FSG Blog
September 8, 2022

The Futility of Expert Prediction

Patrick Marren
Principal

“The Futility of Expert Prediction” is the first installment in a series of FSG blog entries marking the 20th anniversary of our founding. 

The twenty years since the founding of the Futures Strategy Group in 2002 have seen a series of large-scale global disruptions. Historians will reassure us that these upheavals are certainly not as great as the disruptions of the previous century. We have not seen a world war, or a depression of the depth and devastation of that in the 1930s. 

But the disruptions of this new millennium do appear to have some things in common. First, they caught the world, and its leaders, unprepared. Second, they could not have been predicted to have happened at the time and the precise way they did, at least with any degree of reliability. And third, they all could have been anticipated – and prepared for. 

It should be said that for every “unforeseen” event that occurs, there will always be someone somewhere who will claim to have “predicted” it. The problem is that there is no way to tell whether these folks actually knew what was coming, or were just lucky. Outside of the realm of hard science – which is, it must be admitted, constantly, if slowly, expanding – there is very little evidence to show that prognosticators can offer anything like certainty to slake the insatiable thirst of decision-makers to know what will be. 

The Costly Failures of Expert Prediction

None of the major upheavals of the 21st century could have been predicted.

9/11, like most terrorist atrocities, was, prior to the event, a needle in a haystack composed of needles. Even if the FBI and CIA had somehow overcome their mutual cultural and jurisdictional antipathies, there were so many other plausible-sounding plots for them to monitor that it is doubtful that they could have predicted the precise character and timing of the actual attacks.

The Iraq War of 2003 was predicted to be a disaster for America, before it was launched, by a number of people. For better or worse, however, many of the same people had predicted that the Gulf War of 1991 would also turn out to be a strategic disaster for America, which it manifestly did not. And even earlier, many of these same people had warned that Vietnam would not go well. When prognosticators repeat the same general prediction about every instance of a certain phenomenon, sometimes “rightly,” sometimes “wrongly,” it is difficult to fault decision-makers for dismissing their input. (The same, of course, goes for those who always predict smooth sailing in wars.)

The Global Financial Crisis of 2008, likewise, was predicted by a number of people. Michael Lewis, in his excellent book The Big Short, identified a small group that not only predicted the crash of the mortgage-backed securities market, but also bet on that eventuality. Some of them made billions off of the disaster. But timing the crash was of the essence. Many others, such as Paul Krugman, warned of the possibility of a housing bubble and its bursting, but he began doing that in 2002. By 2005, other economists were mocking him for his “failed prediction.” 

(And it should be noted that all the profits made by those who DID predict the bursting of the bubble, and who happened to time their investments correctly, depended upon there being a large number of people on the wrong side of that bet. If everyone had predicted the bursting of the bubble correctly, no one would have made a cent off of it – and what is the sound of a nonexistent bubble bursting?)

Of course, financial markets are not characterized by hard-science-style causal predictability, because the ultimate object of such prediction will not be physical atoms, but the opinions resident in the brains of millions of people. So it will always be in the nature of phenomena like bubbles to persist long enough to convince the mass of people that they have been suckers to think that they would ever end. What Yuval Noah Harari calls “intersubjective realities” are not subject to simple rules of timely cause and effect. Comeuppance is often delayed. 

But when intersubjective realities – “collective hunches,” as the noted economists Lily Tomlin and Jane Wagner put it – begin to turn, they are also not subject to the sort of time delays that characterize physical phenomena. Fortunes can be lost nearly instantaneously. We might paraphrase Langston Hughes: What happens to a “correct” forecast deferred? Does it dry up like a raisin in the sun? …Or does it explode? (Or both?)

Pandemic Foresight

Then came COVID. Once again, many people had predicted a pandemic. Science journalist Laurie Garrett and the technology titan Bill Gates both had warned of the possibility. Yet again, however, timing came into play. Knowing something is inevitable is not the same thing as preparing for it, and when that inevitable event takes more than a century to strike, politicians will be hard-pressed to persuade the public to pay taxes to prepare for it. 

