6 What Comes After Growth: or demise and continuity
What comes after growth? The answers are determined by its subject and by the time spans under consideration. The fan of possibilities that opens with the end of growth is wider than one might think before trying to come up with a typology of postgrowth trajectories. In the biosphere, the extremes range from the nearly instant death of individual ephemeral microorganisms (and only a slightly deferred demise of many short-lived invertebrates) to the collective perpetuation of life. Apoptosis, the programmed death of cells, and obliteration on the individual organismic level (decomposition of bodies and recycling of compounds and elements) is just a part of astonishingly long perpetuations of species across long evolutionary time spans. In the case of bacterial and archaeal colonies, this span is of a near-planetary age. The ancestry of today’s microbes can be traced perhaps as far back as 3.95 billion years old to the sedimentary rocks in Labrador in which Tashiro et al. (2017) discovered traces of biogenic graphite.
And in cases of some higher animals this perpetuation has proceeded in such a conservative manner that the living species look and function very much as they did hundreds of millions years ago. The horseshoe crab (Limulus polyphemus, now on the list of vulnerable species) is one of the few well-known examples of this conservative evolution (IUCN 2017a). Few vertebrates live longer than humans. Long-lived animals include a few species of fish, sharks and whales and the giant tortoise, while the Greenland shark (Somniosus microcephalus) is most likely the longest-living vertebrate: the largest known animal was found to be 392 ± 120 years old (Nielsen et al. 2016). But the evolutionary direction of vertebrates in general, and mammals in particular, has been not only toward advanced behavior but also toward a longer life span, with our species achieving particularly impressive gains (Neill 2014).
The life cycles of many energy conversions, artifacts, and manufacturing techniques have normal distributions, ranging from nearly perfect bell-shaped fits to asymmetrical (or interrupted) declines whose progress is accelerated or slowed down by historical or national specificities. What comes after these instances of growth peaks is thus a fairly orderly decline whose rates closely mirror the gains that prevailed during the (sometimes not-so-distant) ascent. I will describe some notable example of both completed trajectories (when the product or practice has ceased to exist) as well unfolding declines in different stages.
Not too many people are interested in the demise of traditional wheat-harvesting processes or old steelmaking techniques. In contrast, studies of the decline and demise of populations, cities, societies, empires, economies, and civilizations are never out of fashion. Sudden or violent endings—including the Thera eruption (affecting ancient Minoan culture), the end of classical Maya states during the 8th and 9th centuries, or the collapse of the Romanov empire in 1917—attract particular fascination. The collapse of power structures has been a historical constant as long-enduring empires disappear or as artificially assembled states cease to exist—but the term economic collapse is a modern hyperbolic statement. Unlike empires and states, modern economies do not cease to exist, but they can experience periods of severely reduced output that causes hardships, starvation, and deaths, and that may require long periods of recovery.
In chapter 1, I cautioned against indiscriminate use of logistic curves in forecasting growth trajectories of artifacts, processes, and systems, and here is the apposite place to do the same as far as the forecasting of decline and demise based on a normal distribution is concerned. Given the ubiquity of life-cycle distributions that not only resemble a bell shape but often show a near-perfect correspondence between the actual numbers and ideal mathematical expressions, it is not surprising that this regularity seems to offer an excellent forecasting tool. The fate of compact cassettes or CDs could have been very accurately predicted by charting their future course right after their sales reached their peak. But caution is advisable even in those cases where the downslope of a normal curve has been well established and where further, and highly predictable, decline appears inevitable: following that trajectory could lead to some spectacular failures.
Perhaps the most consequential recent example of making this (understandable) error has been the mechanistic application of a normal distribution trajectory to US crude oil extraction. That application was made famous by M. King Hubbert (1956), who predicted, correctly, the peak of American crude oil output in 1970, and his forecast seemed to be of enduring value. If in 1980 you entered the US crude oil extraction data for the entire 20th century and calculated the total annual output expected in 2008 you would have made only about a 6% error as US crude oil extraction continued its long-predicted decline (figure 6.1, top). But the year 2008 turned out to be a turning point as subsequent output began to rise rapidly thanks to the innovative combination of horizontal drilling and hydraulic fracturing of America’s plentiful hydrocarbon-bearing shales.
US crude oil extraction forecast based on the 1900–1980 trajectory, and actual 1900–2018 performance. Data from USBC (1975) and USEIA (2019).
In 2015—when output, if it were to follow the normal curve decline, was due to be back to the level reached in 1940—US crude oil production rose nearly 90% above the 2008 low and missed the 1970 extraction record by only about 2%, resulting in a new bimodal extraction curve (figure 6.1, bottom). This great reversal of US oil-producing fortunes led not only to much-reduced imports but also (in December 2015) to the lifting of the 40-year-old ban on crude oil exports from the US (Harder and Cook 2015). Already by far the world’s largest exporter of refined oil products, the US has ramped up its sales of crude oil to such an extent that by November 2018 its exports surpassed its crude oil imports and put the country ahead of all OPEC exporters except of Saudi Arabia and Iraq. The US, seen just a decade ago as a steadily declining producer, would now qualify to join the organization of crude oil exporters!
And I will cite just another excellent example of the perils of forecasting based on a seemingly preordained normal curve, the growth of Greater London’s population, which can be charted reliably thanks to decennial censuses since 1801 (Morrey 1978; GLA 2015). Fitting the best trend to these data since their inception in 1801 to 1981 produces a near-perfect Gaussian curve (R2 = 0.991), peaking in the 1940s at less than 9 million people and indicating only some 2.1 million people by 2050 (figure 6.2, top). Adding three decades of population data to the trajectory (the era of great immigration and internationalization of the global city) changes the best fit to a four-parameter logistic curve with the inflexion point in 1877 and population stabilizing at about 8 million people during the first half of the 21st century (figure 6.2, bottom).
Population of Greater London: best fits based on 1801–1981 and 1801–2001 totals. Data from Morrey (1978) and GLA (2015).
Answers to the question of what comes after growth fall along a continuum ranging from fairly generic and highly regular outcomes (that are also readily quantifiable and, to a great extent, predictable) to many idiosyncratic, highly space- and time-specific results that present concatenations of many uncertainties and unknowns. Many growth phenomena belong to the first category and this allows us to make highly accurate forecasts, be they typical weights of mature domestic animals heading for slaughter, yields of staple grains to be harvested from intensively cultivated croplands, or heights of schoolchild cohorts. Uncertain outcomes are more common than is generally believed, and (as the recent US hydrocarbon revolution demonstrates) they can be encountered even after decades of consensus have pointed in the opposite direction and can have momentous consequences.
On the global level, there is no greater what-comes-after-growth uncertainty than the very fate of modern high-energy civilization with its still continuing population growth, high material requirements, and commensurately high environmental impacts. All of these long-lasting trends will have to end, deliberately or involuntarily. There is no possibility that we will be “saved” by an early coming of the Singularity or by an early terraforming of Mars. Such fictions make great news headlines but are worthless for dealing with civilizational challenges. Dealing with them is further complicated by continued profound uncertainties as we still have only approximate ideas about tolerable biospheric boundaries (Steffen et al. 2015) and hence of the timing of deliberate epochal shifts onto a new existential plane. If the biosphere is already in an overstressed mode, can we carry on for another five generations before correcting the course, or is the mid-21st century the latest time to act? Will we be able to come close to a genuine planetary equilibrium that would protect the biosphere from any global unraveling, or will our transformations come too late?
We have no way to predict the actual course, not only because of our imperfect understanding of complex interactions but above all because a wide range of future options has not been (yet) foreclosed. What will happen a generation or a century from now remains contingent on our intervening actions. Wrong decisions can accelerate decline and demise, while proceeding cautiously could greatly limit most (if not all) undesirable biospheric and social outcomes. Acting in radical ways could open new prospects for global civilization. What comes after the growth our civilization has experienced during the past two centuries will determine if we will prosper as a species not only during the next two centuries but perhaps for millennia to come.
Life Cycles of Organisms
For the simplest creatures—single-celled organisms without a nucleus (prokaryotes), including bacteria and archaea, another ancient domain present in many terrestrial and aquatic environments—rapid division perpetuates the species but is followed by the swift death of individual cells (excepting those that can remain cryptobiotic for extended periods of time). All prokaryotes divide rapidly (with division cycles measured often in minutes) and grow rapidly and hence the notion of their life span is fundamentally different from the existence of plants and animals. Most single-cell organisms live only briefly. The most abundant single-celled photosynthesizing organisms in the ocean, cyanobacteria belonging to the genus Prochlorococcus, live, on average, less than two days.
Pelagibacter ubique is an even more abundant microbe but it is a chemoheterotroph, it survives by consuming dead organic matter dissolved in the ocean, and it also has the smallest genome known for a free-living microorganism (Giovannoni et al. 2005). The aggregate mass (usually known as standing biomass) of single-celled photosynthesizers present in the ocean at any time averages only about 3 Gt C. But the rapid turnover of these phytoplanktonic cells means that the annual net primary productivity of marine photosynthesis (at 48 Gt C) is almost as high as that of terrestrial ecosystems (56 Gt C) that store perhaps as much as 200 times more phytomass (Houghton and Goetz 2008). Moreover, many prokaryotes—and a few other microorganisms, including Tardigrada, those remarkable miniature (0.1–1.5 mm) water bears (Jönsson and Bertolani 2001)—can enter cryptobiotic states where the capacity for life can be preserved for extended periods of time (up to 108 years) and in extreme environments as all metabolic processes get suspended inside protective spores (Clegg 2001; Wharton 2002; figure 6.3).
Tardigrada, one of the near-indestructible forms of life capable of cryptobiotic existence. Photo available at wikimedia.
The after-growth trajectory of many archaea and bacteria could be thus succinctly described as prompt postdivision death of individual cells but an unequaled longevity of species or communities (microbial communities often include complex assemblages of many archaeal and bacterial species, commonly also in extreme environments). The rapid regeneration of prokaryotes has preserved, in some cases in almost unchanged forms, the world’s oldest living species that appeared for the first time more than 3 billion years ago. The life cycles of photosynthesizing macroorganisms are very different. As already explained (in chapter 2), several tree species survive for more than 1,000 years but these oldest-living terrestrial organisms achieve that feat by being largely dead.
Most of the tree phytomass is locked as cellulose and lignin in structural polymers and in cell walls—but these tissues, indispensable for support, protection, and conduction of water and dissolved nutrients, are not alive. The radial extent of the cambial zone (where the tree growth is generated) is not easy to delimit due to the gradual transition between living tissues (phloem) and dead tissues (xylem). While some parenchymatic cells may remain alive for many months, years, and even decades, the share of living phytomass in the total standing forest mass is almost certainly no higher than 15% (Smil 2013a). The growth trajectory of trees could be thus succinctly described as a mass-scale production of short-lived cells required to ensure structural integrity of the organism and the extraordinary height and longevity of some species.
In plants, animals, and in humans, regenerating cells keep organs and organisms alive for extended periods of time even after the cessation of overall growth in mass or height—but some key organs have cells that last a lifetime and either are not renewed by any subsequent growth or have only a minimal renewal rate. The life span of cells in human tissues can now be accurately determined by measuring 14C levels in modern DNA (Spalding et al. 2005). Cells lining the gut last no longer than five days, red blood cells last four months—but the average age of an adult body’s cell is 7–10 years, and skeletal cells average about 15 years. Cellular regeneration keeps organs and organisms alive, and that includes many brain tissues—but analysis of brain cells taken from the adult occipital cerebral cortex showed 14C levels of their genomic DNA corresponding to the individual’s age and hence indicating the absence of postnatal cortical neurogenesis in humans.
14C studies have also made it possible to establish the age of cardiac muscle cells (cardiomyocytes) in humans (Bergmann et al. 2009, 2015). Nearly all cardiomyocytes were already generated in utero, their final number (3.2 × 109 ± 0.75 × 109) is reached just one month after birth and it remains constant during the entire human life span. The cells have a very low turnover that decreases with age, from the highest rate of 0.8% during the first decade of life to just 0.3% at age 75, and about 80% of the cardiomyocytes will never change after 10 years of age, no matter what the ultimate life span might be. In contrast, endothelial and mesenchymal cells increase into adulthood. The growth trajectory of heart muscle cells could be thus succinctly described as a near-complete acquisition of their total number before birth followed by growth in size until the end of puberty and then a steady decades-long functioning accompanied by very limited, and declining, regeneration before eventual death.
Jones et al. (2014) collected data on relative mortality and standardized the age axis to start at the mean age of reproductive maturity and to end at a terminal age when only 5% of adults are still alive. Their most surprising finding is that the predictable pattern of increasing mortality (and declining fertility) with age after maturity is not the norm, and that the relative mortalities follow a range of trajectories that include not only the predicted increasing trend, but also constant, humped, bowed, and even decreasing patterns for both long- and short-lived species. A barely rising trend during early maturity followed by a steep increase of relative mortality with advanced age is characteristic of humans (figure 6.4), and other species with a relatively steep mortality rise in the later stages of life include water fleas, guppies, mynahs, and lions.
Mortality curves for various species as functions of age. Simplified from Jones et al. (2014).
