Based on W. Arthur Lewis, The Evolution of the International Economic Order (Princeton University Press, 1978)
W. Arthur Lewis has been called a pioneer and a founding father of development economics. In 1979 he won the Nobel Prize in Economics for his work in this field. He was the first Black person to win a Nobel Prize other than the Peace Prize.
Lewis had a deep interest in economic history. He believed that there were strong parallels between the economic development of his own time and the economic development of the late nineteenth and early twentieth centuries. His book on this history, The Evolution of the International Economic Order, was described by William Diebold Jr. as having “more ideas than pages,” high praise in the fussy, cross-every-t-twice world of academics.
How did the world come to be divided into industrial countries and agricultural countries? Did this result from geographical resources, economic forces, military forces, some institutional conspiracy, or what?
Lewis opens his book with these questions. This division occurred in the final decades of the nineteenth century, when countries adopted specialized roles within an increasingly integrated world trade system.
Long-distance trade has a very long history — the Roman Empire imported silks from China — but the volume of trade was small before the nineteenth century. Trade was restricted to sugar, spices, slaves, precious metals, and a few luxury goods such as porcelain and silk. These things had high value relative to their bulk, making them profitable to move despite the high cost of transportation.
Two revolutionary modes of transportation, the steamship and the railroad, led to drastically lower transportation costs in the late nineteenth century. The replacement of sail with steam made for speedier voyages. In 1883 the clipper Samuel Plimsoll sailed from England to Australia in 72 days. Just five years later the steamship Australasian made the same passage in 50 days. As well, the replacement of wood hulls with iron hulls allowed bigger, more fuel efficient ships to be built.1 These two factors, coupled with the continuing optimization of ship design, caused the cost of ocean transportation to fall by two-thirds between 1870 and 1900. Refrigerated ships, which first appeared in 1876, allowed perishable commodities to be traded for the first time.
Railways were built in Latin America and Asia a few decades later than they were in Europe. As in Europe, the railways substantially reduced the cost of land transport. Goods that had previously been trapped in the interior could now be cheaply carried to ocean ports.
High transportation costs had not been the only thing that had impeded long-distance trade. The industrialized part of the world had had little to gain from external trade before the late nineteenth century.
The leading industrial countries — Britain, the United States, France, and Germany — were, taken together, virtually self-sufficient. The raw materials of the industrial revolution were coal, iron ore, cotton, and wool, and the foodstuff was wheat. Between them, these core countries had all they needed except for wool. Although many writers have said that the industrial revolution depended upon the raw materials of the Third World, this is quite untrue. Not until what is sometimes called the second industrial revolution … did a big demand for rubber, copper, oil, bauxite and such materials occur. The Third World’s contribution to the industrial revolution of the first half of the nineteenth century was negligible.2
In the late nineteenth century, however, the opportunities for trade between the industrial countries and the non-industrialized countries were rapidly expanding.
Not Following the Leader
One hundred years after the Industrial Revolution, the non-industrialized countries far outnumbered the industrialized ones. Britain was the only country in which less than half of the population worked in agriculture. North America and Western Europe had industrialized, but there was little industrialization in Eastern and Central Europe, and still less in more distant places. Why had so few countries followed the example of the leading countries?
Lewis believed that most countries were capable of industrialization: “The new ideas were ingenious but simple and easy to apply. The capital requirement was remarkably small, except for the cost of building railways.”3 He might have been overly sanguine on this point. Pollard has argued that in Continental Europe, which shared a scientific and technical culture with England, the inventions of the Industrial Revolution were either adopted without change or not at all (here) — and even adoption without change was very difficult.
The first trial of a new machine or a new process, even when using British equipment or British specialists, almost invariably failed, even though it might have been operating in Britain for a generation or more — the extreme example, perhaps, is coke-smelting, started up at Le Creusot by William Wilkinson, a leading British ironmaster, in 1785, about seventy-five years after the first successful experiments and over forty years after its widespread adoption in Britain; it failed, and it was over thirty years before it was tried again in France, even in areas where costs were favourable. Textile machinery failed with monotonous regularity at first trial, and again many years might elapse before another trial would be attempted, and a further delay would occur before the new [technology] had firmly taken root.4
Technology transfer to, say, China, with its altogether different environment and fleeting contact with Europe’s technical culture, would have been even more difficult.
