Foreign Trade and the Industrial Revolution
Cause, Catalyst, Condition, Consequence?
The Industrial Revolution coincided with an unprecedented boom in overseas trade. By 1900, 30 percent of Britain’s national income was traded, up from 8 percent two centuries before. By the middle of the nineteenth century, Paul Bairoch estimated that the island was exporting a staggering two-thirds of the world’s “new manufactures,” especially cotton textiles. Shipments of industrial goods purchased the raw materials and food imports needed to feed Britain’s factories and workers—goods that she would otherwise have had to grow on a limited land area. In the workshop of the world, American cotton was spun and Swedish ore forged for Canadian loggers and Bengali farmers. From the Peace of Utrecht to V-E Day, British warships patrolled every ocean and, under their aegis, British freighters carried the goods of every nation.
Trade was clearly an important part of the Industrial Revolution. But in which direction does the causal chain run? One “school” of thought, recently revived, suggests that commerce started (or permitted) the “big event” in Britain. Kenneth Pomeranz, Immanuel Wallerstein, and Eric Williams, among others, have sought to show that the foreign connections of the Industrial Revolution stain it with the sins of imperialism, exploitation, and slavery. To smother the “internalist” narrative of endogenous technological progress, they stress British dependence on tropical raw materials acquired by forced labor and on export markets forced open by Royal Navy gunboats. “Whoever says Industrial Revolution says cotton,” says Hobsbawm, and whoever says cotton says slavery. Others, usually of a more conservative or libertarian bent (e.g. D. McCloskey), have tried to show the opposite—that Britain’s titanic overseas trade naturally followed from industrial success, and that its effects were in any case overrated with respect to national income.
In short, arguments stressing the central importance of international trade to the Industrial Revolution don’t all ask the same question. We can perceive several distinct roles for the foreign sector from 1700 to 1870, and each has different consequences for the “externalist” and “internalist” views.1
Trade as cause. Did the expansion of commerce during the eighteenth century lead to the technological advances of 1760-1830?
Trade as condition. Did trade save Britain from plunging into Malthusian catastrophe amid a massive demographic expansion, or at least provide the basis for economic expansion beyond the constraints of agrarian society?
Trade as catalyst. Was trade necessary for Britain to expand its domestic product after 1830, by allowing her to export textiles and import food in the international division of labor?
Trade as consequence. Did foreign trade merely follow from the pre-existing productivity of British industry as the best way of maximizing Ricardian comparative advantage once the Industrial Revolution was already underway?
Externalists, especially “global historians,” world-systems theorists, and new historians of capitalism, are inclined to support the first two causal sequences (trade leads to industrialization); internalists the latter pair, though they will tend to downplay the true size of (3). Disentangling these analyses from the assertion that the Industrial Revolution would not have happened without raw materials and export markets is essential to understanding what really happened. Demand stimulus from foreign consumers is not the same as price competition in Global South textile markets; the high-wage explanation suggests a different mechanism than “coal and colonies.”
And, as ever, the answer is bound up in the eternal controversy over what and why the Industrial Revolution was—why, in the words of David Landes, “we are so rich and they so poor.”2 Commercial causes point to the role of British militarism, violence, and avarice—to the growth of the naval-industrial complex vaunted by Patrick O’Brien; to plantation slavery in the Americas; to the Opium Wars in China. They tend to suppress Anglocentrism, to place the Revolution in the context of a great Smithian international specialization, and to downplay the role of the “great inventions” (or stress their foreign antecedents). At bottom, a trade-based Industrial Revolution suggests that there wasn’t much special (or at least admirable) about Britain; she simply took violent hold of fortune’s winds during a turning point in the world’s technological history. Maybe others could have come first, had Britain’s battleships not ruled the waves and her colonists not exploited slave gangs south of the Mason-Dixon.
How to decide? Examine the theories, of course.
Trade looks like an obvious cause for the Industrial Revolution. We all know about Britain’s eighteenth-century prowess as a naval and commercial power, and about her role in the “triangular trade” linking American agriculture, British industry, and African slave labor. It was just as obvious to the first historians of England’s transformation, writing at the end of the nineteenth century. In Joseph Inikori’s historiographical survey, foreign trade appears as the first mooted cause (of many) of the Industrial Revolution. Arnold Toynbee, for example, saw trade as “the all-corroding force” that swept away “the old simple conditions of production and exchange.” The French historian Paul Mantoux, meanwhile, cited the commercialization of England—“a nation of shopkeepers”—as an antecedent of industry. Views of history (following Marx, to an extent) as a progression through a commercial phase to an industrial one naturally (post hoc, ergo propter hoc) led to the obvious inference that trade played a causal role.
