Skip to main content

View Diary: Is Obama now the frontrunner? On the importance of Iowa and New Hampshire. (135 comments)

Comment Preferences

  •  Ah Stata... (2+ / 0-)
    Recommended by:
    Nulwee, poblano

    My own best friend and worst enemy. I'm currently a SAS man myself.

    Good diary. Tipped and rec'd.

    You know we live in strange times when hearing something as simple as the truth almost seems shocking.

    by redhaze on Mon Dec 10, 2007 at 04:53:49 AM PST

    [ Parent ]

    •  The Problem (2+ / 0-)
      Recommended by:
      Rolfyboy6, kywddavid
      The problem with this diary is that if you'd done the same calculations in '04 at the same point of the race, you'd have concluded that the chances were overwhelming that the nominee would be Dean or Clark.

      Whoever wins Iowa this year is going to win New Hampshire, and is going to win the nomination.  If you want to predict this race, predict Iowa.

      •  Actually it would have predicted Kerry (11+ / 0-)

        This is what the model says if I plug in the 2004 data from just before the Iowa caucus, including an Iowa poll I found laying around.

        Kerry 45.5%
        Dean 38.3%
        Clark 8.9%
        Gephardt 4.4%
        Edwards 2.9%

        Remember, the model thinks that national polling results are really a mixed bag.  They can help you after New Hampshire, but on the other hand, they may be a sign of negative momentum heading into IA and NH, if your results in the early states are lagging behind your national averages.

        •  In fact, the 2004 numbers are sort of similar to (4+ / 0-)
          Recommended by:
          taylormattd, wystler, jrooth, Nulwee

          2008 -- Dean was 15 points ahead of Kerry nationally, 6 points ahead in NH, but had fallen behind in IA by 5 points in the last pre-caucus poll.  

          Now, Howard Dean is a very different candidate from Hillary Clinton, and this is a very different race (just 3 as opposed to 5-6 viable candidates, for one thing), but still those polling margins are very close to what we're seeing right now.

          •  The Same Point In the Cycle... (0+ / 0-)
            Try plugging in numbers from 3 and a half weeks before Iowa in '04, rather than on the eve of the caucuses, and I continue to maintain that you would have shown a Dean nomination, with Kerry far, far behind.
            •  I'll grant you that (4+ / 0-)
              Recommended by:
              taylormattd, kpardue, snout, redhaze

              There's no doubt that Dean would have been way ahead if we plugged in the results from something like December 1, 2003.  At the same time, the model really only advertises itself to work for the polls just immediately before Iowa, so I'm not sure if it undermines the validity of the model so much we're arguably using the model improperly.  Hence the disclaimer about "if the elections began today".

              At the same time, there is evidence that preferences are much harder this year than they were in 2004, at least in the early states.  So while I think that nearly everyone underestimates the extent to which things can change on a dime, I think those changes are somewhat less likely than they were four years ago.

              •  Dem Primary Voters Like All Three Major Candidates (0+ / 0-)
                At the same time, there is evidence that preferences are much harder this year than they were in 2004, at least in the early states.

                I'm not so sure about that.

                Most of the polling I've seen indicates low attachment to choices in the early states.  In addition, all three of the major candidates have very strong fav/unfav ratings in all of the early states.

                Specifically, check how positively all three major candidates are viewed in NH.  Iowa results are going to utterly remake the New Hampshire race, which I think a lot of folks aren't fully understanding quite yet.

          •  Good diary but... (0+ / 0-)

            this isn't like previous primary seasons.  The landscape has changed, everything is more compact now.

            How do you think that will effect things?

            •  It's definitely different -- every year is (1+ / 0-)
              Recommended by:
              flubber

              And I don't know how it will affect things.  When I take my statistics hat off and get into more subjective forms of analysis, I find that I'm perhaps too likely to see the Obama scenarios through rose-colored glasses.  

              But some of these findings are pretty robust.  For instance, there's pretty much no example since 1968 of the results of New Hampshire not shaking things up very significantly, at least not on the Democratic side.  This is different than for the Republicans, where in each of '92, '00 and '76 you had results that didn't move the needle much nationally.

            •  Understanding Front-Loading (0+ / 0-)
              The landscape has changed, everything is more compact now.

              Increased front-loading increases the importance of both IA and NH.

              Moving up the NH primary from 7 days after IA to 5 days after IA increases the influence of IA on NH.

              If you want your candidate to win this nomination, root for them to win IA.

            •  Two differences (1+ / 0-)
              Recommended by:
              taylormattd
              1.  The front loading makes Iowa and New Hampshire more important.
              1.  There is significant evidence already that the interaction between Iowa and other early states is happening before the Iowa Caucuses themselves.  See the Mason-Dixon polling this weekend that shows deadheats in SC and Nevada.
        •  I'd like to see other years too (0+ / 0-)

          if you don't mind.  Republican results would be interesting as well.

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site