Has Our Obsession with Quantification Caused Our Downfall?
At the height of the credit boom, one of the pieces of mythologogy doing the rounds was that you could quantify and measure risk so precisely that it pretty much ceased to be . . . well, risky. Banks and hedge funds employed all kinds of pointy-headed types, from mathematicians to theoretical physicists, to produce ‘risk-management’ algorithms that they claimed made it possible to make money all of the time.
Like every other kind of ‘sure thing’, from infallible gambling systems to simple ways to make millions with nothing more than a home computer and a whole lot of faith, this one too proved to be little more than the latest version of snake oil.
The interesting point, however, is not so much that another infallible system failed, as why people seem to be so prone to believing in the myth of quantification as a substitute for common sense?
Misunderstanding numbers
Part of the problem, I believe, comes from a basic misunderstanding of numbers. Numbers represent a precise language in which it is perfectly possible to express the most total nonsense.
Words sometimes have quite imprecise meanings, but a number is a number. Two never means three, or ‘nearly two’ or ‘somewhere between one and five’. You can add numbers together a zillion times and always get the same answer. They are so reassuringly precise on the surface that they lull us into believing that whatever comes in numbers is as precise as they are.
That, of course, is not true. While I can set the odds on something to the fifth—or fiftieth—decimal place, and verify my computation as many times as I like, all that super-precise number is telling me is the outcome of a process that might well be full of unsupported assumptions, logical holes and guesswork.
The numerical precision of the outcome—often enhanced by the fact that it was produced by an ‘infallible’ computer (which is not infallible either and whose precision is merely mechanical)—seems to allow people to forget that what it represents is no more that a statement of someone’s thinking: a human product as likely to be full of holes and mistakes as any other.
Human nature, numbers and risk
Human beings are both uncomfortable and incompetent when it comes to dealing with risks—even more so when that risk is mixed up with the impenetrable jargon of possibilities. Yet risk is what leadership is all about. If what to do is clear and there is little or no uncertainty about the outcome, who needs a leader?
Strategy, by definition, deals with trying to estimate what to do for the best in situations of extreme uncertainty and ambiguity. Numbers don’t work well to express ambiguity. Their calculation is too unreliable when there are multiple sources of ambiguity and no way of deciding between them. Put simply, in ambiguous and unquantifiable (precisely!) situations, numbers are both misleading and often plain impossible to produce with any accuracy.
Still, the big bosses want numbers, so the little workers dutifully churn them out, even if they know they’re nonsense from the start. By they time they reach the executive suite, they’ve been sanctified by so many committees and middle managers that no one will question them.
Leadership should be based on judgment, not computation
It is exactly in such ambiguous situations that too many executives put their greatest trust in numbers—even though that’s when numbers are at their least accurate or useful.
I suspect that it’s the precision of numbers that causes the problem. They seem so reassuring when everything else is vague and impossible to get a hold of. The same long-term, uncertain aspects of strategic decisions that make us so insecure increase the surface attractiveness of quantified data. Indeed, the more complex they appear to be—and the less we understand them—the more ‘scientific’ we’re tempted to assume they are.
People who feel afraid that they may not be doing the right thing, or making the correct choices, want to see some numbers to lull themselves into the belief that they have a solid basis on which to decide. That’s make-believe, of course; any half-competent statistician can make those numbers say whatever he or she wants them to say. But it’s such wonderfully reassuring make-believe.
Omens, auguries and statistics
The ancient Romans used to sacrifice animals and scan their livers for signs to help them decide on the right course of action for future. We laugh at such ‘unscientific’ ideas, but are we so very different?
They had some very precise guidelines for what did or did not count as an omen. Those who made these auguries did so according to careful observations and a clear set of ideas about what each spot or blemish meant. The emperor and his courtiers weren’t expected to understand such esoteric ideas. That’s what they employed augurs to do for them. The emperor’s job was to listen to what the augurs said, then decide, basing his decision on whatever they told him.
How is that so different from the CEO who listens to some manager droning through a PowerPoint presentation that lists the precise risks and benefits of a plan. That CEO is just as unlikely to have anything save the sketchiest notion of what on earth lies behind all those calculations; let alone be able to decide whether they are based on anything but the craziest estimates of future conditions.
I began my working life as a very junior person producing statistical projections for my betters. Did I always know what I was doing? Hell, no. Much of what I did was based on assumptions those same people gave me to work with. In essence, I was turning their guesswork and gut-feel into numbers. At the end of the process, what came out was no better than what went in (sometimes worse, if I miscalculated). However, since it was now safely expressed in ‘scientific’, quantified ways, it commanded a level of respect it never would have had before.
Let’s stop kidding ourselves. The future cannot be computed. Risks cannot be fully quantified. It’s the job of leaders to stick to reality and acknowledge they cannot know in advance what will happen—then use their best judgment anyway. Guesses and assumptions masquerading as scientific calculations are still guesses. They’ve only changed their clothing.
A risk is always a risk. A big one is riskier than a small one. Acknowledge that and you’re at least forewarned that your strategy may well go wrong. Pretend you’ve found a way to make it a near certainty and you’ll probably bet the farm on it—then have to run to the tax-payer to bail you out.
Technorati Tags: strategy, strategic thinking, quantifying risks, executive decisions, leadership thinking, managing risk, planning the future