The Meaning of "Uncertainty" in Economics


We all know what uncertainty means in everyday speech. In some ways the use of the word in economics is not that different, but there are two kinds of uncertainty in economics that should be differentiated.

The Famous Rumsfeld Quote

At a press briefing in 2002, then Secretary of Defense Donald Rumsfeld offered an opinion that was much discussed. He distinguished two kinds of unknowns: the unknowns we know we don't know about and the unknowns we don't we don't know about. Rumsfeld was afterward mocked for this apparently eccentric observation, but in fact the distinction had been made in intelligence circles for many years.

The difference between "known unknowns" and "unknown unknowns" is also made in economics with respect to "uncertainty." As with unknowns, it turns out there's more than one kind. 

Knightian Uncertainty

University of Chicago economist Frank Knight wrote about the difference between one kind of uncertainty and another in his stock-market-oriented economics text Risk, Uncertainty and Profit. 

One kind of uncertainty, he wrote, has known parameters. If, for example, you put in a buy order on a particular stock at [the current price - X], you don't know that the stock will fall far enough for the order to execute. The outcome, at least in everyday speech, is "uncertain." You do know, however, that if it does execute it will be a your specified price. This kind of uncertainty has limiting parameters. To use Rumsfeld's remark, you don't know what will happen, but you do know that it will be one of two things: the order will either expire or it will execute.

On September 11, 2001, two hijacked airplanes struck the World Trade Center, destroying both buildings and killing thousands. In the aftermath, the stocks of both United and American Airlines plummeted in value. Until that morning, no one had any idea that this was about to happen or that it was even a possibility. The risk was essentially unquantifiable and until after the event there was no practical way of stating the parameters of its occurrence -- this kind of uncertainty is also unquantifiable. 

This second kind of uncertainty, an uncertainty without delimiting parameters, has come to be known as "Knightian uncertainty," and is commonly distinguished in economics from quantifiable certainty, which, as Knight noted, is more accurately termed "risk." 

Uncertainty and Sentiment

9/11 focused everyone's attention, on uncertainty among other things. The general drift of many respected books on the subject following the disaster is that our feelings of certainty are largely illusory -- we only think certain events won't happen because to date they haven't. This view, however, has no plausible rationale -- it is simply a feeling. 

Perhaps the most influential of these books on uncertainty is Nassim Nicholas Taleb's "Black Swan: The Impact of the Highly Improbable." His thesis, which he proposes with many examples, is that there is an innate and largely unconscious human tendency to draw a limiting circle around a given reality, and to think of whatever is in that circle as all there is and either to think of everything outside the circle as an impossibility or, more often, not to think about it at all.  

Because in Europe, all swans were white, no one had ever considered the possibility of a black swan. Yet, they're not that unusual in Australia. The world, Taleb, writes, is filled with "black swan events," many of them potentially catastrophic, like 9/11. Because we haven't experienced them, we may believe they cannot exist. In consequence, Taleb further argues, we're prevented from taking preventive measures to avoid them that might have occurred to us if we'd considered them possible -- or considered them at all. 

We're back in the briefing room with Rumsfeld, facing two kinds of uncertainty -- the kinds of uncertainty we know are uncertain and the other kind, the black swans, we don't even know we don't know about.