Professional Forecasters Have Been Getting It Wrong Since the Beginning
Three economists started competing forecasting businesses in the early 20th century. They all got the Great Depression wrong.
Author’s note: I regularly find myself assailed by questions about the future from business leaders, colleagues, and friends. They ask about companies, elections, technology, their own lives. Their questions imply that the future is knowable, hidden like the Wizard of Oz behind a curtain that only the fearless or gifted can pull aside. This strikes me as magical thinking, not least because research shows that the time horizon for accurate forecasting is shrinking: 400 days if you’re rigorous, closer to 150 days if you’re not.
This is a huge change whose impact few individuals or organizations have fully absorbed or understood. Yet even as our foresight shrinks, our addiction to prediction has grown.
A craving for certainty extracts a high cost: constricting our understanding, our imagination, and our freedom. In both our private and our working lives, how much more might we be able to do, to see and to be if we kicked the habit?
I wrote UNCHARTED to understand how forecasting fails us and to explore better ways of living, working and thinking. Since the beginning of human life, we’ve been tormented by uncertainty. But when and how had we fallen prey to the illusion that we could — or should — eliminate it?
Who knew what was in the air? Enjoying the waters just beyond Narragansett Pier, Rhode Island, the economist Irving Fisher kept swimming. A happy marriage, two daughters, and a full professorship at Yale had the future looking as dazzling as the summer ocean. But looking back to the shore, he was surprised how far the current had carried him. It took all his energy to regain the beach. He arrived at last exhausted, unnerved by the speed with which his glorious future had turned precarious.
For the rest of his life, Fisher would wonder whether the difficulty of that swim in the summer of 1898 had been an early warning sign. By the autumn, Fisher scarcely recognized himself. Everything tired him and every afternoon he ran a fever.
The eventual diagnosis: tuberculosis. At the time, TB was the single greatest cause of death in the Western world. Fisher knew the danger he faced: his father had died from the disease. Now age 31, what kind of future did he face?
By 1898, educated people knew that tuberculosis was an airborne pathogen, but there was no vaccine and no certain cure. Nor was there any reliable prognosis: the disease could lie latent for years, even a lifetime, or you could be dead in a matter of weeks. So had Fisher received a life sentence — or just experienced some mild discomfort that would never return? As painful as the disease was the doubt.
Although the TB bacillus had been isolated in 1882, no cure was known until the discovery of streptomycin in 1944. Until then, patients remained suspended in crisis: fearful of the future and desperate for any remedies or signs that might foretell a long life. For Irving Fisher, uncertainty was not an abstract idea but a visceral reality.
He was not alone. Even a casual scan of events at the start of the 20th century reveals a concatenation of wars, terrorism, political assassinations, earthquakes, royal suicides, epidemics, and famine. Whole new countries took shape while scientists struggled to understand the impact of four-dimensional geometry, new moons, gases, and the new science of relativity and quantum theory. Whether through fear of the unknown, or hope to capitalize on new trends, a large, eager, and susceptible market arose, desperate to know what the future might hold. And for the first time in history, technology provided tools that promised to make forecasting scientific. The telegraph and telephone enabled the collection of large amounts of up-to-date information. The emergence of statistics as a rigorous mathematical discipline, together with the growing sophistication of economics, facilitated serious data analysis. An ever-expanding railway network could distribute newsletters, newspapers, and magazines to an eager market of punters and pundits.
The 101 Biggest Business Moves of the Pandemic
From the admirable to the audacious, the highs and lows that have defined the last eight months
Astrology became a big commercial business at this time, too, but it was in financial markets that forecasting first became an important industry. Panics in the United States in 1893, 1896, 1901, and 1907 had exposed how little reliable information consumers, investors, and managers had about the health of companies, industries, or the economy at large. Into that vacuum rushed three men: Irving Fisher, Roger Babson, and Warren Persons. All were eager to sell reassurance, inspiration, and advice. All three believed that, through data, they could discern future trends in the markets and hoped to build important businesses doing so. And all three, carrying the diagnosis of TB, viscerally understood the pain of uncertainty and sought to alleviate it.
The density of collective unknowns left it up to the forecasters to choose what mattered to whom, why, and when. And they had a field day.
