Senate Election Preview

I know everyone's talking about the House, and I will get to that later, but the Senate has quietly become very interesting, and if I had to call it now, I'd say the Republicans will win it back. They need to gain 10 seats, and here's how things look right now:

Arkansas, Indiana, and North Dakota: stick a fork in them, they're done. The Republican in Arkansas is up by an unbelievable 38 points. Net gain of 3 seats.

After this, the Republicans must win seven of the following seats. They are listed in order of GOP strength, and the numbers after each show the latest poll.

Delaware           48-37    up 11
Pennsylvania     47-37    up 10
Washington       52-45     up 7
Colorado            49-45     up 4
California          48-47     up 1
Wisconsin          47-46     up 1
Illinois               45-45       tie
Nevada              44-45   down 1
West Virginia     42-48   down 6
Connecticut       40-47   down 7
New York           32-51  down 19

A few points:

• I believe polls are underestimating Republican turnout, perhaps dramatically (some polls don't even determine whether someone is likely to vote). Three million more Republicans have turned out to vote in primaries this year. In 2006, three million more Dems voted. Six million is a GIANT swing, and the differential is the largest in 80 years. The "enthusiasm gap," measured by pollsters, is running on the order of 25 points, a level never seen before. Any Republican inside of 5 points on election day should win. By that metric, the Republicans pick up 11 seats (maybe even 12, as the West Virginia poll is quite stale, and that race has been trending towards the GOP).
• Every possible thing is going right for the Republicans, politically, and it's difficult to see how that will change before November. Virtually every one of these polls has been trending GOP. Look for the usual October Surprise from the Dems, though (remember Bush's DUI?).
• The results will be highly correlated. In other words, Republicans will do well as a group, or they won't. This will likely work in their favor. There are four seats that the GOP holds that the Dems are thought to have a fighting chance at (Alaska, Florida, Missouri, Ohio), but the closest the Dems are on any of these is 6 points (Alaska, Ohio). If this is really a "wave" year, none of these will be close.
• While Harry Reid in Nevada is winning by a point, he is only at 45%, well below the all-important 50% number for incumbents. That means that 5% of Nevadans have decided to vote against him, they just haven't figured out who the other guy is yet (in this case, woman).
• Connecticut is thought to be safe for the Dems, but I don't think so. Polls have been closing and the Republican, McMahon, has deep pockets. Also, financial reform is unpopular in Fairfield County, which is where all the money comes from.
• New York may be a stretch, but Gillibrand isn't particularly popular, and she's barely above 50%. Further, the Republican challenger isn't even known yet. Look for this race to close after the Republican primary on the 14th.

Intrade currently has the odds of a GOP Senate takeover at 28%. I would put it at 55%.