How the TV networks may pervert the Iowa result.

A mostly political Weblog.
Dec. 17 2007 8:55 PM

Manufacturing a Winner?

How the TV networks plan to distort the Iowa vote.

It's come to this: Counterproductive overspinner Chris Lehane and his firm get $100,000 a month  (according to S.F. Chronicle's Matier & Ross) to craft mindlessly combative sound bites for Hollywood studios in their dispute with the Writers Guild. Sample sound bit (after union president Andy Stern severed ties with Lehane):

"The real issue here is that Stern needs to do some explaining on how it is that he is fighting for people who make more than doctors and pilots against the interest of real working-class people (set workers and others who have been sidelined during the strike) - and less time punching at shadows."

I dunno. That one was worth maybe only $99,000. ... But hey, you have to hit back! It's the Lehane way. Ask President Gore. [So now that people who buy you dinner are on strike, you're suddenly pro-union?--ed No. Just Anti-Lehane.] 5:50 P.M.

___________________________

New Clinton ad: "Hillary's mom lives with her." But does her husband? Mickey's Assignment Desk: Has anybody updated Patrick Healy's May, 2006 story and calculated the number of days Bill Clinton has spent in Hillary's Washington, D.C. house in the past year (now that it's been officially designated as the place where you live when you live "with" Hillary)? If you're going to flaunt your home life then people are entitled to examine your home life. ... Assigned to: Healy. Hillaryland already hates him. He might as well take all the flak.  ... 5:19 P.M.

Advertisement

___________________________

The TV networks are screwing around with the already-absurd Iowa caucuses again, using an "entrance poll" of only 40 precincts (out of more than 3,500) that threatens to manufacture a misleading result. Ah, but it's all justified because of the valuable information the network poll will gather! Politico's Roger Simon reports

Though the actual questionnaire that will be handed to voters is a secret, Kathy Frankovic, the CBS News director of surveys, told me it would probably include 12 to15 multiple choice questions asking such things as when the voters decided on whom to support, how they feel about the Iraq war, whether they are in a labor union, their political philosophy (i.e., liberal, conservative, etc.), and age, income and level of education.

Armed with this information, a network analyst can say: "Obama got 53 percent of the anti-war vote, while Clinton got 47 percent of the labor vote and Edwards got 36 percent of those who made up their minds in the last two weeks."

a) I deny this information is that useful. If Obama wins, I bet he got most of the anti-war vote! I don't need an entrance poll to tell me; b) Network polling place surveys have a history of humiliating error. Ask Presidents Gore and Kerry; c) The information, even if accurate, is likely to be deceptive. If Obama does get 53 percent of the antiwar vote, that might mean antiwar voters shopped around and found Obama the most anti-war of the candidates. Or they might have liked his smile. They might even have liked Obama first, before thinking about the issues, and then become antiwar voters because that's what Obama talked about; d) Mainly these unenlightening little correlations let network news divisions fill time--because the real news (who won) comes at an inconveniently late hour and then only takes about 10 seconds to report;  e) The conceit of the "caucuses" is that voters meet, argue with their neighbors, listen to speeches, and then vote. But the entrance poll records their preference before the arguing and speeches; f) Worse, the entrance poll results threaten to have a Heisenbergish outcome-distorting effect, since they may be known before the caucus votes are finished and will instantly flash on everyone's Blackberry, cell phone, etc.. If Obama is barely edging out Edwards in the (possibly inaccurate) entrance poll, with Hillary third, will Hillary order her supporters to switch over to Edwards in order to deny Obama a win?  I don't think that's too far-fetched. ... 

Backfill: The networks wouldn't have to resort to a questionable "entrance" poll if Iowans voted at normal hours using, say, easily-countable ballots. But that's not the Iowa way. For a preview  of the state's near-identical vote problems from four years ago, see "The Four Votes of Iowa."  Key point: