The impossibility of infinite growth

[next_migration]

Eliza Daley has just posted an interesting essay, reprinted on Resilience.org, that I’d like to both recommend and talk back to. Since her view of the situation in overlarge cities and the need for deurbanization is tangential to mine, I’d first like to hop back to the topic of unsustainability and post a few more sections of my general view of the problems our society is facing. Climate change is the most urgent, because in some places it’s already bad and heading rapidly towards disastrous. However, it turns out that this is just a special case of a larger phenomenon—what I call Tier 2 collapse—which says that a society built on growth, like ours, must eventually find some consequence of its way of life putting the squeeze on it. The people who have examined this most rationally believe that even if there was no such thing as climate change—or ocean acidification—decline in the relatively near future would still be unavoidable under any business-as-usual model. Here’s a brief summary of their conclusions.

 

 

In the early 1970s, a team of systems analysts funded by the Club of Rome created a computer model to project possible trajectories of world economic growth. Results were widely known as the Limits to Growth study. Three leading members of that team, Donella Meadows and colleagues, in 2004 published an update with additional analyses called Limits to Growth: The 30-Year Update. This book, which is not overly technical, is highly recommended reading for anyone who wants to understand the trap of globalized industrial capitalism. [See a very brief synopsis here.]

The Limits to Growth analyses were inspired by the observation that human population had skyrocketed through exponential growth in the industrial era, and especially the fossil fuel era. (The population problem has worsened by billions since the original study.) Industrial production, fossil fuel consumption, and added carbon dioxide in the atmosphere all showed similar patterns of initially gentle increases that escalated rapidly. Our governments and financial industries became so used to seeing constant economic growth in the fossil fuel era that many nations created economic arrangements in which the stable functioning of the economy requires constant growth, and if growth stopped, parts of the financial system would jam up severely.

But infinite growth isn’t possible. A widely repeated quotation, attributed to economist Kenneth Boulding but apparently of uncertain origin, says: “Anyone who believes exponential growth can go on forever in a finite world is either a madman or an economist.” Conservative economist Herb Stein made a similar point: “If something cannot go on forever, it will stop.” There must be limits. The purpose of the Limits to Growth study was to explore what those limits might be and how they might interact with our current trajectory of growth. It did this by creating a computer model of the world economy, called World3, which represented the world as a collection of basic stocks and feedback loops.

Stocks were such basic categories as population, nonrenewable resources, cultivated land, industrial capital, service capital, and pollution. Positive and negative feedback loops showed the relationships between these categories. For example, if more industrial capital is directed toward the goal of improving agricultural yields, food produced will increase; more food per capita can lead to reduced mortality, which leads to higher populations; if population increases, food per capita decreases.

These categories are obviously oversimplified, e.g., treating oil and iron ore as all one thing, “resources”, and a range of estimates of the strength of feedback relationships would be possible. Meadows and colleagues were therefore careful not to claim that the speed and magnitude of change over time in runs of the model predicted what would happen in the real world. Still, that simplicity makes the general patterns shown by the model hard to ignore. We might dispute how strong the relationship between health services and life expectancy is, or how badly certain levels of pollution would affect agricultural yields, but we wouldn’t argue with a straight face that health services have no effect on health or that any amount of pollution whatsoever would be harmless.

The model’s output showed graphs of stocks and measures of economic and human welfare over time, beginning in 1900 and ending in 2100. The model’s baseline Scenario 1 used numbers that the authors, in 1972, considered reasonable for the stocks and relationships among them. Results of Scenario 1 depicted industrial output continuing to increase until about 2020, when it first leveled out, then began to decline. At that point, the highest-quality nonrenewable resources had been used up. More capital had to be put into extracting lower-quality resources, making less available for other functions like supporting agricultural and industrial growth and maintaining industrial capital, and it became impossible to keep extracting more resources every year. The result was that by 2050 industrial output was little higher than in 1950, and in 2100 it was little higher than in 1900. Agricultural productivity followed a similar but somewhat shallower curve, and pollution followed a lagging curve that peaked years after output. Population also peaked in the 2030s then declined, but much less dramatically than production, which meant that life expectancy and food, services, and consumer goods per person all declined dramatically. The average human welfare index at the end of the 21^st century was lower than it had been in 1900.

We are now past 2020 and we are reaching a point where more and more resources need to be put into resource extraction, e.g., getting fossil fuels via fracking and mountaintop removal rather than the easier, cheaper drilling and mining of olden days. However, setting aside widespread pandemic-induced shutdowns, industrial output has not yet obviously or definitively peaked. There have been waves of unusual shortages and bottlenecks in the global economy, but so far those are still being attributed to the disruption of the pandemic. Therefore, we may suspect that Scenario 1’s estimate of available resources was too stingy. To address the question of what would happen if the world’s supply of resources was more generous, the next scenario that Meadows et al. explored doubled the starting amount.

