Part B: The Path from the Pre-Industrial World




-VIII. The First Global Economy: Production and Trade-



Return for a moment to the conceit that opened this book: a fourth-millennium professor teaching his or her class (or, perhaps more probably, downloading materials into xe’s students’ personal AIs), with a very limited amount of time and attention span in which to flag key concepts and ideas. What would such a professor think were the things that students should remember about the world economy at the start of the twentieth century?

Four things:

In the long run, however, the patterns of migration, of international investments, of the international division of labor, and of economic growth established in the decades before World War I would not last. They were destroyed by wars, politics, and changes in technology in the three decades after 1914. When the global economy was knit back together in the decades after World War II, it was knit back together in a different pattern–and now it is reweaving itself into yet a different pattern still.



A. Production and Technology

1. The first industrial nations

In some ways the world economy at the start of the twentieth century was still remarkably preindustrial. Most human beings still earned their bread out of the earth by the sweat of their brow. Most human beings could not read. Most human beings had not seen a steam engine up close, or travelled in a railway train, or spoken on a telephone, or lived in a city. For most human beings life expectancy was still low–little higher than it had been in most parts of the world since the neolithic revolution.

Even where things were different, as they were in the rapidly-growing half-industrialized core of the world economy, there was still a sense in which what we would call modern life was a thin and new crust on top of older patterns that still owed much to traditional agrarian and commercial patterns.


Social Statistics during the British Industrial Revolution


  1780 1820 1870 1913
Life Expectancy 35 39 41 53
Literacy 50% 54% 76% 96%
Primary School Enrollment   36% 76% 100%
Secondary School Enrollment     1.7% 5.5%
Agricultural Employment 45% 35% 23% 12%


In Great Britain alone was the economy primarily industrial at the turn of the twentieth century. And even in Great Britain the veneer of modernity was still only a veneer. It was true that the share of the labor force employed in agriculture was dropping toward 15% at the start of the twentieth century. The eve of World War I saw agriculture account for only twelve percent of the British labor force; while manufacturing and construction accounted for 38%; and distribution and services for 32%.

But a quarter of Britons were still illiterate as late as 1870. Primary school enrollment did not become universal until the eve of World War I. Life expectancy at birth was still fifty years or less. And less than five percent of the population went to secondary school. Britain’s–precocious–decline in the share of the labor force in agriculture suggested an economy more advanced and more industrialized than was in fact the case.

And Britain was by far the most advanced and industrialized of the world’s economies.


In the United States, and in Europe outside of Britain, farmers still made up the largest single occupational group. More than half the population still lived in the country, farming the land or providing the basic goods and services that farmers needed. Agriculture was still a very substantial share of GDP in the late-nineteenth century. It was only halfway through its long decline to its present role as a very small share of economic activity in industrialized economies.

In the American west, and in the other countries that Arthur Lewis named "regions of European settlement"–Canada, Australia, New Zealand, and Argentina–farming was not only the core of the economy but farmers were relatively rich, both compared to those dwelling in the cities and compared to those who had remained in Europe.

The eve of World War I still saw more than one out of three Americans at work at work in agriculture, and one in thirty at work in mining. And with the exception of Belgium, other European countries were much closer to the American than the British pattern in their distribution of the labor force between town and country, and among sectors. This turned out to have powerful implications for politics as World War I drew closer: too much political influence was still exerted by agrarian landlords who saw themselves as the descendents of knights who fought for their kings with their swords, and proved their worth through battle.


[Figure: American midwestern agriculture, ca. 1900]

In some other ways the economy at the start of the twentieth century was remarkably far advanced. On the eve of World War I, in 1913, Britain burned 194 million tons of coal. In 1987 it would only burn 116 million tons–and the total coal-equivalent energy consumption of Britain in 1987 would be only some two-thirds greater than on the eve of World War I. Energy would be used much more efficiently at the end of World War I than at the beginning, but the growth in total energy used at the leading edge of the world economy would be substantial but not spectacular.


[Picture: South Bank of the Thames at London, ca. 1900]


The U.S. railroads carried passengers some 35 billion miles in 1913: that's 350 miles per person, which indicates a sizeable use of modern technology for travel even then. (But by 1987 U.S. airlines would carry passengers some 400 billion miles in a year.)


Relatively cheap ways of making the fundamental building material of the twentieth century–steel–emerged in the second half of the nineteenth, with the invention Þrst of the Bessemer process and then of the Thomas-Gilchrist process. World steel production was some 70 million tons a year by 1913, and would grow to 170 million tons by 1950.

And people definitely saw what was coming. On the centennial of the storming of the Bastille during the Great French Revolution, France held a universal exposition. At the center of it was not some tableau of the martyrs of the French Revolution, but a construction of steel: the tower designed by and named after Gustave Eiffel that has dominated the Paris skyline ever since. As Donald Sassoon writes, the French Revolutionary centennial was transformed into a "… consecrat[ion of]… commerce and trade, modernity, and the wonders of technology exhibited in the Galerie des machines… Under the banner of modernity, progress, and the peaceful pursuit of wealth, the French people would regain national pride and unity after the humiliating defeat of 1870…"

Chemistry as we know it emerged from German universities and laboratories in the second half of the nineteenth century; by the eve of World War I Germany produced a full quarter of the world’s chemical output; rayon became an established competitor to silk in the Þrst decades of this century.