Trajectories for baboons and deer are far less steep, and mortality of 8–20 years old chimpanzees rises only slightly above the linear trajectory. Near-linear but much less steep increases also characterize mature mortality trends for the freshwater crocodile and common lizards, while relative mortalities for such different species as hydra (Hydra magnipapillata), red abalone, collared flycatcher, and the great rhododendron follow flat, or nearly flat, trajectories with increasing age. And those for red gorgonian, netleaf oak, desert turtle, and white mangrove show slight to pronounced mortality declines with age (figure 6.4). Lack of data for most species as well as for quantifying intraspecific variation precludes making any grand generalizations.
Animal growth is often followed not by any extended period of maturity but by sudden death. That is, of course, the norm for large numbers of animals whose life is cut short by predation and by the billions of birds and mammals we kill annually for meat. Recent annual global totals have been on the order of 60 billion chickens, 1.5 billion pigs, and 300 million heads of cattle (FAO 2018). Some animals are slaughtered long before they reach maturity. The life expectancy of wild pigs is as long as two decades, their typical natural lifespan is 10–12 years, and even in ecosystems with high rates of predation they can average half a dozen years—but animals reared for meat by intensive feeding are now slaughtered at 5–6 months of age, suckling pigs (to be roasted whole, cochinillo asado being a great Spanish favorite) commonly at just 4–5 weeks (Frayer 2013).
In contrast to the much-shortened life spans of domestic animals, the quest for longer human life expectancy has achieved some notable results. Humans are exceptionally long-lived when compared to animals of a similar body mass or to other primates. Pigs can live for up to 20 years, chimpanzee females almost 40 years, orangutans in captivity more than 60 years but the mean is about 40 years—but in 2015 the highest nationwide means for humans were 83.7 years in Japan and 83.4 years in Switzerland (WHO 2016), and in 2015 Japan had more than 60,000 centenarians (Japan Times 2015). What is remarkable about increasing human life expectancies is that once the vigorous organismic growth is over, the period of maturity in healthy individuals can now span decades with little apparent deterioration of basic mental and physical functions.
Some individuals can also maintain their mature weight (within small fluctuating bounds) for more than half a century, and extremes of physical exertion keep rising: 80 years for the oldest climber to reach the peak of Mount Everest, 85 years for completing a marathon in less than four hours (Longman 2016), 92 years for finishing it in less than six hours. The span of mental achievements has been also lengthening: using a large sample of Quebec professors, Gingras et al. (2008) showed that the average annual number of papers by active (publishing) professors declined only minimally between 50 and 70 years of age, and since 1950 the average age of all Nobel laureates rose by 13 years.
But long life after growth is far from universal. Combined life expectancies are below 60 years in at least 25 African countries, and Russian males have a shorter average life expectancy at birth not only when compared with males in China but even with males in India: between 2010 and 2015 the respective means were 64.6, 74.2, and 66.2 years (UN 2017). And every year, death still comes prematurely to tens of millions, preventing them from accomplishing their growth. Too many babies still die in the first year of their life: infant mortalities are nearly 100 per 1,000 live births in some countries of Saharan and sub-Saharan Africa (Mali, Chad, Niger, Angola), about 40 in India, and 20 in Egypt. In only one large populous country, Japan, there are now below 2, about the lowest level attainable in a modern society (PRB 2016). Nearly 5 million children now die every year before reaching their first year of life. And even in the richest countries there are hundreds of deaths among young adults (20–24) who had just reached their maturity, with the US male rate on the order of 1,400/million due mostly to accidents, overdoses, or suicides (Blum and Qureshi 2011).
Even without any specific diseases, lifelong stress on vital organs is particularly high. Normalized entropy stress (with rest of the body equal to 1) is 37 for the heart, 34 for the kidneys, 17 for the brain and 15 for the liver (Annamalai and Silva 2012). The heart is thus under the most severe stress and, not surprisingly, heart disease is the leading cause of death for both men and women (in the US accounting for roughly one in four deaths), with coronary heart disease being the most common type (CDC 2017). For these reasons and also due to other limits (already outlined in chapter 2), it is unlikely that we will see any further substantial increases in the human life span in the near future. One of the most convincing images supporting this conclusion was published by Marck et al. (2017). They plotted maximal ages at death of the oldest women and men, as well as maximal ages for the oldest male and female Olympians since the beginning of the 20th century. The first two trends have shown hardly any change during the 20th century, strongly suggesting that a plateau of around 115–120 years is the upper limit for human longevity (Marck et al. 2017). Maximal life spans of the oldest Olympians showed notable gains until the 1970s but have since leveled off at around 100 years.
Retreat of Artifacts and Processes
Although they attract no media attention and there are no new publications devoted to extolling their indispensability, we rely on a very large number of artifacts that have seen no fundamental change—be it in terms of their basic design or their outward appearance—ever since their optimized forms conquered their respective markets. Similarly, there are many industrial and manufacturing processes whose efficient operation relies on deploying such ancient inventions as sawing, grinding, polishing, casting, annealing, and welding. These periods of remarkable stability are often measured not only in generations (for example, electric current transformers) but in centuries (mass-produced screws) and even millennia; belt buckles are a good quotidian example in that enduring category—the lost wax casting process for metals has been known for more than 5,000 years. Of course, there have been many upgradings of materials and production processes but basic designs and sequences have endured because of their capability and reliable performance.
Many of these necessities of life are durable products with many of their market segments being quite limited (how many hammers will an average family buy in half a century?); others need to be made in vastly expanded quantities in order to meet new market needs. One of the best (and completely hidden) examples of new output growth of an old artifact has been created by the explosion of portable electronic devices that have to be charged. As a result, we have seen an unprecedented growth of the global stock of tiny transformers that step down the AC grid voltage (100 V in Japan, 120 V in North America, 220 V in China, 230 V in Germany) to 5 V DC for mobile phones and tablets and to 12 V–19 V DC for laptops: by 2017 their annual production was about 2 billion units (Smil 2017d).
This impressive growth did not require any change of the artifact’s basic design but it necessitated ongoing miniaturization—while many other long-established artifacts have seen only quantitative (output) growth and no qualitative changes. The long-term trajectories of such artifacts often have a fluctuating growth rate—for example, think of new waves of consumer products creating new demand for screws and other fasteners—but no annual peak of their output is imminent as irregular gradual slow growth is the norm. In contrast, many postgrowth trajectories of technical advances follow one of two similar courses of extended performance plateaus.
First, for decades or generations after reaching the limit of its growth, the upper asymptote can remain horizontal with hardly any attendant changes. The classic black Bakelite hard-wired rotary dial telephone combining the transmitter and receiver in the same unit was introduced during the late 1920s and the long rule of telephone monopolies ensured that hundreds of millions of its copies were added during the next four decades (figure 6.5). Push-button dialing was introduced only in 1963 and during the 1960s phones also began to change color and be available in many new shapes (Smil 2006b). The telephone was thus on an extended no-growth plateau, but once it left it, its development has seen some of the fastest sequential growth waves (portable phones, cellphones, smartphones) in modern industry.
Black bakelite phone. Image available at oldphoneworks.com.
Second, performances can dip a bit from the highest level and settle at somewhat lower ratings, as happened with oil tankers or with the capacities of largest commonly installed steam turbogenerators. The largest crude oil tanker was Seawise Giant, built in 1979, then enlarged to more than 560,000 dwt and relaunched after war damage as Jahre Viking (1991–2004), used as an off-loading unit off Qatar, and in 2009 sold to Indian ship-breakers (Konrad 2010). Building a tanker of 1 million dwt was technically possible but Seawise Giant remained an oddity and in 2015 there were only two ultra-large crude carriers (441,000 dwt) in operation. For a variety of reasons—docking and off-loading options, passages through canals, insurance costs, rerouting flexibility—most of the world’s oil is transported in very large crude carriers (160,000–320,000 dwt) that can take onboard between 1.9 and 2.2 million barrels of crude oil (USEIA 2014).
Similarly, after years of vigorous post-WWII growth, by 1965 the installed capacity of America’s largest steam turbogenerators reached 1,000 MW (Driscoll et al. 1964) and with electricity demand doubling every decade many utilities began to order units of 600–800 MW. But the subsequent slow-down of electricity demand and system considerations (as a rule, in order to maintain system stability in case of a sudden outage, the largest unit should not be more than 5% of overall capacity) led to declining modal ratings, with most post-1970 installations being in turbogenerators and gas turbines with capacities between 50 and 250 MW (Smil 2003).
New developments can sometimes bring very abrupt ends to further growth of well-established but suddenly outdated techniques, but in many cases the retreat of established artifacts and processes has followed normal-distribution curves. I will note here just a few important examples illustrating the Gaussian fate of energy extraction (coal mining), prime movers (farm horses and steam locomotives), mass-produced artifacts (music recordings), and industrial processes (open-hearth furnaces in American steelmaking). And to bring in entirely different, indeed the world’s most destructive man-made products, I will look at the rise and decline of nuclear warhead stockpiles in the Soviet Union/Russia.
We now have two completed coal production trajectories, with the end of the Dutch one preceding the British one by four decades (figure 6.6). WWII disruption aside, the Dutch trajectory resembles a blunted normal curve with a steeper decline resulting from a fairly rapid closure of the country’s coal mines once the abundant and inexpensive natural gas from the giant Groningen field became available during the 1960s (Smil 2017a). The complete British extraction curve (with accurate data going back to the beginning of the 18th century) reflects many national specificities of the world’s pioneering large-scale coal extraction. Major oscillations between the extraction peak in 1913 (when 1.1 million miners produced about 290 Mt from more than 3,000 mines) and the mid-1950s were the result of wars, strikes, and economic downturns.
Complete trajectories of British and Dutch coal extraction. Plotted from data in de Jong (2004) and DECC (2015).
Nationwide output was still above 200 Mt during the 1950s and 130 Mt in 1980, but a protracted coal miners’ strike in 1984 and switch to natural gas accelerated the fuels demise and by the year 2000 the total fell to only 31 Mt and 11,000 workers. The last British deep mine (Kellingley pit in North Yorkshire) was shut down in December 2015, ending more than half a millennium of British coal extraction (Hicks and Allen 1999; DECC 2015; Moss 2015). If it were not for assorted disruptive factors, the completed trajectory would be close to a slightly asymmetrical bell-shaped curve with the bulk of the output extracted between 1860 and the year 2000 (figure 6.6). Unlike British coal, British steel remains a going concern but its annual output has been reduced to around a third of the peak value and the trajectory conforms (with many expected fluctuations caused by economic downturns and wars) to a normal distribution (figure 6.7).
British steel output, 1900–2020. Fluctuating output reflects economic downturns, expansions, and wars and hence the normal curve is not a particularly close fit (R2 of 0.79). The production peak came in 1970, with the 2015 output below the level attained first in 1936. Data from https://
Innovation waves in steelmaking resulted in the formation of two asymmetrical bell-shaped curves: Bessemer furnaces (the first modern method of steelmaking introduced during the 1870s) quickly conquered the market by 1880 and, in turn, they were rapidly displaced by open-hearth steelmaking, which dominated the US industry until the 1960s (Smil 2016b). Departures from the normally distributed trajectories of their American output (fast expansion between 1910 and 1930, fast post-1970 decline) reflect the extraordinarily high demand for steel in early 19th-century America and the belated introduction of basic oxygen furnaces during the 1960s (figure 6.8). That displacement was completed in all Western countries as there are no operating open-hearth furnaces anywhere in North America, the EU, or Japan (WSA 2017).
Steelmaking transitions in the US, 1850–2000. From Smil (2005) and WSA (2017).
No animate prime mover has made a greater historical difference than the horse, and American historical data allow us to follow the rise of draft horses from fewer than 2 million animals in 1850 to the peak of 21.5 million in 1915 (in addition there were also some 5 million mules), followed by an expected decline to 10 million by 1940 and to just over 3 million by 1960 (USBC 1975; figure 6.9). Steam locomotives were introduced to the US soon after their English debut during the early 1830s. Their total reached 30,000 just before 1890 and then doubled in just two decades; the descent began during the late 1920s with the adoption of diesel engines, and by 1960 there were fewer than 400 operating machines (USBC 1975; figure 6.10).
Number of American draft horses between 1850 and 1970 conforms closely to the normal curve trajectory. Data from USBC (1975).
Number of US steam locomotives: a very good Gaussian fit for the nine decades between 1876 and 1967. Data from USBC (1975).
Car ownership continues to grow rapidly in all low- and medium-income Asian countries, but the total number of US passenger cars shows clear signs of saturation. If the trajectory were to follow the projected Gaussian fit, the country would have no more than about 65 million vehicles by the year 2100 compared to nearly 190 million in 2015 (figure 6.11). This decline could be further accelerated by convenient on-demand availability of future autonomous vehicles but I suspect that this innovation will make a substantial difference much later than is now widely assumed. But it is certain that car ownership declines will be much steeper in countries experiencing relatively fast population decline, above all in Japan.
The historical growth of the US passenger car fleet can be fitted quite well into a normal curve peaking around 2030. Data from USBC (1975) and from subsequent volumes of the US Statistical Abstract.
