« Comments Are Now Operational | Main | Umberto Eco, Meet Samuel Kernell »

The Washington Post: If You Don't Get It, You Don't Get It

I was late getting started this morning because the Post contained such a bonanza of bloggables:

  • College students consume more alcohol on football game days than on such well-known drinking days as New Year’s Eve and Halloween, according to a study published in the November issue of Addictive Behaviors. See the abstract here.
  • “It’s an immutable law of political physics that those who prevail in Iowa will hurtle toward New Hampshire with bulked-up poll numbers, gathering blinding momentum on the path to nomination.” Keep reading Howard Kurtz’s column here.
  • The Democratic presidential candidates can’t agree about where “upper-middle class” stops and “rich” starts. See Joel Achenbach’s piece here.
  • Even if John Wilkes Booth hadn’t killed him, Abraham Lincoln may have died of cancer within a year.

And my favorite, because it features such a weird way of testing a new social psychological theory:

  • Using magic markers, leaders are more likely to draw the letter “E” in the opposite direction (inside-out) on their foreheads from the way that followers do it (outside-in). See Shankar Vedantam’s write-up here.

Posted by Lee Sigelman on November 26, 2007 02:22 PM |

Comments

Howard Kurtz says “It’s an immutable law of political physics that those who prevail in Iowa will hurtle toward New Hampshire with bulked-up poll numbers, gathering blinding momentum on the path to nomination” but it never hurts to look at the actual results.

For all that hurtling towards New Hampshire, only 5 of the 11 winners of a contested Iowa caucus from 1976 to 2004 have won New Hampshire. (I’m excluding as uncontested: Reagan 1984, Bush 1992, Harkin 1992, Clinton 1996, and Bush 2004.)

Ever since 1984, NH has followed IA by only 8 days (it had been 36 days in 1976 and 1980), and during that time, the stampede effect is even less pronounced: 2 of 7 winners of a contested Iowa caucus went on to win NH.

So I disagree with the notion that the Iowa winner is going to turn into an instant steamroller.

Kurtz could have found this information in five minutes on Wikipedia. I guess it was easier for him to talk off the top of his head, as he so often does.

Posted by: low-tech cyclist [TypeKey Profile Page] | November 26, 2007 05:47 PM

I haven’t checked your math, but I think you’re exactly correct. I led with that statement from Kurtz’s column precisely because I hoped it would open up this topic to informed commentary. I’ve often found Kurtz’s columns insightful, but he seems not to have done his homework this time around

Posted by: Lee Sigelman | November 26, 2007 07:46 PM

Does anyone know of a good large-n study on this?

I have an undergrad who has collected data for all primary contenders since 1972. Data include (1) national pre-Iowa public support, (2) Iowa results, (3) national post-Iowa, pre-N.H. public support (with missing data), (4) N.H. results, (5) post N.H. public support, and (6) whether they got the nomination.

We just started looking at the data. Controlling for pre-Iowa public support, we find that how one does in Iowa influences how they do in N.H. and that how one does in N.H. influences probability of getting the nomination, again, controlling for pre-Iowa public support.

Posted by: JMA | November 29, 2007 03:04 PM