Lewis looks elsewhere for an explanation. One possibility that he immediately discards is that imperial powers discouraged industrialization in their colonies. They did do so, of course, but this factor alone doesn’t explain the evidence.
When the coffee industry began to expand rapidly in Brazil around 1850, there was no external political force from Europe or North America that made Brazil develop as a coffee exporter instead of as an industrial nation. Brazil, Argentina, and the rest of Latin American were free to industrialize, but did not. India, Ceylon, Java, and the Philippines were colonies, but in 1850 there were still no signs of industrialization in Thailand or Japan or China, Indo-China or the rest of the Indonesian archipelago. The partition of Africa did not come until 1880, when the industrial revolution was already a hundred years old. We cannot escape the fact that Eastern and Southern Europe were just as backward in industrializing as South Asia or Latin America. Political independence alone is an insufficient basis for industrialization.5
Lewis instead argues that an industrial revolution requires a prior or contemporaneous agricultural revolution.
The distinguishing feature of the industrial revolution at the end of the eighteenth century is that it began in the country with the highest agricultural productivity — Great Britain — which therefore already had a large industrial sector. The industrial revolution did not create an industrial sector where none had been before. It transformed an industrial sector that already existed by introducing new ways of making the same old things. The revolution spread rapidly in other countries that were also revolutionizing their agriculture, especially in Western Europe and North America. But countries of low agricultural productivity, such as Central and Southern Europe, or Latin America, or China, had rather small industrial sectors, and there it made rather slow progress.6
An agricultural revolution served two functions. First, it allowed labour to be shifted from agriculture to industry, so that the industrial sector could expand. Second, the demands of affluent farmers created a significant domestic market for manufactured goods. Nascent manufacturers could begin by selling their goods in this market, perhaps protected by a tariff wall. When they had become efficient and low-cost producers, they could begin to contest foreign markets.
The other prerequisite for industrialization, according to Lewis, is a capitalist sector. This sector could be either industrial or agriculture. Its defining feature was entrepreneurs who hired labour in order to produce goods for sale, “men who think in terms of investing capital productively.”7 The capitalists provided a pool of savings to finance industrialization, and they also developed the financial and legal institutions necessary for an industrial economy.
Western Europe had been creating a capitalist environment for at least a century; thus a whole new set of people, ideas and institutions was established that did not exist in Asia or Africa, or even for the most part in Latin America, despite the closer cultural heritage. Power in these countries — as also in Central and Southern Europe — was still concentrated in the hands of the landed classes, who benefited from cheap imports and saw no reason to support the emergence of a new industrial class.8
The ideological divide between industrialists and landowners was evident in the American debate over tariffs during the nineteenth century. The industrialists of the North wanted tariffs to protect new industry, while the landowners of the South saw tariffs as simply raising the prices of the goods they bought. The United States ultimately chose to maintain a substantial tariff wall against imports, pleasing the North and providing the South with another source of dissatisfaction.
By 1870 the world had split between manufacturers and primary producers — and another split was yet to come.
Rich Temperate Countries and Poor Tropical Countries
For the countries that had not industrialized by 1870, trade with the industrial countries was a promising opportunity. The costs of transportation were falling, and the industrial countries were increasingly interested in importing both food and raw materials. The impact of trade varied across the non-industrialized world. It was determined, Lewis says, by two large-scale migrations that provided the labour needed to satisfy the industrial countries’ burgeoning demands for primary products.
About 50 million people left Europe for temperate countries, notably the United States, Canada, Argentina, Chile, Australia, New Zealand and South Africa. These are the countries that Alfred Crosby called “neo-Europes” in Ecological Imperialism. Their climates are so similar to those of Europe that European agriculture can be practiced in them — and a great number of the migrants did exactly that. The single biggest source of migrants was Britain, where farms yielded 1600 pounds of wheat per acre in 1900. British farmers would only migrate if the farms in their new homes generated a higher standard of living than the farms in Britain. This requirement was met in the United States, where somewhat lower yields were offset by greater mechanization, and it had to be met in other destinations for those destinations to be competitive with the United States.
