Happy New Year! Well, this time of year traditionally brings out the “January 1 quarterbacks,” poking fun at the wrong predictions for the year just past, awarding dunce caps, particularly (deliciously) to famous people.

This punditry is widely read, and sometimes published in respectable places. Some of it is just year-end fun, and nothing wrong with that. But there is also a failed-forecast “nyah-nyah” that is corrosive to the foresight field in general, which demands answers. So at the risk of giving the 20/20 hindsight artists undue oxygen of attention, here are a few thoughts:
Consider Foreign Policy’s “10 Worst Predictions for 2008.” (Dec, 2008). Highlights include:
“If [Hillary Clinton] gets a race against John Edwards and Barack Obama, she’s going to be the nominee. Gore is the only threat to her, then. … Barack Obama is not going to beat Hillary Clinton in a single Democratic primary. I’ll predict that right now.” —William Kristol, Fox News Sunday, Dec. 17, 2006
“Should I be worried about Bear Stearns in terms of liquidity and get my money out of there?’ No! No! No! Bear Stearns is fine! Do not take your money out … —Jim Cramer, responding to a viewer’s e-mail on CNBC’s Mad Money, March 11, 2008 [Bear Stearns was sold to J.P. Morgan Chase at about a 90% discount to it market capitalization at the time of the forecast]
“The possibility of $150-$200 per barrel seems increasingly likely over the next six-24 months.” —Arjun Murti, Goldman Sachs oil analyst, in a May 5, 2008, report [Oil was then around $130 a barrel. By late December it was below $40.]
Or this one from Business Week’s list of 10 (December 24, 2008)
“Existing-Home Sales to Trend Up in 2008″ —Headline of a National Association of Realtors press release, Dec. 9, 2007 [On Dec. 23, 2008, the group said November sales were running at an annual rate of 4.5 million—down 11% from a year earlier—in the worst housing slump since the Depression.]
The Future Savvy question is: how should we think about predictions like this? And how should we think about failed-forecast spotting?
1. Failed-forecast spotting is not remotely “scientific”
This should be obvious, but somehow never is. Purposefully extracting the failed forecasts from the total set of forecasts says nothing about the quality of the set in general. Many did predict Obama; did predict the downturn, etc.
2. Failed-forecast spotting raises a healthy skepticism, but runs to nihilism
Despite not passing any credible test of knowledge, at least failed-forecast spotting stokes apprehension about forecasts and the wisdom of experts. At base this is healthy. Prediction is hard, and it is mostly done poorly. And experts often transgress the boundaries of their expertise. (Typically, in this instance, they know a lot about their field, but often don’t know more than the next Joe about the future of their field, often because their expertise is wedded to existing practices and assumptions.)
Prediction skepticism is fine. What happens, however, is that tempts a “nobody can predict anything” nihilism. This is its own failing because many predictions are in fact excellent, producing good foresight, which is a key strategic and competitive advantage.
3. Often the short term future is harder to see.
This is the trickiest insight of the lot. As everyone knows, it’s impossible to accurately predict the future (which is not the same as usefully predict the future, see arguments in other posts). The world is just too complex, too chaotic. But there’s a wrinkle. It should be that the further we look into the future the harder it is to see. The world will change more – there is more time for unpredictable things to happen. The short-term future (one year, say) is closer to us, it should be more like today and we should be able to anticipate it better.
In fact, short-term foresight is the most impossible task: a casino game. In the longer term (10-20 years, say) strong trends can be relied on to have had their impact. For example, the move away from fossil fuels, or effective nanotechnology engineering, or simple domestic robotics, can be reliably forecast. But while the sweep of these and other similar evolutions are reliable over time, the short-term picture will suffer lags or reversals that follow no pattern at all. (It’s no accident that is this is just like the stock market. In the long term the market will go up, in the short term it can go anywhere.) Also short-term predictive failure is compounded by the fact that the standard to which it is held is higher – we expect specifics: dates, places, numbers, players, winners – that are not demanded of a long-term view. In other words, near-term predictions are all about “point forecasts,” and there’s nothing more impossible than a point forecast unless you believe in tea leaves and crystal balls.
The take away: short-term point forecasts really are a mugs game and the skeptics are right. Medium-long forecasts, when well done, are worthy of our strategic and competitive attention.
Tweet This Post