Monday, October 4

Guest Blog: Bill Heald

When he submitted this, I wondered why anyone would send me a piece titled "Battle of the Ukeleles." Yep. The weekend in Tahoe definitely killed a few brain cells. Here's Bill's excellent contribution. Feel free to comment.
by Bill Heald

Amidst all the chatter about how the upcoming election will be so tight that it will be a horse race, I think I should point something out: horse races often aren't even close.

Sometimes the long shot bolts out of the starting gate like a thunderbolt and wins by ten lengths. In that same race the odds-on favorite might toss their jockey, jump the fence and then wander around in the paddock looking for sustenance. Most horses finish somewhere in between these two extremes, but the really close contests are fairly rare.

Despite this, the horse race metaphor is used ad nauseum to imply the Bush-Kerry contest will be a neck and neck, photo-finish race. The media machine lives for such an election, but truth be known the actual race is beside the point. The election is over (usually) in a day, and it's the months-long buildup that matters. Nothing must interfere with the day-to-day hype of a close, 2000-style race for it keeps talking heads talking, analysts analyzing, and cable news chat shows narrowly surviving in the gigantic ratings shadow of SpongeBob SquarePants.

In order to maintain this drama, the scribes and broadcasters must have data to back up their "statistical dead heat" rhetoric. Much like a racing analyst that pores over past results to try and divine what a horse will do in its next outing, pundits and journalists rely on the only set of statistics they really have: the polls.

Daily tracking polls are the bread and butter of political prognosticators, but there's a serious problem with this form of intelligence gathering. While the major polling companies use good math and sound statistical analysis, their results are suspect because they're trying to randomly sample humans about their opinion instead of, say, a lemming population for pierced ears.

You can't get a true random sample of a voting population these days because only a specific, nonrandom sector of the population will participate. Phone polls only get data from willing participants in the polling process, which prevents a true random sampling. Adding insult to inaccuracy, many times these polls focus on "likely" voters, which are typically determined through surveys on past voting behavior. This is critical, because past elections are largely irrelevant when considering the upcoming contest in November.

Why is this election different? Simple. A huge army of human beings that typically doesn't pay too much attention to politics or voting is getting involved this time around. There's great motivation afoot. Well-known individuals who stayed on the fence in the past (like Bruce Springsteen) are speaking up, and in the Boss' case, singing out. This should not be taken lightly. Mr. Springsteen stated he had to abandon his neutral stand of the past because, "this year the stakes have risen too high to sit this election out."

Indeed. A few years ago, if you asked a music critic if an icon like Springsteen might make strong political statements and risk alienating some of his fans they would have said, "unlikely." But things have changed. Whether he meant to or not, Bruce was expressing the sentiments of many Americans. Unlike recent elections, issues like war and deception are stirring the pot considerably. While the punditocracy rambles on about how this presidential race will be determined by Swing Voters or Undecideds, they're missing a more significant group: the Unlikelies.

I know these people, and they don't show up in polls. They are elusive and difficult to locate, like timber rattlers in Connecticut or pale, pasty skin in San Diego. Unlikelies are rarely home to talk to pollsters, and when they are in-house they screen their calls to avoid such annoyances. Many don't even use their home phone anymore, preferring to communicate via cell phone or email. Political matters are typically not a big concern, and they ignore cable news and talk radio. But things like soldiers dying on the streets of Baghdad, health insurance woes, job insecurities and environmental concerns have stirred them to come out from their various hobbit holes and secured locations. They are angry, and fully intend to do what pollsters consider an unlikely thing for them to do: vote.

Since these Unlikelies clearly want a change from the current regime, they lean strongly towards the challenger in this election. At the same time, the incumbent has set out to find his own Unlikelies such as the Amish in Pennsylvania or evangelicals in Ohio that didn't vote in 2000. He's also trying to get Vietnam veterans to vote against one of their own, a bold but dangerous strategy. Will Bush's Unlikelies level the playing field with Kerry's Unlikelies, and make this a photo-finish horse race after all, like all the polls are telling us?