Energy and the Industrial Revolution
This week, I am continuing a series on energy as a driver of economic growth. When we think of the historical importance of energy, the industrial revolution, beginning in Britain in the late 18th century, is likely to be the first historical process to come to mind. Today we will consider how important energy has been in driving industrialization, and what that means exactly.
Next week I will be out of town, and I will probably not have a post two weeks from now. If that is the case, then my next post will be in four weeks.
The Growth of Energy Consumption
That industrialization is associated with a rapid increase in energy consumption is well-established. A well-known Our World in Data chart makes this clear.
To understand the process a bit better, let us focus on the energy and mineral consumption history of the United States, as presented by Wang et al. (2021). Though the industrial revolution did not start in the United States, the U. S. is one of the early countries to industrialized, and the U. S. became the world leader in industrialization from the late 19th century to the present day. It is therefore a good country to illustrate the process.
Wang et al. use the four phase model of industrialization. I find it better to understand industrialization as a continuous process that is not naturally divided into distinct phases, and the idea of a “fourth industrial revolution” is especially problematic because we lack historical perspective to understand contemporary trends. But periodization helps us organize a complex topic. The following dates are based on the American experience, and work focusing on other countries will choose different dates.
First industrial revolution: around 1790 to 1860. This period is characterized by increasing use of steam power in manufacturing. Mechanization, especially in textile manufacturing, substituted energy for human labor.
Second industrial revolution: around 1860 to 1920. This period is characterized first and foremost by the introduction of electric grids. Oil production became significant. Incandescent lighting was important, as were major advances in transportation technology, led by the invention of the internal combustion engine.
Third industrial revolution: around 1940 to 1980 (yes, we skipped some years). This period is characterized by the invention of electronic computers, the jet engine, astronautics, and atomic weapons and power.
Fourth industrial revolution: vaguely around 2000 to some time in the future. Wang et al. suggest that quantum computing and other novel computing technologies, networking technologies such as blockchain, and Internet of Things might be decisive technologies of this time. We will see. See also Ross and Maynard (2021) for more speculation on the fourth industrial revolution.
From 1800 to 1860, coal consumption in the U. S. grew from about ~100K tons to 17.52M tons for an average annual growth rate of 9.3%. From 1860 to 1920, coal grew from ~18M tons to 584M tons, a far more impressive growth rate in absolute terms, but in percentage terms, down to 3%. It was only in 1885 that coal displaced firewood as the nation’s leading energy source, a sobering reminder that the timetables of energy transitions are measured in generations, not years. Coal production peaked in the late 1940s, and during IR3, oil and natural gas surpassed coal as a share of primary energy. For electricity generation, coal rebounded in the U. S. from the 1960s, reached a new peak in 1998, and declined precipitously after 2008, having fallen by about half by 2023, due in no small part to the rapid development of hydraulic fracturing.
Oil production grew from basically zero to 2823 petajoules, or 12.5% of the energy supply, during IR2, whereas the first natural gas production occurred in 1880. During IR3, oil consumption grew from 8187 PJ to 36,088 PJ and natural gas consumption grew from 2,812 PJ to 21,350 PJ, for annual growth rates of 3.6% and 5.1% respectively.
IR3 has been characterized by slower energy growth rates compared to IR1 and IR2. Using 2011 dollars, Wang et al. find that a $1000 increase in per-capita GDP induced a 33.6 gigajoule increase in energy consumption during IR2, and that was down to 8.7 GJ during IR3. Energy consumption has further slowed since 1980, with fossil fuel consumption having reached an apparent peak in 2005.
Much has been written about the failure—so far at least—for nuclear power to take off in the way that coal, oil, and natural grew in the 19th and 20th centuries. A while back, I discussed Jack Devanney’s hypothesis that ill-conceived regulations and a complacent nuclear establishment are to blame. Others point to public opposition, whether rightful or not, as the industry’s main headwind. The above statistics points to another explanation. It is much easier for a technology to build market share in a growing market, when it is satisfying new demand, than in a stagnant market, when it is displacing established alternatives. Nuclear power has not enjoyed as strong of market growth as was enjoyed by coal, oil, and natural gas. Modern renewables of wind and solar are facing an even stronger demand headwind, and for enhanced geothermal and fusion, should those sources ever become a technical reality, the headwind will be greater still.