But even with more economic and business data than ever before, no one understood what kinds of economic conditions were needed to temper an overheated market, grow a slow one, or stabilize volatility. The density of collective unknowns left it up to the forecasters to choose what mattered to whom, why, and when. And they had a field day.
Each of the forecasters built commercial businesses selling their special take on the future. Fisher, whom Milton Friedman considered the greatest economist the United States has ever produced, was one of the first to try to analyze national economies, seen through the lens of the money supply. His working assumption was that too much money in circulation would produce an inflationary boom, too little a recessionary bust. At the time, the government didn’t measure money supply, which left Fisher trying to do so. So he created indexes, aggregates of data he hoped would reveal overall patterns in economic activity.
What his homemade indices revealed was volatility — in prices and in markets. So Fisher became obsessed by a search for stability: Where did it come from, what influenced it, and what sustained it? The more data he collected, the more questions emerged, all needing answers before Fisher could hope to anticipate where the economy was going.
Giddy with new insights, Fisher became one of the world’s first economic pundits. His Index Number Institute, syndicating indexes and forecasts through newspapers and newsletters, made Fisher famous for financial commentary and an immense capacity for data analysis. He was easily drawn into commenting on a whole range of topics from Prohibition to simplified English spelling and calendar reform. But his fortune was made when he sold his card index system to the Rand Kardex Company for $660,000 (approximately $8 million today). Fisher commanded attention and credibility for his mathematical rigor that promised to bring economic forecasting one step closer to a science.
Economics has long suffered from physics envy and nowhere was that more explicit than in the early days of forecasting. One of Fisher’s rivals, Roger Babson, believed that almost everything in life could be reduced to Newton’s laws of cause and effect. Like Fisher, Babson had contracted TB as a young man and devised his own eccentric, ascetic health routine: freezing air and a strict diet. Babson was on a mission to redress a power imbalance. As a young bond salesman, he had discovered that banks held a monopoly on business information; investors knew only what institutions told them. So Babson brought to his new business an evangelical determination to empower individuals with data as sound and thorough as any bank’s. He became famous for his Babsoncharts, spectacularly baroque sheets of graphic design on which he tried to display the full complexity of an economy: stock and commodity prices, manufacturing data, railroad traffic, agricultural production, building construction, business failures, and other indicators of economic output. Onto this data he placed what he called the “normal line,” indicating periods of expansion and recession. Ever the fervent Newtonian, he put his trust in the Third Law of Motion: for every action, there is an equal and opposite reaction. He believed that a period of depression was always matched by a period of prosperity and that the steeper the decline, the faster the market would recover.
Babson imbued economics with morality; booms were the product of wasteful exuberance that needed to be purged by sensible self-discipline.
Babson imbued economics with morality; booms were the product of wasteful exuberance that needed to be purged by sensible self-discipline. These views made him a contrarian: When markets went up, he foresaw extravagance and urged healthy restraint. Profiled as “the man who refused to die,” his success and fame conflated his apparent victory over tuberculosis with the moral lessons implicit in his market predictions. By 1910, he, too, had become a national pundit, called on to pronounce on everything from markets to medicine, education, diet, and religion. No matter the topic, whenever Babson made predictions, he always looked for excess that needed reining in, or restraint that demanded a bigger push.
Both Fisher and Babson became irrepressible entrepreneurs, financing their forecasting businesses with their own money and running them from home. By stark contrast, Warren Persons built his forecasting business, the Harvard Economic Service, right inside the university that funded it, hoping that, far from Wall Street, it could remain aloof from punditry and secure a solid reputation for scholarship and objectivity. A masterful statistician, he was skeptical that any theories fully captured the complexity of economic markets. The best you could do was watch and measure what was in front of you, and ask if you had seen such correlations and patterns before. The Harvard Economic Service was the world’s first economic advisory business to serve a worldwide market of the elite. In the 1920s, the Service began to collaborate with Keynes and Beveridge’s London & Cambridge Economic Service, but for all these academic credentials, a problem lay at the heart of Persons’s approach. Even if he believed himself immune to theory, didn’t his attention to some trends over others imply a theory? In denying his assumptions, did he risk being blind to them?
