Election Day is here at last. Below you'll find a guide to watching the results come in tonight, with my predictions and notes on what it means when I get things wrong. Let's get the easy calls out of the way first:
Mitt Romney will easily win West Virginia, Kentucky, Tennessee, South Carolina, Georgia, Alabama, Mississippi, Louisiana, Arkansas, Missouri, Texas, Oklahoma, Kansas, Nebraska, North Dakota, South Dakota, Montana, Wyoming, Utah, Arizona, Idaho, Alaska, and Indiana (the last was won by Obama last cycle and several of the others were in play -- there is no chance that Romney will do as poorly as John McCain did). President Obama will just as easily win Maine, Vermont, New York, Massachusetts, Rhode Island, New Jersey, Maryland, the District of Columbia, Illinois, New Mexico, Washington, Oregon, California, and Hawaii. This gives each candidate a floor of 191 electoral votes.
Democrats will easily hold Senate seats in California, Delaware, Maryland, Hawaii, Michigan, Minnesota, New Jersey, New Mexico, New York, Rhode Island, Washington, West Virginia, and, for most purposes, Vermont (socialist incumbent Bernie Sanders is an Independent but caucuses with Democrats). Republicans will easily hold Senate seats in Mississippi, Tennessee, Texas, Utah, and Wyoming.
Electoral vote calls are in bold, with O+x indicating that Obama will add x to his total and R+y indicating that Romney will add y to his total. Senate calls are in italics. All times are Eastern (note that some states span more than one time zone, and these are times when the last polls close; don't use these times as a guide for when to vote, especially if you live in Florida outside the panhandle).
7:00 pm -- Polls close in Indiana. Public opinion polling is difficult in Indiana due to strict regulations targeted at telemarketers, so there's been relatively little of it; there's some evidence that the race has moved against Richard Mourdock since his defense of banning abortion in the case of rape, but with so few polls it's hardly a slam dunk. I'll bet on the underdog in this one: Mourdock wins (Rep. hold). If I'm wrong: If Joe Donnelly wins, it will a triumph of a narrative of Republican extremism triggered by Todd Akin's much, much more objectionable comments on the same topic. This narrative has been unfair to Mourdock, but then politics is often an unfair business.
Polls also close in Virginia. Jim Webb doesn't seem to have found the Senate to his liking, and is retiring after just one term; polls show Democrat Tim Kaine slightly ahead of Republican George Allen in the race to replace him. Kaine will likely win (Dem. hold). The presidential race is a dead heat. I've been hearing troubling things about the quality of the GOP ground game in Virginia, so here begins a pessimistic theme: I'm guessing that, after a long night of counting, Obama will win Virginia (O+13). If I'm wrong: Romney's hope of victory is probably contingent on winning Virginia, so if he pulls it out, a Romney victory remains in reach. And if Allen can beat Kaine in the Senate race, it could be excellent news for Romney, because it means we may be seeing a systemic pro-Democratic bias in state-level polls.
Such a systemic bias isn't all that implausible; the polls are regularly turning out Democrat-heavy samples that raise questions about whether there's a skew in the way they're treating party identification. I'm less inclined than some observers to ignore the polling averages on this basis -- party ID ratios that deviate from the norm are a good reason to ignore individual polls, but in this case almost all the polls are pointing in one direction. it's certainly not impossible for pollsters to all be wrong at once, and there are colorable hypotheses as to why that might be happening in this cycle. For an extensive discussion of this, see this post by Ted Frank.
7:30 pm -- Polls close in North Carolina, which Obama carried in 2008; a couple of pollsters show it close, but more of them show it as a relatively easy win for Romney, which it almost certainly is (R+15).
Polls also close in Ohio. In the Senate race, incumbent Democrat Sherrod Brown leads challenger Josh Mandel in every poll, sometimes quite comfortably, so expect him to win (Dem. hold). The presidential race is closer, but the latest polls all show a late break toward Obama. It'll be tight, but I expect the president to win (O+18). If I'm wrong: The Senate race in Ohio, as in Virginia, may give evidence for or against the theory that the polls are understating Republican strength. If Brown isn't beating Mandel comfortably, it's good news for Romney. It's even better news for Romney, of course, if he wins the state -- though if Romney pulls it out in Virginia, he has paths to victory both going through Ohio and also going around it (more on that below).
8:00 pm -- Polls close in DC and in sixteen states. Among them:
Maine. Moderate Republican Senator Olympia Snowe is retiring, the Independent former governor Angus King is a lock to replace her, and King will almost certainly caucus with the Democrats; put an asterisk on it if you like, but essentially that's a Dem. gain.
Connecticut. Joe Lieberman is retiring, and while a month ago it looked like Republican Linda McMahon had a shot at replacing him, the polls are now breaking strongly toward Democrat Chris Murphy (Dem. hold).
Massachusetts. Elizabeth Warren looks poised to unseat Scott Brown (Dem. gain). If I'm wrong: It'll be a huge coup for the UMass/Boston Herald poll, the only one showing Brown slightly ahead.
Missouri. Claire McCaskill looked like one of the most vulnerable Democrats in the Senate -- until Todd Akin mused about "legitimate rape," and then ignored Republicans begging him to step aside. Akin is a legitimate disaster, and the electorate has ways of shutting that down (Dem. hold). If I'm wrong: Large numbers of anti-McCaskill voters were embarrassed to admit to pollsters that they'd vote for Akin (actually quite possible).