One simple version of trade-as-genesis is the “Gilboy thesis,” which—among other factors—points to foreign demand as a stimulus for growth.3 It’s easy to see how this Keynesian mechanism would work: demand curve shifts out, production rises, and so do incomes. But leaving aside the relatively small proportion of goods exported during the eighteenth century, the theory doesn’t fit the facts. Increasing demand should lead to rising or at least stable prices. But Britain’s terms of trade fell throughout the Industrial Revolution, collapsing by 55 percent from 1780 to 1830. This means that industry supply curves were rushing out even faster than was demand. Assuming that world income grew as quickly as England’s from 1780 to 1830 (135 percent, a debatable assumption indeed), price movements imply that supply increased by 290 percent.4
An entirely different perspective is that of Eric Williams, later Prime Minister of Trinidad, who argued that the profits of the Atlantic slave trade and plantation cultivation in the West Indies provided “the capital which financed the Industrial Revolution in England” in an anecdotal survey of large investments made by planters and traders. Barbara Solow (1985) added quantitative rigor to the Williams Thesis, suggesting that slave wealth grew from .3 to 1.5 percent of the national income. Ronald Eltis and Engerman (2000), however, note that since wealth was fully half of the total income, surely a score of other professions could have been equally important to industrial development. They conclude on this basis that “sugar cultivation and the slave trade were not particularly large, nor did they have strong growth-inducing ties with the rest of the British economy” (p. 123).5
The World-Systems and Dependency schools, associated with the late sociologist Immanuel Wallerstein, have a similar focus. They argue that the free-labor economies of the “core,” of which Britain was the leader, exploited a poor periphery through government-backed long-distance trade, which in turn led to capital accumulation and “good institutions.” The Global South, forcibly prevented from transforming structurally, was unable to develop and restricted to feeding raw materials to the core. In a devastating essay, however, Patrick O’Brien (1982)6 demolished Wallerstein’s model, memorably declaring that “the periphery was peripheral.” Under absurd assumptions about savings rates (30 instead of the actual 12-14 percent), trade with the periphery could have financed no more than 15 percent of the initial Industrial Revolution. Ronald Findlay (1990) and William Darity (1992) elaborate on the Williams/World-Systems theses using trade models. They critique O’Brien’s “small ratios” approach (dividing trade by national income), but while both discover significant forward and backward linkages from slave and periphery trades, neither is comfortable assigning commerce a causal role. Moreover, it’s clear that focusing on capital accumulation and profits is wrong-headed. The Industrial Revolution neither began with nor ran on investment, but rather innovation on British technologies that, especially in textiles, were relatively inexpensive to start. Big pools of savings don’t explain the flying shuttle, spinning jenny, or power loom.
One final, more familiar interpretation is that of Robert Allen, presented well over a decade ago in The British Industrial Revolution in Global Perspective (2009). He posited that exceptionally high wages, stimulated in part by commerce and imperialism, incentivized the invention of labor-saving, energy-using technologies in Britain alone. “[High wages] led British firms to invent technologies that substituted capital and energy for labour... Britain’s wage and price structure was the result of the country’s success in international trade, and that owed much to mercantilism and imperialism” (Allen 2011, p. 357). By driving urbanization in London, trade also led to the exploitation of coal reserves in the northeast for home heating, creating a new power source for the new industrial economy. Furthermore, the existence of a large foreign market decreased the per-unit R&D costs of developing new technologies.7 This point is substantiated by the work of Desmet and Parente (2012). They show that larger markets permit a more diverse range of goods to be produced, increasing their substitutability, reducing markups, and forcing firms to expand. Process innovation occurs as firms can spread fixed innovation costs over a greater range of output. Allen’s thesis is immensely popular among economists, and I have generally been positive about most aspects. Nevertheless, his data have come under scrutiny8 and he never deeply substantiated the link between foreign trade and the wages being substituted in the Midlands.