Turning to small consumer products, detailed US data make it possible to follow the waves of music recordings. Sales of vinyl records (singles and long-playing records (LPs) introduced in 1948) reached their peak in 1978 with 531 million units, declined to less than 10 million units by 1999 and 5 million units by 2004, but then revived a bit to 17.2 million records in 2016 (RIAA 2017). Cassette tapes were introduced in 1963 in Europe and in 1964 in the US; their sales peaked in 1988 and essentially ceased by 2005. Compact disc (CD) sales began in 1984, peaked in 1999, and by 2016 were below 100 million units as the last one of these three successive audio techniques, whose rise and retreat followed normal curves, became dwarfed by music downloading (figure 6.12). But its dominance was even more fleeting. American downloads peaked at over 1.5 billion units in 2012 before retreating to just over 800 million by 2016. Another unmistakable Gaussian trajectory has formed rapidly as downloading was replaced by streaming: by 2016 streaming revenues were more than twice as high than for downloading (RIAA 2017).
Successive normal curves chart the US sales of recorded music: vinyl records were displaced by cassette tapes, cassettes were displaced by CDs, and those were largely eliminated by music downloading and then by streaming. Plotted from data in RIAA (2017).
But in some cases commercial demise does not mean the absolute end of old techniques, processes, or machines: techniques linger, forming very long but barely noticeable asymmetrical right tails of normal distribution. Most surviving steam locomotives are now in museums—but some are still used for a few vacation fun rides. Even in affluent Western countries, some small- farms still use horses. And while the last open-hearth furnaces were shut down in Japan in 1980 and have been absent from the Western mills for a generation, the outdated process has lingered in the post-Soviet Ukraine, whose open-hearth furnaces still produced nearly half of all steel in the year 2000 and nearly 23% even in 2015 (WSA 2017).
The growth and decline of warhead numbers in the two nuclear superpowers shows two different patterns (Norris and Kristensen 2006). The Soviet/Russian trajectory forms a nearly perfect and pointed normal curve with a peak total of 40,000 warheads in 1986. Thanks to the dissolution of the Soviet Union, this was followed by an almost instant retreat to levels agreed upon by bilateral treaties. In contrast, the initial US growth during the 1950s was much faster than in the Soviet Union (reaching a peak of 32,040 warheads in 1967) and it was followed by stepwise reductions resulting in a highly asymmetrical distribution (figure 6.13). Of course, unlike open-hearth furnaces or steam locomotives, nuclear warheads are still around in large numbers. In 2017 the combined total of deployed American and Russian warheads was nearly 4,000 and, barring a most unlikely global nuclear disarmament, the right-hand tail stretches as far as we can see.
Totals of US and Soviet/Russian nuclear warheads, 1953–2010. Norris and Kristensen (2006) and Arms Control Association (2017).
Depending on how things get counted, long right-side tails are common with many once-ubiquitous artifacts. The demise of commercial wind-powered shipping has been absolute as there are no sail ships used for intercontinental trade and hence no shipyards fashioning sleek and fast clippers. But sail ships are still used by navies for training and they are increasingly popular for small-capacity warm-ocean cruising. And there is a long list of practices and products that have ceased to have any economic importance but survive as archaic exceptions: some people still make swords by hand in forges, there are still farriers—but how many people could now explain the meaning of such words (still very common at the beginning of the 20th century) as hames, martingale, or crupper?
Finally, before leaving the world of artifacts and processes, a paragraph about the end of Moore’s law. Its progress has been kept alive by steady advances in microchip design but it was always clear that this process has its physical and economic limits: at 5 nm the width of transistors in the latest IBM design is only about 20 silicon atoms, and new fabrication facilities now cost on the order of $10 billion (Rupp and Selberherr 2011; Rojas 2012; IBM 2017; Dormehl 2018). But this does not mean any abrupt end to Moore’s law, just notable deceleration of growth whose pace will depend on other adjustments we can make: introducing better algorithms and software, relying on more specialized chips, introducing new materials and new (3D) configurations and, more distant possibilities, deploying photonic and quantum computing. Feynman’s (1959) famous dictum of having plenty of room at the bottom remains in force.
Populations and Societies
Many low-density pastoral societies disappeared without leaving any long-lasting marks on the landscape where they herded their cattle, horses, or sheep. The demise of traditional agricultures—many based on deforestation (often, as in the Mediterranean and North China, on a very large scale), extensive terracing, and sophisticated irrigation arrangements, and producing increasing densities of permanent settlements—has been easier to follow. While carbon-dating of crops and artifacts may be fairly reliable and the times of decline or disappearance may be narrowly constrained, correctly identifying the reasons for gradual retreat or sudden collapse remains challenging. Controversies about the collapse of Mayan societies (Culbert 1973; Webster 2002) are an excellent example of these continuing uncertainties. In contrast, most sizable cities of early antiquity left at least some archaeological evidence, and during late antiquity and the Middle Ages the material record can be augmented by increasingly abundant written records, making an accurate reconstruction of urban trajectories possible.
Populations
While global projections remain uncertain, we are on much more solid ground in forecasting the long-term trajectories of many nations with very low fertility whose populations are not only rapidly aging but actually already declining. This demographic shift has many social and economic consequences. The emerging demographic deficit is expressed either as the overall dependency ratio—the quotient of the economically active population, usually aged 15–64 but in many countries now more accurately 20–69, and all dependents (0–15 and 65+)—or as the elderly dependency ratio (16–56/65+). In the EU the elderly dependency ratio will rise to 51% by 2050 as the number of economically active per people older than 65 years is halved from four to two. For the OECD countries as a group, the trends of incoming (20–24) and outgoing (60–64) working-age cohorts crossed in 2016.
Leaving aside the unlikely resurgence of fertilities, there are only two ways to reverse this demographic deficit and lower the dependency ratios: large-scale immigration of young people ready to join the labor force (a condition that is not necessarily met by some indiscriminate mass movement of refugees with disproportionate shares of small children, and women without skills and education) and working longer. The economic and social implications of aging societies and depopulating regions and countries will combine universal concerns with country- and region-specific challenges, and we cannot anticipate all possible outcomes.
The first category includes security of pensions, provision of adequate health care, coping with unprecedented numbers of mentally ill old people, and maintenance of expansive infrastructures, while more specific concerns arise from exceptionally rapid regional aging exacerbated by rising income inequality. Some of today’s assumption may turn out to be wrong. Working longer may not be enough to prevent the collapse of pension systems. And even if most people were willing to work past their normal retirement age, there would still be shortages of labor in health services as well as in occupations that (even if highly mechanized) will always require some demanding physical exertions. And a world dominated by the elderly may not be necessarily more peaceful (Longman 2010).
Declining fertility rather than changing mortality is the dominant factor in contemporary population aging (Lee and Zhou 2017) and in a concurrent rise of old age dependency ratios. But forecasts are also complicated by uncertainties associated with aging. For example, forecasts of the western European population above the age of 80 have 95% probability intervals of 5.5–20.7% by 2050 and 5–4.8% by 2100, a surprisingly large difference (Lutz et al. 2008). Japan has been at the leading edge of the massive aging wave that has begun to engulf nearly all affluent nations. Its average fertility fell from the post-WWII peak of 2.75 in the early 1950s to below the replacement level by the late 1970s, and the total population peaked at 128.08 million in 2008 and had declined to 127.7 million by October 2017.
This trend could be reversed only by turning to large-scale, Canadian- or Australian-style, immigration and admitting annually many hundreds of thousands of newcomers, still a highly unlikely choice but one whose adoption cannot be excluded in a more distant future. Forecasts of a future depopulation of Japan keep changing. At the beginning of the 21st century the official forecasts were for about 121 million by 2025 and only about 100 million people by the year 2050 (NIPSSR 2002). In 2012, the forecast was for 86.74 million by 2060 (with 40% of all people above 65), and in 2017 it was changed to 88.08 million by 2065 (NIPSSR 2017).
Using these projections to plot Japan’s population trajectory between 1872 and 2065 produces a good Gaussian fit (R2 = 0.965) and the protraction of that normal curve points to about 58 million people in 2100, equal to the total of the early 1920s. The aggregate decline is only a part of the aging-and-declining progression: the share of people 65 years and older will rise from less than 20% in the year 2000 to 35% by 2050 as Japan’s age-sex population structure will rest on a much-narrowed foundation and its cudgel-like profile will contrast with a barrel-shaped form by the end of the 20th century and with the classic broad-based pyramid of the 1930s and the early 1950s (figure 6.14).
Age-sex structures of the Japanese population, 1930, 1950, 2000, and 2050. Simplified from Smil (2007).
Perhaps the most astonishing outcome of Japan’s population aging is that by 2050 the country is likely to have more people 80 years and older than children up to 14 years of age and become the world’s first truly geriatric society (UN 2017). That would be the first time such a lopsided ratio had been experienced by any population during the long span of human evolution. A cascade of socioeconomic consequences resulting from these new demographic realities is self-evident—but no society is prepared to deal with it, particularly not one that still continues to reject any substantial immigration.
The European Union’s population is expected to grow until 2025 before it begins to decline, but populations of several member states, including Estonia, Latvia, Hungary, Bulgaria, and Romania, had already peaked in the 1980s or 1990s and (compared to 2015) they and other EU nations will see large aggregate declines by 2050. Populations are projected to decrease by 28% in Bulgaria, 22% in Romania, 14% in Poland and Estonia, 11% in Greece and Portugal, and by 8% in Germany (UN 2017). As Demeny (2003) noted, the fate of a depopulating and aging Europe could be contemplated with equanimity only if it were an island and not a continent under enormous population pressure from its high-fertility neighbors. The EU’s southern and southeastern hinterland includes 27 exclusively or predominantly Muslim states situated between the Atlantic Ocean and India. In 2015 their population was about 800 million compared to the EU’s 508 million, but their total fertility rate was 2.8 compared to the EU’s 1.6 (World Bank 2017).
Recent immigration from Europe’s Muslim hinterland is already adding up to the greatest influx of people the continent has experienced in more than a thousand years (Smil 2010a). And as shown by the sudden mass influx of migrants to Germany in 2016, by their inflow from North Africa to Italy and Spain, and by their record numbers in Sweden (in relative terms), this migration is already having major social and political consequences. Some EU member countries have refused to share the refugee burden and attitudes to immigration have hardened even in previously relative welcoming Germany, Italy, and France. But this might be only a prelude. By 2080 the EU’s population (without immigration) is forecast to fall by about 60 million people to less than 450 million (Eurostat 2015) while the total population of the continent’s Muslim hinterland will rise to more than 1.3 billion.
Notable population decline is also forecast for Russia (almost 10% drop between 2015 and 2050). Russia’s total fertility rate declined, with a temporary uptick during the 1980s, from about 2.8 during the early 1950s to only 1.25 at the beginning of the 21st century, and by 2016 it had risen only marginally to 1.46. When the Soviet Union disintegrated in late 1991, Russia’s population was about 140 million, by 2017 the total had risen to about 144 million (largely due to immigration from former Soviet republics), and the UN’s medium-variant forecast is for 132.7 million in 2050 (UN 2017). The most remarkable component of this process has been Russia’s dismal male life expectancy.
By the late 1960s belated health and nutrition improvements raised Russia’s combined life expectancy to within about four years of the EU’s average, but by the century’s end the gap widened to more than 12 years due to dismal male prospects. Between 2000 and 2005 Russia’s male life expectancy at birth was less than 59 years and although it rose to nearly 64 years between 2010 and 2015 it was (as already noted) still below not only the Chinese mean (slightly over 73 years) but even lower than in India (UN 2017). China’s challenges are different, as the country’s population prospects are rooted in what I called an uncommonly twisted foundation (Smil 2010a).
The traditionally excessive preference for sons in China became more intense due to the one-child policy and the country now has a highly aberrant sex ratio at birth, much higher than in other Asian countries with a high preference for boys. While the normal ratio of newborns is 1.06 (106 boys for 100 girls, as in Nigeria or Indonesia) and the global mean between 2010 and 2015 was 1.07, China’s ratio was 1.16 compared to India’s 1.11 and Pakistan’s 1.09 (UN 2017). In some Chinese provinces the ratios have been above 1.2, and a study found 20 rural townships in Anhui province with an incredible ratio of 1.52 (Wu et al. 2006). The consequences of this twisted growth are obvious even when the fundamental moral question of the mass-scale abortion of girls is ignored. The ensuing sex disparity prevents large numbers of men from ever marrying, and creates the well-known challenges of unattached lives, more crime- and violence-prone males and more common female abductions.
While populations will decline, their old age dependency ratios will grow. Between 1960 and 2015 the ratio rose from 15% to 22% in the US (4.5 workers for every old person), but it soared from just 9% to 43% in Japan, where it could reach (depending on demographic assumptions) at least 60% and as much as 82% (just 1.2 people working for every retiree) by 2050 (UN 2017). And thanks to China’s rapid aging, the 2050 old age dependency ratio may surpass 40% compared to just 13% in 2015, a shift for which the country’s economy is unprepared.
But, once again, these aggregates do not tell the whole story because the depopulation process in every nation is characterized by large regional disparities. Perhaps the two best examples of regional depopulation in affluent countries are the recent histories of the former East Germany (German Democratic Republic, the state which existed between 1949 and 1990), and Tohoku, the northeastern portion of Honshu, Japan’s largest island. With the exception of the capital and its surroundings, every region of the former German Democratic Republic is expected to lose people in the coming decades even as some of the country’s western regions keep on having slight overall gains. Japan’s Tohoku has seen migration of young people to the capital and other large southern cities for decades, and the region’s depopulation and aging were accelerated by the Fukushima nuclear disaster in March 2011.
