ABC News
2010 Senate Rankings, Revised and Extended

Time to start over. Rather than giving you the third and final installment of our 2010 Senate Race Rankings, I’m going to re-do the project from scratch, ranking all 37 potential contests from top to bottom based on their probability of changing parties.

The reasons for this are twofold. Firstly, there’s been quite a bit of news in any number of these contests, meaning that things we thought we knew a couple of weeks ago may no longer apply. And secondly, there are enough contingencies in so many of these races (Will so-and-so retire? Will so-and-so run for office?) that a more systematic treatment of them is in order. What I mean by this is that it’s worth going through and calculating the conditional probabilities. In California, for example, the race will look much different if Arnold Schwarzenegger enters than if he does not, which means I’m going to show you something like this:

75%    Generic R vs Boxer (D)                       5%
25% Schwarzenegger (R) vs Boxer (D) 30%
--------------------------------------------------------
Combined Probability of Republican Pickup 11.3%

What does this mean? The numbers along the left-hand column indicate the probability of various scenarios occurring. In this case, I think there’s about a 25% chance that Schwarzenegger will run for the seat, and a 75% chance that he will not. The numbers in the right-handed column, then, represent the chances of the opposition party picking up the seat if each of these contingencies occurs; in this case, I think the Republicans have about a 30% chance of unseating Barbara Boxer if Schwarzenegger is the nominee, but only a 5% chance otherwise. I then multiply all these numbers together to give you the joint probability: Republicans have about an 11% chance of winning California’s seat, combining those times that Schwarzenegger runs with those times that he does not.

Where, you might be asking, do all these probabilities (30%, 5%, 25%) come from? They’re my educated guesstimates — nothing more and nothing less — based on a combination of approval ratings, the strength of each party’s organization in that state, rumor, gossip, innuendo, and various other intangibles. You should not regard the numbers as being especially precise or meaningful; they are ballpark estimates, a codification of my subjective knowledge.

One final reminder and caveat: the races are ranked strictly in order of their likelihood of changing parties. This is not quite the same thing a ranking of the most vulnerable incumbents, because it accounts for potentialities like retirements, as well as inferences about the probable strength of the opposition candidates. In Iowa, for instance, I assign the Democrats about a one-in-four chance of picking off Chuck Grassley’s seat, but nearly all of that stems from the (distinct) possibility that Grassley will retire — his seat should not be especially vulnerable if he decides to run for re-election.

1. Kansas (R-Open; Brownback retiring)

Believe it or not, this might be the most likely seat in the country to flip parties. And it’s all because of Governor Kathleen Sebelius, who announced Saturday that she was withdrawing herself from consideration for Barack Obama’s cabinet. This comes as a mild surprise; the scuttlebutt I’d heard is that Sebelius, who is term-limited in 2010, would have liked the opportunity to serve in Obama’s cabinet. Sebelius enjoys politics, and that she’s turned the opportunity down suggests she has something else up her sleeve. Unless that something is becoming DNC Chair, that probably means a senate run.

Sebelius is no shoo-in if she runs — Kansas hasn’t elected a Democrat to the Senate since the Great Depression, and KS-1’s Jerry Moran, who has already announced his intention to run for the seat, would be a credible opponent. But with her sky-high approval ratings, Sebelius is probably the favorite to win the seat. As Sam Brownback is likely to run for governor, it might also help her in terms of voter psychology if Kansans simply feel as though they’re swapping the two, rather than turning a traditionally Republican seat over to a Democrat.

In the event that Sebelius has other plans, the Democrats’ odds are daunting, as Kansas would likely revert back to its red-leaning steady state, though current lieutenant governor (and ex-GOPer) Mark Parkinson might have a fighting chance.

70%    Sebelius (D) vs Generic R                   60%
30% Generic D vs Generic R 5%
--------------------------------------------------------
Combined Probability of Democratic Pickup 43.5%

2. Pennsylvania (R-Specter)

Approval/Favorability Ratings: Arlen Specter (R)
Poll Date Approve Disapprove Net
Rasmussen 12/2 60 35 +25
Quinnipiac 11/26 56 23 +33
Strategic Vision 11/1 53 34 +19

There are a lot of ways that Arlen Specter could lose his seat.

First, obviously, his health is not good.

