Yet More On Barack Obama, Class, and the Democratic Primaries

Let me flesh out something I mentioned before, that the electoral coalition for all Democratic insurgents starts from the top of the education/income scale and works its way down, but that Barack Obama has worked his further down than most challengers. Back in the halcyon days of the New Hampshire primary, Granite Prof computed the "elite score" of many prominent Democratic contenders by comparing their performance in working-class versus "elite" areas of the Granite State. Crudely, a candidate with an elite score above 1.0 is a "wine track" candidate, while one below 1.0 is a "beer track" candidate. Here's the chart:

As you can see, yes, Obama is the wine track candidate, as are all insurgent challengers. But he is less wine track-y than almost all other challengers. Only Gary Hart had a better balance of working-class and elite support, and of course he nearly won the nomination. Which, along with the increasing upper-middle-class presence in the party and Obama's tremendous support among African-Americans, explains why he's able to fare better than the Bill Bradleys and Paul Tsongases of the world.

Comments

[ikl3]

Blue collar people in the northern tier, (New England, Upper-midwest / old Northwest, Northwest) usually like Obama just fine (at least if they are Democrats, they do) and this has been pretty clear since NH (the expections have been in urban southern New England and in New York state). Blue collar whites in the deep south or Appalachia, not so much. Unfortunately, most political commentators act like all white folks are just about the same politically no matter where they live, which is manifestly false in the US. I'm tempted to add some snide remark about such commentators being out of touch, but I'm not sure if that is really the problem rather than just intellectual laziness.

Also, it would be interesting to see whether Edwards is more or less wine-tracky than Obama in NH.

[Nicholas Beaudrot]

Edwards was down at like 0.71 ... more beer track than anyone else on the chart.

[Ron]

I usually hate the blogosphere's obsessions with "frames", but isn't even talking about beer track vs wine track in effect buying into conservative and MSM frames that Democrats are elitists who don't understand "ordinary" Americans? We need a better way of talking about this demographic distinction that doesn't push the notion that certain candidates are people you want to have a beer with which helped bring the country our current commander-in-chimp.

[Nicholas Beaudrot]

Yeah, I don't like the frame either. That's why I said "crudely".

[weboy]

I still think your argument here isn't doing you any favors... you'd probably be better off arguing that what's changed is that there's simply more "wine track" voters to count; that the Party is wealthier, better educated, and attracting more of them, which is helping Obama succeed where previous "wine track"types tended to trip up. As it stands, you're relying on data from New Hampshire, which in many ways is ages ago (I'd be curious how that number has moved; I'd guess Obama's moved up the line, myself), and on New Hampshire results that don't reflect well the nation as a whole. I'm not saying, by the way, that Clinton looks better in this... but I think there's a troubling divide this season between the educated, monied elite and the less educated and less well earning. It's a problem. And what concerns me is that neither candidate seems well able to bridge the gap. If Obama can, he really hasn't shown it yet.

[ikl3]

First of all, Obama would move to the right in any place with a large black population - "beer-track" Democrats are disproportionately black. Look at Kerry's results with blacks, for example.

Also Obama did better with working class white folks in places like VT, WI, and WA than he did in NH. He did worse in OH, MA, RI and most of the south. Regional differences matter - perhaps more than class differences (because of high mobility, upper middle class white folks around major cities are probably more homogenous than most other demographics in the US).

The Edwards result suggests that the whole metric is pretty off though - Edwards' support in NH tended to be slightly more affluent than average if I recall exit polls correctly, so he shouldn't really be further right than Clinton. I'm not sure if geography is such a great predictor of class background in NH other than for a few rich towns on the coast.

[DTM]

weboy,

As the other poster suggested, Obama has done quite well among lower-income white voters in many regions. Indeed, he did so in Iowa, and that has continued on in places in the Midwest, like Wisconsin, or further West. He also tends to do better among non-Democratic lower-income white voters.

So, what you are actually seeing is that Clinton has a stronghold among lower-income white Democrats in certain regions (the Northeast Corridor, Appalachia, and the South). This is not so surprising, perhaps, for a Democratic Senator from New York who was also First Lady of Arkansas. And the fact that Obama has indeed been able to "bridge the gap" outside of Clinton's core areas suggests to me this effect is a result of her popularity in those areas, rather than of any systematic problem with Obama.

