Monday, November 19, 2012

Catch of the Day

To Nate Cohn, who previewed the vote counting and reminded us before the election (as Andrew Sprung reminds us today) that Barack Obama would gain considerable ground on Mitt Romney after election day -- a lot after election day.

The Cook Report's David Wasserman has been keeping tabs on the current vote totals, and finds today that Obama's lead is up to 3.1% nationally, and still rising. That's up from 2.3% the morning after election day; I'm not sure what it was when the networks closed up shop on election night. That's a pretty large swing! It seems pretty likely that Obama's lead will wind up a full percentage point higher than it was at that point. Of course, no one really cares that much is a national lead goes from 2.3 to 3.1 percent, but they sure might if it were to go from a 1% GOP lead on Election Day to, say, a half a point lead for the Democrat when all was said and done.

Meanwhile, Colorado was at 4.7% on the morning after and is now at 5% even (actually a 4.98% lead, just below Obama's 5.02% lead in Pennsylvania as of now).

So a five point shift to Romney, with uniform swing, would have put Romney in the Electoral College lead on the morning after the election, with a 0.3% margin in Colorado, and with a fairly massive 2.7% national vote lead. Two weeks later, Obama would have emerged as the winner by taking Colorado by the narrowest of margins, while still trailing the national vote by just under 2%. Yeah, that's gonna cause some trouble. Indeed -- Iowa, New Hampshire, Wisconsin, Nevada and Minnesota would all have been very close, and in this scenario all of them would be subject to recounts, allegations of voter fraud, challenges to provisional ballots, and the rest of it.

I'm not sure there's much to be done about any of this, but Cohn deserves the Catch for pointing it out in advance, and it's really something that everyone should be extremely aware of in advance next time around.

Of course, another reasonable reaction would be to find a way to make the mechanics of elections a lot smoother.

Anyway, all of this is at least a mess just waiting to happen.

Nice catch!


  1. Of course, another reasonable reaction would be to find a way to make the mechanics of elections a lot smoother.

    What would you recommend? Voter-verified paper trail? Mandatory paper ballots? Federalization?

    I, myself, wouldn't mind a federal take-over of elections (at least those elections where federal offices are on the ballot). But that would probably take a Constitutional amendment so I'm resigned to the fact that it ain't gonna happen.

  2. My proposed solution never seems to get the support it deserves: Jarvis as philosopher-king.

    1. Matt, sounds good to me, sign me up.

    2. Praises be to Jeff, the first citizen of Jarvistan!

    3. Can I have a duchy? I will accept any currently vacant duchy.

  3. The popular vote was virtually tied when the networks called the election for Barack Obama.

  4. "Anyway, all of this is at least a mess waiting to happen."

    You know, I have this really vivid memory that it already did happen -- in FL, NM and WI.

    Am I a precog?

  5. The Nate Cohn link is here.


  6. Well, I care if Obama's lead goes up. I have a bet riding on the outcome. I took Obama -2% in the popular vote. It looked awfully close last week.

  7. This comment has been removed by the author.

  8. Nate Cohn has does a great job with explaining the digital strategy behind Obama's campaign. Big data and products that are able to best decipher it, like SAP's, are going to continue to refine how elections are carried out.

    What's possible, but hard to believe now, is what a number of statisticians and Big Data gurus have alluded to -- the growing accuracy of data mining techniques could actually render the physical act of voting unnecessary. Of course we won't stop going to the polls, but the Tuesday suspense may no longer exist.


Note: Only a member of this blog may post a comment.

Who links to my website?