What Impressed Me This Week: the Clinton Upset

I met a lady who had traveled three times round the world in order to escape circumstances, but she always came to a world where there were still circumstances. -- Carl Jung

I don't know what hurt worse Wednesday morning: That Clinton-Obama-Edwards was the reverse of the order I was rooting for, or that my predictions were so embarrassingly wrong. In case you forgot: I thought the polls were underestimating the size of the Obama wave. Remember that the next time I predict something.

Last Tuesday Night

So Deb gets home from work and we walk over to a nearby school to vote. There's a Hillary hanger on our apartment door. We meet some Hillary people on the walk over. At the polling place, only Hillary people are waving signs.

We vote, and then decide we're going to walk over to Martha's, the brewpub on Main Street, with the idea that we're going to make a cheap if not very healthy dinner out of the half-price appetizers they serve until six. Turns out, that's where the Obama people are, having a pre-victory-party party. The guy sitting next to me at the bar is some kind of freelance political operative who tonight is a driver for Congresswoman Carol Shea-Porter, the rep from New Hampshire's other congressional district. He claims to recognize my pseudonym (Pericles) on DailyKos, so we're best buddies. (Tip for political types everywhere: Don't try to bribe bloggers, just treat them like they're important. We respond like geeks who suddenly find themselves hanging with the cool kids.)

Deb spots HBO star Larry David over on the other side of the room. At some point I turn away from the TV above the bar and see Senator Dick Durbin standing over my shoulder, watching the returns come in just like me, except that I've got a seat. Other nearby people are on the phone. One says, "Manchester's going for Hillary. That can't be right."

Another local guy at the bar reports voting for Hillary, but he can't for the life of him figure out why. He seems genuinely puzzled, like the victim of a post-hypnotic suggestion. A female high school teacher steals my seat while I'm in the bathroom and winds up sitting between me and Deb. She keeps talking about "Chillary", a nickname she claims to have coined herself. I think she voted for Edwards.

The actual victory party -- assuming there's a victory, which is starting to look questionable -- is supposed to be at one of the Nashua high schools. But the rumor is that Obama is coming here at eight. Most polls closed at seven, so that should have allowed plenty of time for the networks to project him the winner. (McCain's victory is called by then, and that race was supposed to be close.) But Clinton's early 2-4% lead barely wavers as the returns keep coming in.

By nine, Obama still hasn't shown. My new best friend went outside to take a phone call and hasn't been seen since. Shea-Porter and Durbin might have vanished into puffs of smoke, if smoking hadn't been illegal in New Hampshire bars for a year now. Larry David is looking glum. I really shouldn't have drunk that last scotch ale somebody bought me, and if I hang around much longer I'm going to get glum too. Time to go home.

Did Hillary Really Win?

Yes. A bunch of blogs have speculated that the vote might have been hacked, and some unscrupulous newspapers have picked up the story, but it doesn't seem very likely. The big reasons to suspect fraud are (1) the surprise of Clinton's win; (2) that Clinton won in precincts counted by machine while Obama won in precincts counted by hand.

A couple of good analyses of the situation are here and here. The main points are:

• Clinton's win was a big surprise based on the pre-election predictions of an Obama landslide, but the exit polls pointed to a much closer race.
• New Hampshire votes on paper ballots, not on those awful touch-screen voting machines with no paper trail. "Machine counted" means that the paper ballots were tabulated by running them through optical scanning machines. The machines are made by Diebold and are known to be hackable, but they aren't networked. You'd have to hack them one by one.
• Machine-counting happens in the cities, hand-counting mainly in the small towns. It's not really surprising that they might favor different candidates in a close race.
  

Anyway, the paper ballots are still around and Kucinich has asked for a recount -- not because he thinks he won, but because he wants to insure the credibility of the process. I'm glad he did, because I think the recount (by hand) will show that the counting was honest.

So What Did Happen?

There are two questions here: Why were the results different from the pre-election polls? And why were they different from the Iowa caucus results? The polling question is technical and I find it less interesting, so I'll point you to this reference. But what changed between Iowa and New Hampshire?

