Years of research in psychology have taught us how important it is for well-being to live in a world in which we can predict and control the things that matter. As Martin Seligman, among others, has shown, lack of predictability produces anxiety, and lack of control produces helplessness. Being able to control events in our lives means we can predict them (if you do X, Y will happen; if you flip the switch, the light will go on). But even when events (e.g.., the weather) are uncontrollable, being able to predict them enables us to make plans or to mitigate adverse effects.
Of course, the world we live in is uncertain, so let me clarify what predictability means. Imagine a container with 100 balls, half red and half black. Blindfolded, you reach into the container to retrieve a ball. If you pick a red one, you win $10. If you pick a black one, you win nothing. What are the odds of winning? It’s a fifty-fifty proposition, like a coin flip. Now consider a second container, also with 100 balls, each either red or black. When you reach into this container, what are the odds you’ll pick a red ball? You don’t know. If all the balls are black, your chances of winning are zero, if all the balls are red, your chances of winning are 100 percent, and since there could be any number of red balls in the container, from 0 to 100, your odds of winning are somewhere between 0 and 100 percent.
Daniel Ellsberg famously presented this contrast in an article many years ago. Either container confronts you with uncertainty. The first container provides “probabilistic” uncertainty (that is, you can specify precisely the odds that you will win). The second confronts you with what Ellsberg called “ambiguity,” but that I prefer to call “radical uncertainty.” Economist Frank Knight made a similar point years earlier, distinguishing probabilistic uncertainty from what he called “true uncertainty.”
People typically much prefer the first container to the second. Few of us relish uncertainty, but we can tolerate it if we at least know the odds. And importantly, a world of probabilistic uncertainty is predictable in a way that a world of radical uncertainty is not.
A world of probabilistic uncertainty is predictable in a way that a world of radical uncertainty is not.
What kind of uncertainty do we actually face in everyday life? If we’re flipping coins, picking cards out of a deck, or playing table games at a casino, we can specify the odds of various bets paying off quite precisely. These games are designed deliberately to make that possible. We don’t have the same precision when we are playing poker, which requires both knowing about the probability of various hands and about the strategies being pursued by our opponents. We also don’t have the same precision when we’re betting on sports, but we’re not completely in the dark. If the home team has won two-thirds of the time in NBA games in the recent past, we have some justification in believing that the odds of the home team winning tonight will be about two to one in favor. And if your doctor tells you that 80 percent of patients with your kind of cancer beat it, there is good justification for your being optimistic. Your odds look good.
What about the case of financial investment? Which kind of “container” is that? On the one hand, a lot of people make a very good living attaching probabilities to various future market fluctuations (like the first container). On the other hand, lots of people think of financial markets as what are called “random walks” (like the second container). On the third hand, historically the annual rate of return on stock investment has been quite good. I don’t know what kind of container the stock market really is. But I am pretty confident about what kind of container it has become as a consequence of President Trump’s on-again, off-again tariff proposals.
The famous Hollywood producer Jack Warner is reported to have said, when asked how his studio judged which of the possible movies they were considering would be successful, “Nobody knows anything.” This seems to capture well the current state of our (and the world’s) economy. It has become close to impossible to predict the financial future at all. Will Trump change his mind … again? Will he negotiate deals? With which nations? Will he raise tariffs even more than he has already threatened? Will countries in addition to China retaliate? It is hard to imagine any enterprise building a new factory in the United States or developing a whole new supply chain that avoids China at this moment of such radical uncertainty. Even investment professionals, who almost always encourage their clients to add more risk to their portfolios, are telling people to put their money in CDs and just wait and see.
It is one thing to take risks and make decisions in a probabilistically uncertain world; it is quite another to make such decisions in a radically uncertain world.
It has been suggested that risk analysts underappreciate that even though extremes of the distribution of possible future financial conditions are rare, when they occur they can have massive effects. I don’t know whether Trump’s tariffs will put us in the tail of a distribution of possible market outcomes or whether, as I suspect, the tariffs will make it impossible even to calculate what the tails of the distribution will be. I find it laughably precise when a financial expert says something like “Trump’s tariffs have changed the probability of recession from 20 percent to 75 percent.” But I have no doubt that the chances of recession have increased.
I am certain that the ever-changing threats of tariffs have turned an already uncertain matter (our collective financial future) into a much more uncertain matter. As others have noted, this increase in uncertainly will have negative effects on our financial future in and of itself, whatever other effects it may have. It will inhibit expansion of existing ventures, starting of new ventures, buying of new cars and appliances, going on vacations, and the like. It is one thing to take risks and make decisions in a probabilistically uncertain world; it is quite another to make such decisions in a radically uncertain world.
In their book, Radical Uncertainty, two distinguished British economists John Kay and Mervyn King (a former governor of the Bank of England) made the argument that many, perhaps most, decisions we face in life involve uncertainty to which we can’t assign probability. This, they suggest, is true even in fairly circumscribed financial situations, in which risk (probability) assessment is the coin of the realm. One of their chapter titles is “Radical Uncertainty Is Everywhere,” which conveys a sense of what they think is the magnitude of the problem.
Nonetheless, people do assign probabilities to outcomes, even radically uncertain ones. Do they realize that such probability assignment is virtually fictional? If so, why do they (we) do it? Perhaps there is some comfort and assurance to be derived by pretending that the world is more predictable than it really is. But even if, two months ago, it was an exaggeration to say that that radical uncertainty is “everywhere,” it is not an exaggeration now.
And so, as we try to reckon the costs (and benefits?) of our current tariff roller coaster, we should be sure to include in those costs not only the effects of tariffs on the value of our 401(k) and on the weekly tab at the supermarket but also the costs of creating a world in which predictability and controllability of important events is beyond our grasp, leaving us anxious and helpless as a result. The financial costs may pale in comparison to these psychological costs.
Disclosure: Barry Schwartz is a member of Behavioral Scientist’s advisory board.
