If memory serves me, I have to write some variation of this column every election cycle.
Just 10 years ago, public opinion surveys for elections were a mysterious mix of science and technology. It held little interest except to the political campaigns that lived and breathed by its numbers. But in the last decade, political polling has become a sport. Make that a blood sport.
Some readers will recall the late TV sportscaster Howard Cosell. He was brilliant, but as he put it in the title of his own autobiography, "I Never Played the Game." These days, for every one pollster, striving to be consistently accurate in a rapidly evolving industry, there are three Howard Cosells. They've never conducted a poll, but they will tell you how to do it anyway. By the time the actual elections roll around, the pollsters have become as controversial as the candidates who are vying to become governors and senators. The whole thing becomes a circus.
The truth is that technology has turned the world of polling upside down. Years ago, it was considered a given that the "gold standard" of public opinion polling was to get the data exclusively from live, person-to-person interviews, on "landline" telephones.
The only problem was that sometimes these surveys produced disastrous results. Often the problem was in the demographic "weighting" of the poll. But sometimes things boiled down to mistakes in some faraway call center, where a worker on an hourly wage might mispronounce the name of a candidate -- or maybe fudge the number of calls completed.
Then came the "IVR" system, which uses automated phone calls in place of live interviews. This new way of polling was roundly denounced by the big news organizations (some of whose own polls were frequently off the mark). The political Howard Cosells dismissed IVR polling outright.
Things have only accelerated, leaving many naysayers behind. We live in a land with fewer landlines and more and more cell phones, smart phones and online communications. The effect on the polling world is that no one seems entirely certain as to what mixture of these new- and old-communications media can produce a valid and reliable survey.
Hence we have the polling equivalent of the War of the Roses. Things got so bad recently in one state that one pollster who does good work released a poll so contrary to other surveys that it led to a widespread official denouncement by one of the state's major parties. There was more detail in that public attack on polling than in most press releases about the opposing candidates!