Well - the Ames Straw Poll has once again come and gone. Huzzah! And, as we all know now and as we all expected, Mitt Romney had a big win.

Whatever. I think that the straw poll is largely one of those media-created phenomenon. It is a product of the media's desire to find a story in August when there is just nothing happening. This is not to say that it does not have value, but it is to say that its value has been greatly overestimated.

I hate to be a spoil sport by pointing out the following. In the four previous straw polls, only once has the clear winner of Ames gone on to win the nomination. And, more importantly, only twice has the clear winner gone on to win Iowa. So, I'm not putting very much stock in these results - though I would not deny that this is all great publicity for the Romney campaign.

The coverage of the straw poll is thus more evidence of the press' tendency toward Type I error. Type I error is the error of the false positive, i.e. the error of identifying something that does not really exist as existing. I believe that the incentive structure of the press - namely the demand of the news cycle for "news" every day regardless of whether there is real news to put out - is what systematically induces this bias. The press is always and everywhere compelled to over-report events, to over-estimate the importance of the events it reports on, and to over-think about the implications of those events. They have to. There is time and space that needs to be filled with "news," whether the news is real or not.*

The response to Ames is an excellent case in point of this systematic methodological error.

I'll just make three observations about this exercise in news creation:

(1) My hat is off to the Iowa Republican Party. What a fundraising coup they have managed to put together for themselves!

(2) When John McCain and Rudy Giuliani decided to drop out of the race - nearly every journalist or pundit who saw fit to comment upon their demurrals boldly proclaimed that the straw poll is accordingly rendered meaningless. Interestingly, nearly all of those same journalists and pundits were in Ames over the weekend, or writing afterwards about the meaning of the vote. Do we require further proof of the existence of this Type I error? The Ames contest thus recalls to my mind one of the major problems with mainstream media analysis of politics. Nobody remembers anything they wrote or said previously, nobody is "keeping score," and thus nobody is making good use of their past mistakes to learn how to get things right next time. Karl Popper would be most disappointed.

(3) Too bad for Mike Huckabee. To do better than expected, but not to beat a top-tier guy in so doing.

-----
* - Unfortunately, the press' tendency toward Type I error is matched only by its tendency toward Type II error. Type II error is the error of the false negative - or failing to identify as existing something that does in fact exist. The press makes robust use of this error, too! For the nature of the press' incentive structure means that it systematically fails to appreciate fully any causal factors that have any kind of permanence. Political events of today are overcaused by the events of yesterday and undercaused by the events of the last decade.