How much credibility should you attach to that political or economic expert on a television interview?
Philip Tetlock wanted to put this to the test. Between 1987 and 2003, he asked 284 such expert commentators for advice on political and economical trends. He asked them to make 82,361 predictive judgements about events in the world by answering questions such as ‘Would there be a nonviolent end to apartheid in South Africa?’ or ‘Would Gorbachev be ousted in a coup?’ and so on.
His findings were interesting. Tetlock asks us to picture a bunch of chimpanzee throwing darts at a board full of predictions. These dart throwing chimps, in Tetlock’s words, would be more accurate than the experts. The experts had done worse than pure chance!
These professionals brought deep domain expertise on topics that they commented upon – with masters and PhD degrees. In fact, knowing about a domain made their predictive power worse. Yet, this isn’t their fault – the domains of politics and economics are just as unpredictable as a spinning roulette wheel or a flipped coin. Even if you ‘study’ coin tosses for 10 years, it makes no difference to your ability to predict the next toss.
As early as 2005, the ability of experts to forecast political and economic events was shown to be non-existent. After we know this, one might expect them to lose their jobs enmasse. Yet, such experts are only more prevalent today than they were 20 years back. Why is this so?
When we tune into news on significant global events, we like to think of ourselves as being interested in the truth. However, what we actually seek isn’t the truth, but somebody who replaces uncertainity with an interesting narrative, even if that narrative is false.
We are interested more in a story that seems certain than the uncertain truth. We watch the news partly to satisfy this craving.
Observations 