Saturday, polling an open primary state is hard becasue you don’t know what your pool of voters will look like. But the trend lines should be very encouraging to Santorum supporters.
And Romney does measurably worse in Open Primary States. I think Mittens is gonna lose Michigan by 6 to 8 points with the undecideds (11% in this poll) breaking heavy for Santorum and the Gingrich supporters going to Santorum to drive a stake through Romney’s chances.