Pennsylvania. For a while it looked like Republican Tom Smith was gaining enough to make this close, but now incumbent Democrat Bob Casey looks like he'll win easily (Dem. hold). In the presidential race, Romney has made a push to expand the playing field into Pennsylvania and the Obama campaign has felt compelled to defend it, but the same thing has happened in previous cycles and the state always seems to stay just outside Republicans' reach. I'd bet on that happening again (O+20). If I'm wrong: If Romney takes Pennsylvania, he's probably going to win (and I was probably wrong about some of the pessimistic calls above). If Casey isn't ahead as comfortably as expected, it could be another indicator of pro-Democratic sampling errors in the polls.
Florida. In the Senate race, Republican Connie Mack was briefly ahead of incumbent Democrat Bill Nelson in the summer, but Nelson retook the lead in mid-August and has held it since; he will likely be reelected (Dem. hold). In the presidential race, Romney has led in public polling averages for the past month; some Republicans looking at internal polls sound surprisingly nervous behind the scenes, but I suspect Florida will stay red even if it's a closer shave than some of us were expecting (R+29). If I'm wrong: If Obama has Florida, he's winning by a fairly comfortable margin.
About those internal polls, by the way: Because campaigns (and party committees and outside groups) will spend more money on polling than media outlets will -- allowing pollsters to do things like target a particular geographical balance in a sample -- the average internal poll is more accurate than the average public poll. But the average internal poll usually isn't the one we see. Campaigns leak numbers that send the message they want to send, not numbers that accurately reflect the aggregate of the data they have access to, so internal polling released to the public is significantly less accurate than public polling. Yesterday Toby Harnden of the Daily Mail got an exclusive look at some internal polls that look very favorable to Romney -- but there's a very good chance that those numbers are outliers. I haven't been shown any internal polls, but I have had informal conversations with people who have access to them; paradoxically, this is probably more informative than seeing the actual numbers (unless you're seeing them every day). Those conversations influence some of my predictions.
New Hampshire. This one's very, very close, but Obama has a slight lead in public polls, and GOP insiders with access to internal polling are pessimistic, so I'm calling it for Obama (O+4). If I'm wrong: New Hampshire can swing the election one way or another if, for example, Romney wins Virginia and Ohio but loses the states west of Lake Michigan that Obama won in 2008. That doesn't align with my predictions, but it's not a crazy scenario, either.
9:00 pm -- Polls close in thirteen states. Among them are Arizona, where retiring Republican Senator Jon Kyl looks set to be replaced by Jeff Flake (Rep. hold), and Nebraska, where retiring moderate Democratic Senator Ben Nelson looks set to be replaced by Republican Deb Fischer (Rep. gain).
Polls also close in Michigan and Minnesota. The presidential race has at times looked close in these states, but not close enough to flip them; both will stay blue (O+16, O+10) Minnesota has really only been part of the race because its media markets overlap with parts of Wisconsin. And Wisconsin really is in play.
In fact, Republicans who've seen internal polling actually seem to feel better about Wisconsin than they do about Ohio, which isn't what you'd guess from the public polls showing Ohio as the tighter race. I think Obama will probably hang on in the Badger State (O+10), but it isn't certain. Tammy Baldwin beating Tommy Thompson in the Senate race is probably a somewhat safer bet (Dem. hold). If I'm wrong: As mentioned above, if Romney can win Virginia, he has a path to victory that goes through Ohio, and one that goes around Ohio; the latter path requires winning Wisconsin. As you might have heard once or twice or eighty times this year, no Republican has ever won the presidency without Ohio, but it wouldn't be shocking if Romney were the first to manage it. Romney winning Wisconsin's electoral votes could come in tandem with a Thompson upset in the Senate race, so watch the returns in both races.
Colorado is a coin flip. I say it comes up Romney (R+9). If I'm wrong: The path to victory for Romney that includes Virginia and Wisconsin but not Ohio does not work without Colorado.
10:00 pm -- Polls close in Utah (which we've already covered), Montana, Nevada, and Iowa.
In Montana, Republican Danny Rehberg has already been elected statewide -- he's the at-large Representative -- which makes him a formidable challenger to incumbent Senator Jon Tester. I think Rehberg wins (Rep. gain). In Nevada, Republican Dean Heller probably wins the Senate race (Rep. hold), but the Silver State stays blue at the presidential level; even the rosy internals leaked to the Daily Mail concede that Obama wins Nevada (O+6).
In Iowa, Obama underperformed his polls in 2008; the RealClearPolitics average had him up by 15.3, but he only won by 9.5. The RCP average currently shows Obama up 2.4. I think Romney pulls it off (R+6). If I'm wrong: As with Colorado, the scenario where Romney wins without Ohio does not work without Iowa.
11:00 pm -- Polls close in North Dakota, where Republican Rick Berg looks poised to replace retiring Democratic Senator Kent Conrad (Rep. gain).
In the Senate, that comes out to a net gain of one seat for Republicans, yielding a 52-seat Democratic majority (counting Democratic-caucusing independents). In the presidential race, it comes out to 288 electoral votes for Obama and 250 electoral votes for Romney.
One last thing: It's very likely that Romney narrowly wins the popular vote, especially if I've got a state or two wrong in the electoral vote forecast.