Second, he could get primaried out. As the Democrats made fairly massive voter registration gains in Pennsylvania throughout 2008, what remains of the Republican primary electorate is the conservative base, and the very conservative Pat Toomey is threatening to challenge Specter again after having lost to him by just 2 points in the Republican primary in 2004.

Third, the Democrats might be able to beat Specter straight up. Although Specter’s approval ratings remain strong, the fact that Chris Matthews has been able to poll close to him in some recent surveys suggests that much of that support is soft, making him vulnerable against the right Democrat (which might or might not be Matthews). Specter is in an unenviable position: if he obstructs Barack Obama’s agenda, he’ll anger Democrats in this structurally blue state, but if he sides with Democrats too often, he’ll be vulnerable to a primary challenge.

60%    Generic D vs Specter (R)                  22.5%
40% Generic D vs Generic R 70%
--------------------------------------------------------
Combined Probability of Democratic Pickup 41.5%

3. Ohio (R-Voinovich)

Approval/Favorability Ratings: George Voinovich (R)
Poll Date Approve Disapprove Net
Strategic Vision 11/1 53 37 +16
SurveyUSA 10/18 51 39 +12
Quinnipiac 8/8 51 32 +19

This parallels the Specter race in certain respects, winnable either through retirement (Marc Ambinder has suggested that Voinovich might hang it up) or through a straight-up takeover (Voinovich is in fact a bit less popular than Specter). But there are a couple of mitigating factors. Firstly, there’s no indication that Voinovich will face a serious primary challenge, which removes at least once concern for him. And secondly, if Voinovich retires, the Republicans have a fairly strong potential alternative in Rob Portman, who has suggested he might be interested in the position.

With that said, Voinovich still probably represents the Republicans’ best chance of holding onto their seat. The structural advantages of incumbency are worth something, whereas Voinovich’s mediocre approval ratings likely reflect the tough climate for Ohio Republicans in general right now rather than anything about Voinovich in particular; those cooties would probably transfer to any potential Republican alternative, in other words. Portman, meanwhile, while a capable administrator, is relatively unproven electorally.

60%    Generic D vs Voinovich (R)                  30%
20% Generic D vs Portman (R) 50%
20% Generic D vs Generic R 60%
--------------------------------------------------------
Combined Probability of Democratic Pickup 40.0%

4. Florida (R-Open; Martinez retiring)

I expected Florida to rank somewhere in the top two, but upon further review, Jeb Bush’s potential entry into the race complicates things for Democrats, as Bush has star power that no Democrat in the state can match and as he left Florida’s governorship with his popularity largely intact. Furthermore, I indeed think Bush is fairly likely to run, as among other things he might feel compelled to restore dignity to his surname.

67%    Generic D vs Bush (R)                       30%
33% Generic D vs Generic R 50%
--------------------------------------------------------
Combined Probability of Democratic Pickup 38.3%

5. Kentucky (R-Bunning)

Approval/Favorability Ratings: Jim Bunning (R)
Poll Date Approve Disapprove Net
SurveyUSA 10/18 40 44 -4

It’s easy to understand why Democrats are salivating over this opportunity. Bunning nearly lost in 2004, a relatively good cycle for Republicans, against a fairly marginal Democratic opponent; he’s now six years older and no more popular than he was before. Still, this is one state that moved away from Democrats in 2008 (although they retain a party identification advantage) and one state in which campaigning against Obama might pay dividends. Bunning’s goal, in other words, is to turn this into a referendum on the President rather than a referendum on himself.

The Democrats’ odds of a pickup improve significantly if Ben Chandler decides to enter, although Chris Cillizza thinks he’s somewhat unlikely to. Republican odds might go up, frankly, if Bunning decides to retire, although for now he appears game for another go-around.

60%    Generic D vs Bunning (R)                    30%
20% Chandler (D) vs Bunning (R) 50%
10% Generic D vs Generic R 25%
10% Chandler (D) vs Generic R 50%
--------------------------------------------------------
Combined Probability of Democratic Pickup 35.3%

Approval/Favorability Ratings: Harry Reid (D)
Poll Date Approve Disapprove Net
Research 2000 11/24 38 54 -16
Mason-Dixon 8/14 46 43 +3

I wanted to rank this race higher, since Reid’s approval ratings are just awful, and are unlikely to be improved by the flak he’ll take as the Majority Leader. But structurally speaking, the Republicans aren’t in the position that they’d like to be in the Silver State. Nevada is turning blue very quickly, with the Democrats having made huge voter registration gains in 2008, and the Republicans have recruitment problems, as among their two most likely candidates, John Porter just lost his House seat, and as Lieutenant Governor Brian Krolicki was just indicted. The most probable outcome is an ugly, expensive win for Reid, although I give the Republicans’ slightly better than a one in three chance of finding a compelling candidate somewhere and taking Reid down. Reid also might be vulnerable to a primary challenge, which could either help or hinder the Democrats’ chances depending on the circumstances.