In the high-resource Scenario 2, rather than facing a resource crisis, the simulated world of World3 faced a pollution crisis. [The graphs of this model are included in the synopsis linked to above.] Industrial output and population kept climbing until the 2040s. (Population was predicted to peak at 8 billion in 2040; in real life we are already at 7.8 billion, having well exceeded the Scenario 1 prediction.) But the consequence of the continued ability to extract and pollute was that pollution, produced in amounts greater than the Earth could absorb, built up to a far greater level, peaking around 2090 at five times the Scenario 1 peak. Higher pollution reduced land fertility, and as more resources were put into struggling to maintain food production, industrial output declined. Food per person and life expectancy plummeted in the mid-21^st century, with a slight rebound thereafter; the human welfare index declined; and the human ecological footprint peaked in 2050 at a level gigantically higher than that of 2000.

The Limits to Growth team did not try to predict exactly what the effects of pollution would be. It could be, Meadows et al. (2004) wrote, “through soil contamination by heavy metals or persistent chemicals, through climate change altering growth patterns faster than farmers can adapt, or through increased ultraviolet radiation from a diminished ozone layer.” The general principle is that if we dump harmful things into the environment faster than it can cope with them, we must someday find that the fruitfulness of that environment is harmed. Though it is not relevant to my topic of interest, we must also end up eating, drinking, and breathing those things. One might speak of the probable contributions of industrial chemicals to rapid increases in obesity and some cancers, early-onset dementia, declining sperm counts, and reduced cognitive function in children: toxicity that impairs our health and diverts resources towards medical care must reduce productivity in other sectors of the economy.

So here is where climate change, our Tier 1 catastrophe, becomes just a special case of the underlying fact that exponential growth in consumption and pollution must, somehow, someday, run into a limiting feedback loop. Burning more oil and coal means more droughts and heat waves, less irrigation water available from shrinking reservoirs or vanishing glaciers, less fodder for grazing animals in desertifying areas, and in wetter areas, more unexpectedly flooded farmlands. All of these things can lead to reduced agricultural yields, in the U.S. and globally, to say nothing of their direct impacts on human health. Likewise, resources put into rebuilding homes and businesses after hurricanes or creating huge water projects to maintain supplies for cities in drying areas will be resources that can’t be used to maintain other infrastructure or other forms of capital.

You might query Scenario 2 as well. Perhaps Meadows et al. underestimated our ability to avoid the consequences of both depletion and pollution. The team created further scenarios to look for one that would avoid an ugly decline in human welfare. Resources could be invested in ever-improving anti-pollution technologies, then technologies to increase agricultural yields faster, then erosion protection to keep the yield improvements from destroying the farmland faster. In each scenario, at some point the costs of business as usual became too high to sustain, and declines became inevitable, though less extreme than in Scenario 2. The reason is the phenomenon of overshoot and collapse: if the level of consumption is allowed temporarily to rise above what can possibly be sustained long-term, then at some point it must fall.

In Scenario 6, investments in efficiency were added to all the others. This allowed population to level off rather than collapsing, though goods and services per person declined. You might also ask, what about population control? A scenario in which the world agreed en masse to limit fertility to 2 children per woman, starting in 2002, did virtually nothing to change the catastrophic Scenario 2; nor did a global choice to freeze industrial output per person make much difference. The only scenario Meadows and colleagues found that allowed a comfortable leveling out of population, life expectancy, food, goods, and services per capita, and human welfare, with no declines, required the combination of global agreement to pursue stable population and industrial output per person, plus widespread application of all of the footprint-reducing technologies.

Obviously, that didn’t happen in 2002. What are the chances that it will happen in the near future, when it might still make some difference? Effectively zero, it seems. Therefore, the very likely pattern of the coming century will be one of some kind of overshoot and decline, made inevitable by our species’ excessive numbers and consumption. Again, the exact trajectories depicted by the Limits to Growth team’s models are not to be taken as predictions. If we appear to be heading into something like Scenario 2, we shouldn’t imagine that we know exactly when the population peak or agricultural output peak will be, or how severe the fall will be afterwards. Those will depend on a thousand factors that broad general models couldn’t begin to incorporate.

What we should be convinced of is that the pattern of the world’s economy over the past two centuries, continuing to get bigger and bigger every year while global corporations do as little as possible to clean up their own messes and a few naysayers grumble about the harms of pollution or the mass extinction of species on despoiled lands, can’t continue unabated for another century. Decline is unavoidable.