by David Orrell
Agent-based models are an interesting area of research activity, and they can yield insight into the underlying dynamics of markets, as well as expose flaws in the orthodox theory. They can also be used to do “what if ” simulations for particular situations—for example, to see how a change in a company’s transportation network might affect delivery times. However, it is questionable whether they can help to make more general predictions. It is one thing to produce market-like behaviour in an abstract system, another to forecast what will happen in the real market.61
One phenomenon poorly explained by orthodox economic theory is the apparent excess volatility of financial assets, which is inconsistent with the hypothesis that volatility is caused purely by external shocks. Comparisons with biological systems suggest reasons why this is the case. The metabolism of a cell is controlled by many different proteins. The number of molecules of each protein varies in a random way, owing to stochastic effects. The cell also misreads cues from the environment. The resulting variation can be interpreted as a kind of prediction error: the cell produces too many of some proteins and not enough of another, and it also reacts to its own errors. The result is that even in the absence of external shocks, the rate of metabolism will fluctuate randomly. Organisms, like the yeast in Chapter 5, have evolved complex feedback loops that reduce but do not eliminate this natural variation.
Similarly, an economy can be seen as a kind of super-organism and its output a measure of metabolism. The economy takes in resources from people and the environment, transfers them into material wealth, and expels waste products. Value circulates through the economy, changing form from matter to money to labour in a ceaseless flow. The net output is controlled by a large number of individuals and companies, each of which is constantly adjusting output in accord with its predictions about the future. Because the predictions are erroneous, control is incomplete, and the response of each actor is inherently creative, net output will fluctuate randomly of its own accord, adding to volatility.62 The market is therefore not a dead thing, at equilibrium; as traders like to say, it has a life of its own.
Just as biological systems dislike excess volatility and incorporate complex feedback loops to reduce it, sophisticated economies develop mechanisms to damp out volatile fluctuations. An example is insurance. Let’s say a disaster such as a crop-destroying drought has driven individual farmers, whose weather predictions were wrong, into bankruptcy. This has knock-on effects on the rest of the economy, which amplifies the disaster’s impact through positive feedback. But with insurance, the loss is dissipated through financial structures that are less vulnerable to collapse than an individual farmer—negative feedback. Similarly, tools such as options or futures contracts allow businesses to smooth out fluctuations in prices and increase their own value. These effects can be simulated (but not precisely predicted) using simple top-down models that capture overall function.
The biological analogy suggests that while the orthodox theory is right to state that the market is practically impossible to predict, it is right for the wrong reasons. The EMH assumes that market fluctuations are the result of random external shocks, and that its response is governed by rational laws. In other words, it treats the economy as a dead object that can be modelled like a falling stone. However, a model that views the economy as a kind of super-organism would ascribe fluctuations not just to external causes but to the market itself. What makes its response unpredictable is to a large part its own inability to predict. And like a living organism, the economy represents a shifting, dynamic balance between opposing forces—positive and negative feedback loops, buyers and sellers— so models are sensitive to changes in parameterization.
In its insistence on rationality, the EMH is therefore a strange inversion of reality. Its primary aim, it appears, is not to predict the future, but to make it look like we all know what we’re doing. This is dangerous for two reasons. The first is that because the EMH views the market as “normal,” it gives an illusion of control, and at the same time it tends to underestimate the real risk of future financial storms. This is especially a concern when the results are incorporated in policy. As the economist Kenneth Arrow, who worked as a weather forecaster during the Second World War, put it: “Vast ills have followed a belief in certainty, whether historical inevitability, grand diplomatic designs, or extreme views on economic policy. When developing policy with wide effects for an individual or society, caution is needed because we cannot predict the consequences.”63 There is a curious disconnect between the consistent inaccuracy of the forecasts and the confidence with which politicians, banks, and business leaders regularly use them to make important decisions. The second danger comes from the insidious idea that “the market is always right,” that it is some kind of hyper-rational being that can outwit any speculator or government regulator. This view of the economy, enshrined in the EMH, turns the market into a deity who is watched over, granted legitimacy, and explained to the rest of us by the economic priesthood. It leads to what George Soros has described as a “market fundamentalism,” which is as dangerous as any other kind of fundamentalism and gives a kind of carte blanche to empathy-free corporations to do what they want, under the pretext that they are just being efficient.64 People or countries that fail under this system have been judged by the market. But the market is no more rational than the square root of two. It is just the net effect of our own stumbling as we try to find our way in an unpredictable universe.
THREE SIBLINGS
As we’ve seen in these last three chapters, the scientific approaches to making predictions in the areas of weather, health, and wealth share much in common. Atmospheric and economic forecasts have always been linked, even more so when agriculture played a larger part in the economy. This was acknowledged by the Victorian scientist Stanley Jevons when he attempted to predict the business cycle by monitoring sunspots. Francis Galton used his statistical methods to study inheritance, but they have proved equally useful— and much less controversial—in economics. Today, the techniques used by scientists are essentially the same in all three areas, and all have their roots in nineteenth-century astronomy.