Probability of Republican Pickup                  35.0%

7. Missouri (R-Bond)

Approval/Favorability Ratings: Kit Bond (R)
Poll Date Approve Disapprove Net
SurveyUSA 10/18 51 42 +9
PPP 8/20 44 43 +1

I may have underrated this race originally, as various trial heats show any of several Democrats polling within the margin of error against Bond. Retirement is also a possibility. Still, Missouri is not quite the swing state it once was.

80%    Generic D vs Bond (R)                     27.5%
20% Generic D vs Generic R 50%
--------------------------------------------------------
Combined Probability of Democratic Pickup 32.0%

8. North Carolina (R-Burr)

Approval/Favorability Ratings: Richard Burr (R)
Poll Date Approve Disapprove Net
PPP 6/17 27 26 +1
SurveyUSA 11/2006 44 42 +2

Burr occupies North Carolina’s buyer’s remorse seat, which has changed parties five times in the past five elections; his tepid approval ratings suggest that he might not escape the curse. Still, Tarheel Democrats won’t have the benefit of Barack Obama’s coattails in 2010, nor are they likely to run into another campaign run as ineptly as Elizabeth Dole’s.

Probability of Democratic Pickup                  25.0%

9. Iowa (R-Grassley)

Approval/Favorability Ratings: Chuck Grassley (R)
Poll Date Approve Disapprove Net
SurveyUSA 10/18 63 29 +34
Selzer 2/19 67 18 +49

This has moved up slightly on retirement rumors, and also rumors that Grassley might get a serious challenge (e.g. from Tom Vilsack or Bruce Braley) even if he doesn’t drop out. I remain steadfastly unpersuaded that Democrats have much of a shot at winning this seat unless Grassley leaves it open; getting a +34 approval score with the way the US Congress is perceived nowadays is no small accomplishment. With that said, if someone like Vilsack were to form an exploratory committee, he might conceivably bluff Grassley into retiring (or at least into lining up with Barack Obama on cloture votes).

70%    Generic D vs Grassley (R)                    5%
30% Generic D vs Generic R 70%
--------------------------------------------------------
Combined Probability of Democratic Pickup 24.5%

10. New Hampshire (R-Gregg)

Approval/Favorability Ratings: Judd Gregg (R)
Poll Date Approve Disapprove Net
Strategic Vision 10/28 53 37 +16
UNH 7/16 53 23 +30
Dartmouth 4/30 38 27 +11

With Jeanne Shaheen having defeated John Sununu on November 4th, Gregg is now the only Republican among New Hampshire’s four-person congressional delegation, leading a lot of Democrats to assume that he will be their next victim. Gregg’s approval ratings, however, are considerably stronger than Sununu’s were, and the Democrats don’t offer any opponent with Shaheen’s name-brand. New Hampshire, moreover, maintains its libertarian streak, and Gregg’s fiscal conservatism may play well in the current economic environment, although his having voted for the bailout could prove a liability.

Probability of Democratic Pickup                  22.5%

11. Delaware (D-Open; Kaufman retiring)

Republican hopes probably rely on At-Large Representative Mike Castle running for the seat, something which I haven’t heard one way or the other about, but which might not be entirely implausible since Castle is too old to wait for a better opportunity. If Beau Biden throws his hat in the ring for 2010, as is somewhat likely, that might be a deterrent to Castle.

