2010 Midterm Elections – FINAL Senate Prediction(Bonncaruso)
I am going to keep this analysis short and very direct.
SENATE: (statistics only) D51 / R49
SENATE: (with gut feeling) D52 / R48
In the first comment on this map are two tables I have created. The first one shows the trend shift between my last map and 10/29 (2 weeks). The second table shows the trend shift over the last 5 days, from 10/29 to today. The commentary to those tables is very enlightening.
Any polls that may come in today are NOT in my calculations. Frankly, any last minute hoping over a +/-0.01 variance is not going to change anything, even in the three closest races.
We knew all along that this would be a republican year, and indeed it will be, and it will be a big year for the GOP. And, in accordance with my promise to be fair and impartial, I am now reporting that the GOP wave will be the biggest I have seen in my lifetime, if the statistics hold.
It has also been a year for some very dishonest and misleading polling - of this I am 100% sure. And after the results are finalized, I will be doing, as I indicated on the thread to my last map, a major statistical analysis comparing the pollsters figures and the actual results, both for 2008 and 2010. First came the R2000 scandal, which dismayed me greatly. Then we have the fact that one pollster, RASMUSSEN, literally covered 1/3 of all polling in this year, which has caused an undeniable "house-effect" skew. Whether a pollster drives the narrative or whether the media frenzy drives this all is, however, unimportant at this time.
Nonetheless, it is going to be a massive republican blowout, and let me tell you why:
1.) Although I am no huge fan of generic ballot statistics, when Gallup, of all pollsters, shows the GOP going into election day with a +15 lead, 55-40, then you know that the wave has become a tsunami of unprecedented proportions for the GOP. If it were only Gallup to show the GOP with a double digit lead over the democrats, that would be one thing. But of the seven most recent generic matchups, the GOP is in double digits in four of them: Gallup (+15), Rasmussen (+12), CNN (+10) and FOX (+13). I think we can say with fairness here that two of these pollsters have been more favorable to the republicans in the past and two of them have been more favorable to the democrats in the past. When these four agree with each other this much, then you know that something is afoot. Even in years where republicans have done very, very well, the parity was usually 0 or maybe 1% for the democrats. With a double digit lead for the GOP looking very, very likely, then the wave is sure to function according to the following principle:
"A rising tide lifts all boats".
2.) OHIO. Ohio is without a doubt the quintessential battleground state and was considered a real DEM pick-up opportunity. I also hitched OHIO onto my predictions the entire time and I stick with this technique. Fisher has given up in OHIO, Portman will be the next Senator from the Buckeye state and the current margin mean is a BLOWOUT +18.30% - 3.30% above the Gallup final generic figure, which is about right. Assuming the nominal 3 point lead for a seat where the current senator is a republican, this figure matches what we are seeing coming out of the generic polling. Also, FLORIDA, which is showing GOP +15.75 (yes, I know, a three man race, but...) makes yet another state in the trifecta that is moving boldly red. Which causes me to think that Toomey will win bigger in PA than the stats say, as PA is the third state in the Trifecta.
3.) RARE JAZZ CONGRESS is now projecting 81 GOP and 7 DEM pick-ups in the house, making for a net shift of 74 seats in the House of Representatives, also indicating an almost mirror image of the current power structure in the house.
Therefore, for the Senate, I have decided to stick 100% with the end statistics and those end statistics show 8 GOP pick-ups, in descending order of probable margin: ND, AR, IN, WI, PA, NV, IL, CO.
The democrats will maintain a razor-thin majority in the senate.
Of the remaining seats in question: we have seen the DEMS pull away in CA - Boxer is holding at a +6.0 mean, outside the battleground zone, and Brown will landslide in the gubernatorial - and this in spite of a massive GOP wave going on in most of the rest of the Union. Both DEM seats in NY and the seats in VT, OR and HI were never in play; the seat in DE and the seat in CT are both safe. Even if the GOP wave were to crest to +20 (generically speaking) today, those seats will still be DEM seats.
WV is moving demonstrably to Manchin. Even Rasmussen has Manchin over 50% now. It will be a lean win, but Manchin will be the next senator from WV.
Remaining is WA (Murray vs. Rossi). It has been close the entire time. However, of the 22 polls since Labor Day, Murray has lead in 15, one was a tie, and Rossi has lead in 6: 4 of them were RAS polls, one was Fabrizio (R) and interestingly enough, by 2 points in the last PPP. But even more fascinating is that the FOX poll shows Murray in the lead by 2 points. Complicating all of this is that fact that 30% of all WA ballots have already been turned in. In King County, it is by 38%. In Pierce County, it is 43%. Since those ballots from overwhelmingly democratic counties have already come in before the polls in WA tightened again, the statistical probability is very high that Murray narrowly retains her seat. And there are certain pollsters who know the region best. Elway knows WA better than any other pollster, just as Selzer knows IA, IL and IN better than the rest. Murray's internals show her up by 8, which would concur with stats showing her sweeping King County 58-42. I would say that enough of the vote is locked in to ensure a lean victory, much as was the case in FLORIDA in the 2008 GE.
Finally: Alaska. Folks, there is just not enough empiric data to make even an informed guess. What we do know about Alaska is that it is usually a very, very strong GOP state where GOP candidates also chip off a good percentage of the democratic vote. I think it would take a miracle for MacAdams (D) to surge through the center and take the election. The probability is much higher that Murkowski or Miller wins.
So, in this election we will see two curses broken: the North Carolina curse and the curse of both houses of congress flipping when the house of representatives is bound to flip.
The stats say that the GOP will pick-up 8 seats. I think the real zone is between 7 and 9. Yes, the GOP could eek out a win in WA, or on the flip side, there is also the statistical probability that it will not be able to flip all three of those narrow, narrow states: NV, CO, IL. But like I said, I have decided to stick 100% with statistics pure on this one and let the chips fall where they fall. Note that Nate Silver says likewise. He predicts a 52-48 senate, but his map shows a 51-49 senate.