The Cayman Islands traditionally has been a political polling-free zone, but the serpent may be about to enter the garden.
Polling, especially of the political stripe, bears all of the hallmarks of pseudo-science. If polls came with warning labels, almost no one would commission them and, certainly, almost no one would believe them.
Think for a moment of the Brexit vote. On the day before the election, nearly every poll indicated U.K. voters would opt to remain in the European Union. And nearly every poll was wrong.
Voters in Britain, and vote-watchers nearly everywhere, woke up “the morning after” to the inexplicable, if not the unthinkable: Voters had defied both the polls and the pundits, who, in turn, were heavily influenced by the polls (round and round we go).
Or, consider this: On the eve of the Democratic primary in California (Hillary Clinton vs. Bernie Sanders), nearly every poll declared the race “too close to call.” That waffle was then amplified by nearly every mass media outlet, from newspapers to televisions networks.
Then came voting day. Ms. Clinton won by nearly 13 percentage points, a political landslide (to mix metaphors) of tsunami proportions.
Pollsters famously “missed” the strength of Conservatives in Britain’s 2014 elections, the popularity of Prime Minister Benjamin Netanyahu in Israel’s recent elections, and even the Republican Party sweep of the House and Senate in 2014.
How does this happen?
Listen to New York Times columnist Cliff Zukin, a Rutgers University professor and past president of the American Association for Public Opinion Research: “Election polling is in near crisis … So what’s the solution? There isn’t one. Our old paradigm has broken down, and we haven’t figured out how to replace it.”
Readers should not be surprised. Almost all polls are almost always wrong. By definition, they have to be. Consider the tale of the two watches:
If an individual owns one watch, he always knows the time. If he owns two watches, he never knows the time.
It is the same with polls.
If Poll “A” has a candidate leading his opponent by, say, 5 percentage points, and Poll “B” is reporting a 10 percentage point lead, by definition, one of the polls MUST be wrong. If there are 10 polls, all asking the same question but reporting different results, all (but one) MUST be wrong.
None of this would greatly concern us if the news media – the business we’re in – were not implicated. Polling has become the false god to which nearly all news media now genuflect.
Every major news network now conducts its own poll (ABC/Washington Post, CBS/New York Times, NBC/Wall Street Journal, the BBC Poll Tracker, Fox/News Dynamics), and their newscasters pass on the results as if they were accurate and legitimate news: “Hillary surging by 10”; “Trump in free fall”; “Swing states still undecided.”
Frankly, this kind of coverage is more of a commentary on the state of the media than it is on the inexactitudes of polling. Polling has become such a crutch for the media that it’s not certain it could “walk” – or talk – without it.
Ironically, despite all of our concerns about the limitations and misapplications of political polling, it may actually have application in Cayman as we move toward the 2017 elections.
To meet the provisions of single-member constituencies, our electoral map has been subdivided into multiple “mini-districts” with fewer than 1,000 voters in each. Such small numbers make sampling for polling purposes a much more precise process – which, if employed, would likely yield far more accurate forecasts.