[Picture: Carnegie Steel]


On the eve of World War I some 1.7 million passenger cars were registered, and some 100,000 goods-carrying vehicles were registered in the United States. But the United States was far ahead of other countries in its use of internal combustion engine vehicles: there were only some 132,000 passenger cars in Britain.

[Picture: Model-T Ford

The years before World War I saw a large increase in education, as at least elementary school became the rule for children in leading-edge economies. And years of education grew as well. In countries like the United States that made the creation of a literate, numerate citizenry a high priority–and that encouraged those with richer backgrounds, better preparations, and quicker or better trained minds to go on to higher education–industrialists and others soon found the higher quality of their workforce more than making up for the taxes to support mass secondary and higher education. The U.S.’s edge in education was a powerful factor in giving the U.S. an edge in productivity–and Germany’s edge in education was a powerful factor in giving Germany an edge in industrial competitiveness as well.


[Figure: secondary education in America]



In the United States in 1910 some 355,000 were attending college, making up nearly Þve percent of their age cohort. In Germany in 1910 some 1,000,000 students were enrolled in post-elementary education.

Still, not all education was formal education. With the possible exception of Great Britain, a lack of technically trained and educated workers does not seem to have been a major contraint on economic or industrial growth in the years before World War I. Much necessary technological knowledge could be (and can today still be) learned on the job. And the higher wages and salaries paid to trained engineers and craftsmen induced the boom in education.

The new technologies of the end of the nineteenth century were associated with the rise of the modern corporate enterprise. By the end of the 1880s, industrial enterprises found themselves in the middle of a web of ocean steamship, land railroad, and telegraph communication systems that greatly multiplied their ability to order materials and ship products. The vastly expanded potential of the delivery system led to a vast expansion in the size of the enterprise. The expanded firms were more capital intensive than their predecessors, and their continued profitability required that the expensive capital be used to the utmost: coordination of the flow of inputs from suppliers and of output through distributors as well as of product through the factory itself. Such coordination could not just happen: it required professional management. The modern managerial enterprise–the profession of management itself–was born with the twentieth century.

2. Standards of living: the rich

Many of the processes that have blossomed since to make our industrial–post-industrial–economy were clearly underway by the start of the twentieth century. But they were for the most part only seedlings. In what matters most, in the warp and woof of everyday life, our counterparts in the industrial core of the world economy around 1900 still had more in common in their styles of life with their predecessors of 1600 or 1700 than with us today.

In 1902 an anonymous college professor wrote a four-page article for the Atlantic Monthly in which he pleaded for more money for college professor salaries, and claimed to be vastly underpaid. The Þrst thing to note is his salary: he claimed that the "average college professor’s salary"–the salary that he saw as clearly inadequate and unfairly low–"is about $2,000."

Yet $2,000 was four times the average of GDP per worker at the turn of the century. In order to match turn-of-the-century professors in terms of income relative to the national average, a professor today would have to make an academic salary of $220,000–a height rarely attained even by Nobel prize winners, and far above any average.

The second thing to note is that our professor sees himself as a reasonable man. He is not asking for what he would see as the "large salar[y], commensurate with what equal ability would bring in other lines of work ($10,000 to $50,000)"–or 20 to 100 times the then-current average level of GDP per worker. Today, 20 to 100 times average GDP per worker would be between $1,100,000 and $5,500,000 a year. The top 7,000 households in the United States today have incomes that average $500,000 a year. At 50 times average GDP per worker (roughly the mid-point of G.H.M.’s range, corresponding to a salary of $2.5 million a year), we are down to less than 1000 households in today’s United States.

That an ordinary professor could feel that his talents ought, in some sense, to earn such an enormous multiple of the average income is a sign of how unequal an economy and society the turn of the twentieth century U.S. was. Yet as this professor goes through his budget, he expects his readers to nod and we modern readers do indeed nod that his family is indeed strapped for cash.

The Þrst large expense he lists is for personal services: "We must pay $25 a month for even a passable servant" and add to that $10 a month for laundry, for the regular "servants will do no laundry work," $1 a month for haircuts, and $2 a month for a gardener. Already, on personal servies alone, we are up to $445 a year–roughly the average level of GDP per worker in 1900. But the professor sees himself as having no choice but to make such large expenditures on personal services. If he doesn’t, his household will fail to make a properly upper-middle-class impression: the lawn must be trimmed, the house dusted, the clothes cleaned, and the children washed. He has no gasoline-powered lawnmower, no electric hedge clippers, no vacuum cleaner, no dishwasher, and neither a washing machine nor a dryer. Consumer durables take the place now of what took servants’ sweat (at least for college professors’ households) a century ago.

Food bills average $55 a month on food–enough to buy 170 lbs. of veal cutlets (present market value perhaps $800), or 500 pounds of chuck roast (present value perhaps $1000), or 1000 lbs. of bread (present value perhaps $1100). Note that $55 a month works out to be $660 a year, once again considerably more than a year’s average GDP per worker on food alone. In general it was hard to economize on food at the start of this century: food and fuel consume almost half of consumer expenditure for the average household in 1885, but only a Þfth of consumer expenditure in 1987.