50%    Generic R vs Biden (D)                       5%
25% Generic R vs Generic D 7.5%
15% Castle (R) vs Generic D 50%
10% Castle (R) vs Biden (D) 40%
--------------------------------------------------------
Combined Probability of Republican Pickup 15.9%

Approval/Favorability Ratings: Ken Salazar (D)
Poll Date Approve Disapprove Net
PPP 8/6 39 36 +3
SurveyUSA 11/2006 56 36 +20

This remains one of the theoretically more viable pickup opportunities for Republicans, especially if the background political environment beings to swing in their favor, but Colorado is another state that’s turning blue in a hurry, and there’s no one Republican opponent that Colorado Democrats seem particularly worried about. Ex-Gov Bill Owens might be the closest thing to an exception, but he’s declined his last couple of chances to run for national office.

Probability of Republican Pickup                  15.0%

13. Illinois (D-???)

It’s hard to get good information on what Rod Blagojevich is likely to do with his senate appointment, although one source suggests that Blagojevich is inclined to pick an African-American. If so, Blagojevich would be taking a significant risk if he appointed Jesse Jackson Jr., who might be the plurality choice of Illinois Democrats, but who has little support outside of his base. That said, the Illinois Republican bench is incredibly weak, and even Jackson would probably be a favorite to win re-election.

70%    Generic R vs Generic D                     30%
30% Generic R vs Jackson (D) 10%
--------------------------------------------------------
Combined Probability of Republican Pickup 15.0%

14. Texas (possibly R-Open or R-??)

This is Kay Bailey Hutchinson’s seat. I’m going to put this on our radar screen because on Friday, Hutchinson launched an exploratory committee to challenge Rick Perry for governor, and is better than even money to resign from the Senate to pursue that opportunity; one TV station goes so far as saying that she’s “confirmed” a bid for the governorship. This makes all the sense in the world for Hutchinson: she’s quite likely to beat Perry in the primary, and if she prevailed in the general, would become a star of her party as governor of the largest red state in the land. If this happens, Democrats have a couple of potentially intriguing candidates: TX-17’s Chet Edwards, and Houston Mayor Bill White, who was thought to be contemplating a run for governor but who could re-direct his efforts to the Senate as Hutchinson represents a more formidable opponent than Perry.

33%    Hutchison keeps Senate Seat                 --
67% Generic D vs Generic R 20%
--------------------------------------------------------
Combined Probability of Democratic Pickup 13.3%

15. Arizona (R-McCain)

Approval/Favorability Ratings: John McCain (R)
Poll Date Approve Disapprove Net
Research 2000 10/29 58 41 +17
SurveyUSA 11/2006 61 34 +27

This race was originally to have rated much higher, but Janet Napolitano appears to be out of the running after being nominated to head Homeland Security, whereas John McCain now says he’ll be up for re-election. It is unlikely that an up-and-coming Democrat like Gabrielle Giffords would want to risk her stature by running against McCain, especially when Arizona also has an open governor’s seat.

With that said, some Republican observers were surprised by the speed with which McCain declared his interest in the 2010 race. His relationships with the Senate Republican caucus are lukewarm, and McCain may find he has less interest in the banalities of the Senate after spending the better part of a year away from the chamber. Retirement, in other words, remains a possibility, in which case the race should draw a credible Democratic challenger, although popular Republican congressman Jeff Flake would likely be the favorite.

75%    Generic D vs McCain (R)                      5%
25% Generic D vs Generic R 35%
--------------------------------------------------------
Combined Probability of Democratic Pickup 12.5%

16. Arkansas (D-Lincoln)

Approval/Favorability Ratings: Blanche Lincoln (D)
Poll Date Approve Disapprove Net
Arkansas Poll 10/11 54 20 +34
SurveyUSA 11/2006 55 37 +18

This is another race that I planned to rank higher originally, but the buzz I’m hearing is that Mike Huckabee is disinclined to be interested in the seat. In theory, this should be an intriguing seat to Republicans even if Huckabee does not run, as Arkansas is turning substantially redder. But since Lincoln’s approval ratings are strong and since the Republicans just let Mark Pryor run unopposed, I wouldn’t hold my breath.

90%    Generic R vs Lincoln (D)                     7%
10% Huckabee (R) vs Lincoln (D) 60%
--------------------------------------------------------
Combined Probability of Republican Pickup 12.3%

17. California (D-Boxer)

Approval/Favorability Ratings: Barbara Boxer (D)
Poll Date Approve Disapprove Net
SurveyUSA 10/18 49 42 +7
PPIC 9/13 44 39 +5
Field Poll 5/22 48 31 +17

An Arnold Schwarzenegger entry would certainly attract a lot of fanfare, and a poll taken a year ago had him running even with Barbara Boxer. However, the Governator’s approval ratings have since slumped significantly, and there’s no tangible sign that he’s interested in the position. Apart from Schwarzenegger, there isn’t another obvious top-tier Republican opponent.

