Well, it's that time again. On Tuesday, everybody will vote, and that means that I need to put myself out there again with some predictions. Last time around, I did quite well. I messed up only one Senate race (in Montana) and ten House races, one of which was Dem-on-Dem, but correctly predicted a Democratic gain of 8 in the House. This time, I'm predicting the following:
Democrats lose 7 in the Senate
Democrats gain 3 governorships
Democrats lose 6 in the House
I'm going to begin with the Senate, where quite a few races are nailbiters. Races which some predicted becoming competitive (Minnesota, Virginia, Oregon, and New Mexico) have been put away by solid Democratic incumbents and B-list or worse Republican challengers. Terri Lynn Land's abysmal campaign in Michigan has locked that one up for Democrat Gary Peters as well. Right now it appears the biggest question is whether Peters outperforms Obama 2008; he appears likely to exceed Obama's 2012 performance. Finally, there are three rock-solid Republican pickups: Montana, West Virginia, and the surprisingly competitive South Dakota, where Mike Rounds has at least had to fight for his right to win with a likely vote total below 50%.
Now let's move on to the real battlegrounds. I will compare my ratings to those of other prognosticators:
Tilt (50-70%)
Lean (70-90%)
Likely (90-98%), although I personally don't do Likely races
Lean Republican:
Arkansas (Mark Pryor vs Tom Cotton): My personal prediction is Cotton by 8. Pryor has run a mediocre campaign. For that matter, so has Cotton. In fact, I think Cotton's has been worse. But this is Arkansas in 2014, where there's a good chance the entire state goes red this year for the first time in over a century. Pryor's last real glimpses of hope were in late spring and early summer, and since then it's been all downhill.
Of the 8 raters I care about (Cook, Sabato, Rothenberg, 538, NYT, RCP, RRH, and DK), the median rating is: Lean Republican.
Louisiana (Mary Landrieu vs Bill Cassidy): This race will go to a runoff, and on Tuesday I expect Landrieu to receive between 44 and 48 percent of the vote. If I had to make a hard prediction I'd go with 46. To win, Landrieu will need a map like the following:
I just can't quite see it with Louisiana's current racial polarization. But with a runoff, you never know. And if anyone can do it, it's a Landrieu. Median Rating: Tilt Republican
Kentucky (Mitch McConnell vs Alison Lundergan Grimes): This race has barely budged all cycle, with McConnell having a stable 4-7 point lead. I'll go with McConnell by 6. Median Rating: Lean/Likely Republican
Georgia (David Perdue vs Michelle Nunn): The runoff makes for a really uphill battle here, but most importantly, Nunn can get 46%, potentially even 47%, and still lose outright on Tuesday. I think she basically does how Jim Martin did in 2006 and manages to keep this race going until January. Median Rating: Tilt Republican.
Lean Democratic:
New Hampshire (Jeanne Shaheen vs Scott Brown): Shaheen by 5 sounds right. New Hampshire tends to swing fast and hard, so who knows? But I think it's pretty clear Brown doesn't have enough gas in the tank to beat the incumbent here. Median Rating: Lean Democratic
North Carolina (Kay Hagan vs Thom Tillis): Tillis is not too popular, which has hurt his odds in a purple state, and being outspent didn't help. Hagan by 3. Median Rating: Tilt Democratic. A map would look like this:
Watch how Hagan does in Forsyth County (Winston-Salem). If she's winning there, she should hold on. Franklin and Nash Counties northeast of Raleigh should be nailbiters, too.
Tilt Republican:
Alaska (Mark Begich vs Dan Sullivan): Almost nothing would shock me here. I'll say Sullivan by 3, with Begich's ground game helping but not making the difference, but there are lots of question marks up north. Median Rating: Tilt Republican
Kansas (Pat Roberts vs Greg Orman-I): Orman has a one point lead or so in the polls. With about 15% undecided and no party infrastructure, that's a recipe for defeat in a red state. I say he loses by 4. Median Rating: Pure Tossup.
Iowa (Joni Ernst vs Bruce Braley): Boy was I wrong here. I initially figured there was no way Braley could lose to a far-right state legislator and pegged his win at 6-10 points. Now, I'm forced to say he will lose by 2. There's just not much for me as a Democrat to bank on here. No untapped reservoir of minority voters, no history of polling error, no cause for the polling average to be wrong. Median Rating: Tilt R.
