The great fallacy of prediction

The great fallacy of prediction

by Tony Trewavas
article from Monday 12, March, 2018

WHO DOESN’T want to know the future? Mankind has a strong aversion to uncertainty and the future beyond the next few months is the least certain in life. Astrology, palmistry, crystal balls, oracles or even animal entrails filled this need in the past. But predictions were often random, believed when they came to pass and ignored when they didn’t. In this day and age an abundance of pundits in newspapers, magazines, blogs, books, television and social media attempt to assuage our need to know.  What’s the future of the economy, our personal health, energy and food prices or the meanderings of the stock market, progress of love – to name a few? The world is incredibly complex. Unexpected contingencies throw all predictions into a melting pot; the great majority are simply wrong and luck accounts for most of those that weren’t. A few indicate the extent of error.  

In 1929, the economist Irving Fisher famously predicted, nine days before the crash, that stock prices had "reached what looks like a permanently high plateau."  The noted economist Maynard Keynes, thought it would have little effect on the UK economy, which instead fell into serious depression for a decade. In December 2007 months after the credit crunch began, Business Week surveyed 54 economists who predicted the US economy would experience slow but steady growth. The financial and economic meltdown remained unrecognised. In 1984, The Economist magazine asked four Finance Ministers, four London dustmen, four Multinational Company chairmen and four Oxford University economic students to make 10-year forecasts of economic growth, inflation rates, oil prices, exchange rates, etc.  1994 showed how awful the predictions were. The dustmen tied the chairmen for first place.  President Carter in 1977 stated the USA was running out of oil; in 2018 the USA is the largest oil producer world-wide.

In 1968, Paul Ehrlich in “The Population Bomb” stated “the battle to feed all of humanity is over. In the 1970s and 80s the world will undergo famines, hundreds of millions of people are going to starve to death”. “Nothing can prevent a substantial increase in the world’s death rate.”  The UN recorded continuous declines in death rates throughout the last 50 years and the mass starvation didn’t happen.

War is the primary cause of any present starvation.  In 1972 the Club of Rome in its “Limits to Growth” made various gloomy predictions concerning industrial development, population growth, malnutrition, depletion of non-renewable resources and a supposedly deteriorating environment. They predicted a massive population crash in a starving, polluted world. 46 years later all have proved seriously and unsurprisingly wrong.  Their mistake; using an ecological concept called “carrying capacity” applicable to wild animal and plant populations but not to human reproduction. In 1987 the Rand Corporation Think Tank said that the answer to world food problems was… “aquaculture”.

These above predictions have been made by serious investigators. Brills Content a now defunct magazine had fun comparing the predictions amongst the lesser ranks of US pundits with Chippy a chimpanzee who made his by choosing amongst flash cards. Chippy matched or beat the best in the business.

The excuses for failure are usually “I was nearly right” or “it’s just around the corner”.  Without specifying in what year any corner actually is, such claims are nonsense.  Another alternative is to try to avoid failure is using probabilities as with weather forecasts. At 20 per cent chance of rain but 80 per cent sunshine; whatever you experience can be claimed a valid prediction.  

Philip Tetlock, a University of California psychologist, gathered together 284 experts, political scientists, economists and journalists and by throwing them many questions accumulated some 27,000 predictions.  The average was only a whisker better than a dart-throwing chimp. But Tetlock noticed vast differences in accuracy. Those who did badly were uncomfortable with the complexity and contingency in human affairs; they were over confident in predictions, simple and dramatic in response. “Inevitable” and “impossible” were commonly used phrases.  The more popular they were in the media, the less accurate any prediction made. Those more accurate in prediction were much more thoughtful, drew information from many sources to synthesise an answer. They were aware of how interactive contingencies can throw any prediction and were acceptive of failure, often not believing prediction is possible.

How does science fare in prediction? By recognising how future contingency, technological invention and developing knowledge cannot be predicted, few will offer anything over decades. No amount of research can eliminate future uncertainty. Climate change is full of future temperature predictions but good climate scientists will always say theirs is a projection; a prediction only if nothing else changes.  But of course it does change. The less scrupulous fail to add that qualification.

What looks like tomorrow’s problem today is rarely the real problem when tomorrow comes around.

ThinkScotland exists thanks to readers' support - please donate in any currency and often

Follow us on Facebook and Twitter & like and share this article