Dear Reader - A rant on the utter ineffectuality of modern polling.
Voodoo math? Crack for addled statisticians? You'd do just as well with a drinking game to throw darts at a political map.
In three out of the last four national elections, most polls have not been at all predictive of the general outcomes. Excuses are rampant, but the results are inarguable. It’s getting tiring. The next time some inexplicably arrogant pollster looks down their nose at the rest of us and pontificates, smugly, that their polls were actually spot on if the three and a half pages of small print (like that which comes with an extended warranty from a telemarketer) are taken into account, I'm going to lose it - Falling Down style. Don't even be in the same time zone with me if this happens. I kid you not.
There is no shame in being wrong. Being wrong is such a common occurrence that there are no less than a hundred synonyms for it in any thesaurus. Being wrong can also be also quite useful. By analyzing why things went wrong in any post mortem, one may address mistakes made with the goal of not being wrong again in the future. The world improves when things are set right that once went wrong.
If, that is, you recognize that there's room in your game for some improvement.
Striving for redemption seems beyond the ken of most pollsters. Virtually no poll anywhere predicted anything other than a “red wave” in this election. Pollsters can weasel around this by pointing to error bars the size of the Empire State Building all they want, but they just weren't accurate in predicting the results. This is not a one-off either. It's a genuine trend.
The difficulties of polling in the modern age have been well-documented. Most people no longer have land lines and are difficult to reach by phone. Many of those who can be reached by phone don't want to spend fifteen minutes of their life that they can't get back talking to a stranger about things they may well regard as none of anyone else's business. Other methods of contacting potential voters are difficult to sufficiently randomize or get a large enough polling sample to be useful.
I get that it's a daunting prospect. I get why valid samples are difficult to obtain. I understand the challenges. What I don't get is why pollsters just aren't more upfront about the fact that given the obstacles, their predictions are basically educated guesses - and that the odds for improvement aren’t looking great without, say, engaging big tech to spy on us even more than they already do.
This happens, by the way. Statistics on clicks, page views and comments aren't tracked for nothing. This might be the next polling frontier.
Most of this musing is rhetorical. There are several reasons why pollsters still receive enough oxygen to joust with windmills despite lousy outcomes. The first is because they are an established part of the advertising-driven quest for eyeballs and ears that is the media. The second is that both sides of the political divide in America see polling as a way of motivating their voters. The third is that both political parties use polling to spend gobs of campaign money.
As Leonard Leech observed in God Bless You Mr. Rosewater, one should always be on the lookout for “situations where large amounts of money are about to change hands.” That’s one area in which many polling firms succeed spectacularly. Money is catnip to pollsters.
In complete fairness, the polling this time wasn't uniformly terrible – especially in individual races where the outcomes were close. The problem is that all of the major polls got the overall trend quite wrong. Apparently they failed to read their own fine print and decided to call things for Team Red when there was simply not enough reliable data to make that determination.
In any other industry, anyone with a recent track record on par with most pollsters would be given an umbrella and a bag of beef jerky before being shipped out on the next balsa wood raft to the South Pole. It's only when one is in deep with the media that being a perpetual screw-up carries no real consequence.
Just ask Don Lemon.
Associated Press and Idaho Press Club-winning columnist Martin Hackworth of Pocatello is a physicist, writer and retired Idaho State University faculty member who now spends his time with family, riding bicycles and motorcycles and playing guitars.