Tilt Democratic:
Colorado (Mark Udall vs Cory Gardner): The lone bright spot of the big four unknowns (AK, KS, IA, CO) for Democrats will be out in Colorado. While Gardner has run a very strong campaign and Udall hasn't, I think Udall should win by 1. And it may be all thanks to vote-by-mail. I'm going against the prognosticators here, who have their Median Rating at Tilt R. If Udall does win, he'll need this map:
Watch whether he wins suburban Jefferson and Arapahoe Counties, as well as Larimer County, home to Fort Collins. If either candidates sweeps all three, they will win.
Median Projected Senate Pickup: R+7.5
My Projection: R+7, for a 52-48 Republican Senate
The statehouses are, if anything, even more difficult to project.
Off the Board: Georgia (newly moved), Arkansas, Rhode Island, Minnesota, Oregon, Maryland, Pennsylvania, Arizona, New Hampshire
Leans Republican:
Michigan (Rick Snyder vs Mark Schauer): Rick Snyder has basically been in the same boat as Walker: close race, strong labor opposition, light blue state, and narrow polling leads. He has not trailed in a non-Internet poll since September 10. I say he takes it by 4 points. Median Rating: Tilt Republican
Leans Democratic:
Colorado (John Hickenlooper vs Bob Beauprez): Hickenlooper has had a strange cycle: elected easily in 2010 over split opposition, a controversial gun control bill, loss of two state legislators in recalls, an aura of doom, and now, a strong chance at victory. He mirrors Scott Walker in these respects. I think he can beat Beauprez by 4. Median Rating: Tilt Democratic
Florida (Rick Scott vs Charlie Crist): I've had this race at Lean D all cycle; I figured Crist would bounce back when he started spending money. And he has. Amazingly, he has not trailed in a telephone survey since September 22's SurveyUSA poll, a streak of ten. He won't win big, but I'll say 3 points. Watch Volusia County; a Crist win there would be great news for him. Median Rating: Pure Tossup
Alaska (Sean Parnell vs Bill Walker-I): Walker isn't really a Democrat, and he's not really even center-left. But on two big issues for the Alaska Democratic Party (oil taxes and Medicaid), he's on their side. And come January, I think he'll be in Juneau. Alaska is hard to poll, but it seems clear Parnell is down. I think he'll lose by 6. And to think Charlie Cook just moved it to Tossup! Median Rating: Pure Tossup
Maine (Paul LePage vs Mike Michaud, competitive thanks to Eliot Cutler-I): I've never worried about Michaud, and I can't say I worry now. But it appears LePage could get closer to 43% than 39%, and that must worry Team Blue at least a bit. I say Michaud by 5. Median Rating: Pure Tossup
Tilts Republican:
Massachusetts (Charlie Baker vs Martha Coakley): I really didn't think that Coakley could blow another race, but it appears decently likely. Regardless of final outcome, this and the Iowa Senate race were probably my biggest misses of the cycle. Massachusetts is still a Democratic state, however, and Democratic states tend to have undecided voters who vote Democratic at the last minute. So I will say Baker by 2. Median Rating: Tilt R
Wisconsin (Scott Walker vs Mary Burke): Scott Walker has generally held small leads, with the occasional tied race. That's a recipe for victory, and I'll say he takes it by 3 points. If Burke takes all 5 counties in the state's southwest corner, she should take her place in Madison in two months. Median Rating: Tilt Republican
Tilts Democratic:
Kansas (Sam Brownback vs Paul Davis) This is an inverse Massachusetts, and Brownback should close strong. I project a one-point loss for the incumbent. People just hate his tenure too much. Watch Sedgwick County, home of Wichita. Median Rating: Pure Tossup
Illinois (Pat Quinn vs Bruce Rauner) and Connecticut (Dan Malloy vs Tom Foley): I put these races together because they are stunningly similar. Blue state that is okay with GOP executives? Check. Income tax increases that weren't popular? Check. Multimillionaire, out-of-touch GOP nominees? Check. Polling deficits that have turned to small leads? Check yet again. So what is my projection? Quinn by 3 and Malloy by 4. Median Ratings: Pure Tossups for both
My Projection: Democrats pick up PA, ME, KS, FL, and AK, while losing AR and MA. Net D+3.
Median Rating: Democrats pick up PA, lose AR and MA, with FL, KS, ME, IL, AK, and CT as tossups. Net no gain.
Apparently my projections are quite different, perhaps based off the fact that I don't view this year as a Republican wave at all, but rather one where the swing states get split pretty evenly and a lot of Republican governors are vulnerable.
The House is of course the toughest to project.