While celestial objects are happy to obey dictates like the law of gravity, and are amenable to modelling by equations, systems like the weather and the economy appear more anarchic. The rules they obey are local and social in nature, rather than global. As a result, in all three areas of prediction, scientists run into the same problems. The underlying system is uncomputable, so models rely on parameterizations that introduce model errors. As the model is refined, the number of unknown parameters increases. The multiple feedback loops that characterize such models also make them sensitive to even small errors in parameterization. As a result, the models are highly flexible and can be made to match past data, but accurate predictions of the future remain elusive. The models are often most useful as tools for understanding the present function of the underlying systems.
The three areas of scientific forecasting—weather, health, and wealth—are like siblings. They have the same origins, grew up together, and hung out with some of the same people. Each has its own character. Weather, the eldest, is the one the others look up to, because it is closest to the stars and knows physics. Health, the youngest, used to get in trouble, but it’s flush with optimism as it prepares to come of age. (In its school yearbook, it was voted most likely to find a cure for cancer.) Wealth is the narcissist, spending its days preening in front of the mirror, in thrall to its own beauty and efficiency. In the final part of the book, we find out how these would-be clairvoyants fare as they join forces to take on the greatest challenge of all—a long-term prediction for the planet.
FUTURE
7 THE BIG PICTURE
HOW WEATHER, HEALTH, AND WEALTH ARE RELATED
What a chimera, then, is man! What a novelty, what a monster, what a chaos, what a subject of contradiction, what a prodigy! A judge of all things, feeble worm of the earth, depositary of the truth, cloaca o
f uncertainty and error, the glory and the shame of the universe!
—Blaise Pascal, Pensées
Past performance is no guarantee of future results.
—Mutual fund prospectus
CASE HISTORIES
The previous three chapters looked at short-term predictability in atmospheric, biological, and economic systems. Long-range forecasts, the subject of this last part, differ in that the aim is not to predict exactly what will happen at some fixed date, but to estimate major future effects. This may seem a fundamentally different task, but really the only things to have changed are the scales in time and space. Instead of predicting the local weather, averaged over an afternoon, some days in advance, long-range forecasters want to estimate the regional climate, averaged over a number of years, some decades in advance. In medicine, a long-range forecast might address the likelihood of large-scale pandemics emerging in the global population, while in economics, it could be concerned with the scope for, and consequences of, continued growth.
These systems are, of course, not independent of each other, especially over longer time periods. Global warming, for example, is a function of carbon-dioxide emissions, which depend on economic activity. To make a prediction for our civilization and the planet, we need to consider physical, biological, and social effects as, to use Keynes’s expression, an organic unity. The outcome also depends on the choices we make. Let’s look, for example, at two cases where older civilizations have tangled with their environment, to mixed effect. The first is notorious, the second less so.
Case history A is the nicely named Easter Island. This small island in the Pacific is a little off the beaten track, 3,700 kilometres from the coast of South America, but it’s famous among tourists and archaeologists for the amazing stone figures, the moai, that line the shores. When the Polynesians colonized it around 400 A.D., the island was a subtropical paradise alive with forests, birds, and animals, its seas rich with fish and dolphins that the islanders caught from canoes. Over the course of hundreds of years, a kind of small civilization grew up. The population reached around 10,000 and divided into clans and classes. Despite what we see as their remoteness, the islanders, like the citizens of Delphi, believed themselves to be at the centre of the universe. (The name of one spot translates to “navel of the world.”) They honoured their ancestors by carving the giant moai out of volcanic rock. Transporting and erecting the incredibly heavy moai required a great deal of ingenuity, and large numbers of log rollers. Between this, the clearing of land for agriculture, the use of firewood for heating, and other effects, the island was by 1400 completely deforested. The birds in the forests went extinct; exposed soil blew away into the ocean; there was no wood to make canoes; streams and lakes dried up; crop yields collapsed; wars were waged over the island’s remaining resources; everything went Malthusian. By the time the island was encountered by Europeans in 1722, there were only a couple of thousand people left, and they had taken up cannibalism. In the end, the islanders even turned against their stone gods, toppling and destroying them until not one was standing. Perhaps their promises, or predictions, had not come true. In 1900, after most of the remaining population had been ravaged by smallpox introduced by the Europeans, only 111 people remained.1
Case history B is the still smaller island of Tikopia, located just east of the Solomon Islands. Only 4.6 square kilometres in size, Tikopia was settled earlier than Easter Island. It was heading the same way until about 100 A.D., when it seems the population was converted to the benefits of orchard gardening and sustainable lifestyles. (I imagine an early, hard-core version of the Green Party.) Taboos developed that regulated both procreation and the consumption of food. Zero population growth was policy. It was enforced by the usual methods of celibacy and birth control, but also by more extreme techniques, such as abortion and infanticide (usually suffocation). Young men were sent out to sea on highly risky fishing expeditions, with the knowledge that only a few would return.2
The initial conditions in both cases were similar, but over time scales of centuries, the outcomes were completely different. Easter Island is a hit with tourists and a fright show for environmentalists. Many of the moai statues have now been restored, and they stare out from the covers of books and magazines as a kind of warning. No one will ever, ever mindlessly pollute there again. Some trees are beginning to return.3 Tikopia still supports around a thousand souls; zero population growth is assured because the young people tend to leave.
