I was listening to the Today programme last week, and was struck by an interview between the Business presenter, Dominic O’Connell, and BP’s Chief Economist, Spencer Dale. (It’s around 24 minutes in if you want to listen for yourself.)
The economist was talking about BP’s research into demand for oil. Currently we use about 100 million barrels a day, and the prediction was that this would peak at about 110m barrels, then decline to 85m by around 2040. Interestingly, the expected decline was less due to the rise of electric vehicles and more due to expected efficiency improvements from our technology in cars and elsewhere.
Dominic O’Connell asked “What sort of confidence do you have around these predictions?” and Spencer Dale gave this reply, which I found both arresting and enlightening:
“No confidence whatsoever. Any prediciton of the future, any single point prediction of the future, will be wrong. The whole point […] is not to predict the future, it’s to better understand the uncertainty you face. You do that by thinking about a whole range of different scenarios, and then you get a sense of that uncertainty. Don’t try and predict the future. Think about the uncertainty you face and then set your business in a way that is robust to that range of uncertainty”
This is a great insight into managing risk and uncertainty. I’m personally pleased he used the phrase “point prediction”, as I’ve previously written about “point risks”. These are predictions of the future that are so specific it’s totally unreasonable to expect them to happen exactly as stated. Likewise, trying to deal with point risks (or point predictions) won’t be productive. Much more useful is to understand the broader landscape of uncertainty and change the way you operate to ensure you can handle that kind of uncertainty in the natural course of events.
It’s great to hear someone talk about uncertainty in a very practical way on a very public platform. I hope it was useful to a few others listening, too.
2 thoughts on “Predicting an uncertain future”
I suppose you meant M barrels (Mega), and not m barrels (milli).
They said “million barrels”. I’ll update the text to avoid confusion.
Comments are closed.