A word about exit polls: The early ones suck. I do lots of parsing of the first, post-poll-closing wave of numbers, but they are faulty and they get massaged as the night goes on. I anticipated a closer-than-my-prediction race in Pennsylvania when the early exits had him winning the Philadelphia suburbs, but as the night went on Obama's 16-point lead there vanished. So if, for example, the first exit polls show Obama tying Clinton among west North Carolina whites, assume the data sucks.
Indiana (7 p.m.)
The Democrats. If Rev. Jeremiah Wright had greeted the new year with a self-imposed exile to Tibet, or if—even better—he'd turned off the cameras in Trinity United Church and never recorded himself saying "God damn America," Barack Obama would be knocking Hillary Clinton out of the race today. Sure, she would have tried to stagger on after losing Indiana and North Carolina. As Phil Klein points out, her "incredible resilience" is mostly a function of the fact that she's Hillary Clinton and started this race with legions of delegates in her pocket, one of the biggest fundraising lists on the planet, and a list of owed favors that approached Santa Claus naughty-or-nice-list-length.
But if not for Jeremiah Wright, Obama would be crushing her in both primary states. In its widely-circulated post-Super Tuesday spreadsheet of coming delegate fights, Obama's campaign predicted a 7-point win in Indiana. It was even more fertile territory than Wisconsin (they predicted a 7-point win and won by 17) or Virginia (they predicted 2 points and won by 29). It borded Illinois, and voters in the 1st and 8th congressional districts already knew and liked Obama. It was lousy with college towns. It had a small but energized black population that would pad his margin. Obama only needed to win about 40 percent of the white vote to carry the state, and he'd done that in Missouri, a demographically similar state.
Obama is weaker now and will probably lose what he called (in April) a "tiebreaker" state. He won't lose by the 10-point margin of Ohio or 9-point margin of Pennsylvania, because Indiana is more Midwestern than either of those states, and it doesn't touch on the Appalachian Mountains—the single strongest region for Clinton in the entire country. A win wouldn't be impossible, actually, because (as Robert Novak pointed out), Obama could pummel Clinton in the five most vote-rich counties and win, as long as he wasn't totally blown out in the other 87 counties. I think he'll win more than five counties but lose anyway. Clinton 53.5, Obama 46.5, with Clinton netting 5 delegates over Obama.
– Counties to watch: Hamilton, the wealthy, fast-growing county north of Indianapolis. If Clinton's winning easy there, Obama's coalition is coming apart. The Politico suggests that Howard County will be the swing area.
– Demographics to watch: Whites, of course, with a quick glance at the black vote. (The pollsters that show Clinton winning handily show her recovering ground with black voters, but they showed that in Pennsylvania and she lost them by 80 points.) In Pennsylvania Clinton won white Democrats by 30 points. If Obama closes that number to less than 20 points, he wins. If it expands, she wins easily.
– Voters to watch: Republicans. The impact of Rush Limbaugh's campaign (joined by some local hosts) to get Republicans to vote Hillary is really hard to measure, especially now that Obama's image has been damaged by Wright. Republicans went for Obama by 44 points in Wisconsin; they split 50-50 between the candidates in Ohio. (They weren't allowed to vote in the Pennsylvania Democratic primary.) Anecdotal evidence is that the Limbaugh voters are matched by the good-faith Clinton and Obama Republican voters.
– Issue to watch: The gas tax holiday. If exit pollsters ask about it, let's see if the Clinton pander worked.
The Republicans. It's really hard to tell what McCain's margin will be here, as thousands and thousands of Republicans will be bolting their primary to vote for Clinton or Obama. Huckabee, Romney, and Paul are all on the ballot. In 2000, Bush only got 82 percent of the primary vote even though he'd defeated McCain weeks earlier. So I'll guess McCain 78, Paul 10, Others 12.
North Carolina (7:30 p.m.)
I simply don't believe the tightening polls in this state. Not the ones that show it becoming a toss-up.Todd Beeton, a Hillary-leaning blogger, points out something in this expectation-setting post: Southern polls have always underestimated Obama's support as they underestimate the black vote gap between Obama and Clinton (they usually peg it at 50 points, and it ends up around 80) and lowball Obama's white vote. Of course, we haven't had a Southern primary since Wrightgate, and Clinton has worked North Carolina harder than any Southern state since Tennessee, which she won easily. Obama 55, Clinton 44, with a net gain of around 9 delegates for Obama.
– County to watch: Watuga. John Vaught LaBeaume explains why: It's the kind of rural whites-plus-blacks-plus-college kid county that Obama used to dominate.
– Voters to watch: Independents. They've padded Obama's margin in all of the pre-Wright states. What are they doing now?
– Issue to watch: Jeremiah Wright. Apparently voters were split 50/50 on whether they factored him into their vote. What did whites think?
The Republicans. This could actually get interesting. When John McCain criticized the state GOP for running an ad linking Obama and Wright to the state's Democratic gubernatorial candidates, he won his usual dollop of national praise and pissed off a lot of Republicans back here. McCain 72, Paul 12, Others 16.
Other races to watch: The 3rd and 4th district congressional races in North Carolina, and the 5th district race in Indiana. The former to see if Ron Paul endorsees are winning their elections, the latter to see if the grassroots can purge a well-fed insider Republican.