And how will the trajectory of the global population look when seen in a very long-term perspective? When choosing a 16,000-year scale—from 8,000 BCE (roughly since the beginnings of agriculture and sedentary existence) to 8,000 CE—we know that the trajectory of human population will be a nearly flat pre-1500 line followed by a slow rise and then a rapid post-1850 ascent. But we do not know if there will be a partial or a similarly rapid (though not necessarily symmetrical) descent to a rate that could be sustained for millennia of a new, postgrowth civilization or if humanity will destroy itself or if it will be terminated by planetary or cosmic forces.
Cities
The history of cities after they reach their apogee falls into many categories distinguished by the longevity of post-peak existence and by the manner of demise (gradual decline vs. collapse or violent disbanding). Many ancient cities—including such famous early mud-brick and stone settlements as Mesopotamian Eridu, Uruk, and Lagash—continued to exist for millennia before their complete, or near-complete, destruction. Uruk and Lagash are now just mounds of rubble, Nineveh had some excavated walls rebuilt, and under Saddam Hussein’s regime Babylon had extensive reconstruction atop its ruins but then suffered further damage when the site was occupied by the US troops (AP 2006).
Some cities had managed to survive many successive waves of invaders before they almost completely disappeared. Merv (in today’s Turkmenistan) is an outstanding example of this endurance: it survived ancient Greek, Arabs, Turks, Mongols, and Uzbeks before it was razed by the Emir of Bokhara in 1789 (WHS 2017). Other cities had only relatively brief periods of glory followed by destruction or abandonment: Old and New Sarai, the successive capitals of the Mongol Golden Horde on the Akhtuba River (northeast of Astrakhan in the lower Volga watershed), are perhaps the best examples of this ephemeral existence, with the latter one razed in 1556.
The death of some cities begat instantly the existence of very different successors: invading Spaniards erased most of the Aztec culture and transformed Tenochtitlán to their image. But the city was left in the same place, with the same physical vulnerabilities, being too close to two volcanoes, Popocatépetl and Iztaccíhuatl, and prone to repeated major earthquakes whose impact is made worse by the quake-induced liquefaction of lacustrine soils left behind by Lake Texcoco (Tenochtitlán was on an island in the lake, which was later drained by Spaniards). The modern megalopolis (metropolitan area of more than 21 million people in 2017) in the Valley of Mexico is thus a direct descendant of Moctezuma’s lake city (Calnek 2003).
There are many examples of once-great cities that, after losing their dominance or formerly great importance within their respective nations, survived as major settlements, even continued to prosper within the new constraints. Two great eastern capitals are outstanding examples of that trajectory: Xi’an (ancient Chang’an), capital of 10 dynasties (between Qin, 9221–206 BCE, and Tang, 705–904), and Kyoto, capital of the Tokugawa shogunate that ruled Japan between 1600 and 1868 (Stavros 2014). With some 14 million people, Xi’an, the capital of Shaanxi province, is China’s ninth-largest metropolitan area, and besides its cultural riches it is now the country’s leader in the aerospace and software industries. Kyoto is Japan’s seventh-largest city, almost as large as Kobe, and although it remains famous for its temples, gardens and crafts, it also has headquarters of such major companies as Kyocera (electronics), Nintendo (games), and Omron (automation).
But some cities also offer excellent examples of growth followed first by protracted decline (even coming close to a complete unraveling) and, many centuries later, by a remarkable resurrection. Rome is certainly the most outstanding case in this revival category (Hibbert 1985). The steady urban rise of republican Rome during the five centuries up to the beginning of the common era was followed by three centuries of imperial dominance over the Mediterranean world. After the imperial capital was moved east (in 330), Rome entered a long period (more than a thousand years) of marginalized existence (Krautheimer 2000). A slow reversal of this long stagnation began only with the transformation of the city into the great capital of the Renaissance and Baroque architecture, but Rome’s population remained far below even the late imperial count until the late 19th century, when the city became the capital of the unified Italian state in 1870. Subsequent exponential growth brought its population to just above 2.8 million people by the early 1980s and after a temporary slight decline that total was reached again by 2016.
And at the beginning of the 21st century urban decline is, once again, a major European concern as it affects large regions of the continent, above all in the entire southeast (especially in Romania, Bulgaria, and Albania), in the Baltic States, in northwestern Spain and in Portugal, and in most parts of Germany. A remarkable interactive map provides a high-resolution record of this decline (and some continuing growth) between 2001 and 2011 for all EU countries as well as for Turkey (Berliner Morgenpost 2015). Cities and towns in Latvia, Lithuania, and Bulgaria are in particularly rapid retreat (some with annual population losses of more than 2%), while in the former German Democratic Republic only some suburbs of the capital (Berlin) have seen some expansion while depopulation has been affecting not only cities with less than 100,000 people and some industrialized urban centers with more than 200,000 people (Chemnitz, Halle) but even such old prominent cities as Leipzig and Dresden whose 2016 population totals were, respectively, about 20% and 12% below their pre-WWII counts.
Japanese depopulation (in some places going back to the 1970s) has been fastest among smaller municipalities, but Osaka, the country’s second-largest city, now has (within its designated boundaries) about 15% fewer people than in 1960, and Kitakyushu, the most prominent industrial city in the southern part of the country, had its population peak in 1980. The Russian population will decline by a quarter by 2050 compared to the early post-Soviet era (148 million in 1991, 111 million expected in 2050). And in 2015 the country had 319 single-industry towns established or expanded during the Soviet era that are now at risk of economic collapse, which would bring massive social dislocation as they contain about 10% of Russia’s population (Moscow Times 2015). Even some of the major cities have seen recent population declines, including Nizhny Novgorod and Samara, respectively the fourth- and fifth-largest urban areas in the country.
Societies, States, and Empires
The postgrowth typology of simpler societies that were not organized in states, of states (small and large), and empires (when the word is understood in its wider, sociocultural, meaning, not as a specific political entity) resembles that of cities. Not a few societies, states, and empires reached their apogees and then unraveled rapidly. Other polities lingered on after they ceased growing and their periods of decline became eventually much longer than the periods of their ascent. Others yet ceased to exist because of political intrigues and military conquest or survived as greatly diminished and largely impotent entities only to be restored or resurrected as circumstances changed in their favor and as they embarked on yet another territorial or economic expansion.
There is no shortage of examples of rapid societal demise, ranging from small islands to expansive empires. Rapa Nui (Easter Island) has presented perhaps the most persistent case of postgrowth mysteries. Why was a society that erected hundreds of moai, impressively sized stone sculptures, reduced to a small number of inhabitants? In a much-cited explanation, Diamond (2011) attributed this societal collapse to reckless deforestation of the island. This might have satisfied uncritical readers and TV audiences but Hunt (2006) identified the introduced Polynesian rats as the leading destructors of the Jubaea palm forests, and Puleston et al. (2017) estimated that dryland sweet potato cultivation could have supported more than 17,000 people. The eventual population collapse was not triggered by deforestation or starvation but by the introduction of infectious diseases (following the first contact with Europeans in 1722) and by enslavement. Middleton (2017) reviews this history among other modern myths of ancient collapses.
And sudden disintegration of states creating smaller national entities has been relatively common in recent history, often violent (Bangladesh seceding from Pakistan in 1971, the breakup of Yugoslavia, which started in June 1991 and resulted in the formation of five new states), sometimes peaceful (the dissolution of Czechoslovakia in 1993). But large size (of territory or population) has been no protection against sudden (or relatively sudden) unraveling, and histories of rising and collapsing empires and civilizations have had a prominent place in Western narratives. This interest is exemplified by the popularity and durability of such famous works as Gibbon’s influential analysis of the decline and fall of the Roman Empire (Gibbon 1776–1789), writings on the long-lost glories of imperial China, with its apogee under the emperor Kangxi (Spence 1988), and the unraveling of Western civilization narrated first in an influential fashion by Spengler (1918). Gibbon’s history and the demise of the imperium Romanum resonated strongly in late 19th- and early 20th-century Europe as the British and French empires reached their apogees.
Tainter’s (1988, 193) definition of collapse embraces all essential aspects of a sudden loss of existing arrangements:
Collapse is fundamentally a sudden, pronounced loss of an established level of sociopolitical complexity. A complex society that has collapsed is suddenly smaller, simpler, less stratified, and less socially differentiated. Specialization decreases and there is less centralized control. The flow of information drops, people trade and interact less, and there is overall lower coordination among individuals and groups. Economic activity drops to a commensurate level, while the arts and literature experience such a quantitative decline that a dark age often ensues. Population levels tend to drop, and for those who are left the known world shrinks.
Some inevitable exceptions aside, this definition describes well a variety of recent collapses. When applied to the collapse of the Russian Empire in 1917, the definition is almost perfect in all respects, except for the statement about less centralized control (true only in the earliest chaotic postrevolutionary months) and about a dark age ushered in by a decline in the arts (there was no shortage of new creative activities). In contrast, while the collapse of the Soviet Union (a highly centralized de facto empire of continental proportions) led to a large decline in economic activity and to population decrease, it ended the era of Communist control and propaganda and it brought personal freedoms and (despite continued centralizing tendencies) a much freer flow of information.
Popular as it has been, the idea of the death of a civilization that formed a powerful state or a long-lasting empire often does not correspond to reality as languages may survive: for example, even as the original states were obliterated centuries ago, Central America still has a million Maya speakers. More generally, as outlined by a leading proponent of social history, many cultural specificities and beliefs continue to exercise their influence (Sorokin 1957). Indeed, Pitirim Sorokin’s place of birth (the czarist empire) is an excellent illustration of these continuities. Russia’s Eastern Orthodox civilization appeared to be wiped out by the Communist revolution and by subsequent mass-scale persecution of that ancient religion, but today’s Russia abounds with symbols of its prerevolutionary past, state power is allied with the old religion (with its new magnificently restored, and rebuilt, churches), and the continuity of old concerns and fears is unmistakable.
Roman civilization, the foremost and an unceasing subject of collapse studies, is an even better example of remarkable continuity as some of its features were preserved by the Holy Roman Empire for nearly a millennium (962–1806). After a prolonged period of decentralizing turmoil and of two world wars, the gradual unification returned in the form of the European Union, set up symbolically by the Treaty of Rome (March 25, 1957) and eventually extending to every part of Europe that used to belong to the Roman Empire with the sole exception of Switzerland. More than a millennium and a half after the empire’s gradual takeover by outsiders, its enduring components range from governance and legal structures (the parliamentary systems of many countries include senates and powerful senators and are influenced by Roman civil law) to architectural styles, and from an immense linguistic heritage (not only in all Romance languages, as more than half of English words are Latin or Greek-via-Latin) to many cultural commonalities and shared attitudes.
In contrast, there is no shortage of historically ephemeral, albeit in the meantime enormously destructive, expansions that lasted only a few decades, or just a few years. Imperial demise can be swift and powers that appeared to be invincible may be returned to their pre-expansion borders, be severely curtailed, or even cease to exist in a matter of years. Napoleon’s European empire is among the best examples: it reached its territorial apogee before the invasion of Russia in July 1812 (directly controlling most of the western half of the continent, and indirectly much of the rest all the way to the western borders of the Russian and Ottoman empires). The empire suffered an enormous setback with the catastrophic retreat from Moscow—in December 1812 only about 10,000 men of the initial army of 422,000 crossed the Neman River of Lithuania—and then came the defeat at Waterloo; by March 31, 1814, Russian troops had occupied Paris (Leggiere 2007) and just three years after escaping from Russia Napoleon began awaiting his death on St. Helena.
Another prominent 19th-century example is that of the Taiping rebellion led by Hong Xiuquan. Starting in 1850 the followers of this self-styled Heavenly Son conquered a large area of South China, but by 1864 the rebellion was over (Spence 1996). And while British India had been in retreat for decades (perhaps even for nearly a century counting from the Mutiny in 1857), its final territorial unraveling in 1947 took place too fast, resulting in the enormous human suffering of hasty partition (Khan 2007). Japanese and German aggressions are the best modern examples of this short-lived (yet immensely destructive) growth-and-collapse sequence.
Japan’s quest for an empire began with the attack on Qing China in 1894–1895 (resulting in the annexation of Taiwan). The next step came with the annexation of Korea in 1905, in 1933 Japanese armies occupied Manchuria, and just four years later they began their conquest of eastern China. The attack on Pearl Harbor in December 1941 (figure 6.15) and occupation of most of Southeast Asia extended imperial control from Burma to the tropical atolls in the Pacific and from the cold and foggy Aleutians to New Guinea. The empire’s territorial peak was reached in early 1942, and Japan still controlled most of the conquered lands by the end of 1944—but devastating defeat came swiftly and Japan’s capitulation was signed on September 2, 1945: the imperial adventure thus lasted almost exactly 50 years (Jansen 2000).
The beginning of the end of Japan’s short-lived empire: USS West Virginia sinking in Pearl Harbor on December 7, 1941, after the Japanese torpedo and bomb attack. Photo from the Library of Congress.