3. Standards of living: the rest

But before we start to worry too much about the plight of Professor G.H.M, who cannot afford to live within walking distance of campus and cannot afford to keep a horse and carriage and so must bicycle to work–we should reflect that working-class families at the turn of the century lived much more differently from Professor G.H.M. than working class families do from his successors today.

[Picture: Homestead, Pennsylvania]

A large proportion of households then had (largely male, unrelated) boarders sleeping and eating in the house: this provided a way for the household labor of the housewife to bring income directly into the household. It also made her life more difficult, and multiplied the amount of work she had to do. Few households had running water or a hot water heater. Water came in buckets from a faucet in the street into the house, and then heat it on the stove. In the–relatively prosperous for its time–factory steel town of Homestead, Pennsylvania at the start of the twentieth century, only one in six working class households had indoor bathrooms in 1910. Half of "Slav" and "Negro" families lived in one or two room houses. Most white families lived in four room houses.






American Housing 1900-1990. Share of Households with:

  1900 1990
Boarders and lodgers 0.25 0.02
Over 1 person per room 0.49 0.06
Over 3.5 persons per sleeping room 0.23 0.06
Running water 0.24 0.99
Flush toilets 0.13 0.97
Central heat 0 0.84
Gas or electric light 0.12 1
Refrigerator 0.18 0.99
Washing machine 0.05 0.7
Vacuum cleaner 0 0.92
Radio 0 0.96
Television 0 0.99
Telephone 0.05 0.91

But even if you have a four room house, can you afford to heat more than one room of it?

The diets of workers in Homestead, Pennsylvania at the turn of the century were composed primarily of meat of widely variable quality, bread, butter, potatoes, oatmeal, and tea and milk–with luxuries such as sweets added in more or less regularly. We would Þnd the diet somewhat monotonous (however, a lot of time and effort went into Þnding different ways to make potatoes).

Almost the Þrst luxury that a working-class family moving up would purchase would be the services of a laundress. Since laundry was expensive and difÞcult, few working-class families could maintain upper-middle-class standards of cleanliness. How often would you take baths if the water had to brought in from an outside pump, and then heated on the stove? How often would you wash your clothes if everything had to be washed out in the sink, if the fabrics were three times as heavy and the detergents one-third as powerful as the ones available today, and if as a result the laundry was a full day’s chore? Laundry was not a two hour a week but a ten hour a week task.

As a rule married women did not work outside the home–unless they were African-American, in which case they might well do their own family’s housework and be paid for doing a share of some white family’s housework as well. Meal preparation was not a one-hour-a-day but a four-hour-a-day task.

Barring a shift toward larger-scale communal or cooperative living–a shift which simply did not happen even though anticipated, hoped for, and worked for by many feminists–within-the-household production and maintenance soaked up one-third the potential adult work hours. It made it next to impossible for married women (unless they were quite rich, or quite poor) to have independent careers and still fulfill the social expectations of household maintenance.

Those who could afford the resources to maintain bourgeois styles of cleanliness flaunted it. White shirts, white dresses, white gloves are all powerful indications of wealth in turn of the century America. They said "I don't have to do my own laundry," and they said it loudly.

Infant mortality at the turn of the century was high. One in Þve babies in Homestead, Pennsylvania died before reaching his or her Þrst birthday. Adult men died, too, like flies (and adult women faced substantial risks in childbirth). Accident rates in the factory were such as to leave 260 injured per year–30 dead–out of a total population of 25,000 and a steel mill working population of 5,000. Each year, Þve percent were injured enough to miss work for some time (although only one percent per year were permanently disabled), and 1/2 percent per year were killed in factory accidents.

You can do the math.

Start to work for U.S. Steel when you are 20. There is one chance in seven that the factory will kill you before you reach 50, and almost one chance in three that the factory will disable you. Is it any wonder that life insurance–disability insurance–group lodges that provide beneÞts (because the company provides few)–loom so large in American working class consciousness at the turn of the century? And is it any wonder that the Þrst component of the welfare state put into place, in many parts of the United States, was workmen’s compensation?

Of course, in 1910 Homestead (or in 1930 Detroit, or in Los Angeles today) the most arduous and difÞcult jobs were done by minorities: in 1910 Homestead by Slavs, in 1930 Detroit by Blacks, and in 1990 Los Angeles by Hispanics. At the micro level, such groups are concentrated in the most arduous and lowest-paid jobs because they are poor, because they have limited other options.

Most of the Homestead workforce only worked six days a week: for four out of Þve workers, the mill was shut on Sundays. U.S. Steel viewed this–shutting most of the mill on Sundays–as a major concession on their part, a concession that they hoped would produce large public relations beneÞts. From U.S. Steel’s perspective, each hour that a modern plant like Homestead stood idle was tremendously expensive. Variable costs–wages, raw materials, and transportation–made up perhaps 2/3 of total costs. The remainder were Þxed: capital costs on the construction of the plant, and maintenance that had to be performed whether the plant was operating intensively or not.

