Kevin Drum and Dana Goldstein find something curious in the New Hampshire exit poll results. Dana: "Clinton edged out Obama, 41 to 34 percent, among those who want the troops out "immediately." And Obama beat Clinton, 51 percent to 24 percent, among voters who want to keep troops in Iraq! But wait, you protest, Obama opposed the war from the beginning!" Kevin mentions this, and notes the further point that: "Voters who approve of the war prefer Romney by a small margin, while those who disapprove of the war prefer McCain by a landslide. Again, huh?"
Most voters don't form their impression of a candidate by first looking at the candidate's issue positions. Rather, their impressions of the candidate's issue positions arise from the caricature of the candidate that they get from the media, which is what they vote on anyway. The media has been portraying Clinton as a crazy liberal forever, so of course she's the candidate for those who want to get out of Iraq quickly. Obama and McCain are presented as bipartisan uniter types, so of course they're the candidates for independents who vote in one party's primary while deviating from the party line. And Romney is probably the most straightforward GOP establishment choice, so he's a plausible pick for standard pro-war Republicans. These candidate stereotypes are, of course, completely deceiving.
As Dana can tell you, people's political decisions are rarely guided by policy:
Of course, I'm overstating the case -- we all realize (I hope) that unlike committed denizens of the political blogosphere, most Americans have no clue who Mark Penn is or what he does. But in all seriousness, talking to caucus-goers in Iowa, I saw up close just how uninformed the average voter is when it comes to policy. (And remember, an Iowa caucus-goer is part of an elite group of just 11 percent of eligible voters!) I heard that John Edwards seemed to really love his wife. That Hillary Clinton would bring Bill back to the White House, where he belonged. That Barack Obama was charismatic. I would press voters to tell me what policy proposals they were drawn to. This question was often met by a reassertion of just how honest, hard-working, or attractive Candidate X was. Other times, people told me they liked Candidate X's "health care" plan. What about it did they like? Well, it was a plan. And health care in America is messed up.
And that's one of the reasons I've spent much of the last three years telling people to vote for John Edwards. With his Southern accent and small-town roots, Edwards had the superficial features designed to convince a uninformed general election swing voter, particularly in the white working class, that he was just a good-hearted regular guy from your town. He could take on some pretty solid left-wing policy positions without anybody regarding him as a fringe candidate. He was perfectly set up not only to win the general election while holding solidly left-wing policy positions, but to make those positions seem moderate afterwards.
Edwards set himelf up to play exactly that part. After repenting of his Iraq War vote, he moved left of Obama on foreign policy with his criticism of the 'war on terror', and his full-throated opposition to Lieberman-Kyl. On all manner of domestic issues (particularly health care) his early policy proposals were so progressive as to bid up the price of left-wing support for anyone else. Forcefully arguing for these positions didn't damage his general election poll numbers one bit.
That's what drove my constant posting about why John Edwards should get the Democratic nomination. Blogging is a good way of getting in touch with highly informed progressives, and I left no argument unmade in convincing that audience that Edwards was the candidate for them. Edwards' clear victories in the last few Daily Kos straw polls and his general blogosphere success suggest that a large proportion of progressive voters, if informed at the level that bloggers are, would support him. It was up to myself and others in the information-spreading side of the progressive movement to make sure people were informed. Otherwise, the same stereotype that made Edwards such a good general election candidate would doom him in the primary. Poorly informed voters would see the white Southern guy next to the woman and the black man, and assume that he was the conservative one.
On most political matters, do you consider yourself:
| Category | % Total | Biden | Clinton | Dodd | Edwards | Gravel | Kucinich | Obama | Richardson | U |
| Very liberal | 18 | 7 | 24 | 3 | 16 | 0 | 4 | 40 | 5 | 2 |
| Somewhat liberal | 36 | 3 | 25 | 1 | 25 | - | 1 | 36 | 9 | 2 |
| Moderate | 40 | 6 | 31 | 1 | 22 | 0 | 0 | 33 | 6 | 1 |
| Conservative | 6 | 3 | 22 | 2 | 42 | - | 1 | 21 | 6 | 3 |
In the end, the stereotypes won and the policy lost. Edwards did amazingly well among conservatives, and badly among those who described themselves as "Very liberal". The "Very Liberal" results look nothing like a Daily Kos straw poll, with perennial single-digit straw poll candidate Hillary Clinton leading Edwards among the most liberal voters. (Obama's success among the 'very liberal' here is probably because of huge college student turnout in a caucus environment where only 11% of voters show up. And he won the moderates too.) The NH results show similar trends, though in a more muted form.
To quote commenter Tony V:
the Iowa caucuses bore out the central thesis of Edwards supporters: moderates and conservatives voted for him because of his identity politics, and despite his liberal policies. Edwards won conservative-identifying caucus goers, while Obama's victory was due to overwhelming support among liberal-identifying caucus goers.
Edwards is up 21 over Obama in the last pre-Iowa Daily Kos straw poll, while Obama beat him among Iowa veryliberals by 24. That's a 45% swing. Multiply that by the 18% veryliberal makeup of the Iowa caucuses, and you get 8.1%, which coincidentally is Obama's margin of victory. The distance between Edwards and victory in Iowa is the distance between 'very liberal' Iowa caucusgoers and Daily Kos readers.
Not that there's any way I could've made the difference short of going to Iowa myself (which I would've considered if I didn't have philosophy job interviews in the days after Christmas) and undergoing several iterations of Parfit-style fission to turn myself into maybe 1,024 pro-Edwards volunteers. But while the Iowa results bear out the thesis that Edwards could turn conservatives into supporters of a progressive candidate, they also show that the task of informing progressives was a failure.
This post, as you may have noticed, is written mostly in the past tense. It'll take a collection of highly implausible events for Edwards to get the nomination, and there are limits even to my wishful thinking. As I prefer Obama to Clinton, I want Edwards to stay in at least until SC, to take away some white voters over 30 and help Obama win that state. And there's a bunch of articles to be written about the Edwards campaign's beneficial effects on our politics. Recently, I've enjoyed his tendency to turn every concession speech into an infomercial about the evils of the health insurance industry. But it looks like the greater pleasure of nominating a candidate so perfectly suited to move America left will be denied to us.