These migrants were not the wretched refuse of teeming shores: they were entrepreneurs who made calculated decisions. The prices they received for their goods had to be high enough to justify production.
Since Northwest Europe needed first their wool, and then after 1890 their frozen meat, and ultimately after 1900 their wheat, it had to pay for those commodities prices that yield a higher-than-European standard of living.9
The second migration also involved about 50 million people. They came from India and China to work as indentured labourers on plantations, mines, and infrastructure projects in the tropics. For most of these workers, the alternative to migration was farming in the home country, and there, the yield was only 700 pounds of grain per acre. They became indentured labourers to obtain a standard of living somewhat better than they could get by farming at home.
A European would migrate to a temperate country to obtain an income somewhat better than he could get farming in Britain, where the yield was 1600 pounds of grain per acre. An Indian or Chinese labourer would move to the tropics to obtain an income a little better than he could get farming at home, where the yield was 700 pounds of grain per acre. Since a British farmer cultivated more land than his Indian or Chinese counterpart, the income demanded by a European migrant was several times greater than that demanded by an Indian or Chinese migrant. This difference was reflected in the wage rates of the time:
In the 1880s the wage of a plantation laborer was one shilling a day, but the wage of an unskilled construction worker in Australia was nine shillings a day.10
The cost of labour largely determined the prices of primary products. In the temperate countries, the prices of wool and wheat had to be high enough so that their producers lived as well as British grain farmers. In the tropical countries, the prices of peanuts and rubber and tea only had to be high enough to ensure that the plantation workers lived as well as Indian or Chinese farmers. Lewis stresses this point:
If tea had been a temperate instead of a tropical crop, its price would have been perhaps four times as high as it was. And if wool had been a tropical instead of a temperate crop, it could have been had for perhaps one-fourth of the ruling price.11
Temperate countries got high prices for their exports and gained a great deal from trade with the industrial core. Tropical countries got low prices for their exports and gained much less.
Whether a good became a “tropical export” or a “temperate export” was often determined by the differential labour costs. Temperate exports were goods in which the temperate countries’ productivity advantage outweighed the tropical countries’ labour cost advantage. Tropical exports were goods in which the tropical countries’ labour cost advantage outweighed the temperate countries’ productivity advantage, or in which the tropical countries had the productivity advantage.
The productivity advantage was sometimes innate. Rubber was exported from countries in which the rubber plant grows readily, and the same was true of tea and coffee. Beef and wool were exported from countries in which grazing land was readily available, and timber was exported from countries in which dense forests were still standing. In other cases the productivity advantage reflected differences in technology or in capital intensity. North America had an advantage in grain production, for example, because it had great expanses of arable land that were worked with large amounts of machinery. There was a synergy between these two advantages: large farms could not have been worked without machinery, and machinery on small farms would have been unprofitable.
Meat, wool, timber, and grain became temperate exports because cheap labour could not offset the productivity advantage. Crops that were grown on plantations became tropical exports because they required large amounts of labour.
The tropical countries shared the advantage of cheap labour, so within this group of countries, a particular good would only be exported by the countries with the greatest productivity advantage. Many tropical countries relied on the export of just one or two goods, as the table below shows.12 These countries would be badly hurt by the sharp decline in commodity prices that occurred in the early twentieth century.
There were a few goods that were produced in both temperate and tropical countries. Although mining was a labour-intensive enterprise, a mine’s profitability was largely a matter of ore grade, and mining occurred wherever the ore grade was sufficiently high. Cotton was also a labour-intensive activity, and yet the United States was a major producer. The United States was only competitive because it had its own pool of cheap labour, the Blacks who had been emancipated by the Civil War. They were held in place by Jim Crow laws, the Ku Klux Klan’s brutality (including lynching, which continued into the 1960s), poor access to education and financial capital, and the suppression of their voting rights. Although these things were undoubtedly motivated by the South’s racial animus, they also kept American cotton profitable until it was finally mechanized in the 1950s.