75%    Generic R vs Boxer (D)                       5%
25% Schwarzenegger (R) vs Boxer (D) 30%
--------------------------------------------------------
Combined Probability of Republican Pickup 11.3%

18. North Dakota (D-Dorgan)

Approval/Favorability Ratings: Byron Dorgan (D)
Poll Date Approve Disapprove Net
SurveyUSA 11/2006 75 21 +54

Byron Dorgan is probably too popular to elicit a challenge from Governor John Hoeven, who just got re-elected and can afford to wait for a better cycle. If Hoeven were to run, however, we’d have a barnburner on our hands.

80%    Generic R vs Dorgan (D)                      2%
20% Hoeven (R) vs Dorgan (D) 45%
--------------------------------------------------------
Combined Probability of Republican Pickup 10.6%

19. New York (D-??? née Clinton)

One rumor we’ve heard is that Rudy Giuliani is concerned about his ability to defeat David Paterson and might run for Senate instead if Paterson appoints a no-name candidate (e.g. anyone not named “Cuomo” or “Kennedy”).

50%    Generic R vs Cuomo or Kennedy (D)            2%
35% Generic R vs Generic D 10%
15% Giuliani (R) vs Generic D 40%
--------------------------------------------------------
Combined Probability of Republican Pickup 10.5%

20. Connecticut (D-Dodd)

Approval/Favorability Ratings: Chris Dodd (D)
Poll Date Approve Disapprove Net
Hartford Courant 10/29 43 46 -3
Quinnipiac 6/28 51 34 +17

Ordinarily, an opponent with a negative approval rating would be a prime target, but among the two Republicans with some statewide cachet, Jodi Rell has no reason to leave the governor’s mansion, and Chris Shays just lost his House seat and says he isn’t interested in the Senate. I’m assigning a 10 percent probability of a Republican victory as a default, but they’ll need to find a candidate first.

Probability of Republican Pickup                  10.0%

21. Oklahoma (R-Coburn)

Approval/Favorability Ratings: Tom Coburn (R)
Poll Date Approve Disapprove Net
Sooner Poll 7/21 59 24 +35
SurveyUSA 11/2006 50 38 +12

Probably won’t be competitive, but Democrats have a couple of contingencies to contemplate: (i) Democratic Governor Brad Henry is ridiculously popular and will be term-limited in 2010, although he had indicated in the past that he wasn’t interested in running for the Senate; (ii) Coburn is less than 100% positive that he wants to run for re-election.

70%    Generic D vs Coburn (R)                      1%
15% Henry (D) vs Coburn (R) 30%
10% Generic D vs Generic R 10%
5% Henry (D) vs Generic R 50%
--------------------------------------------------------
Combined Probability of Democratic Pickup 8.7%

22. Wisconsin (D-Feingold)

Approval/Favorability Ratings: Russ Feingold (D)
Poll Date Approve Disapprove Net
Strategic Vision 11/1 57 33 +24
SurveyUSA 10/18 53 43 +10
WisPolitics.com 6/9 59 29 +30

Feingold does not look at all vulnerable now, but there is a chance he could become so if the mood of the country changes. WI-1’s Paul Ryan, who has a history of overperforming in his district, would probably need to be the challenger.

75%    Generic R vs Feingold (D)                    3%
25% Ryan vs Feingold (D) 20%
--------------------------------------------------------
Combined Probability of Republican Pickup 7.3%

23. Hawaii (D-Inouye)

Approval/Favorability Ratings: Daniel Inouye (D)
Poll Date Approve Disapprove Net
SurveyUSA 11/2006 68 26 +42

Inoyue is a local institution and will not be defeated unless he retires or is unable to run for office, although that is not entirely unlikely since he’ll be 86 years old by 2010. The Republicans will then need Governor Linda Lingle to run (who will be term-limited in 2010), who still might might be a slight underdog against Mazie Hirono or Neil Abercrombie in this very blue state. It’s a longshot parlay for the GOP, but not inherently an impossible one.