Were U.S. Steel to move from two 12-hour shifts a day to one 12-hour shift, its output would be halved but its costs would be reduced by only 1/3-so total costs per ton of steel made would rise by 1/3. This was not a margin that U.S. Steel could afford. As long as it could Þnd workers willing to work the night shift, the Homestead mill (depressions and recessions apart) stayed open 24 hours a day on weekdays. And when things did change, they changed all at once-from two 12-hour shifts before and during World War I, to two 8-hour shifts (or three 8-hour shifts) during the 1920s, and during and after World War II.

Yet Homestead jobs–at least Homestead jobs taken by native-born Americans–were good jobs by the standards of the United States. As historian Ray Ginger puts it:

And Homestead, Pennsylvania jobs paid well both by the standards of the United States and the standards of the world economy of the time. White households could make around $900 (of 1910 value) a year, placing them well the upper third of the U.S. population in terms of income per household in 1910. Relative to what could be earned by people of similar skill levels anywhere else in the world, a job in the Homestead mill was a very attractive job. Even the unequal America at the turn of the century was a very attractive place compared to the rest of the world. America was exceptional. In spite of the hours, in spite of the risk of death or injury, in spite of the working conditions, these were very good jobs by international standards: jobs worth moving 7,000 miles for, from Hungary or Lithuania to suburban Pittsburgh. For the economy of the late nineteeth century was for the first time in human history a truly global economy, filled with long-distance trade and migration.


B. International trade

1. Before the late nineteenth century

Even before the U.S. Civil War of 1861-5, the export of cotton from the American south and of sugar from the Caribbean islands had shown how European agriculture could be replaced and augmented. After the Civil War the coming of the steam engine and the iron hull to ocean transport made possible the export to Europe not merely of cotton and sugar–crops which could not be grown in the climate of the European industrial core–but also grain, meat, and wool.


[Figure: Decline in the cost of ocean shipping, 1500 to 1900]


The extent to which the navies and trading fleets of the great European sea-borne empires of the sixteenth, seventeenth, and eighteenth centuries shaped the industrial development of western Europe has always been one of the most fiercely-debated and unsettled topics in economic history. That European expansion in the sixteenth, seventeenth, and eighteenth centuries had catastrophic consequences for the regions of west Africa that were the sources of the slave trade, or for the Indians of the Caribbean, or for the Aztecs, Incas, the mound-builders of the Mississippi valley, or the princes of Bengal is not in dispute. But how much did trade and plunder affect European development? That is not so clear.

The first wave of plunder from the Americas was made up of gold and silver. But gold and silver are not things that you can eat, or things that have much use as raw material for industrial or artisanal production. Most of the gold and silver that arrived from the Americas was promptly coined, and thus its principal effect was to drive up Europe’s overall price level as more money chased fewer goods, according to the quantity theory of money.

An important secondary effect of the large-scale plunder of gold and silver from the Americas was to redistribute purchasing power within Europe, away from northern European countries like Britain, Holland, and Germany and to southern European Spain, where the treasure fleets landed (save when they were captured by the British or Dutch navies). Because the kings and nobles of Spain could use their American gold to support their armies, fund their masses, and purchase their luxuries, Spain became more a country of soldiers and servants–living well, or at least supporting the armies of the Counterreformation, until the gold and silver ran out–while Holland and Britain became more nations of craftsmen, sailors, and merchants.

On the one hand, without American gold the armies of the Counterreformation would never have come so near their goals of suppressing Protestant heresies. On the other hand, demand for northern European products funded by American gold and silver changed relative prices and increased the returns to northern European entrepreneurship and enterprise. Which dominated? Did the arrival of treasure fleets from America in the end accelerate or retard northern European commercial development? I have seen no convincing arguments on either side for significant accelerating or retarding effects.

Another important secondary effect was that gold and silver from American treasure fleets made Europe’s price level relatively high, and thus made Europe’s goods too expensive to be worth purchasing by Asian consumers while it made Asia’s goods–spices, porcelains, textiles, teas–very much worth purchasing by Europe’s consumers. Thus the trade around Africa’s Cape of Good Hope from the sixteenth through the eighteenth centuries was overwhelmingly a trade of precious metals from Europe for useful material goods from Asia. To the extent that American gold and silver made Asia’s goods affordable and thus encouraged Europeans to develop navigational and other technologies that turned out to have large external benefits, American gold and silver may have played a small role in spurring European development.


[Figure: Trade fleet in the Guadalquiver]



But I have seen no convincing arguments that this was important enough to change the shape of Europe’s economy.

The second wave of products from the Americas were the staple plantation crops of tobacco and sugar, the mainstays of the triangle trade: sugar and tobacco to Europe, rum and weapons to Africa, slaves to the Caribbean. But tobacco and sugar were boosts to European consumption, not raw materials for further transformation in industrial enterprises. It is hard to see how this second wave of products might have shaped the structure of the European economies.

Things become different with the early nineteenth century, and the third wave of products from the Americas: cotton. North American slaves and Eli Whitney’s cotton gin together produced a cheap material input to an important manufacturing industry, and this time the industry was dynamic enough and the raw material important and cheap enough that the presence of the Americas did change the shape of the leading European economy, that of Britain.