Even if trade did not cause the Industrial Revolution, there’s still scope for argument that foreign commercial connections were a necessary condition. The most popular explanation cites the limited resource endowments of the British Isles as a limiting factor for continued expansion and the role of commerce in relieving these constraints. Brinley Thomas (1985), for one, famously saw a Malthusian crisis developing in eighteenth-century Britain. A demographic explosion was making increasing demands on a limited land area for bread grains, charcoal fuel, pasture land, ship timber, and construction sites. Agricultural output grew at .44 percent per annum from 1760 to 1801, while the population rose at .83 percent, increasing to 1.36 percent from 1801 to 1851. Short of converting the island to arable, Britain had only one option: “The way out was for England (through a transportation revolution and international trade) to endow itself with the equivalent of a vast extension of its own land base” (Thomas 1985, p. 731). By selling capital- and labor-intensive manufactures to land-rich food producers, Britain effectively received a great “importation of land.” Until transport improvements brought in the necessary food supplies, however, investment rates were stagnant—and declining in industry.9 By 1910, fully 80 percent of wheat consumption came from abroad.
Thomas presaged the far more famous theory of Kenneth Pomeranz, whose seminal work, The Great Divergence (2000), turned twenty last year. His argument revolves around the assertion that the most advanced parts of Western and Eastern Eurasia followed similar development paths until 1800, when his titular separation got underway. Why? Both regions—primarily Britain versus the Yangzi Delta in China—experienced rapid population growth as a result of expanding proto-industry.10 A “vicious cycle” ensued, whereby piece rates fell, leading workers to produce more to acquire food, lowering rates still further. The demographic boom also created “ecological bottlenecks” by taxing local resource endowments:
And more generally speaking, population growth—whatever its relationship to proto-industrialization—could place serious pressure on the land needed for raising fuel, fiber, and other necessities of industrial development. Unless these goods can be acquired by trade, the only way to keep increasing output is by working the land more intensely, which with the technologies then available meant higher farm-product prices, lower per capita productivity, and a drag on industrial growth. (Pomeranz 2000, p. 73).
On the Lower Yangzi, few or no trade partners could be found who would take surplus Chinese cloth or supply the needed food and timber to meet rising demand. Europe, by contrast (and obviously Britain), was able to continue specializing in industry and growing its population without costs to living standards because “the limits imposed by its finite supply of land suddenly became both more flexible and less important.” Pomeranz estimated that the New World supplied Britain with 25-30 million “ghost acres” in addition to her 17 million of arable, comprised of 1-2 million acres of sugar cultivation, 3-4 million acres of timberland, and 23 million acres of pasture (cotton would have had to be substituted for by wool). Crucially, these regions—cultivated by slavery, and allegedly irreplaceable by free labor—were reciprocally purchasers of British goods, which would otherwise have glutted the domestic market and reduced prices.
Whether or not substitutes for the Americas could have been found is another question entirely, though the research (some of which is to be discussed later) suggests that this is so. But was the ecological cork keeping the industrial genie bottled? It depends on how the question is asked. UK imports of cotton really took off after 1800, for example, and by this point, the Industrial Revolution as a technological phenomenon was well underway. Pomeranz dismisses the possibilities for flax and hemp—which would have required only 200,000 acres apiece in 1815 and 500,000 in 1830—as substitutes out of hand, considering them “inferior fibers” and overly labor- and manure-intensive. But the fact that they weren’t successfully planted does not indicate that they could not have been, at moderate cost, in an autarkic scenario. This is granting that wool-spinning would have been impossible to endogenously mechanize. Thomas’s consideration of the cereal constraint is certainly important. But for the early stages of the Industrial Revolution, the percentage of food imported was neither large nor foreign—England “imported” from Ireland, the domestic periphery. Ireland supplied 70 percent of grain, meat, and butter imports in 1815 and 85 percent in 1835, but imports at the latter date made up just 11 percent of total agricultural income.
This leads us to the third role: trade as catalyst.11 England’s external ties may not have started the Industrial Revolution, and it may have begun even without them. But there’s plenty of reason to suggest that the process would not have accelerated so quickly nor continued for so long in the absence of trade. Britain after 1800, and especially after 1850, began to hyper-specialize in the production of several high-value goods, including textiles, iron, and machine tools. She exported these products worldwide in exchange for raw materials and food, exploiting her Ricardian (not absolute) advantage in mechanized, labor- and capital-intensive industry. By 1876-85, Britain alone was exporting 38 percent of the planet’s manufactured goods, leading observers to claim that the island enjoyed a monopoly on the fruits of the Revolution. At the same time, 80 percent of her wheat supply came from abroad. Clearly severing these links would have had dramatic consequences for the structure of the nineteenth-century British economy.