75%    Generic R vs Inoyue (D)                    0.1%
12.5% Lingle (R) vs Generic D 45%
12.5% Generic R vs Generic D 5%
--------------------------------------------------------
Combined Probability of Republican Pickup 6.3%

Approval/Favorability Ratings: Lisa Murkowski (R)
Poll Date Approve Disapprove Net
Hays Research 3/11 63 33 +30
SurveyUSA 11/2006 54 41 +13

If there’s action here, it’s liable to be in the primary rather than in the general election. Most Republicans with aspirations for national office would prefer to be a governor rather than a senator. But Sarah Palin is potentially an exception, as Alaska’s literal and figurative distance from Washington make it difficult for her to expand her network or her brand. Meanwhile, governance is likely to become substantially harder in Alaska if oil prices remain low, as the state relies on oil royalties for the bulk of its revenue.

Lisa Murkowski, in other words, is not being entirely paranoid when she’s warning Palin to stay away from her seat. The potential opening for Democrats is if Murkowski is bloodied but not beaten by Palin, either surviving a tough primary challenge, or perhaps losing credibility by getting in a public spat with the ‘Cuda. This is also, in theory, the sort of seat a third-party candidate might target.

80%    Generic D vs Murkowski (R)                 7.5%
20% Generic D vs Palin (R) 1%
--------------------------------------------------------
Combined Probability of Democratic Pickup 6.2%

25. Georgia (R-Isakson)

Approval/Favorability Ratings: Johnny Isakson (R)
Poll Date Approve Disapprove Net
PPP 11/24 30 25 +5
Strategic Vision 11/1 50 38 +12

Although Isakson’s approval scores are not terrific, this race has been demoted significantly as a result of Saxby Chambliss’ strong performance in Georgia’s runoff this week. The Democrats, moreover, lack strong candidates in Georgia, and will need some time to build a bench.

Probability of Democratic Pickup                  6.0%

26. Louisiana (R-Vitter)

Approval/Favorability Ratings: David Vitter (R)
Poll Date Approve Disapprove Net
SMOR 4/2 52 32 +20

As we saw on Saturday night, the electoral environment is pretty brutal for Democrats in Louisiana right now, and unless I see polling to convince me otherwise, I don’t see any reason to conclude that Vitter is anything other than what he seems, which is a fairly popular Republican incumbent in a red state. LA-3’s Charlie Melancon, the only remaining Democrat to represent Louisiana in the House, is considered the most credible opponent, but unless he sees himself as a potential victim of redistricting (Louisiana is likely to lose a House seat once the Congress is redrawn in 2012), it would be a lot to ask him to give up his seat.

Probability of Democratic Pickup                  5.0%

27. Maryland (D-Mikulski)

Approval/Favorability Ratings: Barbara Mikulski (D)
Poll Date Approve Disapprove Net
SurveyUSA 11/2006 57 35 +22

If re-elected, Mikulski would become the longest-serving female senator in history, which might be the extra incentive she needs to steer her away from retirement; Republicans face long odds even if she quits.

85%    Generic R vs Mikulski (D)                    2%
15% Generic R vs Generic D 10%
--------------------------------------------------------
Combined Probability of Democratic Pickup 3.2%

28. South Carolina (R-DeMint)

Approval/Favorability Ratings: Jim DeMint (R)
Poll Date Approve Disapprove Net
SurveyUSA 2/19 49 37 +12

Although DeMint’s approval ratings are marginal enough to create some wiggle room for a challenge, Democrats have a weak bench in South Carolina.

Probability of Democratic Pickup                  3.0%

29. Washington (D-Murray)

Approval/Favorability Ratings: Patty Murray (D)
Poll Date Approve Disapprove Net
Strategic Vision 11/1 55 37 +18
SurveyUSA 10/18 55 35 +20

Murray is fairly entrenched, and Washington isn’t really behaving as a swing state of late.

Probability of Republican Pickup                  2.5%

30. Vermont (D-Leahy)

Approval/Favorability Ratings: Pat Leahy (R)
Poll Date Approve Disapprove Net
Research 2000 10/2007 67 29 +38
SurveyUSA 11/2006 71 25 +46

There’s a small chance that Pat Leahy could retire, in which case Republican Governor Jim Douglas could run. Otherwise, the seat is untouchable.