[Figure: The Cotton South]



The availability of cheap American slave-grown cotton, as opposed to more expensive and more distant Egyptian and Indian varieties, may have saved Britain as much as four percent of national product in the reduced prices it had to pay for raw material inputs in peak years in the first half of the nineteenth century, and saved perhaps two percent of national product in reduced foreign materials prices in average years.


[Figure: Growth of cotton exports to Britain; growth of cotton consumption]


If the pattern of consumption had been maintained in the absence of slavery in the U.S. south, then this reduction in real national product would have come entirely out of investment–and would have reduced the pace of growth of the pre-Civil War British economy by perhaps 0.3% per year (cumulating to perhaps 15% over fifty years). If the cut had come proportionately out of investment and consumption, the reduction in growth would have been only 1/4 as large (cumulating to perhaps 4% over fifty years).

In an era in which British standards of living and levels of productivity are grew at roughly 1/2 a percent per year, the availability of slavery in the U.S. cotton south could have been responsible for between 15 and 60 percent of British economic growth in output per capita and per worker before the midpoint of the nineteenth century.


[Picture: Manchester 1850]



Yet cotton was a uniquely important good. And British imports of cotton were a uniquely strategic pressure point to apply to a pre- or an early-industrial economy. No other commodity, or set of commodities, had the potential to affect the destiny of any European economy in the years before the late nineteenth century. Trade was simply too small relative to domestic production for it to have been the prime mover or the balance wheel of any of these economies.



2. The first global economy

All this began to change in the early nineteenth century, when canvas sails began to be replaced by iron steam engines. By the end of the nineteenth century wooden hulls had been replaced by iron ones, message sloops had been replaced by submarine telegraph cables, and new methods of packing and preserving made it possible to transport a much broader range of commodities across oceans cheaply. The first transoceanic shipping of perishable organics came before 1850. Well before 1900 Europe’s beef was raised in Argentina, its mutton and wool was raised in Australia, and its butter raised in New Zealand.


Railway Mileage in the Late Nineteenth Century
























































European Russia
















International trans-oceanic trade was no longer limited to luxuries, rarities, drugs–tobacco and tea–and the occasional strategic, bulk, easily-shipped commodity like cotton. Instead, nearly anything could become the object of international trade.


[Figure: Ocean freight rates]


Moreover, the profits–even the profits counterfactually holding oceanic transportation costs constant–from long-distance international trade grew markedly as a result of two sets of changes in the world economy. The first was the industrialization of northwest Europe: all of a sudden northwest Europe had an enormous comparative advantage in making manufactured goods–and thus an enormous increase in its desired trade of its manufactures for the raw materials and agricultural goods of what was to be the periphery of the first global economy. The second such shock was what iron and steam did to the availability of natural resources out on the periphery. All of a sudden copper, coal, coffee, and all of the other mineral and agricultural products could be much more cheaply extracted and then shipped by rail to ports.

Two gigantic comparative advantages–in extracting resources and delivering them to ocean ports where they could be shipped, and in making manufactured goods–had been created from nothing. And lowered intercontinental ocean transport costs suddenly made it possible to take advantage of these comparative advantages through specialization: the core specializing in the manufactures that its superior access to industrial technologies allowed it to produce cheaply, the periphery specializing in the primary products that its new infrastructure allowed it to export.

Moreover, the British Empire meant that the crops that a region could grow were not limited to those that it had traditionally grown. During the nineteenth century the rubber plant came to Malaysia, the tea shrub came to Ceylon, and the coffee tree came to Kenya. The comparative advantages of the regions that were to become the periphery of the late nineteenth century global economy were not so much given as made: made by innovative and entrepreneurial use of mineral deposits and climatic zones.

Given these differential potential comparative advantages–to make manufactured goods in northwest Europe in the late nineteenth century, and to extract materials and grow crops elsewhere–the coming of low ocean transportation costs produced enormous pressure to specialize: to create a region-by-region international division of labor. Exports boomed even half a world away from Europe’s industrial revolution. Between 1870 and 1913 exports as a share of national product doubled in India and in what was to become Indonesia, and more than tripled in China. And in Japan–forced out of two and a half centuries of Tokugawa isolationism–exports rose from practically zero to 7 percent of national product in the generation of the Meiji Restoration.

Thus the second half of the nineteenth century saw the development of the first global economy based largely on an international division of labor derived from Ricardian comparative advantage.

What Arthur Lewis calls the "regions of European settlement"–the United States, Canada, Australia, New Zealand, Argentina, Chile, Uruguay, and perhaps South Africa–soon demonstrated that they could produce and ship staple grains, meats, and wool at vastly cheaper prices than could Europe itself. The steam-driven iron-hulled ship and the steam-driven iron-riding locomotive quickly brought down transport costs. From the 1870s German farmers found themselves with new competitors: not just new world producers, but Russian grain shipped from Odessa. In 1870 wheat cost nearly fifty percent more in Liverpool than in Chicago, meat cost more than ninety percent more in London than in Cincinnati, and iron cost seventy-five percent more in Philadelphia than in London. By 1913 all these prices were within twenty percent of one another.