UPDATE 6:08: The Daily Kos has the only early exits that are ever any good: racial voting. Blacks went against Clinton in both states by about 85 points. That's nightmarish and much worse than in the polls that showed her closing strong–they showed her climbing back into the teens. In North Carolina, this matters. Assuming 33 percent black turnout, the difference between a 70-point loss of the black vote and a 85-point loss is about 4 points overall. She'd have to win about 71 percent of the white vote to overcome that.
UPDATE 7:03: NBC calls Indiana "too early to call," which must mean they have clear exits (or else they'd say "too close"). I'm guessing Obama isn't pulling the votes he needs in the Chicago burbs, which have been bombarded with Wright coverage.
UPDATE 7:06: Early, bullshit exits have Clinton winning by 4 in Indiana as Obama carries everything but the rural south and central parts of the state. I'm sure this'll get massaged.
UPDATE 7:11: On MSNBC, John Kerry is really pushing the "Rush Limbaugh is begging Republicans to vote Hillary" meme. They want to discredit a Hillary win if she takes it by the margin of crossover voters. Indeed, the exits have her winning Republicans.
UPDATE 7:30: North Carolina is called a nanosecond after polls close for Obama. Hey, Mickey Kaus, how's that "Obama by 3 or less" prediction holding up?
UPDATE 7:35: North Carolina exits look good, not great, for Obama. He held but didn't gain among white voters: He'll win about 38 percent of them, his best result outside of Virginia and Georgia. Also—and expect this to change—he's winning every region of the state. Landslides in the Research Triangle and Charlotte, single-digit wins everywhere else.
UPDATE 7:55: Basically nothing has been counted in the NC-3 and NC-4 races, but Jones and Lawson won big in early voting. Jones is up by 17 points, Lawson is up by about 40 points.
UPDATE 8:01: Unless northwest Indiana breaks for Clinton, she's not holding the current 14-point lead. Her vote is coming in early: Counties like Vigo, Jefferson and Floyd are counted, while Marion (home of Indianapolis) is 15 percent in and Lake (home of Gary) is completely out.
UPDATE 8:45: Good news from Marc Ambinder:
Campaign advisers are saying that the gas tax pause debate helped him … though they're not terribly happy that the question wasn't on the exit poll questionnaire
Said one adviser:
"It blunted her appeal with middle and downscale and helped us in the burbs with upper income/college educated voters; it showed why Obama was different than her, which we needed to make people make the leap to vote for us. That didn't happen in PA or OH."
Ha. Also, ha.
UPDATE 8:51: Maybe Paul's biographical links to Pennsylvania made the difference for him, because he's running third behind Huckabee in both states. If he's lucky, though, he'll get 80,000 votes from both states, and cross the 1 million vote mark.
UPDATE 8:54: I don't know why NBC hasn't can't Indiana, as the crucial counties that needed to swamp for Obama are coming in rather weak for him. St. Joseph County (with South Bend) by 6 points, for example.
UPDATE 9:04: I just spoke to B.J. Lawson, who is clobbering Augustus Cho 70-30 with more than 60 percent of the vote counted. "We're very excited," he says, although he won't declare victory until Cho concedes. "We're ready to run a positive campaign based on Constitutional values."
UPDATE 9:15: Hm… Obama altered the Indiana line in his speech. He added the word "apparent" when congratulating Clinton on her Indiana win. MSNBC, at least, is eyeing Lake County and calling it too close to call. (It looks in some ways like Missouri, where Obama was losing, and several networks called it against him, until the final precincts of St. Louis came in. The problem for him is that Gary is not St. Louis.)
UPDATE 9:28: NC-3 is going to take forever to report… all of those hard-to-reach coastal towns. But Jones is widening his lead over McLaughlin. Everyone I know thinks Jones has it won.
UPDATE 9:43: Tim Russert is reiterating the "gas tax hurt Clinton" storyline—it took the focus off Wright at the 11th hour and showed voters the non-pandering Obama they liked in the first place.
UPDATE 9:51: B.J. Lawson wins.
It's a landslide, and a surprisingly big total for a low-turnout GOP primary.
9:56: I can sense the spin moving away from "split decision" to "Obama wins." Even if Clinton wins Indiana now, it's within the margin of Limbaugh-listening spoilers.
10:15: The Daily Kos (yes, I'll link 'em again) begs Clinton to stay in for two weeks in order to reduce the embarrassment of Obama's coming West Virginia (May 13) and Kentucky (May 20) losses. He's on to something. Obama would lose both states even if Clinton dropped out five minutes ago. The question is whether Clinton can take Oregon (also May 20) from his column, something that seems less and less likely as Obama's coalition refuses to budge.
10:39: Clinton speaks! She claims to have "broken the tie" without actually declaring an Indiana victory. Because she can't declare it yet, if she can declare it at all. Sorry if I'm stepping on Ann Althouse's beat, but my God does Bill Clinton ever look red-faced and sad, standing behind Clinton. This is the body language of a mounting defeat.
10:47: With all but 5 (of 17) counties counted, Jones is winning his primary by 20 points. It's over. The pro-war bloc won't get its scalp.