[Nicholas Beaudrot]

Yes, it's just New Hampshire, but that should affect all candidates equally, right? In terms of post-NH challengers, only Tsongas (and Kerrey?) McGovern, and Hart were really viable. And I don't know that we have exit polls from those days; we'd have to rely on county data.

And I'm not sure I believe the exit polls on the Edwards subsamples; at that point you're talking about a very small number of poll respondents. Edwards did very well in the North Country ... a lot of those places went Clinton-Edwards-Obama. In the Southern part of the state, he did roughly the same in college towns (Keene) and working-class towns (Manchester).

[ikl]

17% of the vote in NH isn't that small a sample in my view. And it is consistent with other primaries and with some states in '04. NH exit poll data wasn't some sort of wierd outlier - check out the '04 Wisconsin primary exit poll, for example. Or the '04 NH exit poll for example.

[weboy]

Okay...

First, Washington state has all sorts of problems as an example - it's one of those weird primary/caucus combos that was confusing and many things - including the internal roiling of her campaign, as well as the deep ground organizing for caucusing - probably kept Clinton from having the presence she might otherwise. I'd be more convinced if you had other, simpler examples; it's not "Washington doesn't count", just "Washington isn't really like the others".

Second, Wisconsin doesn't really prove your point - Clinton won high school grads and they were essentially dead tied in people earning under $50K. That may be worse than some of her other performances, but it's hardly a resounding proof that Obama's winning working class voters in substantial (certainly substantial enough) numbers. Also, though Obama "won" people earning less than $50K in Iowa, it's a three way split; in fact, 51% of people earning less than $50K voted for Clinton or Edwards (much the way, when you add it up 74% of white voters in South Carolina went for... a white person), Obama won 34% of Iowa's lower income voters (NYT data).

Third, by "most of the south" you apparently sweep in places like Texas, Oklahoma and Missouri; that's a pretty broad brush (and by the way, is one reason Kentucky, West Virginia and Pennsylvania, not to mention Indiana, look good for her). Even so, you leave out California... and New Mexico. And Nevada. That doesn't strike me as the regional case your trying to limit it to; moreover, there's a "sample size" question here as well, which goes back to the dilemmas of comparing caucuses to open primaries.

Look, I think you've got something here, and it's interesting; I just don't think it makes the case you want to make, that Obama is somehow the outlier of "intellectually appealing candidates." He seems very much in line with the string of names you cited...but he's not failing in the way they did. As I said, I think the bigger story here is a demographic shift in the party as a whole to a point where older, well off liberals - and their college age kids - finally have the critical mass to count in a way they haven't in the past... even as recently as 2004 (and by the way... where is Howard Dean on your chart?). I think the far more curious thing is that working class voters aren't signing on... and as always, we intellectual types (me included) see them as not comprehending why the intellectually appealing guy isn't the best thing for them. I've only lately come around to the opposite perspective... perhaps the question shouldn't be what don't the voters get, but what does the candidate lack. And that, I think, is still where you're not looking.

[Nicholas Beaudrot]

Let's see here.

For the first few sentences of your conclusion, I think we basically agree. I'm not really saying he's an outlier, more on the edge of the eggheadedness of the spectrum. I think it's a combination of the Just Broad Enough appeal with the rising critical mass of

For this business about high-school graduates in Wisconsin, I call goalpost moving; the median education demographic in the electorate is now voters with "some college but no degree" and tying voters making $50K or less is an accomplishment for the egghead candidate. I mean, if you want him to win white working class voters 60-40 you're asking him to win a primary 70-30, which just isn't going to happen in a contested primary. I mean, if we're setting that as the standard, we should start complaining about Clinton's lack of appeal to people with college degrees.

[ikl]

Obama won every county in Vermont and did well (won or was close) in backwoods Maine. He won by big margins everywhere in Washington - not just in greater Seattle. I don't think that is demogrphically possible without doing well among blue collar white folks. The same goes for lots of other places in the interior west. It is very, very hard to win 2-1 in places that aren't especially upper income if you lose badly among working class voters. Even if caucus voters are unrepresentative (his big win in Utah suggests that Obama would have won primaries in most of these places as well). It is not just Washington - it is pretty much everywhere white parts of the interior west, northern New England and the Wisconsin / MN area (including northern IA and IL).

