TL;DR summary: currently a wash presuming continued economic improvement and Tea Party dropoff. D gains in Kentucky, New Hampshire, North Carolina, and Ohio, R gains in Arkansas, Delaware, Indiana, and Nevada, Crist wins in Florida, but Lieberman leaves the caucus.
Back in April, I went through the Senate races and made my predictions on what I expect to see happening in November. I'm still comfortable with my basic conditions - the economy will continue to improve (a natural recovery from the recession, smart moves by the Democrats to head it off, and jitters about Europe convincing people to shift back to the US), not dramatically but steadily, and the Tea Party will prove unable to either draw in new members or sustain its own intensity as the general election season forces Tea Party candidates to disavow their more extremist ideas.
And now, to the races, with my Bold Prediction at the end!
Arizona: John McCain has surrendered all pretense of principle and is running for the right as hard as he can, openly abandoning any hope of Arizona Republicans being able to draw substantial Hispanic support for generations to come. J. D. Hayworth is scaring the pants off of McCain, but it looks like McCain has kept himself in the driver's seat in the primary for now. I think it comes down to a closer-than-expected primary, but McCain ends up safe, then easily wins the general. Even Hayworth only moves this from strong Republican hold to moderate Republican hold; there's just too big a polling gap between even Hayworth and any of the Democrats to expect to overcome (not too big to potentially overcome, but too big to expect to overcome).
Arkansas: With the runoff between Halter and Lincoln still to come, I like Halter's chances pretty well. The question becomes whether Morrison's voters come out to support Halter, or just stay home. If they stay home and voters drop off at the same rate for both Lincoln and Halter, Lincoln could still win. Given Halter's success at closing the gap right before the primary, though, I think Halter takes the runoff and goes up against Boozman. Unfortunately, I'd say it would take a great campaign by Halter, a couple of serious bungles by Boozman, and a strong economic recovery to get Halter over the finish line in first place. Never underestimate the power of a Republican politician to sabotage himself with a rentboy scandal, but I've got to give this one a moderate Republican pick-up.
California: In the depths of the economic downturn, Barbara Boxer had a slump of her own. Now that the Republicans have been beating up on each other and the economy has been picking up, and with the Arizona immigration laws reminding Hispanic Californians why Republican control of anything is a bad idea, Boxer recovers and finishes with a solid (though tighter than usual) win. Moderate Democratic hold.
Colorado: Bennet looks to me like he ought to be able to win his primary, while the Republican primary is quite a bit closer. In either case, though, Bennet has been finding his political feet and getting his numbers moving up. Norton is the standard conservative on the Republican side, while Ken Buck has the Tea Party support and the anti-immigrant fervor. I'm going to say that Ken Buck ends up taking the primary and then flames out in the general, exciting Colorado Hispanics to come out against him just as the Tea Party sputters. Norton makes it a neck-and-neck race to the end. All told, I give it a slight Democratic hold.
Connecticut: Blumenthal's recent hit-job in the NY Times is going to be the Bill Ayers story of the race; the right-wingers will continue to scream about it and convince themselves that everyone is up in arms over it because everyone they listen to is up in arms over it. Then when the rest of the state goes 'meh' over it (and, in the full video, he did quite clearly state his position in the Marine Reserves), the Republicans will be convinced that the election must have been stolen in order for Blumenthal to have racked up such a convincing win. Even an economic downturn wouldn't hurt this one. Safe Democratic hold.
Delaware: With Beau Biden not involved, it looks like Castle has too much of a lead and too much pseudo-incumbency as an existing statewide federal officeholder to overcome even with an economic recovery, and he hasn't been depending on Tea Party support. Hounding him all campaign long on whether he'd support Republican filibusters every week might make good media, but he doesn't have the rabid-right supporters he has to show absolute fealty to all the time, so that could backfire if he makes a show of reasonableness. At 70 years, he's old but not Byrd-old; he could take the seat and have some ideas about holding it for a while if he takes care to maintain a moderate image. Castle could well join Brown, Snowe, and Collins in the 'have to cross over sometimes to appease my Democratic state' Republican caucus. Strong Republican pick-up, unfortunately.
Florida: Oooh, the fun one. Maybe not so fun as it was before Crist made his decision, though. The parallel to 2006 Connecticut isn't as clear as it might be; in Florida Democrats actually have a registration edge while in Connecticut Republicans were badly outnumbered in the first place. If Florida Democrats hung tough, Meek would be favored to win. The problem is, they don't seem to be doing that so far, and Crist looks likely to keep making Democrat-pleasing moves for the rest of his term as governor. If Meek can't get any oxygen and the Tea Party continues to fade as a political force, I see a lot of room for Crist to win - and the question becomes who he'll caucus with. If Meek is still languishing under 25% by July, the debate might start about asking him to drop out and let Crist be the de facto Democratic nominee. If Crist's fundraising dries up, though, then he might fade away and let Meek recover. Still, I think it looks like Crist is the favorite at the moment.
Illinois: Polling has been all over the place while Giannoulias dealt with issues around a bank he was involved with, but as the election proceeds and the economy rebounds, expect to see this one slide off the radar into a solid Democratic hold. If there's any doubt with a month to go, expect to see President Obama show up to use his star power in the state to bring it home. Strong Democratic hold.
Indiana: While it's not as red as it was, and there's certainly still a lot of low-hanging fruit here in terms of voter registration and Democratic organization, Obama's 2008 win doesn't remake Indiana overnight. Coats has lots of room to fall, but Ellsworth would have to run an excellent campaign and still get some blunders by Coats to overcome the lead there and the partisan inclination of the state, even with an economic recovery. Moderate Republican pick-up.