Who would have seen it coming? The course of civilization does not run smooth; ingenuity may not translate into survival skills, and technological achievements outlast their creators. The people of Easter Island and Tikopia have achieved a kind of quasi-balance with their environment. So how will the rest of us fare as we push against the limits of our much larger but still finite island? What type of story will ours be—horror, light-hearted comedy, or difficult European art film that offers no easy answers?
It seems unlikely that a kind of global civilization model, similar to the psychohistory in Isaac Asimov’s fictional Foundation Trilogy (“Q: Can you prove that this mathematics is valid? A: Only to another mathematician”4), can tell us the answer, given that we can’t predict next week’s weather. As Karl Popper asserted in 1957, “There can be no prediction of the course of human history by scientific or any other rational methods.”5 However, in 1968, the Club of Rome’s thirty members, drawn from science, business, and government, had a go. In collaboration with some professors from MIT, they loaded a computer model called World3 onto a mainframe, fed in some data, and stood back to see what would happen. The results, published in The Limits to Growth, were Easter Island: The Sequel. “If the present growth trends in world population, food production, and resource depletion continue unchanged,” the forecasters wrote, “the limits to growth on this planet will be reached sometime within the next one hundred years. The most probable result will be a rather sudden and uncontrollable decline in both population and industrial capacity.”6 The results appeared to imply that world population would peak at around 10 billion and crash to around half that in the middle of the twenty-first century.
The problem with such predictions, as the authors pointed out, is that present growth trends will not be sustained. Nothing in this world is fixed, especially not trends. The results implied that oil supplies would run out in the 1990s, but that didn’t happen (in part because more oil was discovered).7 Similarly dire warnings in the 1960s that the human population would collapse for lack of food did not come true because trends changed (population growth rates fell and food production improved).8 So how far can we see into the future? And can scientific models help?
To answer this question, we must consider long-term predictions of weather, health, and wealth. But we must first set the stage with a brief history of these three intertwined aspects of our lives— the story so far—and then go on to discuss future projections. Climate change is a highly contentious issue, so it gets the most space. Economic predictions that extend more than a few months ahead are more futurology than science, so economic growth is here discussed primarily in the context of how it will affect, and be affected by, climate change. Finally, we take a careful peak at global pandemics.
OUR HUMAN STOCK
The easiest place to begin a prediction is with the historical charts. The earth is billions of years old, and mankind has been around for hundreds of thousands of years—unless you believe in creationism, in which case you probably believe in Armageddon as well, which kind of takes the fun out of prediction.9 But for the rest of us, we join the story about 10,000 years ago, with the invention of agriculture in places such as the Fertile Crescent, in today’s Middle East. This technological and cultural leap was in part made possible by a relatively stable climate. For 3 million years, the climate had alternated between warm and cold periods, driven by subtle oscillations in the earth’s orbit.10 Ten thousand years ago, the last ice age, known as the Younger Dryas, had just thawed out, and average
temperatures had increased from about 0°C to a relatively balmy 14°C. Sun hats were back in fashion.
Agriculture spread slowly, bringing improved nutrition and fuelling a rapid increase in population (from about 5 million to around 250 million at the time of Christ). Societies, which pre-agriculture had consisted mostly of roaming bands or tribes, grew into increasingly complex and stratified civilizations, with distinct classes of priests, soldiers, rulers, and labourers. Civilizations including the Greek, Roman, and Mesopotamian empires, developed money— usually coins of precious metals like gold, silver, and bronze, stamped with the images of gods and goddesses (like Apollo).
We tend to think of environmental problems as a modern phenomenon, but any civilization will grow until it encounters a limit of some type, and often it is a natural limit. Removal of forests and over-extensive agriculture led to droughts, floods, and topsoil erosion, sometimes causing local environmental collapse. Plato was aware of the dangers and famously described deforestation in Attica: “What now remains compared with what then existed is like the skeleton of a sick man, all the fat and soft earth having wasted away, and only the bare framework of the land being left.”11 Despite such warnings, forests continued to disappear across the Mediterranean region.