Let's start with facts and work back to explanations. The election results are the hardest facts, followed by the exit polling about who the voters were and why they voted the way they did.

Turnout. As in Iowa, there was a huge turnout advantage for the Democrats: 270,000-210,000. If this were one big primary, McCain would have finished third behind Clinton and Obama. (In Iowa, Republican winner Huckabee would have finished fourth.) The New York Times' Ron Klain sums up:

In the three decades since 1980, there have been four primary years when both the G.O.P. and the Democratic nominations were contested – 1988, 1992, 2000 and 2008. In all three of the previous elections, there were more votes cast in the Republican primaries than in the Democratic primaries. The G.O.P. margin was almost 40,000 votes in 1988 and almost 80,000 votes in 2000. So to see more votes cast in New Hampshire’s Democratic primary last night than in the state’s Republican one — not to mention 60,000 more votes — is almost as historic as seeing a one-two finish by a woman and an African-American.

This is part of a larger trend in New Hampshire. In 2006 we not only re-elected a Democratic governor with over 70% of the vote. We also put Democratic majorities in both houses of the legislature and turned over our entire Congressional delegation: We replaced our two Republican congressmen with two Democrats. We're a blue state now. Senator Sununu should be worried about November.

Clinton's edge. Clinton had a 2% margin over Obama, 39-37%, or about 7,500 votes. The only thing that makes this primary feel like a rejection of Obama is that expectations were so high going in. Edwards was third with 17%, which has to be a huge blow to him. There's going to be a lot of pressure for the non-Hillary vote to unite behind one candidate, which has to be Obama at this point.

According to the exit polls, Clinton got her margin over Obama from women (46-34%, a 12% margin), older voters (16% margin in the over-65 group), and the less well-off (15% margin among those with household income less than $50K).

Now let's compare to Iowa. Obama got 35% in Iowa and 37% in New Hampshire, so you're not looking at a voters-abandon-Obama story. And Clinton-plus-Edwards got 56% in New Hampshire versus 59% in Iowa. So the question to be answered is where Edwards' support went to Clinton.

The age factor also showed up in Iowa. According to the Iowa entrance polls, Obama had a 46% margin over Clinton in the under-30 age group, but Clinton had a 27% edge in the over-65 group. The interesting difference is that Edwards won the 45-64 age group in Iowa with 31%, but he got only 21% of the 50-64 group in New Hampshire. (I'm not playing tricks with the groupings; CNN is.) Obama got 30% of the 50-64 group in New Hampshire, almost the same as the 27% of 45-65-year-olds he got in Iowa. So one answer is: Middle-aged people shifted from Edwards to Clinton.

Obama narrowly won the under-$50K income group in Iowa (34-32% over Clinton). In New Hampshire he loses it (47-32%). The only notable difference in this group between Iowa and New Hampshire is that Clinton's support shoots up. It comes in dribs and drabs from all the other candidates.

Now let's look at gender. In both New Hampshire and Iowa, women were 57% of the electorate. (There's a corresponding male advantage among Republicans. It's not that women vote and men don't.) But Obama won the women's vote in Iowa 35-30% over Clinton. He lost it 46-34% in New Hampshire. His margin over Clinton among men was almost the same: 12% in Iowa and 11% in New Hampshire. Obama didn't lose female support; Clinton gained it. Edwards+Richardson is 30% of the female vote in Iowa, but only 18% in New Hampshire. Women shifted from Edwards and Richardson to Clinton.

The temptation is to lump the three factors together and imagine that middle-aged working-class women moved in one big lump from the non-Obama male candidates to Clinton. None of the polling data I can find is refined enough to say anything about that small a niche of the population, but let's imagine it. Why would they do that between Iowa and New Hampshire? And why would they poll differently on the weekend than they voted on Tuesday?

I'm not buying the Clinton-humanized-herself-by-crying explanation. I don't know why; maybe just because I haven't bothered to watch the video. But I remember several political conversations I've had with women over the last year. They'd claim to be undecided, but if anybody criticized Clinton they'd bristle as if they'd been criticized themselves. Here's what I think happened during the Clinton-bashing media orgy between Iowa and New Hampshire. At first those middle-aged working-class women got depressed, so when the pollsters called on Saturday they said they were undecided or that they weren't voting. And then somewhere around Monday night they got mad and decided to do something about it.

