Outliers: What Oregon Shows About Kentucky and West Virginia

Another lopsided win in Appalachia, and the chatter resumes about how bad off Obama is among the all important white vote.  Marc Ambinder, at least, posts up some exits from both states that voted tonight, though his focus, like many others, are on those dreadful numbers from Kentucky.  Yahoo’s headline reads, “Exit poll: Whites back Clinton strongly in KY,” and evenChris Cillizza is in on the act.

Now, I was doing what I’m sure few others will be doing, that is, reading the exits from Oregon, when it hit me, everything that is wrong with the white working class argument.

You see, a curious thing became clear when I read the exits provided by CNN; Obama won white people.  He won young white people and middle aged white people.  College educated white people, and non college educated white people.  He won female white people and male white people, and a lot of the church going white people.  About the only white people Obama didn’t win were the over sixty crowd.

But, wait, I thought the conventional wisdom was that Obama couldn’t win white people.  Why did he just about sweep the board on white people when it came to Oregon?

Well, for one thing, all these “questions” about Obama’s appeal to the white working class voter are being raised by his performance in West Virginia and Kentucky.

The more time you stare at polling data, the more obvious the problem should become.  Let’s show some examples.

With 83% reporting in Oregon, Obama has a 16 point lead over Hillary Clinton.  Now, if you had been following all of the polling data that has emerged from Oregon recently, this will not come as much of a surprise. Pollster’s aggregation in the Beaver state showed Obama up in the mid teens.

But, if you were to run off of the most recent poll reported by Suffolk, Oregon all of a sudden becomes some sort of miraculous ascension for Obama given that Suffolk only had him up by four points.  Meanwhile, in Pennsylvania, if you heeded only the final ARG poll put in the field before voting, you would be shocked to see that Clinton failed to win by sixteen points.

Responsible journalists, pundits, political anylists, and bloggers ignored these polls, however, because they were outliers.  They were data points, true, and carried with them the possibility of being correct, but in the face of quite a bit of data to the contrary, there just wasn’t that much reason to put a lot of trust in them.

This didn’t, nor will it ever, stop some from engaging in overly reactionary analysis that treats every outlier bit of data as though it were gospel, which is slowly beginning to bring us back to our point with West Virginia and Kentucky.

Now, as Gallup reported yesterday, what they are seeing nation wide runs contradictory to what we saw happen in West Virginia and Kentucky.  Namely, on a national scale, virtually every core demographic that used to be firmly in the Clinton camp has shifted to Obama’s favor.  This national poll would seem to be backed up by Oregon’s exits which show Obama beating Clinton in many of the demographics she has been touting that only she could win.

In other words, West Virginia and Kentucky are the electoral (or delegatorial if I wanted to make up a word) equivalent of outliers.  They don’t mesh with what’s going on in the rest of the country.

Sure, Ohio and Pennsylvania don’t paint rosy pictures for Obama, nor are they outliers.  But in those states, based on the margin of victory for Clinton alone, there should be no reason that Obama can’t be effective there with some vigorous campaigning, a wise allocation of resources, and a unified party.  Those states weren’t outliers, and show true challenges that the Obama campaign.

But to treat West Virginia and Kentucky as though they were indicative of the country as a whole is downright ludicrous.  It’s the equivalent of saying that a forty point poll on the eve of an election spells assured doom when dozens of other polls show a close race.

All this talk about Obama’s problem with working class white with West Virginia and Kentucky as the banner data points is nothing more than reactionary analysis based upon outlier data, and nothing more.  Does Obama have challenges ahead with working class white voters?  Yeah, he does in some regions, but it’s nowhere near as bad as these two states indicate, and he has plenty of time to fix or circumnavigate the problem.

Thus, the leading purpose for promoting these memes based upon this data is to undermine Obama’s electability status in an intellectually dishonest, or at least ignorant, manner.  There are two more states left that are almost homogenously white, and they are likely to go for Obama, and I wonder if then perhaps there will be a slightly more honest approach to Obama’s problems with working class white voters.

After all, it was the white vote in Iowa that propelled Obama’s candidacy from the long shot to the frontrunner.  The only way to discount that would be if there was mass buyer’s remorse regarding the likely Democratic nominee, and unfortunately for those trying to make the case that Obama somehow doesn’t have the support, there’s just no evidence that that kind of buyer’s remorse exists.