As for mineral resources, Wang et al. (2021) find that many bulk minerals have also peaked in production in the United States. Production of pig iron, a staple commodity in IR1, peaked in the 1970s and has fallen by more than half since then. Manganese, copper, tin, and gold are among the metals have have shown clear peaks in the U. S. Silver, cobalt, and most rare Earth minerals, all important for rapidly growing electronics and renewable energy industries, show no signs of peaking, but their overall production is far less than the staple metals that have peaked.
Causes of the Industrial Revolution
It is clear that industrialization, at least in the early stages, is associated with increased energy consumption. But does energy cause industrialization? As I discussed last week, there is reason to doubt this in recent times for wealthy countries, but maybe it is true in the past and for poorer countries today.
I discussed Ayres and Warr (2005), which argues that energy should be treated as an input for growth in the Solow-Swan model, alongside the usual inputs of labor and capital. Most economists will object that energy production is a capital investment, not fundamentally different from other capital investments, and not merely a fixed resource bestowed by nature. Obviously, the coal reserves in Britain were there long before the 18th century, and so to say that they are responsible for the industrial revolution is to beg the question of why those reserves were harnessed in the 18th century and beyond, but not earlier.
Many answers could be given to the question of what caused the industrial revolution. Simon (1994) attributes the industrial revolution to European population, which most importantly increased the number of people who could potentially contribute the technological advances that industrialization requires. The role of science, and the extent to which the industrial revolution should be seen as applied science, rather than non-scientific tinkering, is debated. Jacob (2014), in The First Knowledge Economy, argues that the industrial revolution was very much a scientific project that built upon the advances of the recent past, whereas Ó Gráda (2016) discusses several studies arguing that the industrial revolution was driven by artisanal knowledge. Mokyr (2008) emphasizes informal social institutions that drove cooperation between inventors and cooperations, deemphasizing a traditional focus on conventional explanations in terms of formal institutions of patenting and property rights. Blinov (2014) attributes the industrial revolution to the surplus resulting from the fact that hunger had been banished in Britain at the time. I could dedicate many full posts to this question, and of course the various answers are not mutually exclusive.
The Dutch Golden Age
To shed some light on the question of energy and the industrial revolution, it is worth comparing to another, earlier episode. Goldstone (2002) observes that the pattern of preindustrial economic growth was generally one of stagnation, punctuated by periods of rapid growth that Goldstone calls “efflorescences”. One such efflorescence was the Dutch Golden Age, a period of invention and growth that lasted roughly from 1570 to 1670.
According to Goldstone (2002), to a contemporaneous observer, the Dutch Golden Age would have looked in 1670 much as the British efflorescence did in 1820: a period of invention and growth that appeared to have run its course. However, the Dutch Golden Age did end, whereas after 1820, the industrial revolution intensified and spread around the world in a process that continues to this day. What is the difference?
As documented by de Zeeuw (1978) (sorry, link is not working right now), exploitation of peat was a central aspect of Dutch prosperity at that time. Peat is somewhat of an intermediate energy source between biomass and fossil fuels; it is the decomposed remains of plants that have been dead for thousands—as opposed to millions—of years. The Netherlands also enjoyed usage of wind power—you can visualize those iconic Dutch windmills—and according to Saelens (2024), coal burning proliferated in the early 18th century, not long after the end of the Golden Age period.
An ecological economic explanation for why the Dutch Golden Age ended and the industrial revolution is still going derives from ERoEI, or energy return on energy invested. According to the U. S. Geological Survey, peat is of a lower quality than coal and thus has a lower ERoEI, and so Dutch society would have had to devote a greater portion of its economy to energy production that Britain a century and a half later. Peat just wasn’t good enough to kickstart industrialization, and coal was. I have seen this argument made in a well-researched blog post that I unfortunately cannot find now.
