Slacking Off

Well, I must report, unfortunately, that I will probably be doing little posting this week. Life -- especially professional life -- has once again rushed in with a vengeance, crowding out other pursuits. So you will be treated to little of the banality incisive commentary that you have come to regret expect from this Weblog in the five years one month since its unholy spawning birth.

Having said that, I must spend just a moment this morning carping about yet another great moment in the art of the opinion survey. The New York Times published its latest poll of the electorate this past Saturday -- which, according to the paper, documented that George Bush had opened up a significant lead (eight percentage points) above John Kerry among registered voters. In what passes for contemporary political discourse, this kind of poll has the power to drive public opinion, since many fence-straddling voters will, in the end, fall prey to the very human tendency to want to back the winning horse.

Unfortunately, looking closely at the Times survey, we find that, of the respondents, some 32% didn't vote in the 2000 presidential election, 36% voted for Bush, and 28% voted for Gore (with 3% voting for other candidates). Put another way, of the portion of the sample that did vote last time around, 54% voted for Bush and 42% voted for Gore -- even though Gore won the election by half a million popular votes.

Can you say "biased sample"?

Yet the reporting of the survey gave no indication of that, and every indication that Bush was "opening up" a significant lead. In a close contest like current one, journalism such as this, especially from an "authoritative" voice such as the Times, crosses the line from the merely slipshod to the outright irresponsible.