And Hitler’s Thousand Year Reich lasted just 12 years, with its collapse proceeding much faster than its ascent (Shirer 1990; Kershaw 2012). Hitler became Chancellor in January 1933, German troops occupied Austria in March 1938, Czechoslovakia was dismembered in September 1938, and Poland was invaded in August 1939—but the territorial limits of German conquest came just three years later, in October1942, when the Reich controlled territory from the French Atlantic shores to the Caucasus and from Norway to Greece. All of that was gone within 30 months, by May 1945. That was both a precipitous and a complete collapse involving the demise of an initially powerful political entity as well as of the social attitudes and cultural norms dictated by its leadership as it tried to realize its pernicious vision of a new Germany.
And, quite remarkably, the demise of a vastly more extensive and incomparably more powerful Soviet empire was accomplished—aside from an aborted coup staged by the Communist old guard in August 19–21, 1991, by removing Gorbachev from power, when just three people died—without any violence. The Soviet Union’s European possessions (under de facto control since the end of WWII) regained their independence during a domino-like sequence that started in Poland in April 1989 and ended in late 1989: the Berlin Wall fell on November 9, the end of the Communist regimes in Sofia and Prague came on, respectively, November 10 and 28, and in Bucharest on December 22 (Fowkes 1995).
The Communist Party’s control within the Soviet Union had weakened steadily during 1990. Russia declared sovereignty with a limited application of Soviet laws in June 1990 and the inevitable unraveling was made legal on December 8, 1991 at a meeting in one of the most unlikely places to dissolve an empire, in a state hunting lodge in one the continent’s last primeval forests, in Belarus near the Polish border (Plokhy 2014). In contrast, as detailed in chapter 5, the much-studied fall of the Roman Empire was an affair that spanned centuries if we look at just the western part, or roughly a millennium when we include the prolonged weakening of the eastern (Byzantine) empire.
Given this range of outcomes, Arbesman’s (2011) analysis is an expected quantitative confirmation of the fact that imperial survival is remarkably idiosyncratic and distinct, and that the aggregate distribution of imperial lifetimes (including the entities whose rule spanned more than three millennia) follows an exponential distribution, with the rate of collapse of an empire independent of its age. Empires thus behave much like species whose probability of extinction is independent of their age but remains constant over the evolutionary span (Van Valen 1973). Thanks to the eponymous queen in Lewis Carroll’s Through the Looking Glass (who explained “Now, here, you see, it takes all the running you can do, to keep in the same place”), this reality is known as the Red Queen effect: longevity confers no advantage, survival demands constant adaptation, evolution, and proliferation merely to maintain one’s place against the onslaught of competing species—or adversarial groups, be they upstart nomadic marauders, neighboring, or even distant, states, or other already established empires.
The Ottoman Empire is perhaps the best example of protracted retreat and Poland is one of the most famous cases of national resurrection. The Ottoman Empire reached most of its peak extent of about 4.3 million km2 after about three centuries of logistic growth that followed the dynasty’s founding in 1299, but its subsequent decline lasted more than three centuries before its dissolution in 1922 (Gündüz 2002; Barkey 2008). Polish statehood was established during the 10th century and the territory controlled by the country’s rulers eventually expanded (after the Peace of Deulino in 1619, in the form of the Polish-Lithuanian commonwealth) from the Baltic to southern Ukraine and from Silesia to Smolensk (Zamoyski 2012). Less than two centuries later the country ceased to exist after three rounds of partitioning (1772–1795) among Russia, Prussia, and Austria but it was reconstituted after WWI and reemerged from the occupation and destruction of WWII in a much-changed territorial form after Stalin took away its eastern territories and rewarded it with new lands taken from Germany in the west.
But China provides by far the most notable example of a modern resurrection trajectory. After millennia of territorial expansion, it became the world’s second-largest state (surpassed only by imperial Russia) and during the early modern era (until the end of the 18th century) it was also the world’s largest economy. China’s retreat from this great power position began in 1842 with the British victory in the First Opium War. The country, unwilling to modernize along Japanese lines, was defeated in 1895 by Japan less than four decades after Japan began its modernization. Its last imperial dynasty unraveled in 1911 and that left the country adrift until 1949, when the Communist Party reconstituted unitary rule. Misrule is the correct term, as its policies caused the greatest famine in human history (1959–1961) that was preceded and followed by a madness of violence of many purges in the 1950s and the grotesquely mislabeled Cultural Revolution of the late 1960s and early 1970s (Schoppa 2010).
Only in 1979, with Deng Xiaoping’s rise to power, did China begin to reclaim its great power status and in 2014, when its GDP is expressed in purchasing power parity, it became, once again, the world’s largest economy (World Bank 2017). But the Chinese leadership should not ignore the lessons of post-WWII Japanese history: there is no comparable example of a country whose global standing was transformed so swiftly from a much admired socially dynamic paragon of rapid economic growth, manufacturing skills, and export superpower to a chronically underperforming economy and a fraying society beset by a multitude of challenges that have no readily deployable solutions. And Japan’s trajectory is of universal interest. We can forecast and assemble alternative scenarios but we will not truly understand the new dynamics of very low or no-growth economies with declining populations until they have existed for some time. Post-1989 Japan offers the earliest glimpse of the challenges of a new, postgrowth, society as the country’s new economic realities have become conflated with an apparently irreversible demographic retreat.
Japan as a Study in Retreat
Japan is thus the first major modern, affluent society to deal with unprecedented adjustments following high-growth decades of its economy and of its population. Many European countries will soon follow, but given the size of its economy, and of its importance for global trade, Japan is a particularly noteworthy case for studying what follows after reaching the peak of economic performance as well as the peak of population. In chapter 1 (in dealing with exponential growth) I noted the admiration that accompanied Japan’s rapid economic progress once it had recovered from WWII destruction. By the mid-1980s the country’s prestige and universal recognition of its technical excellence and economic dynamism appeared to bring ever closer Ezra Vogel’s forecast of Japan as the world’s number one (Vogel 1979).
The achievements of this new economic superpower elicited strong—and, predictably, quite opposite—sentiments on both sides of the Pacific. Perhaps most outspokenly, in 1989 a book entitled provocatively The Japan That Can Say No was published by Akio Morita, the co-founder and chairman of Sony Corporation, and Shintaro Ishihara, a prominent politician and that year’s candidate for the leadership of the country’s ruling Liberal Democratic Party (Morita and Ishihara 1989). Their book described the world dependent on Japan’s innovation in general and on its semiconductors in particular, scolded America’s inferior business practices, and praised Japan’s superior morals and behavior. And Americans did not confront this with determined defiance, as the country seemed to lose its collective nerve and as too many of its leaders began to believe that Japan could really become the world’s economic leader (Smil 2013b).
This fear found expression in reactions ranging from puerile smashing of Toshiba electronics by a few US congressmen to congressional demands that the Japanese automakers adopt “voluntary” export restraints. The US had also imposed 100% tariff on Japan’s electronic exports ranging from TV sets to computer disks, and in 1987 Congress began to fund Sematech, a new industry-government consortium uniting the 14 largest US semiconductor companies in a bid to prevent what appeared to be an inevitable Japanese dominance of that critical economic sector. But even a disinterested observer had to agree that Japan’s economic growth and dynamism of the 1980s were real enough, evident in the country’s large trade surpluses gained from selling a growing variety of high-quality manufactures, in its extraordinarily high saving rates, appreciating currency, and clean, safe and well-functioning cities.
As I wrote on the 20th anniversary of the Nikkei’s peak (Smil 2009, 6):
I will never forget the feel of Japan during the late 1980s, the peak of its power and, even more so, of its confidence and arrogance. In 1988 and 1989 there was no other place on the planet like Tokyo’s Ginza. One could see such congregations of polished large Mercedes SEs and SELs, such elegantly dressed crowds, and such free spending, driven by the worldwide profits of Japan, Inc. and the soaring purchasing power of its currency. The notion that in just a few years this time of shining opulence and unlimited prospects would become known as the time of deluded baburu ekonomi was not on anybody’s mind.
The fall was even more spectacular than the rise. During the 1960s Japan’s GDP grew nearly 2.5-fold, during the 1970s it expanded by another 50%, and during the 1980s, even as Japan became the world’s second-largest economy, it managed another 50% rise. And on December 29, 1989, Japan’s leading stock market index, Nikkei 225, reached its historic peak of 38,915.87 points, concluding a decade that saw it rising nearly six-fold, and almost quadrupling during the preceding six years. When the bubble economy began to deflate, the retreat was first mistaken for a temporary correction. The Nikkei fell below 30,000 in March 1990, then it recovered in just two months before a long slide set in. The index lost nearly 40% in 1990, fell below 15,000 in January 1995 and below 10,000 in September 2001 (Nikkei 225 2017; figure 6.16). As McCormack (1996) noted (hyperbolically, contrasting the previous adulation with new realities), it had taken less than a decade for Japan to move from number one to number zero.
Nikkei 225 average, 1950–2010. Graph based on data available at https://
A new century brought no fundamental relief. Before the global economic downturn, the Nikkei 225 managed to rise to more than 17,000 by July 2007, then it fell and after a fluctuating recovery it stood just above 20,000 at the beginning of 2019, still considerably lower than its peak of three decades ago! And the falling Nikkei was not an exceptional marker of retreat. During the first half of the 1990s, urban land prices in the country’s six largest cities dropped by 50% and by 2005 they were just 25% of the bubble-era level. Japanese manufacturing began shifting to China; Intel, not Hitachi or Mitsubishi, remained the world’s largest semiconductor maker and the industry’s global leader, and in 2016 the US companies dominated with 48% of the global market while Japanese production, the world’s largest during the late 1980s, was just 11% of worldwide shipments (SIA 2017).
Moreover, some Japanese companies have been failing to deliver the quality that has become expected for products coming from the country that based its economic ascent during the 1970s and 1980s on reliable, high-quality manufactures, a wave led by cars and electronics (Smil 2013b). Since the year 2000 a growing number of Japanese companies have sold substandard, even dangerous, products (millions of Takata airbags are perhaps the most notable entry in the latter category) and admitted to falsifying inspection data. These companies have included Toray Industries (textiles and chemicals), automakers Nissan and Subaru, and most notably Kobe Steel and Mitsubishi Materials, whose parts and equipment are used worldwide in airplanes, trains, cars, and electricity generation plants (Wells 2017). GDP growth fell sharply, adding less than 12% during the 1990s and less than 8% during the first decade of the 21st century (World Bank 2017).
And the long-term (1870–2015) logistic fit of GDP growth (values in PPP in 2011 international dollars; see figure 5.29) indicates that the country faces a long period of minimal or no economic growth. Japan’s economic weakness soon translated into notable social shifts, including an obvious rise in homelessness, widespread loss of previously standard lifelong employment, and a decline in labor participation among young workers. And just as the worst post-WWII global economic crisis was coming to its end, Japan was hit by the Tohoku earthquake, a massive tsunami, and a catastrophic failure of three nuclear reactors in the Fukushima Dai-ichi plant in March 2011. This was a setback with enormous social and economic consequences for the region, where even before those catastrophic events the only local employment in many small villages in the interior of Fukushima and Iwate provinces was to grow some vegetables, produce artisanal charcoal, and make small wooden manufactures.
Rapidly changing governments, whose promises of reforms keep falling far short of Japan’s enormous challenges, have not offered any effective solutions. The unique combination of competitive advantages that propelled the country during the 1980s will never return and, perhaps most importantly, the post-1989 economic retreat lasted for so long that it became conflated with Japan’s demographic decline (figure 6.17). Not only is the country aging rapidly—now clearly on its way to becoming the most geriatric society (NIPSSR 2002)—but after a decade of minimal growth its population began to decline and, as already noted, by the fall of 2017 it was down to 127.7 million people, a loss of nearly 1.3 million in six years (SB 2017b), and in 2018 the population loss reached a new record of 449,000 people.
In a photo taken in April 2009 an old man looks from the top floor of Tokyo’s municipal building on the sprawling city: during his lifetime the country rose from defeat and devastation to become a respected, even feared, economic superpower but almost immediately began its gradual economic and demographic retreat. Author’s photo.
The Japanese population, the 11th-largest in 2017, might just squeeze into the top 30 by 2100 (UN 2017). The long-term impact of this demographic retreat may have a relatively limited impact on Japan’s manufacturing. Cohorts of young people best suited for this work will be declining, but significant shares of production capacities have been already moved offshore as Japan followed (with the lag of 10–20 years) the deindustrialization trend so evident in the EU and the US, and further progress of robotization should keep productivity rising: Japan has pioneered this automation trend, it already deploys by far the largest number of industrial robots, and several companies remain among the leading global enablers of plant automation.
But there will be no way to avoid adverse effects on Japan’s admirable infrastructure of public transportation, on the country’s food production, and on its health care. Japan’s dense but aging transportation network will continue to experience high rates of use and hence it will require meticulous maintenance, an imperative perhaps best illustrated by the fact that during peak travel times rapid trains, traveling at up to 300 km/h, leave Tokyo’s main station at intervals as brief as three minutes, and have been doing so with an annually averaged delay of less than one minute (Smil 2014a). Maintenance and reconstruction of transportation networks has become highly mechanized but it still entails plenty of physical exertion not suited for an aging workforce.