Off the board (but formerly on it): MT, NV-3, MI-8, IA-4, WI-6, MN-2, NY-23, MI-1, MI-7, NY-21, IL-13, CA-26, ME-2. You can tell 2014 clearly started tilting to the GOP by the seats which have moved off; they break down as 9 GOP held, 3 Dem held, and 1 GOP pickup.
Leaning Democratic:
NH-2 (Ann Kuster vs Marilinda Garcia), NY-18 (Sean Patrick Maloney vs Nan Hayworth), NY-24 (Dan Maffei vs John Katko): These three are all light blue northeastern seats which could definitely fall if it were 2010, part two, but which look okay for their Democratic incumbent freshmen. Median Ratings: NH-2 Lean D, NY-18 Lean D, NY-24 Tilt D. My projection is Kuster by 7, Maloney by 5, Maffei by 4.
IL-10 (Brad Schneider vs Bob Dold): Like ME-Gov, this is a race I've never sweated. Dem-leaning suburban seats tend to stay blue once they go blue, the inverse of places like NC-7. I think Schneider takes it by 5. Median Rating: Pure Tossup
NV-4 (Steven Horsford vs Cresent Hardy): This one broke later than anything else on the map thanks to horrific Democratic turnout in Nevada. Horsford by 7, thanks to Hardy's lack of cash. This definitely could've been a GOP win, and may be similar to NE-2 last cycle, where a party just straight blew it. Median Rating: Lean D
GA-12 (John Barrow vs Rick Allen), MN-7 (Collin Peterson vs Torrey Westrom): Republicans will win both when the incumbents retire, but potentially not until then. Barrow by 6, Peterson by 8. Median Ratings: Both Lean Dem.
MA-6 (Seth Moulton vs Richard Tisei): Democrats did the smart thing and ditched their underperforming incumbent, and Moulton will be rewarded for his primary success with a 6 point victory. Median Rating: Lean Dem
HI-1 (Mark Takai vs Charles Djou): Think of HI-1 like IL-10, and Djou like Dold, and you'll understand why this is a close race. But also remember there's no real reason Takai won't win. I say he does it by 7. Median Rating: Lean Dem
IA-1 (Pat Murphy vs Rod Blum), IA-2 (Dave Loebsack vs Marianette Miller-Meeks): Personally, I think Democrats are a bit more scared than they need to be that Braley will drag down the whole ticket, but these are definitely competitive. Murphy by 5, Loebsack by 10. Median Ratings: IA-1 Tilt D, IA-2 Lean D
CA-7 (Ami Bera vs Doug Ose): I'm not too worried here because of Bera's strong totals in the June primary. I think he wins it by a close but comfortable enough 4. Median Rating: Tilt Dem
Leans Republican:
AR-2 (French Hill vs Patrick Henry Hays): If Mark Pryor is losing by 8 and Mike Ross by similar margins, I don't see why a non-incumbent Democrat would get elected to the House from the Natural State. AR-2 is probably fool's good. Hill by 5. Median Rating: Tilt Republican
WV-2 (Alex Mooney vs Nick Casey): Outside groups are getting involved, vindicating my prediction that this would be close. Ultimately, it just doesn't seem West Virginia is willing to elect Democrats of any kind outside of Charleston and Coal Country anymore. Mooney by 7. Median Rating: Tilt/Leans Republican
NJ-3 (Tom MacArthur vs Aimee Belgard): This just wasn't the year for Republicans to lose a district like this with a decent nominee. MacArthur by 5. Median Rating: Leans Republican
VA-10 (Barbara Comstock vs John Foust): Comstock's weak summer campaign appears to have improved, and Foust isn't setting the world on fire. In a Romney district, that's enough. Comstock by 6. Median Rating: Leans Republican
CO-6 (Mike Coffman vs Andrew Romanoff): This was another big surprise for me, as I thought Coffman was the most vulnerable GOP incumbent after his weak 2012 performance. It appears nobody agrees. I'm still going to say it's closer than expected. Coffman by 3. Median Rating: Tilt Republican
Tilts Democratic:
MN-8 (Rick Nolan vs Stewart Mills): I don't trust the SUSA poll, and it seems to me Nolan should squeak this one out; I'll say he manages it by a mere point. Median Rating: Tilt Republican
AZ-2 (Ron Barber vs Martha McSally): I expected Barber to lose back in spring and summer, but he seems to have run a more energetic campaign than last time around and I project a 3 point win. Median Rating: Tilt Democratic
IA-3 (Staci Appel vs David Young): Young has inspired enthusiasm from essentially no voters, coming in fifth in his own primary and running a poor campaign with some weird ads. The poor Democratic fortunes upballot are his only hope, but IA-Sen is close enough that Democrats are likely to show up. I think Appel by 4. Median Rating: Tilt Republican