90%    Generic R vs Leahy                        0.01%
6% Generic R vs Generic D 2%
4% Douglas vs Generic D 40%
--------------------------------------------------------
Combined Probability of Democratic Pickup 1.7%

31. Alabama (R-Shelby)

Approval/Favorability Ratings: Richard Shelby (R)
Poll Date Approve Disapprove Net
SurveyUSA 10/18 61 32 +29

The Democrats have some candidates like Arthur Davis that might be compelling in another state, but are highly unlikely to be viable in Alabama, especially against a popular incumbent like Shelby. Shelby could retire, which might give the Dems an outside chance, but likely not more than that.

85%    Generic D vs Shelby (R)                    0.5%
15% Generic D vs Generic R 5%
--------------------------------------------------------
Combined Probability of Democratic Pickup 1.2%

32. South Dakota (R-Thune)

Approval/Favorability Ratings: John Thune (R)
Poll Date Approve Disapprove Net
Mason-Dixon 10/14 64 34 +30
SurveyUSA 11/2006 61 35 +26

One day, At-Large Representative Stephanie Herseth-Sandlin might take a run at this seat, but she has no reason to do so now while Thune remains highly popular, especially when South Dakota will also have an open governor’s seat.

Probability of Democratic Pickup                  1.0%

33. Indiana (D-Bayh)

Approval/Favorability Ratings: Evan Bayh (D)
Poll Date Approve Disapprove Net
SurveyUSA 2/4 58 32 +26

Bayh generally wins his races by large margins, and with Indiana turning bluer, this cycle is unlikely to break the pattern.

Probability of Republican Pickup                  0.5%

34. Oregon (D-Wyden)

Approval/Favorability Ratings: Ron Wyden (D)
Poll Date Approve Disapprove Net
SurveyUSA 10/18 54 34 +20
Davis/Hibbitts 12/2007 45 19 +26

Wyden is quite popular; any credible Republican opposition will gravitate toward the governor’s seat instead, where Ted Kulongoski will be term-limited.

Probability of Republican Pickup                  0.2%

35. Utah (R-Bennett)

Approval/Favorability Ratings: Bob Bennett (R)
Poll Date Approve Disapprove Net
Dan Jones 1/2007 67 18 +49
SurveyUSA 11/2006 58 32 +25

Bennett might retire, in which case the Democrats will find out that Utah is still Utah.

85%    Generic D vs Bennett (R)                  0.01%
15% Generic D vs Generic R 1%
--------------------------------------------------------
Combined Probability of Democratic Pickup 0.2%

36. New York (D-Schumer)

Approval/Favorability Ratings: Chuck Schumer (D)
Poll Date Approve Disapprove Net
Marist 10/22 55 33 +22
SurveyUSA 10/18 60 31 +29
Quinnipiac 8/2 61 23 +38

As this seat is now protected both by a governor’s race and by a second, more vulnerable senate seat, Schumer should face only token opposition.

Probability of Republican Pickup                  0.1%

37. Idaho (R-Crapo)

Approval/Favorability Ratings: Mike Crapo (R)
Poll Date Approve Disapprove Net
SurveyUSA 7/2007 61 26 +35

It’s pronounced Cray-poh, so you can take your Larry Craig jokes elsewhere.

Probability of Democratic Pickup                 0.01%

Summary

Even if momentum has swung somewhat against the Democrats by 2010, they remain in a strong position to gain seats in the Senate. They are pretty much locked into reasonably competitive races in Pennsylvania, Ohio, Florida, Kentucky, Missouri, North Carolina and New Hampshire, will probably get one in Kansas, and could potentially get one in Iowa and perhaps even Texas or Arizona depending on retirements. By contrast, the only near-certain competitive Democratic seat is in Nevada; beyond that, the Republicans are looking at a series of second- and third-tier races like Colorado and Wisconsin, or races that require a white knight like Arkansas, California, North Dakota and New York. Overall, our guesstimate is that the Democrats will pick up 3-4 Republican seats and the Republicans 1-2 Democratic seats, for a net swing of 1-3 seats for the Democrats, potentially placing them beyond the 60-seat threshold. Democratic gains in the Senate, indeed, are possible or even likely even if they lose ground in the House. Although much of this simply reflects the staggered nature of US Senate seats — it’s Democrats who will have their backs to the wall in 2012 — there are some deeper implications as well, which we will explore at a later date.

Nate Silver is the founder and editor in chief of FiveThirtyEight.