The average age of Japanese farmers has surpassed 67 years, nearly all rural areas have been experiencing progressive depopulation (with only a handful of elderly residents left in many villages), the small sizes of typical farms are not suitable for mechanized operations deploying large field machinery, and the country already has the lowest food self-sufficiency rate among all major economies. When measured in terms of overall dietary energy supply, Japan’s self-sufficiency rate (more than 70% in 1965) is now just 38%, lower than in such import-dependent countries as Switzerland or South Korea) and in an entirely different category than more than self-sufficient US, Canada, or Australia (Smil and Kobayashi 2012; Japan Press Weekly 2018). And the shortages of health-care personnel and other workers, bound to get more acute as the population ages, seem to have no easy solution. A new bill passed in 2018 opened the way for the formal acceptance of foreign workers but getting permanent residency will remain difficult.
The aftermath of Japan’s ascent has been so poignant because it was preceded by a vertiginous rise and because the contrast between before and after became so stark within a single decade. Akio Morita died in 1999, spared from seeing his Sony falling from a globally dominant maker of admired electronics to a troubled underperformer. In early 2019 its stock value was only about a third of its peak level of February 2000; the company has cut its labor force repeatedly, lost its high credit rating, and has not had a globally successful product for more than a decade. Ishihara became a long-serving governor of Tokyo (1999–2012), lost his re-election bid as a member of parliament, and kept making provocative statements. But the notion of an assertive Japan dictating its own terms and saying no to the world in general, and to the US in particular, seems now quite risible.
And yet it is not easy to answer the question of what has come after growth in Japan, and what will follow. Many labels—stagnation, chronic retreat, gradual unravelling, creeping decay, spreading shabbiness, return to more realistic expectations—could be used to describe the post-1989 economic, social, and demographic trajectories, but such descriptions do not amount to the complete verdict. At the time when the country was greatly admired for its high economic growth and manufacturing prowess, its housing conditions remained poor in comparison to other affluent nations. That situation has actually improved a bit during the last generation even as a new concern—rising numbers of abandoned properties—becomes a common reality: about 10 million houses and apartments were empty in 2018, and the total is expected to rise to 20 million by 2030.
Looking back, what stands out above all is not any particular statistic of decline but the contrast with the past and expected performance. French or British economic problems seem unremarkable because during the 1980s nobody thought about France or the UK as the future #1, nobody saw Italy as the paragon of economic dynamism and astonishing inventiveness. And although it may now be seen as an economic and demographic has-been, Japan still remains the world’s third-largest economy (no matter if the total is expressed in nominal or purchasing power parity monies) and its per capita (PPP-adjusted) GDP is about the same as in France or the UK (World Bank 2018).
This is what I wrote on the 20th anniversary of the Nikkei 225 (Smil 2009, 5–6):
Nations rise and fall, and they stay in positions of strategic dominance or economic ascendance for different (and unpredictable) periods of time … Japan’s quest for the leading role in the global economy lasted roughly 40 years … and in its final phase it involved an economic confrontation with the United States that engendered a great deal of exaggerated confidence in Japan and deep concern and self-doubt about the nature of America’s resilience and the effectiveness of its response … As it moves into its new (and truly uncharted) demographic and economic era, Japan will not implode and turn into a dysfunctional polity. If the past two decades are any guide, it may not manage its retreat brilliantly, but I believe it will find ways to deal with its new challenges without causing any grave perturbations on the global scene … and without losing its deserved status as a well-functioning society.
Coming decades will show to what extent I got it wrong.
Economies
Temporary setbacks—often relatively short-lived, sometimes lingering for many years—have been recurrent features of modern economic growth. Among the major economies the greatest output declines of the 20th century have included a 26% reduction of US GDP (in constant monies) between 1929 and 1933 caused by the global recession; a 51% decline of Japan’s gross national product (again, in constant monies) between the pre-WWII peak in 1939 and the postwar low in 1946; and Germany’s 29% drop between 1944 and 1945 (Harrison 2000). Recoveries from such declines took years: the US regained the 1929 GDP level in 1936; Japan surpassed its prewar economic performance only in 1953.
But no modern economy has gone into such a long, uninterrupted decline that it could be seen as a new unprecedented trend, and even stagnating economies had their poor performances interspersed with periods of growth. When measured in national currency, Japan’s economy more than tripled between 1975 and 1995 (from ¥153 trillion to ¥512.5 trillion) while between 1995 and 2015 it grew by less than 4% (to ¥535.5 trillion), an enormous downturn from a steep linear rise to a generation-long stagnation (World Bank 2017). But even during those two decades Japan’s stagnating and fluctuating economy declined during nine years and grew slightly in 11 years, avoiding uninterrupted protracted retreat.
What lies ahead for major countries and for the world economy? Some economists reconstruct historic trajectories and see a logistic curve forming—but are content to leave it at that. For example, Boretos (2009) fitted global economic growth into a logistic curve whose full life cycle is about two centuries, eventually reaching saturation level by the beginning of the 22nd century. But there is no talk about saturation in mainstream economic publications, which all assume that the future will not be that different from the past as endless human ingenuity will be able to support many generations of growth.
For decades economists have been engaged in constant forecasting, and global, regional and national versions are now offered by all leading international organizations, including the International Monetary Fund, OECD, United Nations, and the World Bank. Near-term forecasts (1–4 years) of national GDP and of regional and global economic product are the most common, but some forecasts now extend to mid-term (5–10 years), and a few of them look ahead for more than a generation. The World Bank offers short-term GDP forecasts for all of its member countries, while the OECD’s latest forecast is simply the continuation of an exponential curve with an annual growth rate of 2.5%: starting with the global economic product of about US$11 trillion (in 2010 PPP monies) and reaching about $76 trillion in 2018, it brings the total to US$221 trillion by 2060.
How long will a similar trajectory hold: for another decade, another century, or another millennium? Of course, most economists have a ready answer as they see no after-growth stage: human ingenuity will keep on driving economic growth forever, solving challenges that may seem insurmountable today, especially as the techno-optimists firmly anticipate wealth creation progressively decoupling from additional demand for energy and materials. Mokyr (2014) has nearly unlimited confidence about the coming flood of near-miraculous transformations ranging from genetically modified crops (that will withstand rising global temperatures, manufacture their own nutrients, and protect themselves against insects) to a revolution in material science “that may make the synthetic substances of the twentieth century look like the Stone Age by comparison.” In June 2018 a special innovation issue of Spectrum gathered some of these accounts of coming near-miracles—but it also carried my critical (not so fast!) riposte (Smil 2018b).
Such cornucopian views have been further potentiated by many claims about the impending arrival of omnipotent artificial intelligence. A growth of computerization, robotization, and the rising capabilities of artificial intelligence are expected to bring a massive elimination of existing jobs—with the share of occupations at risk of such a takeover as high as 50% (Frey and Osborne 2015)—but to allow continued economic growth. But how will the robots secure the raw materials for their production, how will they be energized? Will the robots organize their own supply infrastructures, their extraction of metals and minerals? Will they design and put in place their own generation of renewable electricity and its long-distance high-voltage transmission, transformation, and distribution?
We are already near global saturation with devices that are essentially powerful portable mini-robots: every mobile phone is a computer whose processing power is orders of magnitude higher than that of the stationary devices of two generations ago—but its ownership is now in billions and more than 1.5 billion of these complex artifacts (built of aluminum, plastics, glass, and precious metals) are now discarded every year. Obviously, such trends cannot continue on a planet that is expected to accommodate some 10 billion people before the end of this century, and hence looking at what might come after the economic growth is not just a matter of fascinating speculations, it should be a key concern as we think about extending the life span of modern civilization.
Decoupling economic growth from energy and material inputs contradicts physical laws: basic needs for food, shelter, education, and employment for the additional billions of people to be added by 2100 will alone demand substantial energy flows and material inputs. True, those inputs will have lower relative intensities (energy/mass, mass/mass) than today’s average rates—but the absolute totals will keep rising (with continued population growth) or will moderate but remain substantial. Ward et al. (2016, 10) confirmed this truism when they used historical data and modeled projections to conclude “that growth in GDP ultimately cannot plausibly be decoupled from growth in material and energy use, demonstrating categorically that GDP growth cannot be sustained indefinitely.” This makes it highly misleading to advocate any growth-oriented policies assuming that such a decoupling, and continued GDP growth, is possible.
And it is similarly misleading to talk about any imminent practice of circular economy. Modern economies are based on massive linear flows of energy, fertilizers, other agrochemicals, and water required to produce food, and on even more massive energy and material flows to sustain industrial activities, transportation, and services. Circularization of the two key flows is impossible (reusing spent energy would require nothing less than abolishing entropy; reusing water used in cropping would require the capture of all evapotranspiration and field runoff), and (with the exception of a few metals in some countries) high-intensity (>80% of total flows), mass-scale recycling of materials (above all construction waste, plastics, and electronic waste) remains elusive.
Daly (2009) summed up the three conditions that would allow continuous economic growth on the Earth: if the economy were not an open subsystem of a finite and nongrowing biophysical system; if the economy were growing in a nonphysical dimension; and if the laws of thermodynamics did not apply. But none of these realities can be evaded, circumvented, or substituted by other arrangements—and hence it is easy to side with Kenneth Boulding who noted (not sparing his fellow economists) that “Anyone who believes in indefinite growth in anything physical, on a physically finite planet, is either mad or an economist” (quoted in US Congress 1973, 248).
Boulding was also one of the early proponents of new economic thinking when he introduced his distinction between the “cowboy economy” and the “spaceman economy”:
the cowboy being symbolic of the illimitable plains and also associated with reckless, exploitative, romantic, and violent behavior, which is characteristic of open societies. The closed economy of the future might similarly be called the “spaceman” economy, in which the earth has become a single spaceship, without unlimited reservoirs of anything, either for extraction or for pollution, and in which, therefore, man must find his place in a cyclical ecological system which is capable of continuous reproduction of material form even though it cannot escape having inputs of energy.
The difference between the two types of economy becomes most apparent in the attitude towards consumption. In the cowboy economy, consumption is regarded as a good thing and production likewise … By contrast, in the spaceman economy, throughput is by no means a desideratum, and is indeed to be regarded as something to be minimized rather than maximized … This idea that both production and consumption are bad things rather than good things is very strange to economists … (Boulding 1966, 7–8)
Questions about the costs of economic growth and its further desirability and arguments in favor of a steady-state economy began to appear during the 1960s and 1970s (Boulding 1964, 1966; Mishan 1967; Daly 1971) and the period’s most widely publicized, and the most influential analysis of what will come after economic growth was The Limits to Growth (Meadows et al. 1972). This short study was a slightly modified version of Jay Forrester’s work on dynamic systems (Forrester 1971) and it modeled global interactions of population, resources (including energy), industrial and food output, pollution, investment, and health based on historical values for the period between 1900 and 1970. The “standard” world model run assumed no major changes in the existing world system and it predicted that both food supply per capita and industrial output per capita would peak shortly after the year 2000.
This was to be followed by fairly steep declines
as resource prices rise and mines are depleted, more and more capital must be used for obtaining resources, leaving less to be invested for future growth. Finally investment cannot keep up with depreciation, and the industrial base collapses, taking with it the service and agricultural systems, which have become dependent on industrial inputs (such as fertilizers, pesticides, hospital laboratories, computers, and especially energy for mechanization). For a short time the situation is especially serious because population, with the delays inherent in the age structure and the process of social adjustment, keeps rising. Population finally decreases when the death rate is driven upward by lack of food and health services. (Meadows et al. 1972, 124)
Overshoot and collapse caused by nonrenewable resource depletion were inevitable outcomes and while the report concluded that the exact timing of these events was not meaningful, it was certain that growth would stopped well before the year 2100 if no major adjustments were made. Because I knew the programming language used to build the model (Forrester’s DYNAMO) I deconstructed the model line by line (not a very difficult task, as their model of the world fit into fewer than 150 lines) and quickly realized the number of indefensible simplifications and misleading assumptions. I still remember my surprise when I saw such key variables as Nonrenewable Resources and Pollution—as if it were possible to lump together the enormous variety of mineral resources (ranging from relatively abundant and highly substitutable minerals to unsubstitutable and fairly rare elements) and all forms of pollution (lumping short-lived atmospheric gases with long-lived radioactive wastes) into single bundles interacting with other complex variables.
A 30-year update of the report (Meadows et al. 2004) left the basic finding unchanged: humanity is in overshoot and the ensuing damage and suffering could be greatly reduced through wise policies. Another retrospective analysis found the standard model to be on track after 30 years (Turner 2008)—but how that can be asserted by actually quantifying the just noted disparate bundles of variables is hard to understand. The original report, as well as all sequels inspired by it, emphasized the obvious (exponential growth is impossible, biospheric flows and capacities are finite), but the grossly simplified modeling approach is not a nuanced realistic analysis of new global complexities—but essentially an exhortation, a call for change based on some solid evidence and some dubious assumptions. As a result, some of its conclusions are unexceptional, and others are questionable.
The report was followed by many kindred inquiries into the capacity of the biosphere and the availability of natural resources to support further economic growth: some were of a cornucopian nature, seeing few limits to further growth (Simon 1981; Simon and Kahn 1984); others, reflecting concerns about the Earth’s carrying capacity, led to the rise of a new discipline of ecological economics (Daly 1980; Costanza 1997; Daly and Farley 2010; Martínez-Alier 2015). This has eventually led to advocacy of not just economies without any growth but ones deliberately trying to reduce overall economic output, a shift awkwardly labeled as de-growth. Book titles convey these sentiments: Living within Limits (Hardin 1992); Beyond Growth (Daly 1996); Prosperity without Growth (Jackson 2009); From Bioeconomics to Degrowth (Georgescu-Roegen and Bonaiuti 2011); The Economics of Enough (Coyle 2011); Degrowth: A Vocabulary for a New Era (D’Alisa et al. 2014). In reality, there are no economies embarking on such paths.