As for the rest, what Nicholas said - this is a contested primary, not a referendum on Obama.

[ikl]

I do think that Obama's weakness / Clinton's strength (I think that there is some of both at play) in Appalachia and parts of the upper south generally is a real argument for Clinton's greater electibility. Ironically while Clinton has done really badly since Feb. 5th, I think there is a case for her greater electibility that was not at all clear in January (when pretty much all of the results suggested that Obama would do better - except maybe SC which is not really in play for either candidate). On balance, though, I'd rather put VA and CO in play than AR and WV both for short and long term reasons.

[weboy]

ikl, again, you're talking caucus (Maine), after caucus (Washington) after one small primary (Vermont) in arguably the most lefty state in the union (including Mass and CA). I'm not even sure there is exit polling from Washington, their process was so convoluted, but it's clear he won there based on earthy crunchy lefties and college kids - no one, really, has disputed that. Ditto Maine, which was much more about his appealing to wealthy white retirees and, again, college kids.

Second, you want West Virginia. There's no reason we should lose WV, none at all. The idea that Republicans have anything, absolutely anything to offer those mining communities is ludicrous. We've long been able to count on WV, and without it, we start to slip.

Finally Nick, no, it's not goalpost moving - Obama's "some college" numbers in WI were good, but he hasn't been winning that category much either (and then we could get into how WI is likely to go Republican in November, like other states where he does best). In states where she's been competitive, she's won some college even if he's won the primary (Connecticut, Missouri). His best high school and some college numbers come when the population of less educated voters is more heavily black (Mississippi, Georgia, etc). I think the question here comes to making Wisconsin so central; it doesn't seem to make a pattern. And no, he wouldn't need 70% margins to win working class voters, too... because, and this is a big part of it, some of what he'd have to do to win working class voters would probably lose some of the intellectual elite. You can't, I agree, expect to win everyone. But he's really not shown an ability to appeal, really successfully, to working class voters, especially working class whites, given a choice. That's the reality, and I'm quite clear the numbers support me on it. And by the way, I have, and I do complain about Clinton's failures to appeal to the educated set. That's a huge problem too. Neither one is broad based enough, if you ask me.

[ikl]

Weboy, look at the maps. You can't win some of the places that Obama was winning in Maine by appealing to "wealthy white retirees" and "college kids." That works on some of the coast and in college towns but not for much of the rest of the state. Caucuses don't magically make Obama demographics appear in places where they don't live. Similarly, Obama won big all over Washington - not just in Seattle (which is also where the biggest university is located). You can't win based on college students were there are no college students. And what the #$%^ does "earthy crunchy lefties" mean? Is it mutually exclusive with working class? How many of such people live in Eastern Washington? Northern Idaho? Nebraska? Etc.
Vermont has lots of white working class folks. Not sure why they don't count. Have you actually visited any of these states? Outside of Seattle and the Maine coast? You seem to have pretty stereotyped pictures of them.

Check out the Utah exit poll:
http://www.cnn.com/ELECTION/2008/primaries/results/epolls/#UTDEM
Obama wins or ties in all income groups and levels of education. No reason to believe that other parts of the interior west where Obama won even bigger in caucuses are really different.

We have now lost WV fairly badly two very close elections in a row. It has 5 EVs. VA is more important. It has 13 EVs and unlike WV is trending blue. Margins were similar in both states in 2004. This isn't a hard choice.

[ikl]

Also, if you are going to exclude working class blacks for identity politics reasons, I think that it is fair to focus in on working class white men since there is a pretty clear identity politics vote among white women too.

Wisconsin comes up at lot because there was an exit poll. There wasn't for MN, CO, ID, ND and various other caucus states where Obama won big. It is hard to believe that Obama won MN 2-1 without doing reasonably well among working class white folks.

[Kaleberg]

And which one of them won? (I suppose we could give Gore a half credit for the popular vote).