A sense of the Senate

With the primary season over, now is a good time to assess the chances that the Republicans will win control of the Senate. To do so, Republicans will need to pick up a net of ten seats.
The good news is that, following Kelly Ayotte’s victory in New Hampshire on Tuesday night over Tea Party favorite Ovide Lamontagne, the Republican Senate candidate is the clear favorite in every state where a Republican is the incumbent. This includes states like Florida, Kentucky, and Ohio that not long ago were thought to be in play.
Alaska may have become a bit more dicey due to Lisa Murkowski’s decision to launch a write-in campaign, which makes her an independent. But I continue to believe that, although Murkowski would have a shot at defeating Republican Joe Miller in an ordinary three-way race, it will be difficult for her to do so where it is possible to vote for her only through the write-in process.
Thus, with Republicans well situated to “hold serve,” the question becomes can they win at least ten seats now held by Democrats.
The bad news (for this purpose) is that Delaware, which looked like a near-certain pick-up before Christine O’Donnell’s emergence, is now unlikely to elect a Repubclian. O’Donnell trails her opponent by double digits, is considered qualified for the Senate by only about a third of the Delaware electorate, and may struggle to maintain even that low level of confidence in the wake of new revelations about her past.
But the Republicans can get to ten without winning in Delaware. The following four states seem poised to switch from a Democrat to a Republican: North Dakota, Arkansas, Indiana, and Pennsylvania. The task is to pick up six more.
The following six potential pick-up states are generally considered toss-ups: California, Washington, Nevada, Colorado, Illinois, and Wisconsin. Although each race is legitimately rated a toss-up, I like our chances in each except for Washington, which I think is 50-50 at best for the Republican candidate. But even if I’m not being too optimistic about each of these races individually, the odds of winning all six, and thus getting to ten this way, aren’t high.
This highlights the impact of Tuesday’s result in Delaware. A Castle win would have created a margin of error in the six “toss-up” races. The O’Donnell victory leaves no margin for error unless the Republicans can pick up a seat that the Democrats are expected to hold.
There are two main possibilities (three if you count Delaware) — West Virginia and Connecticut. In both, the Democrat leads in the polls by 5 to 10 percentage points. In West Virginia, the Democrat is popular, but the state is trending Republican. In Connecticut, the Democrat isn’t popular, but the state seemingly remains blue. (In Delaware, the Democrat is a proven vote-getter in the largest of the state’s three counties, the Republican is not widely viewed as credible, and the state seemingly remains blue).
In sum, the Republicans’ main road to control requires (1) holding all Republican held seats (likely); and (2) winning all six of the “toss-ups” (unlikely but not far-fetched) or (3) winning five of the six toss-ups plus one of West Virginia, Connecticut, or Delaware (unlikely but not far-fetched).
It would be foolish for me to attempt to quote odds on a Republican takeover. Nate Silver, who quotes them professionally, places the probability at 15 percent (down from 26 percent prior to O’Donnell’s victory). My hunch is that this figure is a bit too low.