Published by EH.Net (May 2020)

Vaclav Smil, Growth: From Microorganisms to Megacities. Cambridge, MA: MIT Press, 2019. xxv + 634 pp. $40 (cloth), ISBN: 978-0-262-04283-3.

Reviewed for EH.Net by Philip Coelho, Department of Economics, Ball State University.


Vaclav Smil’s Growth is formidable; he has acquired a wealth of knowledge over his life and career. Its coverage is encyclopedic, yet it is flawed by a lack of basic economic principles and any sense of his limits. Growth’s stated task is to deal “in realities as it sets the growth of everything into long-term evolutionary and historical perspectives and does so in rigorous quantitative terms” (p. xxiv). The stated task is literally impossible; an examination of the growth of everything is just as likely as a perpetual motion apparatus. Still the book has achieved substantial praise earning it an encomium from Bill Gates on the dust jacket, and a favorable mention in the New Yorker magazine (John Cassidy 2020). The book is formidable with a text of over five-hundred pages (divided into 6 chapters, and a coda), plus a separate preface. The first three chapters are devoted primarily to the mathematics of growth, and the growth of organisms/objects/things not usually under human purview. The last three chapters are devoted to the human economy/ecology.

Chapter 1 (“Trajectories: Or Common Patterns of Growth”) deals with how growth is depicted within mathematical models: linear, exponential, logistic, etc. What is lacking in his analysis is any sense that the actual details/history matter. Smil continuously states that exponential growth is a long run impossibility. Contra Smil, suppose that economic growth proceeds at a compound rate of one thousandth of 1 percent per year. Then output doubles every 69,000 years; is this an impossibility, or, if it is a problem, how many current resources should we devote to solving it? The point is that size matters. Details matter; Smil states: “Sails are a good example … [of obsolescence] as their development and deployment … [except for leisure] ended fairly abruptly during the second half of the nineteenth century, just a few decades after the introduction of steam engines” (p .3). This is palpably false; if you search Google’s images with the terms “sail shipping 1900” you will get pictures from New York and other ports showing vast numbers of merchant sail ships. Sail as a method of shipping freight was the most economical for long voyages until after World War I. What eventually led to the end of sail were canals (the Panama Canal in 1914, Suez in 1869) that ended voyages of more than 12,000 miles (except for the Australian wheat trade where sail was marginally economic into the 1930s). Authors cannot be experts on all subjects, so a little humility goes a long way. That virtue is absent in Growth.

In Chapter 1 Smil measures efficiency and growth in physical terms rather than basing it upon economic fundamentals. (This is done throughout the book. The index has twelve different entries under the rubric of “efficiency” and 32 separate page citations; none of them refer to prices or costs.) He discusses measures of output per acre and concludes “… their yield trajectory has been one of minimal gains or outright stagnation” (p. 29). Here he uses “yield trajectory” to measure progress. But the economist asks: why do we want a greater yield per acre? Instead, suppose we could have a greater yield per man-hour of labor in lieu of an increase in output per acre, which is more important? The answer is that it depends upon relative prices; thus, rice output per unit of land is higher in Japan than the U.S., but rice output per man-hour is greater in the U.S. Both societies’ agricultures are efficient given differences in the relative prices of inputs. This illustrates Smil’s fundamental error: he equates “efficiency” in physical terms rather than in economic (value) terms. Stressing the obvious, if you want to profit from your labors, you should attempt to maximize output per dollar spent; and not maximize output per unit of land, labor, or whatnot.

Chapter 2 (“Nature: Or Growth of Living Matter”) and Chapter 3 (“Energies: Or Growth of Primary or Secondary Converters”) deal with biological organisms. They illustrate Smil’s strengths (a compendium of facts and knowledge), and weaknesses (a lack of elementary economic analysis). He worries (p. 120) about the stagnation of China’s grain yield per acre and its “staple food self-sufficiency.” The economics of comparative advantage elude him. Chinese crop yields per acre are not increasing rapidly because labor inputs used in agriculture are falling, even more so when age-adjusted. The growth of Chinese manufacturing and service sectors has attracted low-wage labor from the countryside into cities. Real wages in both rural areas and cities have increased. This is what we expect with economic development: the transfer of resources from low-productivity sectors to high-productivity sectors. Self-sufficiency is a (desirable) casualty of economic growth. More than occasionally, it appears that Smil is advocating autarky, yet, inconsistently, he seems to approve of the growth of the Chinese economy that has lifted one-fifth of humanity from poverty. Economic growth means increased productivity, and increasing productivity demands trade among specialized producers.