[Figure: Total Volumes of world trade, 1700-1900]


By 1900 the harvest west of Chicago affected grain prices in Odessa and Hamburg; the price of lambs in Auckland affected meat prices in London; the agricultural sector was the first to become globalized–and in becoming globalized, to demonstrate that farmers could feed more people better than had ever been possible before. In commodity after commodity, prices drew tther in the years between 1870 and 1913. In 1870 it would have cost you 60 pre to buy wheat in Liverpool than in Chicago. By 1913 the price differential was down to 15 percent. In 1870 copper cost some 33 percent more in Philadelphia than it did in Londos. By 1913 the prices of copper in the two cities were almost exactly the same.





Price Gap by Year

Commodity Markets 1870 1895 1913
Wheat Liverpool/Chicago 0.576 0.178 0.156
Meat London/Cincinnati 0.925 0.923 0.179
Textiles Boston/Manchester 0.137 0.037 -0.036
Iron Philadelphia/London 0.750 0.434 0.206
Cotton Liverpool/New York 0.133 0.112 0.097
Copper Philadelphia/London 0.327 0.136 -0.001
Wool Boston/London 0.591 0.659 0.279
Tin New York/London 0.159 0.053 -0.023

Improvements in food processing and preservation meant that mutton from Argentina and Australia, beef from the American west, grain, and hides for leather could be produced on the praries or the pampas and consumed in the growing economies of Europe. A sharp division of international labor began to emerge: "temperate" settler colonies, like the western United States, Canada, Australia, and Argentina, produced and supplied grain, meat, leather, wool, and other high-value agricultural products to Europe and to the eastern, urban, industrialized United States; "tropical" regions, like Malaysia, Colombia, Cuba, Brazil, or Ghana (and to some degree the U.S. south), supplied rubber, coffee, sugar, vegetable oil, cotton, and other relatively low-value agricultural products to Europe; Europe paid for its imports by exporting manufactured goods--some 75% of Britain's exports were manufactured goods in the years before World War I, and textile exports made up half of Britain’s manufacturing exports.

By contrast, some 40% of United States exports in 1900 were food, feeds, and beverages; and a futher 35% were industrial supplies and materials. Industrial supplies and materials would rise to be fully half of exports by 1910.

Falling transport costs led to an integrated world economy. By 1910 exports of goods and services amounted to more than a quarter of British, and Australian, and Canadian, national product; and to perhaps a Þfth of German and Italian national product. Even the enormous and largely closed economy of the United States managed to import and export roughly Þve percent of national product in the years before World War I.

The social returns to the investments in technology and infrastructure that created this late nineteenth-century world economy were enormous. Robert Fogel calculated that the social rate of return on the Union Pacific Railroad’s trans-North American tracks and vehicles was some thirty percent per year.

The development of the integrated world economy brought new industries and new agricultural specialties to all the corners of the world. As William Ashworth points out, the rubber tree was not introduced into Malaysia, Indonesia, and Indochina until the last quarter of the nineteenth century. But by the end of World War I these three regions had become the principal sources of the world’s natural rubber supply. Tea was introduced to India and Ceylon through a similar process. And the world was covered with railroads: some 12,000 miles of railroads in Africa, 38,000 miles in Asia, and 26,000 miles in South America by 1900; some 40,000 miles of railroads in Africa, 80,000 miles in Asia, and 60,000 miles in South America by 1930.





3. Globalization and living standards

What effect did this enormous expansion in world trade have on standards of living across the globe?

Some believe that the explosion in world trade (and in international migration) put substantial downward pressure on the wages of labor in the labor-scarce land- and resource-rich parts of the periphery (North America, the southern cone of South America, Australia and New Zealand), gave a boost to real wages in labor-abundant Europe, and raised the returns to owners of capital on both the periphery and in Europe–in Europe because the development of international capital markets allowed European owners of capital to take advantage of high rates of profit on the capital-scarce periphery. Yet the proxies for economy-wide real wages assembled by O’Rourke, Taylor, and Williamson do not suggest that workers on the western, peripheral side of the Atlantic lost relative to workers in northwest Europe. American, Canadian, and Argentinian real urban worker wages appear to have grown at 1.0, 1.7, and 1.7 percent per year in the years leading up to 1914–compared to growth rates that averaged 0.9 percent per year in northwest Europe. Only in Australia, where real wages seemed to stagnate in the half-century before 1913, does the logic of economic theory seem to hold: only there does increased trade appear to erode the relative wages of workers in a labor-rich economy.

Outside of the North Atlantic economy the wage-raising impact of expanded international trade and migration on labor-abundant economies is even harder to see. In India and China there were no increases in real wages. And in countries to which workers from India and China were allowed to migrate–whether Malaysia, Kenya, or South Africa–increased potential competition seems to have been assoicated with stagnant or falling real wages in the late nineteenth century as well.

4. Slow diffusion of industrial technology

It is understandable that China, India, and the other regions of what would become the post-World War II third world did not produce and export the relatively high-value commodities like wheat and wool exported by temperate settler economies: agricultural productivity was too low, and climate was unfavorable. It is understandable why–with heavy downward pressure put on wages in Malaysia, Kenya, and Colombia by migration and threatened migration from China and India–the prices of the export commodities that they did produce were and remained relatively low.

