All along the untrodden paths of the future I see the footprints of an unseen hand. – Sir Boyle Roche
We who inhabit the Land of Scenariotopia (as a former colleague termed our little realm) think that the best way to predict the future is not to. That is, you should abandon the search for certainty and explore multiple scenarios of what might be. Many people don’t do this, however. We find that lots of them subscribe, usually semi-consciously, to some or all of the following fallacies:
1. “The future is ever-changing.” No it isn’t. The future cannot change, because by definition it has not happened yet — and it never will. It’s always out there, just out of reach. It’s the past that is actually constantly changing, as we move away from it, the angle of our view is impeded by more recent events, and revisionist historians distort it in order to score points and sell books. (That’s why we stick to working on the future. The past and present are too confusing for us.)
2. “The future is unknowable.” Actually, it’s far worse than that. Even the fact that the future is unknowable is unknowable. It may actually be that someone out there does know everything that is going to happen; but it is impossible in advance to separate any such true seer from the armies of the charlatans or the merely lucky. Certainly not in time to make high-stakes decisions about the future.
3. “Well, the future is uncertain, but at least the past is knowable.” As adverted to above, the past is not knowable either. To “know” the past in any meaningful sense would require that all the workings of cause and effect would be uncontroversial. This is clearly untrue. We still argue about the causes of the Civil War, World War I, the Depression, and why Tiger Woods can’t seem to win anymore. The infinity of phenomena in the world, and the ridiculous complexity of interactions and linked causalities between them, make any such certainty impossible.
4. “If we could just get enough good data, we could predict the future perfectly.” There are no data about the future. Repeat: There are no data about the future. If you have data, it’s about the past. In order for data to predict the future, your hypotheses about the causal relationships between certain quantities, derived from perceived past correlations, would have to hold true in the future. Sometimes they do; sometimes they do not. When your predictions depend upon the vast majority of such correlations persisting, the likelihood of a useful prediction begins to approach zero.
5. “The future is just a continuation of the past, and is qualitatively similar.” This seems to make sense, especially when we start drawing Cartesian graphs where the future is just the portion of the graph to the right of the present. This treatment may work wonders in physics, but it can be terribly dangerous for decision-makers. The past is a buzzing blooming confusion of data and sense impressions available for analysis. Our minds are constructed to look at a squiggly line, no matter how random, and find some sort of story or pattern in the squiggles — a story we expect to continue into the future. But the future is a different realm from the past. We will never come to a final judgment on the past; but at least we have data there and can begin to advance hypotheses and use the data to argue about them. But the future, again, has no data. It is a trackless and featureless waste, like Antactica in a blizzard. There is no particular reason to expect whatever “story” we have inferred from the squiggly line of the past to continue into the future. (Think of the clear patterns evinced by the stock market and housing prices up to 2007. A strong trend line can and does often simply drop off the table — or zoom into the stratosphere.) Thinking that the future is in any way qualitatively like the past is like a mariner coming to Antarctica and expecting his nautical charts and equipment to be of use in crossing Antarctica in a blizzard. Which would you rather know about Antarctica as you step on shore: that there are no coral reefs ahead, or that the entire continent may be riven by crevasses, so you should probably rope up?
6. “If we just do what the experts say, then we’ll be fine.” Expertise IS necessary, but it is insufficient. Expertise can be defined as “knowledge of the if>then statements that seem to have held true in the past.” It can be very useful in generating hypotheses about how the future MAY evolve; but outside the hard natural sciences, it cannot be depended upon to offer an unerring guide to what WILL be. It should be used as a guide to plausibility, but it should never restrict creative and open-ended thought about what COULD be. The best experts are well aware of the fact that the least expected and most consequential changes, especially in competitive contexts, will come not from a 10% error in some single coefficient within a mathematical model, but from a change in the basic form of the equation. IBM did a very good job predicting the growth in total world computing power back in the 1980s. But they failed utterly to predict that some 98+% of the computers sold in the 1990s would be PCs.
7. “Change is inevitable in the future.” SOME change, yes. But not everything changes, and even the stuff that does change often changes far more slowly than we think it will. Many people predicted the bust in dotcom stocks, as well as the bursting of the housing bubble. But they often predicted them too early, which can be as bad as not predicting them at all, because they are then seen as Chicken Littles or the Boy Who Cried Wolf. Paul Krugman did predict the housing crash. But he predicted it in 2003 or so, and announced he was arranging his personal affairs to shield himself from it. He missed out on four years of price increases, and they must have seemed to be four very long years to the good professor. Other things that have been predicted to change, but have not (or changed only after a much longer time than expected):
- US dominance in world affairs has been predicted to disappear since at least the end of World War II. Yet the US still remains at the top of the heap in almost every important respect.
- “Dow 36,000” was published in 1999, just before the big dotcom collapse. Many others have predicted that the Dow would zoom far beyond 10,000. Yet a glance at the last two Dow decimal barriers – 100 and 1,000 – would suggest that it was going to take a while. The Dow Jones Industrial Average first broke the 100 barrier in 1906; it did not leave it behind for good until 1942. Similarly, the Dow first broke 1,000 in 1966; it did not leave 1,000 behind for good until the end of 1982. The Dow first broke 10,000 in 1999. It’s uncertain that it has permanently left behind that particular barrier, even after some 13 years.
- The Chicago Cubs continue not to win a World Series, in a string that will reach 104 years if they fail this season.
Though we ordinarily frown upon assuming anything about the future here at FSG, this reporter, at least, thinks that that last one remains a sure bet for at least a couple of more years.