One interpretation, presaged in Part 1, is incorrect: that rising international demand, expressed in buoyant exports, led British growth. This view is perhaps the most venerable of the four, expressed with clarity during the Belle Epoque by Alfred Marshall: “by the end of the nineteenth century the British economy was heavily dependent on world markets, and the rate and pattern of British economic growth was largely conditioned by the responses of producers and consumers in the rest of the world.” Britain was simultaneously the world’s most open and most advanced economy and the leading sector (textiles) sent two-thirds of its product abroad. Cuenca Esteban (1997) showed that exports accounted for 18 percent of gross industrial output in 1760, 40 percent in 1801, and a massive 49 percent in 1831. “To be sure, rising export shares alone cannot support the view that British industrial growth was export-led in any dynamic sense involving causation,” he wrote. “In the present state of empirical research, however, existing alternatives to the hypothesis of demand-led growth are not demonstrably sounder” (Cuenca Esteban 1997, p. 899).
So why didn’t export demand lead growth? Simply put, because prices fell thanks to technological progress more than foreign consumers shoved them up. Britain won foreign markets because her goods were cheap, not because their denizens were clamoring for shipments of British goods at any cost. Causality probably ran the other direction. Britain needed to export to pay for food and raw materials—for the US cotton shipments that rose from 189,000 lb in 1791 to 21 million lb in 1801 and 93 million lb in 1810. Producing these elements domestically would have sent input prices soaring in Britain, but abundant land reserves in North America ensured that cotton (and wheat) were supplied elastically. New World grain kept the Malthusian demon chained, and New World cotton pulled Britain clear. Elastic overseas demand absorbed excess output, allowing living standards to rise, but at increasing cost to producers as efficient methods passed the gains of the Revolution to foreign buyers.
The findings of Clark et al (2014) reinforce the argument for an Industrial Revolution perpetuated, but not started, by foreign trade. They experiment with the consequences for Britain in 1760-9 and 1850-9 of three scenarios: the loss of North American trade, the loss of “Rest of World” trade, and autarky. In the 1760s, as the Revolution was launched, the consequences of losing both the New and Old Worlds would have been relatively small. While real rents would have increased by 44.9 percent and real wages fallen by 13.9, overall utility would have dropped by less than 4 percent. This is inconsequential even compared with the sluggish growth of the late eighteenth century. By 1850, however, the picture was entirely different. While Britain could have done without North America, experiencing welfare losses between 1.6 and 3.6 percent, the loss of both regions would have plunged utility by 27 percent. Cotton production would have been just 41 percent of 1850 levels, while real wages and profits would both have declined by a third. The Global South could have grown the slave-picked cotton of the American South, but not Britain. Agricultural output would have risen by more than 42 percent, drawing labor and capital into low-productivity, slow-growing occupations.12
Apart from showing the need for foreign cotton, the paper also confirms the previous findings of Harley and Crafts (2000): that rising population (1.3%) in Britain combined with slowly growing agricultural output (.3%) necessitated extensive exports of manufactured goods in order to pay for food imports. The expansion and specialization in industry redounded to Britain’s benefit, as productivity in foreign-facing sectors rose. This circularity accelerated the country’s rise to dominance and efficiency in manufacturing. Britain would not have been impoverished by autarky, but growth after 1830 depended on the ability to exploit comparative advantages in textile, iron, and machine tool manufacturing at the expense of agriculture.
Pomeranz was wrong on timing. The key period for “relieving constraints” was after 1800, once the first Industrial Revolution was underway—and perhaps even coming to a close. Only with the post-1850 “grain invasion” from Eastern Europe and North America was the “ecological bottleneck” alleviated, but by then Britain was a mature industrial nation. Ghost acreages helped Britain to grow faster and longer, but didn’t start—and couldn’t stop—the process. The early twentieth century, which saw a peak in British food imports, coincided with increasing backwardness in domestic textile production and hesitant entry into the science-based industries—chemical, automobiles, and electricity—dominated by Germany and the United States. Excessive focus on cotton imports at this later stage would be misguided.