What is more puzzling is why industrialization did not spread much more rapidly to the future third world in the years before World War I.

After all, the example of the industrial core seemed easy to follow. Inventing the technologies of the original industrial revolution–steam power, spinning mills, automatic looms, iron- and steel-making, and railroad-building–had required many independent strokes of genius. But copying the technologies did not, especially when you could buy and cheaply ship industrial capital goods made in the same New and Old England machine shops that supplied the industries of England and of America.

As industries in the industrial core became more and more mechanized–more and more characterized by mass production–they should have become more and more vulnerable to foreign competition from other, lower wage countries.Throughout the twentieth century, the U.S. had the highest wage level in the world; inside the U.S., Þrms devoted immense time, energy, and thought to redesigning their production processes so that lower-skilled, and lower paid, workers could replace highly-skilled craftsmen. One would think that manufacturing would have fled the United States. If Ford can redesign production so that unskilled assembly line workers do what skilled craftsmen used to do, why can’t Ford also--or someone else–redesign production so that it can be carried out by low wage Peruvians or Poles or Kenyans rather than by Americans, who are extraordinarily expensive labor by world standards?

Industries do migrate, but they have done so surprisingly slowly in the twentieth century. One reason is added risk: political risk of all kinds tends to make investors wary of committing their money in places where it is easy to imagine political disruptions from the left or the right. Moreover, there are substantial advantages for a Þrm in keeping production in the industrial core, near to other machines and near other factories making similar products. It is much easier to keep the machines running. A reliable electric power grid is much more likely to be found in the industrial core. And so are the services of specialists needed to Þx the many things that can go wrong: minimum efÞcient scale for an industrial civilization can be far larger than the apparent minimum efÞcient scale for a plant.

Arthur Lewis hypothesized that barriers to starting up an export-oriented industry were large, that infant industries on the periphery of the world economy had to rely on domestic demand, and that where domestic demand was low because of mass poverty modern industry could not flourish. Thus only a small share of output in what was to become the third world came from the industries of the industrial revolution.

Certainly attempts to industrialize in what is now the third world in the early (and, save for Japan, the late) nineteenth century were not successful. The machines could be bought, shipped, and installed. The domestic market could (in some cases at least) be protected to provide captive consumers to pu the products of industry. Managers could be found. Yet in Egypt, in India, in Mexico–even in Italy and in Spain–attempts at the periphery to copy what was going on in the industrial core were unsuccessful.

But we still understand far too little about why the pace of technological diffusion out of the industrial core was so slow back before World War I: why "peripheral" economies did such a good job at specializing in plantation agriculture for export, and such a bad job at creating modern manufacturing industries.

Gregory Clark has counted the staffing levels–how many operatives for each machine–at textile firms worldwide early in the twentieth cenury, and found enormous differences in how many workers watch, operate, and maintain the same machine across countries and continents. It is not that places where labor is abundant use the same machines more efficiently: it is that it appears to take many times the workforce to achieve the same level of machine performance. Workforces in the industrial core appear to have an acquaintance with machines and how they work, which was very, very hard to duplicate in the periphery. It is as if one’s social capital allows an economy to overleap one or perhaps two stages of technological development, but no more: and the gap between the quality and sophistication of machinery needed to be competitive in the late nineteenth century and the largely-handicraft traditions of the periphery appears to have been just too large.

The World Textile Industry, 1910: Weekly Wages and Staffing Levels

Country or Region

Weekly Wage (Dollars)

Looms per Worker

Ring Spindles per Worker

New England 8.8 2.97 902
Canada 8.8 2.53 750
U.S. South 6.5 2.65 770
Britain 5 2.04 625
Germany 3.8 1.28 327
France 3.7 1.11 500
Switzerland 3.7 1.4 450
Austria 2.8 1.24 403
Spain 2.7 0.91 450
Mexico 2.6 1.15 540
Russia 2.4 1.1 450
Italy 2.4 0.88 436
Portugal 1.72 0.88 384
Egypt 1.69 0.81 240
Greece 1.38 0.46  
Japan 0.8 0.53 190
India 0.78 0.5 214
China 0.54 0.48 168
Peru   1.17 391
Brazil   0.88 527

5. International Migration:

The fall in the price of carrying goods across oceans carried with it a fall in the cost of carrying people across oceans as well. And people did move. International migration has played a signiÞcant role in raising and beginning to equalize material wealth over the past century. Those presently living in Ireland are, compared to Great Britain and the United States, relatively poor. But the descendants of those who lived in Ireland at the start of the nineteenth century are, today, one of the richest groups in the world: less than half of the descendants of the Irish of 1800 live in Ireland today; instead, they are spread throughout America, Britain, and Australia, and they have prospered.


[Descendents of the mid-nineteenth century Irish. Where and how rich]

The half century before 1925 saw perhaps one hundred million people moved from one continent to another in search of a better life. About Þfty million left Europe, largely eastern and southern Europe, for Australia, and the Americas. Perhaps Þfty million (although we are not really sure) left China, India, and other Asian countries for destinations in the Americas, in lands surrounding the South China Sea, and in east Africa. Peru in the late twentieth century could have a President surnamed Fujimori. The author V.S. Naipaul was born not in India but in the Caribbean. The redwood forests of northern California contain shrines to the boddhisatva Guan-Yin.

