Brandon Finnigan:
When I first took a look at the picture for the U.S. Senate, I saw fourteen exciting-to-mildly interesting Democrat-held seats and three Republican-held ones. Since December, it has been clear Senator Collins has no intention of retiring, so with that possibility axed, the number of seats on the GOP worth watching drops to two. Senator Tom Coburn announced his retirement, and the special election to fill his seat will be held concurrently with the scheduled midterm election. His seat is quite safe, so the primary will be where all of the action lies.
Bigger than the retirement or our removal of Collins from my watch list has been the developments in the Democrat-held states. With polling releases and a closer examination of the Gallup surveys (along with available exit poll data from ’08, ’10, and ’12), I am moving several vulnerable Democrats to the endangered list: Mary Landrieu of Louisiana and Kay Hagan of North Carolina are now projected to lose their seats by a close margin. In addition, Republican Terri Lynn Land, for now, appears to have the edge in Michigan, with fund-raising figures and polling strongly suggesting this race to be the best for Republicans in the Midwest. Despite a lack of polling, further review of the election data in 2008 along with fund-raising figures have convinced me to push Mark Begich onto the very vulnerable list: Alaska is now a toss-up. Right now, the Republicans are projected a net gain of seven seats, enough for a 52-48 majority in the Senate if Begich manages to hold on.
In addition, several other races have the potential to become interesting: a review of the polling data released from December onward, along with previous election cycle datasets, has compelled me to move New Hampshire and Colorado from moderate-Democrat to lean-Democrat. Former Senator Scott Brown will likely enter in the former, and has seen close polling in two out of the last three releases (PPP, Purple Strategies, and UNH), while Senator Udall’s weakened rating is the result of a closer examination of the polling data and previous cycles where partisan turnout was evenly matched, as what appears to be shaping up.
Lastly, we have two minor ratings changes.