I tipped my hand in the last section. Trade may have been the child of industry, as McCloskey wrote back in 1981, but that child had to take care of the parent in late maturity and senescence. Substitution would have indeed have ameliorated autarky, but by how much? It’s unlikely that the beer industry could have taken up Lancashire’s slack in 1800, let alone 1850. Commerce was a consequence of the Industrial Revolution’s heroic age. It flourished because British innovations made British goods exceptionally cheap on world markets. To say that trade was merely a result of population growth and industrial expansion, however, would be as disingenuous as to claim that tweets are merely the result of Twitter. And to say the reverse, that England grew just by virtue of her commerce, would be as absurd as saying that Twitter exists because people tweet a lot.13
Trade is an important part of the story of the first Industrial Revolution. How could it have failed to be? The cotton textile sector that led the way exported two-thirds of its output and imported all of its primary input, and there were no comparable substitutes available. So trade is certainly a big reason why there was a Great Divergence and a Great Enrichment. But there would have been an Industrial Revolution without export markets and raw material supplies, and it would still have been British.
Pseudoerasmus has recently elaborated upon this distinction in a Medium post. He has also kindly allowed me to reproduce his own typology of trade in the Industrial Revolution:
1) Demand-side
2) Supply-side
capital accumulation
critical inputs
alleviation of Malthusian constraints
stimulation of technological change… (i) via market size effects or (ii) via high wages
urbanization & service sector infrastructure
Feel free to map this onto my schema as you will.
“[M]ost answers to the question posed by my title fall into one of two lines of explanation. One says that we are so rich and they so poor because we are so good and they so bad; that is, we are hardworking, knowledgable, educated, well-governed, efficacious, and productive, and they are the reverse. The other says that we are so rich and they so poor because we are so bad and they so good: we are greedy, ruthless, exploitative, aggressive, while they are weak, innocent, virtuous, abused, and vulnerable. It is not clear to me that one line of argument necessarily precludes the other, although most observers and commentators have a strong preference in the matter. What is clear is that, insofar as we may want to do something about the gap between rich and poor, each of these explanations implies a very different strategy” (Landes 1990, p. 1).
See H. J. Habakkuk and Phyllis Deane, “The Take-off in Britain,” in Walt Whitman Rostow, ed., The Economics of Take-off into Sustained Growth (London, 1962), pp. 77-78.
From on Findlay and O’Rourke (2007).
Another angle on the Williams thesis is that of former Berkeley economist Ellora Derenoncourt, whose 2018 working paper shows that a 10 percent increase in slaving voyages issuing from a port was associated with a 1.2 increase in population. The agglomeration advantages resulting from shipping and financial services, transport and warehousing, and infrastructure would have been useful across the British economy. Yet it is likely that industrialization could have begun without these boons and developed them endogenously as needed, as occurred outside the port cities.
Before his “Damascene Conversion” to an opposite view, represented in a 1999 article (not ungated) in the New Left Review. “During the eighteenth century, something like 40 to 50 per cent of Britain’s non-agricultural workforce produced directly—or indirectly—for markets overseas. There is no evidence that more than a fraction of these workers could have been absorbed into alternative employment for the home market at other than sharply reduced levels of productivity… Transcontinental trade—in close association with British imperialism created jobs, social overheads, capital and industrial capacity outside agriculture. Commerce beyond the realm supported by state power reinforced a more productive division of labour both within the kingdom and between the kingdom and its empire overseas.”
He did not fail to quote a line so over-used that I hope never to hear again: “It is not worth my while to manufacture your engine for three counties only, but I find it very well worth my while to make it for all the world” (Matthew Boulton to James Watt, 1769).
See the following paper by Judy Stephenson, for example: ‘Real’ wages? Contractors, workers, and pay in London building trades, 1650–1800.
Thomas’s other point revolves around the conservation of land through coal-using technologies, but this is tangential to our discussion (though worth reading alone).
For my own protection, I give Pomeranz’s own definition of “proto-industrialization”: “the massive expansion of nonmechanized industries, mostly composed of rural laborers producing for (often distant) markets through the mediation of merchants.”
I use the word “catalyst” in its biological, not its colloquial, sense: as the accelerant of a (chemical) process.
Clark et al do not estimate the effects of market size on innovation, which would lead the 1750 measure to be understated.
Credit to my dad, the consummate stylist, for this excellent line.