Tension between descendants of peoples whose ancestors had resided in the areas for somewhat longer (after all, ultimately all humans are indigenous to Africa) and descendants of migrants from China and India has dominated the politics of many countries in the twentieth century. And since World War I migration has been tightly restricted by national governments, and population flows have been much smaller as proportions of the total world population.


[Migrations, gross and net, 1870-1930]

But the roughly one hundred million migrants of 1870-1925 made up one-tenth of the world’s population in 1870. Because the migration stream contained relatively few children and few old people, the 1870-1925 intercontinental migration stream amounted to perhaps one one out of every seven people of working age.

Even before 1925, there were substantial restrictions on the migration of Asians to areas that Europeans considered "their" province of settlement. One of the most popular causes in late nineteenth century America was the restriction of immigration from China and Japan. Railroad barons wished to continue the expansion of the Asian-born population in America. Workers and populists wanted the Chinese, Japanese, and (Asian) Indians kept out of California and on the other side of the PaciÞc. The plutocrats like Leland Stanford (the railroad baron and governor of California who founded and endowed Stanford University in memory of his son) favored immigration; the populists favored exclusion–and "Chinaman go home."

By and large, the populists won. Asian immigrants were largely kept out of what Arthur Lewis calls the "temperate countries of European settlement"–the United States, Canada, Argentina, Chile, Uruguay, Australia, and New Zealand. The flow of migrants out of China and India was directed elsewhere, to the tea plantations of Ceylon or the rubber plantations of Malaysia. Arthur Lewis believes that this redirection of the migration stream had enormous consequences for the distribution of income in the twentieth century world. Europe had escaped the Malthusian trap of low living standards and populations high relative to agricultural resources and technology at perhaps the end of the eighteenth century. The availability of resource-rich settlement areas like Canada and Argentina with Europe-like climates provided a further boost to European living standards: industrializing European countries at the turn of the twentieth century found their land/labor and capital/labor ratios, and thus their productivity levels and living standards, rising as migrants left for America.

India and China, through ill-luck and bad government, had not escaped the Malthusian regime. Technology had advanced: the population of China in the late nineteenth century was some three times what it had been at the start of the second millennium, and living standards were no (or not much) lower. But improvements in productive potential had been absorbed in rising populations, and not in rising living standards.

So potential migrants from China and India were willing to move for what seemed to Europeans to be starvation wages.

Thus the large populations and low levels of material wealth and agricultural productivity in China and India put downward pressure on wages in any of the areas–Malaysia, Indonesia, the Caribbean, or east Africa–open to the Asian migration stream. Workers could be cheaply imported and employed at wages little above the physical subsistence level. These workers would be very happy with their jobs: their opportunities and living standards in Malaysian or African plantations would be far above what they could expect if they returned to India or China.

Low wage costs meant that commodities produced in countries open to Asian immigration were cheap. And competition from the Malaysian rubber plantations pushed down wages in the Brazilian rubber plantations as well. The late nineteenth century saw living standards and wage rates become and remain low (although higher than in China and India) throughout the regions that were to come to be called the third world.

Conversely, the restriction of migration to temperate latitudes to European natives meant that the prices of temperate agricultural commodities–like wheat, beef, and wool–would be relatively high because wages had to be high enough to lure Europeans, with agricultural productivity levels three or four times those of China or India, off the farm and across the ocean. Save for cotton (grown by African-American sharecroppers living at standards closer to physical subsistence than the rest of America cared to know about or cares to remember), temperate economies simply did not produce any of the commodities that could be produced in regions open to Asian migration: they could not compete. Instead, the temperate settler economies concentrated on the resource- and technology-intensive agricultural and mineral products that could nnot be produced closer to the equator.

The politically-set pattern of migration ensured that one set of countries would be relatively rich, and another set relatively poor, as of the beginning of World War I. Since 1900 destinies have diverged further. In most vicious and virtuous circles have acted to push them further toward the nearest edge of the world’s relative income distribution. But some have followed aberrant and surprising trajectories through the world income distribution. The countries in the southern half of South America were first world nations in 1900. They are not so today. Japan, with in 1900 a relatively poor developing economy, is now one of the leading industrial powers.

It is impossible to determine what the world would be like if there had not been substantial restrictions on poor people–Asians–moving to rich countries–Europe and its overeas settler colonies–throughout the twentieth century. California would certainly be very, very different if migration from Asia had been allowed to continue up until the general restrictions on all immigration were imposed in the mid-1920s.

It might have been easier for poor people to move to rich economies than it has proven to be to transfer the political institutions and economic technologies from rich to poor economies in the twentieth century. If so, the world would today be a more equal and a richer place if not for the white Australia and analogous policies of the pre-World War I era, and for the tight restrictions on all kinds of immigration imposed from the 1920's on.

Alternatively, the institutions of political democracy and the capitalist economy in the rich settler countries might have collapsed under the strain of coping with more massive immigration flows, and the resulting increased degree of internal inequality–or so has always been the argument of those favoring immigration restrictions.