Wednesday, February 05, 2020

Correlation is not Causation

My memory of Iowa caucuses past is that they were perfectly meaningless and usually overlooked almost entirely in favor of the primary in New Hampshire (First in the nation!*).  Then, before caucus night, and in blissful ignorance of the debacle 4 years ago that pissed off Bernie Bros. so much we got the new and improved caucus rules for 2020 (thanks for the sour persimmons, cousin!), articles started appearing about how Iowa had "picked" the eventual nominee 7 out of the last 8 times (or some such).  It didn't occur to me then to wonder if that was correlation or causation, but yeah:  how many times did Iowa pick an unknown/underfunded candidate who also won the nomination?  I don't know either, but I'm betting it wasn't all that often.

*And as for primaries being scattered all over the calendar and the country (and considering how much time/money was spent in Iowa just to claim a victory that meant few delegates at the convention but "momentum" going into New Hampshire, a momentum that now goes to Buttigieg whether he actually "wins" Iowa or not), I'm with the Boston Globe:

**I'm also old enough to remember when Biden wasn't counting on Iowa, but rather on South Carolina.  Now, of course, he's doomed because he didn't carry Iowa.  I don't care, I'm not that excited about Biden; but a single-day primary would probably end this kind of idiotic reporting/punditry:
Again: I know, Iowa and New Hampshire have picked the nominee for decades now.  But correlation is not causation, and yet losing Iowa or New Hampshire (or both) can mean losing support that might carry you to the nomination.  Why?  Because you lost Iowa and New Hampshire, who ALWAYS pick the winner?  Or is the winner chosen because of Iowa and New Hampshire, and not by them?

No comments:

Post a Comment