Economists define “efficiency” as reductions of per unit cost (“cost” adequately measured to include real non-monetary costs and benefits). Repeatedly Smil uses physical measures of “efficiency.” “[Poultry] crowding is also a perfect illustration of how maximization of profit drives growth even as that growth is not optimal. Studies have shown that broilers convert feed more efficiently, grow heavier, and have lower mortality when given more space” (p. 148). The issue is how to define “optimal?” We are not discussing the ethics of crowding poultry. (Although if you are considering ethical issues, then, perhaps, you should consider what the poultry want. Humans, after all, grow heavier and have lower mortalities in the countryside rather than in cities, yet humans flock to cities.) Suppose instead of talking about poultry we were to discuss the production of micro-organisms, and if the production of these organisms convert energy into biomass at a greater rate when less densely packed, but that the costs of the less densely packed micro-organisms are twice that of the densely packed. Choosing a process that is more costly cannot be justified on an economic basis; it can only be justified normatively (you should use less energy, or you should use more space). If you “should” do something because of normative precepts, then you are not dealing in “realities.”

Chapters 4, 5, and 6 deal with the human economy and its history. The chapters are deeply flawed. I am particularly disturbed by the graphs that are here (and throughout the entire book); the graphs are fitted, smoothed, and sourced. Yet curves frequently extend beyond the present; consequently, they are predictive, not descriptive. The extrapolated data are not from the sources provided. Still the graphs are sourced as if the predictions are derived from the sources; this may mislead the unwary. The smoothing of the data reduces the graphs’ value as historical devices. Curve fitting may be useful as a predictive device, but not as a descriptive device.

In his discussion on sailing ships (pp. 272-73), he is particularly confused. Again, he measures growth (progress) in physical units. But it is not the speed of delivery that matters, but the costs. In the eighteenth century, shipping speeds fell as well as costs relative to those of the seventeenth century. In the 1600s, pirates were numerous and markets small. Bigger ships that were low-cost haulers in the Baltic and North Seas trades were uneconomic in the North Atlantic circa 1650 because they were slow, vulnerable to pirates, and their carrying capacity was too large for colonial markets. Their large cargo holds were useless because colonial traders could not provide a full load for them. Narrow, fast boats sturdily built, carrying cannons, with many sails (and sailors) could make a rapid crossing (reducing their exposure to piracy) and if attacked the sailors could put up a formidable resistance to pirates. Defensive measures were costly both in terms of ship construction, labor, and cargo capacity. Still, given the circumstances, swift vessels were more economical than vessels designed for carrying freight. As the world changed so did the optimal (least cost) method of shipping. With the growth of markets, specialized resources (the British and French navies) were devoted to vanquishing piracy. Increased markets meant that ships with large cargo capacities (wide beams, consequently slow) could load a full cargo quickly. Market production increased; specialized port facilities increased. These developments affected the trade. By 1760 in the North Atlantic colonial trade, port times were shortened by months. The reduction in port times meant that the slower mid-eighteenth century ships in the North Atlantic trade could make more passages per year than their faster mid-seventeenth century counterparts (Walton 1967). Even when discussing the speed of shipping, Smil manages to confuse readers; when assessing the speed of sailing ships he states: “These comparisons imply an approximate tripling of sailing speeds … during the course of 18 centuries, once again a minuscule average gain if it were seen as a steady process of improvement, and even the doubling from 15 to 30 km/h took about 400 years” (p. 274). Certainly, using a cosmic scale diminishes everything to insignificance, but to do so is to embrace nihilism.

Smil does not embrace nihilism, instead he aligns himself and advocates a Degrowth strategy. The Degrowth group believes that continued economic growth is undesirable for a variety of reasons. Among them are environmental deterioration, inequality, racism, and global warming. Smil starts with an outline of “the long-term growth of global population” (p. 303), and on the very next page concentrates on Europe: “The growth during antiquity and Middle Ages was limited by inherent difficulties in securing sufficient [resources].” He blames the lack of urban areas to the costs of producing and delivering food. But there are other interpretations; an examination of the data shows that people lived shorter lives during past epochs primarily due to infectious diseases, many of which were caused by water pollution. Wages/incomes in urban areas were always higher than in rural areas, but death rates in urban areas were typically whole number multiples of those in the countryside. Roman aqueducts, for a time, reduced the urban penalty but the decline and fall of Rome meant a vast reduction in the supplies of potable water. Smil’s economic history leaves much to be desired. He makes clear that he is neither enamored with economic growth nor economics, unfortunately his distaste appears to cloud his judgement.

