They said Donald Trump would lose; he won.
Experts, psephologists, with fancy analytical tools included, disappointed again. Given the frequency with which they fail, perhaps it is time we treated them on a par with astrologers and crystal ball gazers. Or we should treat them as simple entertainers who play with numbers and spice it up with some knowledge of history and good articulation of the commonsensical.
For those intrigued by the ‘expert’ phenomenon – the way they fail spectacularly and the way we keep buying their confident intellectual talk every time – one would recommend the book Future Babble by Dan Gardner. The book deconstructs the phenomenon with ruthless logic backed by research. Here’s what the jacket of the book says:
In 2007, experts said it would be smooth sailing in 2008; then came the global financial hurricane. In 2008, as oil surged above $140 a barrel, experts said it would soon hit $200.; a few months later, it plunged to $30. This is how it always goes. In the 1960s, experts said the Soviet economy would be bigger than the American economy by 1997; in 1997, the Soviet Union did not exist. In 1911, experts predicted declared there would be no more major wars; we all know how it all turned out.
Let’s face it: Experts are about as accurate as dirt-throwing monkeys. And yet every day, we ask them to predict the future – everything from the weather to the likelihood of a catastrophic terrorist attack...
While on monkeys, a report in The Times of India says, Geda, a Chinese monkey with a reputation for predicting things right, got it right on Trump too. Chanakya, the fish doing a similar job in Chennai, also foresaw the victory of the mercurial Republican. All the while, experts, both Indian and American, had been predicting a Hillary landslide. They were busy reducing Trump to a bizarre mix of racism, nativism, misogyny and whatnot, making him look inferior to his competitor, while he was actually connecting with his voters in a more meaningful way than Hillary.
It is possible they were talking about the Trump phenomenon without understanding it at all. They were too busy studying statistics, splicing and dicing data on demography and making excel sheet projections while ignoring the real factor that counts in all elections: the man who votes. Using numbers to predict human behaviour is fraught with risks. It becomes riskier when societies are in a churn, thus unstable. That is one big reason why psephologists and experts are falling short everywhere. We have seen it in India too.
Again, to be blunt, the expert may have more information at his command than the common man but the universe of his knowledge is limited. A lot of his knowledge draws from history – pattern of voter behaviour over a length of time; the distinct voting track record of a community or in a constituency; and the past of the party and the candidate. A bit of it comes from quick analysis of immediate developments – for example, a controversial statement made, a senior leader switching loyalty, a case of communal violence etc. Experts tend to mix both to have a presentable argument to back their prediction.
But as evidence over the last some elections suggest the knowledge fails to capture the reality on the ground or the complexity of voter behaviour. In the US many of the states that have been voting Democratic shifted to Republicans this time. Women, who were expected to be disgusted at Trump’s rather vulgar record with them, didn’t vote overwhelmingly against him. Few experts caught the shift in the first case and all of them wholly misread the so-called anger against his misogyny.
The truth of the matter is media experts have been pulling a con job on their audience. A truly knowledgeable person would be defensive about making a prediction. Like Gardner says in his book: Expertise means more knowledge and more knowledge produces more detail and complication. More detail and complication makes it harder to come to a clear and confident answer. At least it should make it harder...A conman’s greatest strength is confidence. We see a lot of it in experts in the media.
The strategy is simple. Don’t get troubled with the complexities that massive information brings. Like researcher Philip Tetlock, quoted in Gardner’s book, would say: ...Reduce the problem to some core theoretical theme and use the theme over and over, like a template, to stamp out predictions...”
It does not matter if the prediction goes wrong, not once but again and again. We are psychologically so conditioned to seeking predictions that we go back to them always. If we are ready to be conned, there will be people to con us. But that’s another story for another day.