Mike the Mad Biologist

A Case Study in Pundit Over-Analysis

Here’s one example, unintentionally brought to you by NY Times columnist Frank Rich, of how writing political narratives instead of discussing data leads to unsupported conclusions (italics mine):

The continued political import of Iraq could be found in three different polls in the past six weeks — Pew, ABC News-Washington Post and Wall Street Journal-NBC News. They all showed the same phenomenon: the percentage of Americans who believe that the war is going well has risen strikingly in tandem with the diminution of violence — from 30 percent in February to 48 percent in November, for instance, in the Pew survey. Even so, these same polls show no change at all in the public’s verdict on this misadventure or in President Bush’s dismal overall approval rating. By the same margins as before (sometimes even slightly larger), a majority of Americans favor withdrawal no matter what happened during the “surge.” In another poll (Gallup), a majority still call the war a mistake, a finding that has varied little since February 2006.

It’s safe to assume that these same voters did not forget that Mrs. Clinton and Mr. Edwards enabled the Iraq fiasco. Or that Mr. Obama publicly opposed it. When Mrs. Clinton attacked Mr. Obama for his supposedly “irresponsible and frankly naïve” foreign policy ideas — seeking talks with enemies like Iran — she didn’t diminish him so much as remind voters of her own irresponsibility and naïveté about Mr. Bush’s Iraq scam in 2002.

Rich is doing fine until the italicized part. Maybe, particularly if you’re trying to shoehorn facts into a narrative (although to Rich’s credit, it’s a new narrative), Clinton’s criticism reminded Rich of her voting history about Iraq. But an equally plausible explanation, and I would argue, far more parsimonious and straightforward, is that a lot of Democratic voters think that engaging in sincere talks with Iran is preferable to blowing things up and killing people. That seems a far more likely explanation. Rich makes a typical pundit error, which is thinking that his conclusions are broadly shared by others, when, in fact, this is basically a Rorschach Blot test of Rich.

I was tempted to state that what this really shows is Clinton’s Lieberman-esque saber rattling probably didn’t play well to an electorate that was disproportionately anti-war and consequently was accused of treason by the same people who are now urging violence against Iran. Think of it as a dog whistle in the wrong direction.

But that would be projecting my narrative onto the Iowa results…

Update: Tristero makes a similar point about Gail Collins.

Comments

  1. #1 natural cynic
    January 10, 2008

    The voters in New Hampshire had to be in a serious WTF mode in the polling booth. The evidence lies in the exit poll shown in Kevin Drum’s blog. The D voters who wanted to get out of now went for Clinton and the voters who wanted a slow withdrawal/stay in Iraq went for Obama. The R voters were in the same mood, with those disapproving the war favoring McCain and those approving the war went for Romney.

  2. #2 Troublesome Frog
    January 10, 2008

    This is very similar to the financial programs that recap the stock market fluctuations of the day and add a narrative of what happened. “Stocks climbed early today due to optimism about oil prices and shot down sharply in the afternoon as investors reacted to the subprime crisis.” Nothing in the news during that period of time actually supports the analysis, but it’s a nice story and it keeps the paychecks flowing.