One of the faults that he attributes to economic growth is that it has done little to ease economic inequality. “The extent of global inequality (unweighted by population) appeared to change little between 1950 and 1975, but increased during the last quarter of the twentieth century . . . while the population-weighted trend showed a significant convergence of national incomes from the late 1960s, nearly all of it attributable to China’s post 1980 rise and without its gains global inequality showed little change between 1950 and 2000” (p. 433). First the unweighted data are meaningless; I do not know of a legitimate reason that anyone would compute an index of inequality giving equal weights to Timor-Leste (population 1.2 million) and China (population 1.428 billion). Secondly, in 1960 China had the largest number of impoverished people in the world; since then over 800 million Chinese have graduated from the ranks of the impoverished. If Smil had used 2010 as the terminal year rather than 2000, then additional hundreds of millions of Indians would have joined the Chinese in leaving the impoverished. These are the results of economic growth. Smil is correct that economic growth is a cause of inequality. Part of this is due to arithmetic: if incomes increase the right-hand side of the distribution of incomes will move further from the origin. Another reason for increased inequality is that increasing wealth (economic growth) is a result of humans reallocating resources to produce more than the same resources would have produced otherwise. The increase in income/wealth that entrepreneurs get are both rewards and incentives. If you take away incentives and rewards from entrepreneurs, you may reduce inequality, but, on the downside, you may get an economy like Venezuela’s.

Smil is firmly aligned with the Degrowth faction, partly out of aesthetics (there is too much “waste”), partly out of environmentalism, and partly out of technological/scientific pessimism. The first two are normative, the pessimism is predictive. He believes that we have reached an apogee of sorts: “We may never see such a concatenation of technical advances (electricity, internal combustion engines, automobiles, powered flight, chemical syntheses) as we experienced in the 50 years preceding WWI” (p. .423). Pessimism does sell; people and media who are proponents of optimistic scenarios are alternatively, vilified or ignored. Economic history provides an optimistic scenario; given incentives and institutions that protect individual freedoms and properties, we should expect a series of achievements matching or exceeding those of the past. The human economy is analogous to nature’s ecology; the more resources that are available in nature (sunlight, water, warmth, soil, organisms) the more complex is the ecology. Complexity opens ecological niches for new/differentiated organisms. The human economy/ecology is the largest it has ever been and growing ever more complex. In large urban areas where there are multitudes of economic transactions, novel economic niches are propagating ever more rapidly. (Granted there are always costs, among them is that the transmission of pathogens is easier in densely populated areas.)

As for the decline in the rate of scientific progress, that is a possibility, the question is how probable? Since 1950 we have discovered that the mass of the universe is actually 10-fold more the mid-century estimates; the existence of dark matter and dark energy may have vast unforeseen effects. In 1950 quantum effects where blackboard stuff, now they are being extensively exploited even before we have a complete understanding of them, and there is a very real chance that instantaneous quantum communications may be possible. The biological sciences have knowledge and technologies that were science fiction in 1950. These examples are massively incomplete, as always, because as the boundaries of one science expand, the boundaries of other sciences are affected and there are feedback effects. Smil has good company among the technological/scientific pessimists. But there is a lesson here; William Thomson (Lord Kelvin) is equally famous for his intellectual achievements in the sciences, and his risible statements on the age of the Earth, the decline of the industrial economy, the impossibility of powered flight, and the end of the Earth.. Lord Kelvin’s example is a caution to all: you should know and respect your limits.


John Cassidy, “Steady State: Can We Have Prosperity without Economic Growth?” New Yorker. February 10, 2020.

Gary M. Walton, “Sources of Productivity Change in American Colonial Shipping,” Economic History Review (1967).


Copyright (c) 2020 by EH.Net. All rights reserved. This work may be copied for non-profit educational uses if proper credit is given to the author and the list. For other permission, please contact the EH.Net Administrator ( Published by EH.Net (May 2020). All EH.Net reviews are archived at