Skip to main content
Modelling and simulation

Modelling and simulation

An uncertain growth

13 Feb 2020 Susan Curtis
Taken from the February 2020 issue of Physics World.
Sunflowers

When I first picked up Vaclav Smil’s latest book, Growth: From Microorganisms to Megacities, I wondered whether his main argument would be that systems as diverse as nature and demographics follow a universal growth trend. Indeed, it’s true that there are many similarities to learn from: the height gained by sunflowers during the summer, for example, forms the same growth pattern as the average area of American houses since 1990 and the adoption of mobile phones over the past two decades. But Smil – a Czech-Canadian environmental scientist and policy analyst – also takes pleasure in exploring the idiosyncrasies of these different growth processes, and the factors that influence them.

Throughout Growth he presents rigorous quantitative analyses of disparate systems, ranging from biological processes and crop cultivation to economies and changes in population. Many of these follow an S-shaped growth trajectory, in which initial incremental gains are followed by rapid expansion, and then a long tailing off as the system approaches its limit. Other processes may show more exponential growth from the start – such as the rapid adoption of the telephone at the beginning of the 20th century – but the gains always seem destined to level off over time.

Smil exploits these detailed analyses to draw important conclusions about the nature of growth. It becomes clear that even small variations or external interventions can disrupt the neat progression of an expected growth trajectory. The rapid spread of a flu epidemic, for example, can be curtailed by vaccinating just 20% of the population, while only tiny changes in rainfall or temperature can wipe out expected improvements in crop yields. Coupled with that unknown variability is the difficulty of fitting the best growth curve to observed data.

I particularly enjoyed Smil’s discussion of attempts to estimate how many musical masterpieces Mozart could have produced if he had lived beyond the age of 35. Some previous commentators had fitted an S-shaped curve to Mozart’s cumulative musical output, from which they concluded that he must have written 18 unpublished works as an infant, and that his creativity would have been 91% exhausted by the time he died. Smil puts those claims into doubt by presenting four alternative growth trajectories, all of which provide a good fit to Mozart’s back catalogue, with the number of projected compositions at age 50 ranging from fewer than 800 to more than 1300.

The lesson to be learnt, argues Smil, is that growth models are an unreliable predictor of the future. Growth curves often provide valuable insights into the evolution and development of particular systems, and offer some predictive power for repeatable processes that are well understood – such as the growth of bacteria and other living organisms. But extrapolating a trajectory from a few early data points is fraught with danger, particularly when dealing with more complex systems such as cities, economies or civilizations.

Smil is particularly critical of what he sees as wildly optimistic predictions for the development and adoption of new technologies, such as claims that we will all be driving electric vehicles by 2025. Such growth projections, he argues, often mistake initial performance improvements for the early stages of an exponential curve, while most new technologies advance in a more stepwise fashion – where periods of fast growth as new materials or designs are introduced are interspersed with frequent and sometimes long-lived plateaus.

What can seem like disruption is also often the result of previous, more incremental, advances – the recent explosion of digital technologies, for example, has been enabled by successive innovations in optical fibres and long-distance communications systems over the last 50 years or so. Even Moore’s law, which for decades has successfully predicted exponential growth in the number of transistors that can be fitted onto a silicon chip, is likely to slow down, now that transistor linewidths have reached the atomic scale – although Smil notes that computing power will continue to grow rapidly due to other improvements, such as more specialized chips and the emergence of photonic and quantum technologies.

Smil is also sceptical of the role that technological innovation plays in driving economic progress. The key test, he says, is whether a new invention improves productivity or quality of life – and in those terms the most innovative period was the half century before the First World War, when electricity, telephones and motorized vehicles all became widespread. In contrast, the digital revolution of the past few decades may have changed how we communicate and consume information, but it hasn’t delivered any measurable improvement in economic prosperity.

Smil reserves most of his ire for the modern notion that economies are healthy only if they grow

But Smil reserves most of his ire for the modern notion that economies are healthy only if they grow. He has long been a critic of using gross domestic product (GDP) as an economic indicator, partly because it is an aggregate metric that hides as much as it reveals. Nor does it truly reflect economic prosperity, since headline GDP growth takes no account of unpaid work – such as caring for children or the elderly – and can be achieved without any significant improvement in the quality of life. Japan, for example, enjoyed annual GDP growth of around 8% throughout the late 1950s and the 1960s, but wages remained low for most industrial workers.

Even more troubling to Smil is that the GDP metric ignores the impact of economic growth on natural resources and the environment. Alternative measures that take these factors into account – including a study by Smil himself on China’s rapid economic growth in the 1990s – suggest that gains in GDP can be negated or even reversed by the loss of so-called natural capital.

And this is the crux of Smil’s argument: that the growth of human activity and productivity is fundamentally limited by the resources available on our planet, just as the growth of sunflowers is limited by the amount of light available for photosynthesis. Those limits are already being reached, he says, with soil erosion causing crop yields to decline, irrigation depleting deep aquifers faster than they can be replenished, and urban spread and deforestation causing a loss of biodiversity. These and other environmental impacts – including the unpredictable consequences of a rapidly warming climate – are largely ignored in the modern quest for perpetual economic growth.

Smil doesn’t necessarily believe that a climate calamity is just around the corner, but he also rejects the idea that radical technologies – such as terraforming Mars – offer a viable solution for our long-term survival. Instead, he believes we need to recalibrate our expectations of success. Our progress should no longer be defined by economic growth or material consumption, but by our ability to inhabit our planet alongside other species for millennia to come. It’s hard to argue with the conclusion, but easy to question whether those in power will be persuaded to chart a different course.

  • 2019 The MIT Press 664pp £30hb

Related events

Copyright © 2024 by IOP Publishing Ltd and individual contributors