I now return you to the breathless and excitable musings of everyone else.

8 Responses to “Outliers: What Oregon Shows About Kentucky and West Virginia”

  1. Chief says:

    My question re: voting in Dem primaries. How many of the people who voted for Sen Clinton in PA and OH were crossover Republicans? Limbaugh was promoting Operation Chaos.

    Sen Obama certainly has this 65+ year old white guys vote.

  2. Heh. Glad to hear it.

    Off the top of my head, and that’s as good as you’re going to get as I’m getting ready for bed, but I can look it up later if I get time, I don’ think that Republican crossover vote ever got up over a few points.

    I do vaguely recall that in Indiana, where Clinton one by a very narrow margin, the Operation Chaos vote actually extended up towards ten points, but I can’t be sure. What I am absolutely positive about, though, is that Clinton did not win Indiana by greater than the Operation Chaos contribution.

    Yes, Republicans voting to extend the Democratic primary succeeded in that state, but that’s the only state in which I feel safe confirming it.

    Still, as much as I would love to be able to use that in some kind of argument for Obama or against Clinton, I can’t. Fact is, those are votes that Obama really isn’t going to get in the fall, and must be accounted for elsewhere.

  3. Mark says:

    At what point does it become politically correct to point out that the states where Hillary seems to trounce Obama with “white, working class, less educated voters” also happen to be states or areas with really bad records on race relations? By the way – the exit polling data bears this point out, as Sully pointed out last night with respect to Kentucky. But I’ve noticed similar data in a lot of the other states where Hillary has done quite well.

  4. Anita says:

    And how much of that national trend toward Obama is due to the media drumbeat that he’s the nominee? I would bet quite a bit.

    And of course you insert the qualifier “there should be no reason that Obama can’t be effective with ……. a unified party”. Which is precisely the point. He can’t unit the party. If he was capable of doing that it would have happened already.

  5. Manamongst Hussein says:

    Especially when some from the Clinton side will cling to excuse after excuse and make him out to be the boogie man. What SHE is doing now is giving her supporters false hope…”going toe to toe”? Really, every non-clouded and clear thinking person know’s it’s over. The vast majority of her supporters now rightfully will continue to vote for her until she pulls the plug. They won’t allow themselves to be united. Good thing for them this is still a slow moving train out of the station. As each week goes by they will realize HOW BAD a candidate John McCain is. And after that point they’re welcome to vote for him…something tells me that they were the dummies that crossed over and voted for Bush a second time…buying into the elitist crap they’ve been suckered into again by the Clinton camp. in short, we don’t care when you get on the train, just get on.

  6. Chief says:


    That wasn’t my point. “Fact is, those are votes that Obama really isn’t going to get in the fall, and must be accounted for elsewhere.” We were talking about outliers and the overall Dem nominating process. The fact that a certain number of Republicans crossed over and voted for Sen. Clinton makes the race for the Dem nomination appear closer than it really would be if only Dems had been voting. And ‘exit polls.’ Do we really want to believe them?

    I view WV and KY as non-predictive of the Dem outcome in November.

  7. Okay, so we were really kind of talking around each other; understandable as I replied to your original comment at about bedtime for me, which is not my most lucid and cogent.

    Do I really want to believe exit polls? Sometimes. Exit polls in advance of official tallies, not unless they are very solid, ie. Kentucky’s exits were probably pretty accurate. After the tallies come in, sure, as they provide some of the data that we use to analyze what happened.

    But yeah, we’re about at the right place here. Clinton’s support, I believe is being falsely inflated by the Chaos effect in some states, particularly indiana, which, taking it away from Obama, does mean that there is not quite as strong of an indicator to believe that Clinton would perform as well as she suggests in those states.

    And no, WV and KY are not predictive at all, especially considering how few electoral votes they are worth. Still, I think Obama’s going to contest them. I think, assuming he continues to haul in money at the rate he has been, I think he’s going to push every state regardless of how likely he is to win it.

    If you want an example of where the Democratic party really is, I think Gallup’s numbers make an awful lot of sense at this point.

  8. bostondreams says:


    Do you really believe that CLINTON will unite the party? Does it not work both ways?

Leave a Reply

Your email address will not be published. Required fields are marked *

Connect with Facebook