Kentucky: A change of heart here, as Rand Paul's seriously damaging last couple of days makes me rethink my earlier prediction that Kentucky is just too red to elect a Democrat to the Senate. As the biggest face of the Tea Party, coming out in favor of racism will help erode his own movement and alienate Grayson's supporters (even more than they already were; polls say that 43% of Grayson supporters said they would absolutely vote against Paul). As his radical and dangerous ideas about remaking the country get more air time, this race gets easier and easier. Conway with a moderate pickup.
Louisiana: Diaper Dave Vitter isn't seriously challenged in this reddening state. Charlie Melancon is unable to tar Vitter with the environmental disaster (har har har) and Vitter wins easily.
Missouri: Lots of close races in Missouri over the last few years, and Carnahan-Blunt will probably be another one. The polls have shown things close, with Blunt currently in a small lead. With the Democratic tide rising by November, that lead evaporates. Toss-up with the slightest of edges to Blunt given that Obama did not win the state in 2008, a good Democratic year.
Nevada: Too hard, too soon? Harry Reid may have made a strategic blunder in hammering Sue Lowden before the primaries on her 'Chickens for Checkups' plan. If she survives the primary to face Harry Reid, then he stands an excellent chance of retaining his seat. The problem is, she's so badly damaged by her own bizarre statements that I think she loses the primary now. Against another opponent without such easy flaws to hammer on, Reid's poor poll numbers are probably too much for even his huge warchest, clout, and an improved economy to overcome. The presence of a third-party candidate on the right could make things more interesting, but for now I've got to go with moderate Republican pick-up.
New Hampshire: The primaries aren't for a while yet, and the only polling on the race so far has been in what Dean Barker refers to as 'Cloud Hampshire,' where Republicans still outnumber Democrats by a 60-40 margin (summing up the 'Tea Party' and 'Republican' mutually-exclusive options in a recent Granite State Poll and comparing those to the 'Democrat' option). So there's little reliable information on the race, but re-weighting that universe to something closer to parity makes it appear that Hodes and Ayotte run close, while Hodes has a definite lead over any of the other contenders. Ayotte has been losing ground to the Tea Party candidates in the race, and has still fought to keep from identifying her own positions on issues. With popular governor John Lynch at the top of the ticket and a national rebound, Hodes should take the lead, but without any decent polling information so far I'm going to have to say toss-up still, with a slight edge to Hodes.
North Carolina: Burr's approval numbers continue to look pretty awful, hitting a net of 10% lately. With the Democratic primary runoff still to come, I think Elaine Marshall is in the driver's seat - and since she polls near Burr now, an improved Democratic climate in November would put her over the top, with a slight Democratic pick-up.
Ohio: A close race, but with an improved economy taking the headwind out of the picture, Lee Fisher successfully overcomes Rob Portman with a steady reminder that Portman was part of the team that wrecked the economy in the first place. With a currently competitive race, improved national conditions should make this a moderately likely Democratic pick-up.
Pennsylvania: With Arlen Specter off the November ballot, and Toomey running hard to the right, this becomes a likely Democratic hold (and an upgrade over Specter, whose reliability after the primary was going to be doubtful). Toomey is just too much Santorum for Pennsylvania to take, even if the economy flattens out the rest of the year - very likely Democratic hold.
Utah: Not much to see here, and only notable because of the off chance that Bennett decides to run as a write-in candidate. Even if he does so, I ultimately have a hard time seeing him pick off enough Republican votes to let a Democratic candidate sneak through. Easy Republican hold.
General: Again, all of this assumes that the economy will remain on a general upward slope this year, and that the Tea Party fades as its wilder ideas get more notice. If the economy improves even more, or people respond more positively to the improvement we do get than I expect, then we might see as much as a gain of +4 (with Arkansas, Indiana, and Nevada coming back home and Missouri flipping). I don't think those are likely, though. On the flipside, if the economy flops again or if scandals start coming out the woodwork, then Colorado, Florida, Kentucky, North Carolina, New Hampshire, and Ohio might go the other way. I think it'd have to be seriously bad for California, Connecticut, Illinois, and Pennsylvania to end up in the R column. That puts the 'reasonable upside' at +4, with a 'reasonable downside' at -6, but the median expectation at 0.
North Carolina's seat continues flipping back and forth as the state continues purpling, making that one a toss-up in the future as well. Delaware is a distinct mismatch of a Republican Senator in a fairly Democratic state. Ohio remains a battleground for the future as well. Kentucky gains a Democratic Senator in contrast to its own red nature. Florida gets an independent and could go anywhere in the future. Missouri also remains a battleground (though not a bellwether). All the other outcomes generally align national-election inclinations with officeholders.
Bold Prediction (winning party listed, with +0 for hold, +1 for gain):
Arizona: R+0
Arkansas: R+1
California: D+0
Colorado: D+0
Connecticut: D+0
Delaware: R+1
Florida: I+1 (Crist)
Illinois: D+0
Indiana: R+1
Kentucky: D+1
Louisiana: R+0
Missouri: R+0
Nevada: R+1
New Hampshire: D+1
North Carolina: D+1
Ohio: D+1
Pennsylvania: D+0
Utah: R+0
And for my Wild Random Event, I'm going to stick with my 'Lieberman leaves the caucus' prediction from April. Presented with an ultimatum of being stripped of his chairmanship if he doesn't commit to not running as a Republican, he leaves the caucus in a fit of pique.