However, it is not all that clear to me that ERoEI is, in and of itself, a particularly relevant metric. Furthermore, Kedrosky (2021) documents, as noted above, how The Netherlands gained access to the very same coal by the end of the Golden Age that would power the British industrial revolution. While the price of energy rose in The Netherlands by 1650 and peat resources were being depleted, Kedrosky see the problem as more on the demand side than the supply side and cautions against simplistic narratives of the Dutch Golden Age being ended by peat depletion. Prak (2015) tells the story of the end of the Golden Age in terms of demographic problems, though population decline, which according to ten Raa et al. (2009) began around 1670 and stabilized only around 1750, may have been an effect of the economic problems as much as a cause.
Conclusion
Why industrialization began where and when it did is a complex question that I cannot possibly do justice to now. But narratives that the cause of the industrial revolution is that British entrepreneurs figured out how to use coal via the steam engine are far too simplistic and do not shed much light on what happened.
Quick Hits
The last two weeks have been eventful in the news. I think all readers are aware that Pope Francis died last Easter Monday. He had been hospitalized for pneumonia a few weeks earlier, and after his discharge, I thought he would be OK for a while. The funeral Mass can be viewed here. Francis attracted much criticism from more conservative elements within the Church, some justified and most not, but in my view his compassion for the marginalized and emphasis on service outweigh anything that can be said about politics. The conclave to elect a new pope is scheduled for May 7.
Last week, there was a major blackout on the Iberian Peninsula, which J. Guillermo Sánchez León at The Conversation explains. Around 12:33 PM on April 28, unusual fluctuations occurred on the grid, and a rapid cascade of failure followed. What exactly went wrong is still undetermined. Josh Smith warns against jumping to the conclusion that the high rate of variable renewable energy on the Spanish grid is to blame, noting that similar claims about the 2021 Texas blackout during Winter Storm Uri have turned out to be unfounded, whereas Seaver Wang and Alex Trembath are more willing to make that conclusion. Spain has made the decision to begin a phaseout of nuclear power in 2027, to be complete in 2035. A similar decision had bad consequences in Germany, estimated at $12 billion per year by Jarvis, Deschenes, and Jha (2019) and with benefits far less than that. Regardless of the details of what happened, which I emphasize are yet to be determined, the incident is a stark reminder of the importance of grid reliability.
Also last week, Ted Nordhaus discussed the Climate Realism initiative of the Council on Foreign Relations. The brainchild of Varin Sivaram, once an aide to John Kerry, it is a curious mixture of conventional leftwing climate politics and America First realpolitik, and Ted characterizes the initiative as having characteristics of ecofascism, which Nils Gilman wrote about in 2020. As especially eyebrow-raising passage in Sivaram’s piece is,
As greenhouse gas emissions exacerbate hurricanes and wildfires that level whole U.S. communities spanning North Carolina down to southern California, the effects resemble those if China or Indonesia were to launch missiles at the United States. Every tool of the U.S. and allies’ arsenals, spanning diplomatic and economic coercion to military might, should be on the table.
It is left to the reader’s imagination what exactly this might mean. However, one does not have to imagine atrocities, which are not as well known as they should be, that the U. S. government engaged in during the 1960s and 1970s in the name of environmentalism. In 1966, Lyndon Johnson conditioned foreign aid on recipients adopting population control policies. In 1969, Richard Nixon established the Office of Population within USAID, whose head, Dr. Reimert Thorolf Ravenholt, engaged in a sickening campaign of forced sterilization around the world, as Robert Zubrin documents. Elements of Climate Realism represent a dark turn that has been seen before and should be taken very seriously.
Last Wednesday was the 50th anniversary of the Fall of Saigon, which occurred on April 30, 1975, finally bringing an end to the war in Vietnam which had been raging from 1955 to 1975. I have been intending a post on the Vietnam War for some time, mainly to consider the Domino Theory. The anniversary might have been a good occasion for the post, like I did with the Iraq War two years ago, but that will wait for another day. American involvement in the Vietnam War was motivated by the belief that if communists were to take over Vietnam, other countries would succumb to communism, like falling dominoes, and the result would be a less free and more dangerous world. Thus the United States had a pressing interest to support the Western-aligned South Vietnamese government. The communist North did in fact take over, the spread of communism to other countries did not happen, and now relations between the United States and Vietnam are so good that you would hardly guess that the communists won.