IL-12 (Bill Enyart vs Mike Bost): I admit this is the call I'm least confident of, and I will show it by projecting Enyart to win in a recount. Bost just seems too unstatesmanlike. But then again, this isn't 1920 anymore. Anger may play well here. Quinn's likely victory statewide also helps Enyart out. Median Rating: Tilt Republican
NE-2 (Lee Terry vs Brad Ashford): A chronic underperformer, it appears this is the end of the road for Lee Terry. I doubt Omaha will miss him. Ashford by 3. Median Rating: Tilt Democratic
Tilts Republican:
NY-11 (Michael Grimm vs Dominic Recchia): Recchia has managed to make himself seem like the worst option even when the alternative is likely to be a convicted felon by next election. That's very hard to do. Along with IL-12, however, this is my toughest call. I think it's very possible people decide they just can't pull the trigger for Grimm. I'll say he takes it by 4, though. Median Rating: Tilt Republican
AZ-1 (Ann Kirkpatrick vs Jonathan Paton): Kirkpatrick seems to generally outrun expectations, and I might regret making this call. I'll project this as the other race which goes to a true recount. Paton by a quarter of a percent. Median Rating: Tilt Republican
CA-52 (Scott Peters vs Carl DeMaio): I'm gradually losing support for the idea that DeMaio will win, but I still believe he will. It just depends on how much people believe the allegations against him. DeMaio by 2. Median Rating: Tilt Democratic
WV-3 (Nick Rahall vs Evan Jenkins): With Rahall's defeat, the West Virginia wipeout will be complete. Should, he lose, he could be the answer to a trivia question: which congressman who lost re-election to a member of the opposite party outside of a wave had the most seniority? I'm gonna say Jenkins by 5. Median Rating: Tilt Republican
FL-2 (Steve Southerland vs Gwen Graham): I'm just not seeing how Graham takes this one; Republican turnout will be extremely high due to the governor's election and students are the most likely group of the Democratic coalition to drop off in a midterm (this district is home to Florida State). Southerland by 1. Median Rating: Pure Tossup
FL-26 (Joe Garcia vs Carlos Curbelo): Garcia managed to beat a corrupt man to take this seat, but Curbelo is simply another member of Miami-Dade County's strong Cuban GOP machine, and I think he'll defeat the earwax-eating incumbent by 4. Median Rating: Tilt Republican
NH-1 (Carol Shea-Porter vs Frank Guinta): Shea-Porter has never been strong, and New Hampshire loves to toss out its incumbents. I think she'll lose by 3. Median Rating: Pure Tossup
NY-1 (Tim Bishop vs Lee Zeldin): I don't know what to make of this one. Bishop nearly lost in 2012, like Barber. Usually that means high vulnerability, but this hasn't appeared to be considered a top five GOP pickup. Democrats, however, have no reason to turn out, with Cuomo poised to romp at the top of the ticket, and so I think Zeldin takes this by a point. Median Rating: Pure Tossup
Intraparty Races:
WA-4 (Dan Newhouse vs Clint Didier): Establishment vs insurgent can lead to insurgent win in the GOP primary, but not in a Republican-on-Republican general. Newhouse should take this by double digits.
CA-4 (Tom McClintock vs Art Moore): I initially thought McClintock could lose like Pete Stark did. He's not beloved by the median voter in his district, certainly. But I've heard no buzz about Moore, and I'm now inclined to think McClintock takes between 53 and 58 percent.
CA-25 (Tony Strickland vs Steve Knight): Knight may have geographic advantage, but he's appeared to run a weaker campaign than Strickland. Ideology isn't really a factor here. I think Strickland by high single digits.
CA-17 (Mike Honda vs Ro Khanna): [Disclosure: I interned on this campaign over a year ago for Team Honda]. Mike Honda's campaign has not been too stellar, thanks in part to his being a 74 year old who hasn't had to campaign hard in over a decade. Khanna is a fresh face with appeal to the South Asian and business communities in the district. I admit this is the toughest race to make a call in because I have a stronger stake in it, and I may be biased, but I think Khanna goes down by about five.
Prognosticators' projections (including 3 safe Republican pickups and 1 safe Democratic one): R+8
My Projection: R+6