The Limits of The Limits to Growth
Contemplating 1972 predictions of environmental doom, just in time for Earth Day
Forty years ago, The Limits to Growth, a report to the Club of Rome, was released with great fanfare at a conference at the Smithsonian Institution. The study was based on a computer model developed by researchers at the Massachusetts Institute of Technology (MIT) and designed "to investigate five major trends of global concern—accelerating industrial development, rapid population growth, widespread malnutrition, depletion of nonrenewable resources, and a deteriorating environment." The goal was to use the model to explore the increasingly dire "predicament of mankind." The researchers modestly acknowledged that their model was "like every other model, imperfect, oversimplified, and unfinished."
Yet even with this caveat, the MIT researchers concluded, "If present growth trends in world population, industrialization, pollution, food production, and resource depletion continue unchanged, the limits to growth on this planet will be reached sometime within the next one hundred years." With considerable understatement, they added, "The most probable result will be a rather sudden and uncontrollable decline in both population and industrial capacity." In other words: a massive population crash in a starving, polluted, depleted world.
The problem, as the MIT researchers saw it, was exponential growth in all five areas of concern that they investigated. Linear growth is additive—1, 2, 3, 4, 5—whereas exponential growth compounds over time—1, 2, 4, 8, 16.
Earlier this month, Smithsonian magazine ran a short item reporting the findings of a 2008 study [PDF] by Australian physicist Graham Turner. According to Turner, an examination of currently available data reveals that that world economy is right on track to collapse by the middle of this century. Without taking Turner on directly, let's look at what has happened with each global concern highlighted in the original study since 1972.
Industrial development: World GDP stood in real 2010 dollars at about $19 trillion in 1972 and has tripled to $57 trillion today. Average per capita incomes rose in real dollars from $5,000 to $8,100 today. Just to explore how incomes might evolve between 1972 and 2000, the researchers simply extrapolated the current growth, investment, and population growth rates to calculate GDP per capita for 10 large countries. They stressed these were not "predictions" but added that if one disagreed then one was obligated to specify which factors changed, when and why. A comparison of their extrapolations with actual GDP per capita (in 2010 dollars) finds U.S. GDP per capita $56,000 versus actual $44,000; Japan's per capita GDP was projected to be $120,000 versus actual $46,000; the now defunct USSR would be $33,000 versus Russia's $2,200; and China's per capita income was supposed to grow to $500, but was instead $1,200.
Population: The Limits researchers noted, "Unless there is a sharp rise in mortality, which mankind will strive mightily to avoid, we can look forward to a world population of around 7 billion persons in 30 more years." In addition, they suggested that in 60 years there would be "four people in the world for everyone living today." In fact, average global life expectancy rose from 60 to nearly 70 years. On the other hand, the global fertility rate (the average number of children a woman has during her lifetime) fell from about 6 per woman in 1970 to 2.8 today and continues to fall.
World population stood at 3.8 billion in 1972, which means that a four-fold increase in 60 years would have yielded a total world population of 15 billion by 2030. Even the latest U.N. high fertility population projection foresees about 9 billion by 2030. The U.N.'s low fertility variant yields a maximum world population of about 8 billion around 2050, falling back to 6 billion by the end of the 21st century. It turns out that the invisible hand of population control correlates very nicely with economic freedom.
Food supplies: According to the data from the Food and Agriculture Organization, global food production has more than tripled since 1961, while world population has increased from 3 billion to 7 billion. This means that per capita food has increased by more than a third. The latest figures [PDF] from the United Nations show that as world population increased by a bit over 10 percent between 2000 and 2009, global food production rose by 21 percent.
Arable land was proposed as one possible ultimate limit in the MIT model. In one generous model run, pollution was controlled and nonrenewable resources were essentially unlimited. The MIT researchers assumed that as long as industrial production continued to rise in the 21st century "the yield from each hectare of land continues to rise (up to a maximum of seven times the average yield in 1900) and new land is developed." Interestingly, since 1900 American corn farmers have already boosted yields nearly seven-fold from 26 bushels per acre to 166 bushels per acre. A 2010 article in Philosophical Transactions of the Royal Society B argued that available technologies could close the yield gap between first world farmers and developing country farmers even as the world warms. If this is done, the article concluded, "There is a good prospect that crop production will increase by approximately 50 percent or more by 2050 without extra land."
In 1972, the Limits researchers noted that about 1.4 billion hectares of land was being cultivated and projected that if current crop yields did not improve 3 billion hectares would be needed by 2000 to feed a projected population of 7 billion. The Limits analysts did note that if crop yields doubled (which they did not expect) that land devoted to producing crops would only increase marginally—which is what actually happened. The U.N.'s Food and Agricultural Organization reports that since 1960 cropland has expanded from 1.4 billion to 1.5 billion hectares [PDF].
Nonrenewable resources: Probably the most notorious projections from the MIT computer model involved the future of nonrenewable resources. The researchers warned: "Given present resource consumption rates and the projected increase in these rates, the great majority of currently nonrenewable resources will be extremely expensive 100 years from now." To emphasize the point they pointed out that "those resources with the shortest static reserve indices have already begun to increase." For example, they noted that the price of mercury had increased 500 percent in the last 20 years and the price of lead was up 300 percent over the past 30 years. The advent of the "oil crises" of the 1970s lent some credibility to these projections.
To highlight how dire the situation with nonrenewable resources was, the MIT researchers calculated how quickly exponential consumption could deplete known reserves of various minerals and fossil fuels. Even if global consumption rates didn't increase at all, the MIT modelers calculated 40 years ago that known world copper reserves would be entirely depleted in 36 years, lead in 26 years, mercury in 13 years, natural gas in 38 years, petroleum in 31 years, silver in 16 years, tin in 17 years, tungsten in 40 years, and zinc in 23 years. In other words, most of these nonrenewable resources would be entirely used up before the end of the 20th century.