I have no idea whether such an effect would carry over into the next primaries.

Did I Do the Right Thing?

Last week, right before I voted for Edwards, I wrote: "I would like to live in an America where Kucinich is a viable candidate."

I expected to get comments on that, and I did. "Kucinich is viable if you vote for him," someone wrote. "Don’t let the media tell you who is viable. They have their own agenda."

I thought about that, and I wrote an answer so long I decided to make a separate post out of it. You can find it on my own blog or on DailyKos.

Short Notes

Remember the National Intelligence Estimate that said Iran had discontinued its nuclear weapons program? Newsweek claims that Bush is privately telling other world leaders that he doesn't believe it.

We're supposed to give amnesty to the telcom companies for spying on us because they're so patriotic. They are, that is, until the government doesn't pay its bill. Apparently the FBI fell behind on its payments due to some bookkeeping snafu, and the telephone companies cut off their wiretaps.

As the Democrats apparently realize and the Republicans don't, negative campaigning only works in a two-candidate race. Unfortunately, we seem to have arrived there now. TPM-TV pulls it together.

My nominee for the most wrong-headed article of the week is the NYT op-ed We Still Need the Big Guns by Air Force General Charles Dunlap. Dunlap is worried that we're learning the wrong lessons from our success in Iraq -- the whole "hearts and minds" thing. "Starry-eyed enthusiasts [of the new counter-insurgency manual] ... dismiss as passé killing or capturing insurgents." He worries that this attitude might lead to something truly awful: less money for new airplanes. Dunlap points out that success in Iraq has come the old-fashioned way: more troops, more bombs, killing and imprisoning more people. JesseLE on DailyKos asks the obvious follow-up question: "Why not just kill everybody?" This is a good time to plug one of the most popular things I ever wrote, Terrorist Strategy 101: a Quiz. It's from 2004 and I predict an attack that still hasn't happened, but otherwise it holds up very well.

Quoting a study from the London School of Hygiene and Tropical Medicine, The Financial Times says that the United States now has a higher rate of deaths due to treatable diseases than any other industrialized country. We've slipped behind medical powerhouses like Portugal. The authors note that the U.S. slide in the rankings "has coincided with an increase in the uninsured population." Jerome a Paris sharpens the point by comparing our death-due-to-treatable-disease rate to the world's best: "Each year 101,000 Americans die needlessly because they're not French."

The Supreme Court is getting set to rule on one of the centerpieces of the long-term Republican project to keep marginalized citizens from voting: the Indiana law that requires voters to have a picture ID. Slate covers the issue pretty well. This is one of many interesting legal cases covered by Christy Hardin-Smith in her "Between the Briefs" summary.

CNN pundit Paul Begala relates a ridiculous story about Fox News. After Iowa, they reported he was joining the Clinton campaign. It wasn't true, but OK, these things happen. So Begala called Fox reporter Major Garrett to deny the story. Instead of correcting the story, Garrett kept repeating it. He emailed Begala that "the sourcing is strong, very strong, or I wouldn't go with it." So somebody more authoritative than Begala himself was telling Garrett that Begala had taken a new job.

It would be interesting to hear a thoughtful conservative columnist's assessment of the Democratic turnout in Iowa and New Hampshire, which (as I described above) swamped Republican turnout. Is the sky falling on Republicans, or does some other explanation not point to a Democratic landslide in November? Here's David Brooks' thoughtful assessment: It didn't happen. His #1 Surprise of the New Hampshire primary was: "Republicans voted in nearly the same numbers as Democrats." That's why he gets the big bucks, I guess.

I disapprove of Democrats voting Republican to screw up the Michigan primary, but the Democrats for Romney spoof is hilarious.

Two Huckabee links: Huckabee telling MSNBC's Joe Scarborough that Fred Thompson needs some Metamucil. And Dave Sirota urging Huckabee to keep talking about class issues.

Last week I promised a review of Paul Krugman's new book Conscience of a Liberal. Another bad prediction: I didn't get it written in time. I'll try again next week.