As already noted, since the 1990s there have been also many studies of the limits to the growing extraction of mineral resources in general and to an imminent arrival of peak global oil production in particular (Deffeyes 2003)—and, given oil’s importance in the global economy, of inevitable and permanent economic downturn. I labeled this wave a new catastrophist cult, and wrote that the proponents of imminent peak oil “resort to deliberately alarmist arguments as they mix incontestable facts with caricatures of complex realities and as they ignore anything that does not fit their preconceived conclusions in order to issue their obituaries of modern civilization” (Smil 2006a, 22). More than a decade later, the global output of oil keeps on slowly rising and world oil prices remain relatively low.
And all of these concerns have been made more pressing by new forecasts of continued population growth that have not confirmed the earlier conclusions that the global population was unlikely to surpass 9 billion, and predicted a total of 9.7 billion by 2050 (UN 2017). A much-publicized approach to address these matters focused on the notion of sustainable development. The term entered the public discourse with the release of the Report of the World Commission on Environment and Development: Our Common Future, widely known as the Bruntland Report after the former Norwegian prime minister who chaired the commission that prepared it (WCED 1987).
Ever since, the adjective has become one of the most misused descriptors of desirable human actions. The report’s definition of the process is exceedingly loose:
Sustainable development is development that meets the needs of the present without compromising the ability of future generations to meet their own needs. It contains within it two key concepts: the concept of “needs,” in particular the essential needs of the world’s poor, to which overriding priority should be given; and the idea of limitations imposed by the state of technology and social organization on the environment’s ability to meet present and future needs. (WCED 1987, 41)
This leaves all key variables undefined: what are “the needs of the present”? Do they correspond to American, EU, Japanese, Bangladeshi, or Congolese expectations or to some average concocted by a committee? Even essential needs are arguable: they may be easily defined in terms of the nutrition required to meet adequate physical and mental growth (so much of total energy, so many grams of the three macronutrients, and so many fractions of a gram of micronutrients), but it is much more difficult to define them in terms of rewarding employment, adequate living conditions, widely available education, and opportunities for personal development and leisure.
Moreover, the report made it clear that its goals are global:
Thus the goals of economic and social development must be defined in terms of sustainability in all countries—developed or developing, market-oriented or centrally planned. Interpretations will vary, but must share certain general features and must flow from a consensus on the basic concept of sustainable development and on a broad strategic framework for achieving it.
But given the existing disparities between affluent, middle-income and low-income countries, it is difficult to see any universal agreement either on “certain general features” or on “a broad strategic framework” for achieving sustainable development.
Concerns about excessive global warming (a rise in the average tropospheric temperature surpassing 2°C) further strengthened the arguments in favor of limited or “sustainable” growth, and to many it appeared that the deepest post-WWII economic crisis in 2008 and 2009 was the unplanned but unsurprising beginning of global economic retreat. Heinberg (2010, xv), argued that “a good case can now be made that the year 2007 … was indeed the year, if not of ‘peak of everything,’ then at least of ‘peak of many things’ ” and subtitled his book Waking Up to the Century of Declines. He explicitly listed the “zeniths” of worldwide economic activity and global energy consumption, and peaks of crude oil output and of worldwide shipping.
But all of those zeniths have been already much surpassed, with most increases following the temporary economic downturn of 2008 and 2009 being actually fairly impressive. By 2017 global economic product was 60% higher than in 2007 (IMF 2017), by 2016 global primary energy consumption was 14% higher than in 2007, crude oil supplies rose by 11%, and the total of all seaborne cargoes was 25% higher (UNCTAD 2017). Obviously, this cannot be taken as a firm indicator of prospective achievements—but it is yet another proof of the futility of quantitative forecasting of complex affairs. At the same time, there is no doubt that since 1973 (when the unprecedented period of rapid post-WWII growth ended) the world economy has become impressively more energy efficient and relatively less material-intensive—while continuing population growth, further increases of consumption in affluent countries, and fast economic advances in Asia in general, and in China in particular, have translated into relatively strong absolute global growth in both energy and material requirements.
In relative terms (per unit of economic product) the global economy has shifted in the direction of greater sustainability but in absolute terms it has shown no tendency toward deliberately slower growth, and degrowth remains a cherished topic for ecological economists, not a guiding principle for any companies or governments. As a result, we can only speculate when and how we might be able to put an end to material growth and forge a new society that would survive without worshipping the impossible god of continuously increasing consumption: no country has committed to following such a path. Two generations after the concerns entered the public domain, the economic orthodoxy still does not have any more rational model to follow than the one of continuous growth, with special admiration reserved for high rate-gains such as those recorded by China during the first decade of the 21st century. Indeed, the worship of eternal growth has in some ways intensified because we are now promised that truly miraculous solutions will be provided by technological change that will soon reach the unimaginable Singularity, the result of an “exponential growth in the rate of exponential growth.”
A persevering reader of this book—replete with facts and arguments about confined growth, constraints and limits—might have some doubts regarding the likelihood of this specific outcome and hence it might be in order to repeat the key conclusions of Kurzweil’s forecast that were cited in this book’s preface:
An analysis of the history of technology shows that technological change is exponential … we won’t experience 100 years of progress in the 21st century—it will be more like 20,000 years of progress (at today’s rate) … There’s even exponential growth in the rate of exponential growth. Within a few decades, machine intelligence will surpass human intelligence, leading to The Singularity … a rupture in the fabric of human history. The implications include the merger of biological and nonbiological intelligence, immortal software-based humans, and ultra-high levels of intelligence that expand outward in the universe at the speed of light. (Kurzweil 2001, 1).
If this were true, writing this book would have been a monumental blunder, as we would soon get growth at the speed of light (I urge those with a modicum of scientific education to pause and think what this would mean in reality). And the author of these forecasts is swallowing daily scores of pills in order to ensure that he will live to see the Singularity’s arrival: his latest forecasts see machines achieving human levels of intelligence in 2029 and the Singularity still on track for 2045 (Kurzweil 2017). Modis (2006, 112) put it well as he closed his review of Kurzweil’s book by writing that “as science fiction goes … I prefer more literary prose … and less of this science.” I believe that, acting as risk minimizers, we must assume that there will be no singularity-driven technical salvation because there will not be any speed-of-light growth of our knowledge. I agree that “on today’s evidence, technologizing our way out of this does not look likely … the only solution left to us is to change our behavior, radically and globally, on every level. In short, we urgently need to consume less. A lot less.” (Emmott 2013, 184–186).
Hence back to square one: no modern society has been taking any thoughtful, effective steps to find its way toward deliberately very low or no growth even in settings where a relatively high level of average affluence and obviously excessive levels of consumption and waste are all too evident. This means that I can answer the question of what comes after economic growth only by making it scale-specific and contingent on the time spans under consideration. The answers would then range from more growth during periods of years to decades for most of the world’s economies to a nonnegligible probability of some kind of involuntary global retreat—that is substantial prolonged worldwide retrenchment, followed, at best, by greatly diminished rates during a halting recovery, at worst by further gradual decline, that is degrowth not by choice but as a reaction to cumulative (economic, extraction, consumption, environmental) excesses.
Modern Civilization
After millennia of slow and unsteady progress, the two centuries of unprecedented growth—of populations, food production, infrastructures, and of extractive, manufacturing, transportation, and communication techniques—have brought changes that were truly unimaginable at the outset of this transformational process. Modernity has been synonymous with growth and its rewards have been immense and all-encompassing. This achievement is even more impressive given the fact that Malthus did more than greatly underestimate our capacities to provide for growing populations: he also underestimated the future growth of the global population.
Its total, at about 900 million in 1800, not only kept on increasing exponentially during the 19th century, taking 110 years to double to 1.8 billion, but (as already explained) the growth rates accelerated during the 20th century as hyperbolic growth resulted in the next doubling, to 3.6 billion, in only 60 years; and, after the growth rates moderated, the next doubling, to 7.2 billion, was accomplished in not even 45 years, with the 2017 total at 7.3 billion in 2017. And yet never in history has there been such a high proportion of the global population that is adequately fed, well-schooled, lives in a modicum of comfort, and has such a long average life expectancy. Moreover, we have the technical means to eliminate the remaining malnutrition and to raise the living standards of the poorest segments in any society.
But scientific understanding offers no clear grasp of what lies ahead. Utopianism (now in its techno-optimistic-electronic-artificial-intelligence garb) and catastrophism (updated Malthusianism concerned with exhaustion of natural resources and destruction of the biosphere’s capacity to support continued economic growth) are not just labels for contradictory opinions and sentiments of uninformed commentators. The terms correctly describe divergent views that coexist within the mainstream of modern scientific research. Long-established scholarly journals considered to be the most reliable sources of information in their fields have been carrying these contradictory messages for decades, and I have not discerned any toning down of extreme claims in the 21st century.
Are we to look forward to the coming age of plenty where “farmers could feed the world indefinitely” (Fuglie 2013, 26), where “economic and environmental outcomes can be decoupled” (Hatfield-Dodds et al. 2015), and where people using abundant, infinitely malleable materials “will conjure objects as easily as we now play music or movies” (Ball 2014, 40)? Or do we agree that waste production must begin to decline during this century (Hoornweg et al. 2013) and that we should contemplate the peak of everything due to the imminent depletion of many mineral resources (Heinberg 2010; Klare 2012)? Should we then view the future of mass consumption with equanimity or with increasing foreboding? Or, to rephrase it by reusing the labels chosen for the title of a recent book examining the two polar positions, should we listen to prophets of environmental perils or ignore them because the salvation will come from inventive wizards (Mann 2018)?
Growth has brought a number of obvious benefits, from making life easier (including the ownership of all machines and gadgets that make running a household incomparably less onerous that a century ago) to (however ephemeral) feelings of satisfaction and enjoyment by displaying (often almost instantly disposable) pieces of manufactured junk. Compared to these, losses of individual comforts and intangible benefits are of a minor importance, although many individuals value them very highly. For myself, they might include walking on a quiet forest path, looking at a starry sky bisected by the Milky Way, standing alone in front of Las Meninas … The first experience is still common in remote boreal forests and those are also the best places to avoid light pollution. But unless you get a private, after-hours, tour of the Museo del Prado, standing alone front of Velázquez’s astonishing Las Meninas is now, briefly, possible only if you are the first one in the morning lineup and walk fast directly to room 12 on the first floor: in minutes it will fill up with guided masses from Shanghai or Osaka.
Any meaningful cost-benefit analysis of these (real and perceived) personal gains and burdens inherent in mass consumption is impossible as the two effects have no common metric. The judgment falls largely into the realm of value. But (as challenging as it may be) appraising the collective gains and losses of global economic growth and mass consumption raises indisputable concerns, above all because of intergenerational obligations arising from the need to maintain a habitable biosphere. Again, techno-optimists are not perturbed and cite the recent dematerialization trend as a key shift promised to make a new world possible.
But while relative dematerialization, particularly in consumer electronics, has helped to maintain some high growth rates, absolute dematerialization is a different matter. Mass consumption (measured by numbers of people acquiring an item) is also always increased consumption of mass (be it measured by inputs of energy or raw materials). Arguments about the impressive miniaturization (and hence dematerialization) of modern electronics are based on faulty assumptions. Smartphones may be small and light but their energy and material footprints are surprisingly large. Here are my calculations of respective embodied energies in the year 2015 based on the best available data (Smil 2016a).
Inevitably, in absolute terms, a car with a mass 10,000 times that of a smartphone (1.4 t vs. 140 g) embodies considerably more energy, but global aggregates tell a very different story. In 2015 worldwide sales reached 1.9 billion mobile phones, 60 million laptops, and 230 million tablets (Gartner 2017). Their total mass was about 550,000 t, and with conservative assumptions of average embodied energies of 0.25 GJ/phone, 4.5 GJ/laptop, and 1 GJ for a tablet (Wu et al. 2010; Anders and Andersen 2010), making these devices required about 1 EJ of primary energy.
A passenger car (steel, aluminum, and plastics make most of its mass) needs nearly 100 GJ to produce (Volkswagen 2010), which means that the 72 million vehicles sold in 2015 embodied about 7 EJ of energy in about 100 million t of machines. The mass of newly sold cars was thus 180 times that of all portable electronics—but their production required only seven times as much energy. Moreover, portable electronic devices have short life spans (averaging just two years) and their production thus embodies globally about 0.5 EJ per year of use—while passenger cars last a decade and their worldwide production embodies about 0.7 EJ per year of use—that is only 40% more than making all portable electronic devices! This makes for a stunning conclusion: even if my approximate aggregate calculations were to err by 50% in opposite directions (i.e., cars embodying more and electronics requiring less energy than I assume), the global totals would be still of the same order of magnitude and, most likely, they would not differ by more than a factor of two.