They recognized that it was very likely that undiscovered reserves would be found and that technological improvements at extracting resources would occur, so just to be generous they made the same calculations with known reserves increased five-fold. Again at exponential consumption rates, they expected that after a gratuitous five-fold increase in resources there would now be only 15 years of aluminum left, eight years of copper, one year of mercury, nine years of natural gas, 10 years of petroleum, two years of silver, 21 years of tin, and 10 years of zinc.
Based on current consumption rates, the U.S. Geological Survey (USGS) in its 2012 mineral summaries report [PDF] estimates that the world has 130 years of bauxite reserves, which are used to produce aluminum. Similarly at current consumption rates, known copper reserves will last 43 years. Known lead reserves will last 18 years, although the USGS adds that identified lead resources equal 1.5 billion tons and that would mean a supply lasting somewhat more than 300,000 years. Mercury reserves are enough to another 48 years, but the USGS notes, "The declining consumption of mercury, except for small-scale gold mining, indicates that these resources are sufficient for another century or more of use." Current silver, tin, tungsten, and zinc reserves will respectively last 22, 19, 43, and 20 years more.
In 1972, the Limits researchers estimated known global oil reserves at 455 billion barrels. Since then the world has produced very nearly 1 trillion barrels [PDF] of oil and current known reserves hover around 1.2 trillion barrels, a 40-year supply at current consumption rates. With regard to natural gas supplies, the International Energy Agency last year issued a report [PDF] asserting, "Conventional recoverable resources are equivalent to more than 120 years of current global consumption, while total recoverable resources could sustain today's production for over 250 years."
Why does the horizon of mineral reserves never seem to go out further than a few decades? Basically because miners and technologists do not find it worthwhile to find new sources and develop new production techniques until markets signal that they are needed. How this process evolves is encapsulated by the USGS report which notes that in 1970 known world copper reserves stood at "about 280 million metric tons of copper. Since then, about 400 million metric tons of copper have been produced worldwide, but world copper reserves in 2011 were estimated to be 690 million metric tons of copper, more than double those in 1970, despite the depletion by mining of more than the original estimated reserves."
Environment: In most of the Limits model runs, the ultimate factor that does humanity in is pollution. In their model pollution directly increases human death rates and also dramatically reduces food production. In fact, as the world economy has grown, global average life expectancy has increased from 52 years in 1960 to 70 years now. It must be acknowledged that globally, pollution [PDF] from industrial and agricultural production continues to rise. But the model assumed that pollution would increase at exponential rates. However, many pollution trends have not increased exponentially in advanced countries.
Consider that since 1970, the U.S. economy has grown by 200 percent, yet the levels of air pollutants [PDF] regulated by the federal government have fallen by nearly 60 percent. For example, in both the U.S. and the European Union [PDF] sulfur dioxide emissions have dropped by nearly 70 percent since 1990. Recent data suggests that sulfur dioxide emissions even from rapidly industrializing China peaked in 2006 [PDF] and have begun declining. Earlier studies cite evidence for a pollution turning point income threshold (purchasing power parity) of around $10,000 [PDF] for demands to reduce this form of air pollution.
Another pollution concern was world fertilizer consumption that by 1970 had increased five-fold since World War II to 50 million tons. The Limits analysts noted that fertilizer consumption was growing exponentially "with a doubling time of ten years." The concern was the excess fertilizer running off of farms pollutes rivers, lakes, and the oceans. Presumably this doubling time suggests that since 1970, global fertilizer use should have increased to 400 million tons today. In fact, global fertilizer use is currently 150 million tons [PDF]. Around 1980 U.S. annual fertilizer use stabilized at around 20 million tons, yet U.S. agricultural yields have more than doubled since then.
Although, the Limits model measure for what counted as "pollution" was quite vaguely commodious, the Australian researcher Turner decides to use in his analysis carbon dioxide emissions from fossil fuels as his chief pollutant. The Limits analysis was actually pretty good at projecting atmospheric concentrations of carbon dioxide.
However, a recent Reason Foundation study reported [PDF] that the U.N.'s Intergovernmental Panel on Climate Change own analyses find that the scenario in which future temperatures increase the most is also the same world in which the greatest amount of wealth creation occurs. The result is that "by 2100 GDP per capita in poor countries will be double the U.S.'s 2006 level, even taking into account any negative impact of climate change." This means more technology and wealth will be available to cope with any problems that may occur from climate change.
One of the odder features of the Limits computer model is that it basically ignores one of the most robust feedback mechanisms in the world—markets and price systems. The modelers warn against placing our faith in the technological solutions, pointing to the collapse of the whaling industry as an example. They argue that improvements in whaling technology ended up destroying that industry. They completely overlook the fact that whaling occurred in an open access commons in which everyone has incentive to kill as many whales as possible to make sure that their competitors didn't benefit from them. Similarly, today wherever one identifies an environmental problem, one can be sure that it is occurring in the moral equivalent of an open access commons. In fact, the depletion of whales and rising price of whale oil encouraged entrepreneurs to seek new form of lighting; in this case, turning gooey crude oil into kerosene.
A lot of the data over the last 40 years fails to confirm the model's projections. But only time will tell if The Limits to Growth computer model is a true case of "pessimism in" generating "pessimism out."
Ronald Bailey is Reason's science correspondent. His book Liberation Biology: The Scientific and Moral Case for the Biotech Revolution is now available from Prometheus Books