Import Substitution and Late Industrialization
Trade made the temperate countries prosperous, creating a demand for a wide range of manufactured goods. This demand was initially satisfied by imports, but it created a new opportunity for industrialization: new industry could develop to displace imports, perhaps with the help of a tariff wall. Some countries took this opportunity but others did not. Lewis argues that different power structures led to different outcomes:
The contrast between Argentina and Australia is particularly instructive. These two countries began to grow rapidly at the same time, the 1850s, and sold the same commodities — cereals, wool, and meat. In 1913 their incomes per head were among the world’s top ten. But Australia industrialized rapidly, and Argentina did not, a failure which cost her dearly after the war when the terms of trade moved against agriculture…The crucial difference between the two countries was that Argentinian politics was dominated by the old, landed aristocracy. Australia had no landed aristocracy. Its politics were dominated by its urban communities, who used their power to protect industrial profits and wages. The slowness with which industrial classes emerged in Latin America, or Central Europe, North Africa or much of Asia is explained as much by internal social and political structures as by the impact of external forces.13
Economists are, to some extent, storytellers. They look for simple stories that bring coherence to the evidence before them. The story that Lewis tells has two basic elements. The first element is textbook economics: people will only move from one activity to another if the rewards of the new activity are at least as great as the rewards of the old activity. British farmers migrate to South Africa only if they can expect a better standard of living there. Brazilians switch from raising cattle to growing coffee only if coffee growers live as well as ranchers. The second element is that there was such a large pool of potential indentured labourers that the tropical economies could expand without driving up the wage paid to this kind of labour. Since the wage determines the prices of exports, anchoring the wage anchors the prices, limiting the tropical countries’ gains from trading with the industrial core.
Lewis’s story is useful because it explains many things, but there is a little sleight of hand in it. The cost of operating a plantation isn’t just the cost of the labour. There is also the cost of plant and equipment, as well as the cost of the land itself. These capital costs have to be financed, and a plantation owner could find himself competing for financing with English and American industrialists. He must earn a return large enough to cover these costs. As well (the first element reprised), entrepreneurs choose among ventures: they won’t start a plantation if cutting timber is more profitable. These considerations, taken together, imply that fixing the wages of indentured servants doesn’t completely fix the price of plantation crops. These prices had to rise with the tightness of the capital markets and with the profitability of other activities. Lewis probably understates the tropical countries’ gains from trade.
A better understanding of these gains can be obtained by looking directly at the terms of trade, which is what Jeffrey Williamson has done over the last decade or so. His book Trade and Poverty summarizes the state of this research (discussion here).
- Size and fuel efficiency rise together because cargo capacity increases with the cube of the ship’s dimensions, while fuel consumption increases only with the square of the dimensions. ↩
- W. Arthur Lewis, The Evolution of the International Economic Order, p. 5. ↩
- W. Arthur Lewis, The Evolution of the International Economic Order, p. 7. ↩
- Sidney Pollard, Peaceful Conquest: The Industrialization of Europe 1760-1970 (Oxford University Press, 1981), p. 146. ↩
- W. Arthur Lewis, The Evolution of the International Economic Order, p. 9. ↩
- W. Arthur Lewis, The Evolution of the International Economic Order, p. 10. ↩
- W. Arthur Lewis, “Economic Development with Unlimited Supplies of Labour,” The Manchester School (1954), p. 159. ↩
- W. Arthur Lewis, The Evolution of the International Economic Order, pp. 10-11. ↩
- W. Arthur Lewis, The Evolution of the International Economic Order, p. 15. ↩
- W. Arthur Lewis, The Evolution of the International Economic Order, pp. 15-16. ↩
- W. Arthur Lewis, The Evolution of the International Economic Order, p. 16. ↩
- This table is excerpted from Table 4.2 in Jeffrey Williamson, Trade and Poverty (MIT Press, 2013). ↩
- W. Arthur Lewis, The Evolution of the International Economic Order, p. 25. ↩