All three men were personally invested in their competing theories and methods; their businesses and professional reputations depended on them — as, to a large degree, did the future of the forecasting industry as a whole. Persons compared Babson’s ideas to astrology and said that Fisher’s data was unreliable. Where both men’s methods were easily castigated as pseudo-science, Persons’ hundred-page explanations left him open to methodological criticism from rivals and colleagues alike. Across the industry, rivals sniped at each other, trying to prove that they, and they alone, held the key to the future.
The Economists Urging Us to Embrace Uncertainty Instead of Forecasts
The author of ‘Radical Uncertainty’ on why building resilience matters more than predicting the future
The test came in October 1929. At the beginning of the month, Fisher had been buoyant, claiming that stocks had “reached a permanently high plateau.” Stability at last! When the market collapsed on October 24, while conceding there might be a slight price retreat, he saw nothing “in the nature of a crash.” For months afterward, he insisted that a rapid recovery was imminent. His faith cost him dearly: holding on to his shares in Rand Kardex through the collapse, he lost his fortune.
Persons’s Harvard Economic Service was equally blindsided. He later maintained that, with no historic precedent for the crash, a normal and swift recovery must follow. Month after month, the Service kept predicting a recovery that failed to arrive. The Service closed in 1931.
Only Babson, the least scientific of the three men, came out with his reputation enhanced. Ever the contrarian, he had been predicting a crash every year for the past three years. He basked in his triumph, snapping up some of his failing competitors and blaming his rivals for encouraging speculation. But in May 1931, he announced that the market had bottomed out and that it was time to get back into the stock market. He was wrong: the U.S. economy wouldn’t recover for a decade.
What they left behind was the commercialization of a fantasy: the belief that the future is knowable and that there are some special people or processes that can reveal what the future holds.
Persons moved on to consulting, where his statistical punctiliousness was said to dismay clients. Fisher, his reputation in ruins, died broke; his investments had all gone bust. Though Babson was the forecaster who left behind a fortune and a college that still bears his name, his ideas — about markets and medicine — are now discredited.
While they had seen themselves as wildly differentiated rivals, in fact these three pioneers had much in common. What they left behind was the commercialization of a fantasy: the belief that the future is knowable and that there are some special people or processes that can reveal what the future holds. They both ratified our addiction to prediction and proved that fame and fortune could be made from it. But these early pioneers also uncovered three profound problems endemic to forecasts that dog them still today: they are incomplete, ideological, and self-interested.
3 Ways to Prepare for a Future You Can’t Predict
Companies don’t have the luxury of waiting for a stable decision-making environment
The first problem is models. Fisher and Babson relied on simplified versions of the markets they sought to understand. They didn’t have the resources or the means to capture all the data on the economy — and even if they had, they lacked the tools to handle it. So they had to make choices about what to leave in and what to leave out. All economists do this kind of editing, in an effort to s ee more clearly what is really going on. Today’s technology accommodates far more data but the intrinsic difficulty of models remains: the more data is compressed, the more its predictive power is compromised.
The second problem lay in agendas. However much the early forecasters believed themselves to be men of pure, scientific inquiry, they all held cherished, implicit beliefs about how the world worked, about what mattered and what did not.
If boom and bust was good for the forecasting business, how easily might the industry itself, instead of reporting data, start trying to influence events instead?
Finally, all of these early forecasting businesses were commercial enterprises. How far were forecasters driven to please their payers? The first person to raise this question had been John Maynard Keynes, whose early interest in business barometers had led him to work with Persons. But writing to colleagues at the Harvard Economic Service in 1925, he had wondered whether there wasn’t a conflict of interest in the very nature of the business. If boom and bust was good for the forecasting business, how easily might the industry itself, instead of reporting data, start trying to influence events instead?
What Babson, Fisher, and Persons had discovered, and Keynes had identified, was that forecasters gained bigger reputations and could make more money by being bold; they had a vested interest in exploiting our hopes and our fears. What none imagined was that the drama of the markets would evolve so spectacularly from drama to showbiz to hype to propaganda and the manipulation of consumer behavior. Human discomfort with uncertainty, together with a craving for reassurance, has fueled an industry that enriches itself by terrorizing us with uncertainty and taunting us with certainty.