Friday, November 16, 2012

Catch of the Day

To Dave Wasserman, who first of all has been performing a terrific service to everyone by updating the presidential election results when most or all of the major news organization sites haven't been, and second notes:

4,000+ new votes reported in Colorado, now within 0.03% of losing "Tipping Point State" honors to Pennsylvania
Which means that everyone owes a major apology to Team Romney. Even if Pennsylvania doesn't wind up as the state that put Barack Obama over the top, it was pretty close, and Romney's campaign was entirely sensible in putting resources there; indeed, the real question is probably why they didn't do more earlier.

Now, there are caveats. We don't know whether uniform swing is a reasonable assumption. Perhaps if the nation moved four points towards Romney, Pennsylvania would have only moved, say, two points, so that other states would have overtaken it and become more likely to actually flip than the Keystone State. I'm also curious about what effect if any Sandy may have had in PA. If it hurt turnout in a way that helped Romney, then perhaps it wound up closer than people expected.

But at least as the numbers have it now, Pennsylvania was a very reasonable investment for Romney's campaign in the last few days.


  1. So this means Republicans can't win presidential elections without Pennsylvania now? Suggests a Democratic advantage in the EC even more strongly than what we had before.

  2. sweet factoids: Obama exceeded his 08 vote totals in VA (12245) and NC (35737).

    Also, an anomalous increase in LA (26152). Weird.

  3. Except...not really, I think. Here's D voteshare in PA since 1980:

    2012 – 52.54
    2008 - 54.4
    2004 - 50.96
    2000 – 50.61
    1996 – 49.29 (with 9.59 to Perot)
    1992 - 45.15 (with 18.2 to Perot)
    1988 - 48.39
    1984 – 45.99
    1980 – 42.48

    Since Clinton in 1992 Dems have never lost PA. Now, they only lost the national popular vote once in that time (2004), but Kerry did BETTER that year in PA than Al Gore did despite losing USA. I think this mitigates somewhat against "uniform swing" and towards "some states are different than other states." I think Pennsylvania just has a lot of voters that are Democrats and a lot that are Republicans and the Dems outnumber the GOP by a smallish but persistent and difficult-to-shake margin.

    Basically, I think Pennsylvania is more like Oregon and less like Ohio. It's just a lot bigger and everyone always acts like its a swing state so people think it is.

    Another way of saying it: even if the GOP wins the White House in 2016, they will not win PA unless it's a monumental blowout. They will win CO, though.

    1. I think this is right. Pennsylvania, it seems to me, is very stable in its partisan allegiances. I guess there's a leakage towards the Republicans in the western part of the state, but also a leakage towards the Democrats in the Philly suburbs that more or less matches it.

      Pennsylvania seems to be at a fairly stable, slightly pro-Democratic, equilibrium in presidential election years.

    2. No, uniform national swing is pretty solid, I think. You have to account for Nader. John Kerry's absolute vote share and Al Gore's absolute vote share were very close nationally and in PA; you're talking about a shift of a few tenths of a percent. And George W. Bush improved by about 2.8 points nationally and by 2 points in PA:,_2000,_2004

      So PA was maybe off from uniform national swing by a point, but it wasn't as big a difference as you imply, I think.

    3. More precisely, nationally:

      2000, Bush/Gore/Nader, 47.9%/48.4%/2.7%.
      2004, Bush/Kerry, 50.7%/48.3%.
      2008, McCain/Obama, 45.7%/52.9%
      2012, Romney/Obama, 48.8%/50.6%.


      2000, Bush/Gore/Nader, 46.43%/50.60%/2.1%.
      2004, Bush/Kerry, 48.4%/50.9%.
      2008, McCain/Obama, 44.2%/54.4%.
      2012, Romney/Obama, 46.9%/51.9%.

      Don't really think that's far from the uniform swing. Pretty much every time, the Republican nominee does about 1 or 2 points worse in Pennsylvania than nationally, and the Democratic nominee does about 1 or 2 points better.

  4. I live in the part of PA affected by Sandy. Impact on turnout was minimal, as far as I can tell. Turnout was down a bit from 2008 around here, just like it was most places. But not enough to pin on Sandy.

    Romney never had a chance here. I was sure Bush would carry it in 2004, and if Mr. 9/11 couldn't do it then, Mr. 47% wasn't going to this year. It's awfully hard to argue with a +1 million Dem registration edge, at least in a presidential year.

  5. Do you think that +3 swing towards Obama is likely to be real or just an artifact of an incomplete vote count?

  6. Can't reply directly for some reason, but I am mostly convinced by Xenocrypt.


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

Who links to my website?