And what made COVID particularly deadly, in part, was the last of the great unpredictable events of this century: the intense societal division that politicized even science and medicine. “Elites” were deemed untrustworthy; millions of Americans decided they could “do our own research” on topics that experts considered settled science. It is possible that this attitude cost America hundreds of thousands of lives. Even at this writing, with the mass of the country acting as though the COVID-19 pandemic is over, about 400 Americans a day are dying from COVID – a yearly rate of 146,000 dead, far in excess of all but the worst flu years of the past 70 years. Unvaccinated people make up the largest part of this toll; those over the age of 50 are some 14 times more likely to be dying from COVID right now than those over 50 who have received the full modicum of boosters.[1]

Backlash Against “the Experts”

But there is something else besides partisan politics that is behind the newly widespread reluctance to do what experts say. And that is the constant implicit certainty that has been conveyed by experts when predicting the course of the pandemic – and the general arrogance and assuredness of elite experts over the past few decades when predicting all things, whether they be financial markets, terrorism, the outcomes of wars, the effects of free trade agreements, the non-addictiveness of pharmaceuticals, etc., etc.

Prediction of the fundamentally unpredictable, followed by (predictable?) failure to predict, has caused a backlash against expertise among the public, and a generalized disdain for “experts” and “elites” of all kinds – sadly, even when those experts are actually legitimately certain about, for example, the health outcomes of not getting vaccinated. 

Scenario Thinking to Rehab Expert Judgment 

It will take a lot of work to restore this broken trust. Which means that experts should publicly commit now to forsaking prediction of the fundamentally unpredictable, and begin to constantly educate the public as to the true level of uncertainty that is inherent in science (and pretty much everything else). It’s “trial and error,” after all, not “trial and perfection.” And experts should try to anticipate the full range of plausible outcomes, rather than point-predict one brittle and almost certainly inaccurate vision of the future, then walk away from the prediction as if they had not badly misled people. Experts must engage in educated, rigorous imagination and speculation, in that vast sphere of reality that is not amenable to prediction. It’s what FSG scenario planning consultants aim to do in every engagement. 

And then, once they have speculated, if they are serious people, experts need to make serious plans for the full range of plausible futures. They don’t have to execute all these contingent strategies right away. But they should make sure their people are always conscious of that full range of plausible future scenarios, to inculcate what one of our great clients called “a culture of strategic intent.” Ultimately, we who seek to plan for the future should hope to create such a culture – aware of the full level of uncertainty we face, but confident that we have thought through our options in case the unthinkable happens ­– throughout society. 

Imagine a nation that dispenses with specious certainty and illegitimate prediction, and modestly considers every policy undertaken as what FDR in 1932 called “bold, persistent experimentation.” We need to abandon this rigid us/them politics of fear of flip-flopping, and build our flips into our legislation beforehand, to be executed once enough evidence is in. Expert prediction is at the root of many of our political problems: it too often becomes ossified into a macho posturing that cannot be abandoned without suffering serious loss of face. Imagine a politics of confidence and open-mindedness. 

Just imagine.


[1] https://www.cdc.gov/coronavirus/2019-ncov/covid-data/covidview/index.html

Blog Sign-Up

This field is for validation purposes and should be left unchanged.

7 thoughts on “The Futility of Expert Prediction”

  1. Thoughtful, as always. And nice to learn who great clients are.

    I’m a proponent of scenario-based strategy development. I would love to see that fully integrated with strategic planning and system dynamics modeling. Identifying drivers and key dimensions is great, but understanding the relationships among those drivers would (IMHO) greatly enhance the journey to strategic intent.

    Reply
  2. Much obliged for the article and sharing of your FSG reflections on experience given the years of hard work.

    From here in the EU, specifically Copenhagen Business School, there’s a values/norms criterion methods element emerging for research inquiry arising from increased focus on the now fraught concerns over climate crisis, expectations of and for topics ranging from authenticity in employment relations, to diversity and change management, and the employment conditions that are lean-success enabling.

    So I wonder how the shift from predictive fails to such empowered scenario planning might functionally play out? This isn’t exactly scenario planning from a given present, rather scenario planning with deliberate, societally normative intent on the future performative side.

    How might that nuance your scenario work?

    Here is our School’s attempt to help establish educational principles along such lines from a recent, lengthy internal assessment and reflection process. I imagine these would set a stage for, yet also benefit from the scenario notions you so successfully represent and offer – our CBS Nordic Nine:

    https://www.cbs.dk/files/cbs.dk/nordic_nine.pdf

    Best,
    Charlie Tackney, Assoc. Prof.
    Dept. of Management, Society, and Communication
    cta.msc@cbs.dk

    Reply
    • Interesting stuff! Scenarios as we do them normally [no pun intended] leave a normative element out, as the range we examine is varied across dimensions over which we are assumed to have no control (e.g. economic growth weak/robust, level of global integration higher/lower, etc.). Of course some normative intent must enter at some point. Thanks for the info… will read with interest.

      Reply

Leave a Comment