Of course, operating energy costs are vastly different. A compact American passenger car consumes about 500 GJ of gasoline during a decade of its service, five times its embodied energy cost. A smartphone consumes annually just 4 kWh of electricity, less than 30 MJ during its two years of service, or just 3% of its embodied energy cost if the electricity is from nuclear fission or from a PV cell, and about 8% if it comes from burning coal. But the cost of electrifying the net is already high and it continues to rise. In 2013 US data centers consumed about 91 TWh of electricity (2.2% of all generation) and are projected to use about 3.5% by 2020. In global terms, the aggregate demand of information and communications networks claimed nearly 5% of worldwide electricity generation in 2012 and it will approach 10% by 2020. In aggregate, tiny phones leave a not-so-tiny energy—and hence environmental—footprint.
And there are no even remotely comparable dematerialization shifts where the basic modern infrastructures, structures, and now indispensable artifacts (ranging from fertilizers to turbines) are concerned. Two key factors militate against any early decline of global energy demand and material consumption. The first is the continuing growth of the global population with its obvious implications for higher food and energy outputs and expanded industrial production. But the more important reality is that the demand for higher material per capita consumption is still far from saturated even in the world’s most affluent societies and that their achievements act as powerful attractors for all societies on the lower rungs of economic development: the recent quest of China’s nouveaux riches to out-American America in ostentatious consumption is a perfect example of this effect.
Hopes for an early end to this demand are unrealistic because the growth of material consumption is a universal and durable phenomenon: objects of desire change, desire remains. Mukerji (1983) showed that among wealthy merchants and aristocrats the growth of personal consumption began at the very beginning of the early modern era. During the 16th century their homes began to fill with paintings, imported rugs, tea services, and upholstered chairs—and even before the 17th century a growing array of consumer goods found its way to the homes of peasants and laborers. Rugs and tea services may be low on the list of modern desires but far more energy-intensive products requiring rare minerals and elaborate industrial processes and handled by transportation-intensive global production chains have taken their place. In any case, how likely is it that in the future our ingenuity could feed growth that could be embodied solely in immaterial achievements?
Techno-optimists are convinced that technical fixes (those already emerging and those to come in the future in response to critical problems) will solve even seemingly intractable challenges. Anticipations of technical progress have been always affected by unrealistic expectations belonging to several categories of distinct errors. The combination of the early hype and of the replacement hype error is perhaps the most common, with recent cases including the claims of extraordinarily rapid decarbonization of global energy use and, perhaps most notably, the promise of a fourth industrial revolution “that will fundamentally alter the way we live, work, and relate to one another. In its scale, scope, and complexity, the transformation will be unlike anything humankind has experienced before” (Schwab 2016, 1). Impact errors are also common, as economic, environmental, and social aspects of new techniques and processes are underestimated or are naively portrayed as innocuous and easily manageable.
But historical perspectives demand skepticism. I stand by my conclusion that the two generations preceding WWI were the most exceptional innovative period in history and that its contributions have been far more consequential than the advances of the last two generations (Smil 2005). Similarly, Ferguson (2012, 2) based his refusal to believe in the techno-optimistic hype on contrasting recent achievements with past accomplishments and by offering “simple lessons from history: More and faster information is not good in itself. Knowledge is not always the cure. And network effects are not always positive.” In contrast, Mokyr (2016) stresses that progress is not a natural phenomenon but a relatively recent human invention and that the alternative to technical progress “is always worse.” Ultimately, it comes down to the biosphere’s capacity to support an expanding population consuming at higher rates.
And yet, the biosphere’s indispensability and degradation are not among the concerns of those introducing ever-larger information flows and ever-faster communication and they are never mentioned in the Kurzweilian promise of infinite growth. In contrast, half a century after his first apocalyptic warnings, Ehrlich still predicts the bleakest planetary future due to the insurmountable environmental problems: “Environmental problems have contributed to numerous collapses of civilizations in the past. Now, for the first time, a global collapse appears likely. Overpopulation, overconsumption by the rich and poor choices of technologies are major drivers; dramatic cultural change provides the main hope of averting calamity” (Ehrlich and Ehrlich 2013, 1).
In contrast, it is not difficult to offer a very different scenario that, while not highly likely, is not implausible. African fertilities decline much faster than expected. Indian population growth decelerates rapidly. Rest of the world sees population stagnations and declines. Aging populations consume less and this, in combination with relative dematerialization, eases the burdens imposed on the biosphere. Economic growth moderates while advances in energy conversion and storage usher in affordable all-electric or hydrogen economies. Natural ecosystems begin their comeback, as forests have already done in Europe and parts of North America. I wish all of this came to pass as rapidly as possible—but acting as responsible risk minimizers we cannot simply hope for low-probability outcomes.
There is no need to be a catastrophist in order to see what I call the great obverse: all that we have lost as a result of growth in general and mass consumption of artifacts and experiences in particular, the extent to which we have already imperiled the life on Earth, and the potential for further damage resulting from a growing population and rising aspirations. The overall environmental cost of growth is still going up as it spans an enormous range of impacts. Some of them are largely matters of sentiment and preference, some have brought regrets and inconvenience and affected our perception and enjoyment of some aspects of our existence, even our collective health, but have not imperiled civilization’s survival.
As already noted, the loss of darkness (light pollution) is a foremost example in this category: it not only ruins astronomical observations and prevents hundreds of millions of people from ever seeing the great band of our galaxy, it also affects ecosystems and animal and human health (due to the disruption of circadian cycle) and, obviously, it increases energy consumption in a particularly wasteful manner (IDA 2017; figure 6.18). But if it were the only anthropogenic disturbance of our biosphere, our lives would be negatively affected but the future of our civilization would not be fundamentally compromised. Unfortunately, there are too many anthropogenic transformations whose increasing intensity and combined effect have been doing precisely that, and while two or three generations ago the same actions had overwhelmingly local or regional consequences, their impacts are now truly global.
Few nighttime images from space illustrate the extent of anthropogenic light pollution as stunningly as the view of the most densely populated parts of Western and Central Europe. NASA image.
These effects have been abundantly documented by increasingly comprehensive monitoring of the biosphere. Thanks to sensors mounted on orbiting Earth-observation satellites, we now have (adequate to outstanding) knowledge of the anthropogenic insults to the integrity of both terrestrial and aquatic ecosystems. No major biome (be it tropical or temperate grasslands and forest, tundras or wetlands) has escaped extensive destruction or at least modification, with some of these transformations going back thousands of years, others remaining relative subdued but growing at unprecedented rates since the 1950s. Oceans have been affected by changes ranging from gradual storage of heat generated by the tropospheric warming (Wang et al. 2018) to massive accumulation of microplastics (GESAMP 2015) and by declining oxygen content in both open and coastal waters (Breitburg et al. 2018).
The capacity of fresh water resources to deliver reliable supplies, particularly in densely populated Asia where cropping relies heavily on irrigation and where Himalayan glaciers are a key source of runoff, has been reduced, in some regions to a worrisome degree. Changes in the Earth’s gravity measured by satellites indicate large-scale mass losses of groundwater in northern India caused by excessive withdrawals (Tiwari et al. 2009); the Aral Sea has almost entirely disappeared (Usmanova 2013); groundwater levels in the Ogallala Aquifer, underlying the world’s most productive farmland in the US Corn Belt, continue to decline (USGS 2017a); and too many aquifers around the world are polluted with pesticide and herbicide residues, nitrates, and heavy metals.
There is no need to resort to exaggerated claims about species loss to realize that the decline of global biodiversity has been proceeding at rates that, on geological time scales, may already amount to the Earth’s sixth mass extinction wave (Barnosky et al. 2011). I have calculated that during the 20th century the mass of wild mammals was halved (and the mass of elephants was reduced by 90%), while the mass of domesticated animals more than tripled and the global mass of humanity more than quadrupled (Smil 2013a). People and their animals have been steadily marginalizing all wild species. And Darimont et al. (2015), after analyzing a database of 2,125 exploited wild animal populations, found that humans take up to 14 times as much adult zoomass (that is the reproductive capital of animal species) as do other predators, functioning as an unsustainable “super predator.” Some insects are also in retreat: most importantly, some wild and managed pollinators (bees) have been in decline in several regions, a shift with potentially enormous consequences for many crops (Potts et al. 2016).
Losses of high-quality arable land in alluvial regions and destruction of natural coastlands—both done in order to accommodate growing cities, factories, and transportation links—do not make urgent headlines but their extent clearly imperils our capacity to feed ourselves. China now has less arable land per capita than Bangladesh and yet its population is too large to be ever fed by imports. There is not enough grain on the global market to satisfy China’s annual need for rice, wheat, and corn even if China were the only importer: in 2017–2018 worldwide trade in milled rice, wheat, and corn was about 380 Mt while China’s annual grain harvest is now around 570 Mt (FAO 2018; USDA 2017b). The expanding human footprint has greatly reduced areas of contiguous wilderness, with the largest remaining intact forests concentrated overwhelmingly in just three countries, Russia, Canada, and Brazil (Potapov et al. 2008).
All of these concerns, some going back many generations, have been recently both intensified and overshadowed by the worries about the impact of anthropogenic warming, an environmental change with truly global effects. Worldwide emissions from the combustion of fossil fuels (and, a minor addition, from the production of cement) have been reconstructed starting with just 3 Mt C in the middle of the 17th century; 100 Mt C was reached by 1863 and the first billion by 1927; 50 years later the total surpassed 5 Gt C and in 2015 it was just a bit below 10 Gt C, or more than 36 Gt CO2 (Marland et al. 2017). The trajectory fits very closely a symmetric logistic curve with the inflection point in 2010 and yielding values of about 17 Gt C in 2050 (figure 6.19). US post-1800 emissions show some significant temporary departures from a logistic trend but they reached their inflection point in 1967 and it appears most likely that they will not rise significantly above the recent level.
Global CO2 emissions, 1750–2050. Data from Marland et al. (2017).
The logistic trend indicating slightly rising global CO2 emissions is in contrast with the worldwide effort to reduce CO2 emissions in order, as the 2015 Paris Agreement called for, to keep average global warming to no more than 2°C above the preindustrial level (UNFCCC 2015). Indeed, US energy-related emissions have already seen some declines and global emissions barely changed between 2014 and 2016—but resumed their growth in 2017 and 2018 (IEA 2018). The long-term outlook is unclear; after all, a key conclusion of the Paris meeting was that “the estimated aggregate greenhouse gas emission levels … resulting from the intended nationally determined contributions do not fall within least-cost 2°C scenarios but rather lead to a projected level of 55 Gt in 2030” (UNFCCC 2015, 3). As a result, 2030 emissions would be more than 50% above the 2015 level and higher (at almost exactly 15 Gt C) than the logistic forecast for 2050!
Many of these concerns led to a renewed call (the first one was issued in 1992) to curtail environmental destruction (Ripple et al. 2017). The appeal cited increases in population, ruminant livestock numbers, CO2 emissions, and tropospheric temperature, and declines in freshwater resources, in marine catches, in forested areas, and in the abundance of vertebrate species as the key trends that pose a growing risk to sustaining biospheric conditions compatible with the long-term survival of humanity In contrast, Nielsen (2018, 1) assures us that “anthropogenic signatures are characterised by the Great Deceleration in the second half of the 20th century. The second half of the 20th century does not mark the beginning of the Anthropocene but most likely the beginning of the end of the strong anthropogenic impacts, maybe even the beginning of a transition to a sustainable future.”
This is, of course, both expected (as all high growth rates must eventually moderate) and not at all reassuring as far as the state of the biosphere is concerned. To use a key example, growth of global primary energy consumption has been decelerating since 1950, but between 1900 and 1950 (when the growth rate was hyperbolic) annual use of all energy grew 2.3 times while during the century’s second half the decelerated growth increased the annual energy use nearly four-fold. Growth rates are lower than two generations ago but absolute annual use of energy (or materials, food, or water) is considerably higher, a combination that does not reduce the burdens imposed on the biosphere.
At the same time, there are no simple, single-value thresholds that would indicate crossing the lines from worrisome but acceptable levels of deterioration of the biosphere to the realm of catastrophic outcomes. But our understanding of the dynamic links between the state of the biosphere and the fortunes of our civilization makes it clear that all of the trends that have been moving in undesirable directions will have to be, sooner rather than later, curtailed if not reversed. Good life within planetary boundaries is possible even if the global population continues to grow—but not without fundamentally restructured provisioning systems, a shift that would entail substantial challenge to current economic strategies (O’Neill et al. 2018).
There is no possibility of reconciling the preservation of a well-functioning biosphere with the standard economic mantra that is akin to positing a perpetuum mobile machine as it does not conceive any problems of sustainability in relation to resources or excessive stress on the environment. Most economists are either unaware or dismissive of the advances that took place in our understanding of the synergistic functioning of civilization and the biosphere—and yet they maintain a monopoly on supplying their physically impossible narratives of continuing growth that guide decisions made by national governments and companies.
And, as if this economic growth perpetuum mobile were not enough, those who believe in imminent singularity make an even more improbable claim as they envisage acceleration of perpetual growth based on electronics. A small minority of economists, and many historians, environmentalists, and students of complex systems disagree: they recognize the obvious, the impossibility of infinite growth on a finite planet, but the steps we have taken so far have been insignificant and largely ineffective compared to the ubiquity and the scale of the required temporary remedies and eventual long-lasting solutions.