(11/2) Election Forecasts 2010: Senate

I'm not going to do a Governor race forecast, but don't think they're not important. I just don't have the resources. Watch Connecticut to see if Foley(R) will have a come-from-behind win, watch Ohio to see if Strickland(D) can hold his ground and RI, CO, and ME will be interesting. Among others.

In the Senate I've been watching Washington, California, Nevada, West Virginia, Pennsylvania, Wisconsin, Florida, Colorado, and Illinois-- these were the "competitive" races. Since my last forecast, some polls have diverged, and some polls have tightened; some haven't moved at all.

Remember that Republicans almost certainly will gain 3 Senate seats from Democrats to begin with. North Dakota, Arkansas, and Indiana. But one Republican seat is being added to the "competitive" list. Thus, Republicans need to win 8 more from the competitive list in order to take control of the Senate. Therefore we will begin with a baseline of 42 Republican seats and 48 Democrat seats; I see almost no viable way that either party could end up with less than this number of seats.

I have some hesitancy about this, but given a recent poll out of Alaska, I have re-evaluated the race in a post a few hours ago, and on the virtue of it being a 3-way race, I'll add it to the competitive list.

I will now note very likely Republican gains from the competitive list. I predict with confidence that these three races will end up in Republican hands. Florida (Rubio), Wisconsin (Johnson), and Pennsylvania (Toomey). You can tell your Democratic friends that these races are pretty much a lost cause.

And very likely Democratic gains include California. You can confidently tell your Republican friends that California is basically out of their hands.

That leaves us with a new baseline of 45 Republicans and 49 Democrats.

Here's how the rest break down. I wouldn't make a bet on this, but evidence points to these results: Republicans gain Alaska (Murkowski or Miller). Democrats gain West Virginia (Manchin).



The final four races are just too hard to predict from my standpoint. Republicans are favored in Illinois, Nevada, and Colorado albeit by small percentages. Democrats are favored in Washington. I think it's just as reasonable to allocate these seats 3-1 as it is to allocate them 2-2. They all might be very close.

I tend to want to call Washington to Murray(D) and Illinois for Kirk(R). Nevada and Colorado are even harder for me. Nevada's polling option to select "none of these candidates" may end up helping Reid(D), or... Angle(R). Angle may win, but if so, I think there's a good chance it'll be more narrowly than about 2 points, as polls now show. Democrats are going to be desperate to win Nevada as a symbolic victory, though I think politically it would help them to have a change in the face of leadership. I tend to think Tancredo's rise in Colorado might un-motivate more moderate Republican voters, which might help Bennet(D). But all of this is conjecture.

I think Dems will end up with 51 or 52 seats. This is consistent with my other forecasts. The margin will be between 50D-50R and 54D-46R (worst and best case scenarios, depending on your position). Thus, I confidently predict Democrats will retain the Senate absent a defection or two from more moderate Democratic caucus members.

On the other hand, Democrats will very likely lose in between 5 and 9 seats, probably closer to the latter number.

I want to make a quick note about the NC Senate race as well. Watch Burr's(R) winning margin. Marshall(D) has run a weak campaign, but Burr hasn't run much of a campaign at all. If Burr wins by less than 11 points then it will be a sign that Democrats may be doing better than expected. If he wins by much more, brace for significant Democratic losses in NC and elsewhere.