Global Migration, 1846-1940

Author(s): Adam McKeown

Source: Journal of World History , Jun., 2004, Vol. 15, No. 2 (Jun., 2004), pp. 155-189
Published by: University of Hawai'i Press on behalf of World History Association

Stable URL: https://www.jstor.org/stable/20068611

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

World History Association and University of Hawai'i Press are collaborating with JSTOR to
digitize, preserve and extend access to Journal of World History

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 Global Migration, 1846-1940*
ADAM MCKEOWN
Columbia University

Mass long-distance migrations

modern world history, have
but historians have beenbeen
slow to an important part of
acknowl
edge their global extent. Movement across the Atlantic is recognized
as a critical aspect of industrialization and expansion into American
frontiers, but migrations that were part of similar demographic and eco
nomic transformations in north and southeast Asia are largely ignored.
Asian and African migrations, when mentioned, are usually described
only as indentured migration subject to the needs of Europeans or as
peasants fleeing overpopulation pressures, quite different from the free
migrants that transformed the Atlantic world. But migrations across
the globe were broadly comparable in size and timing. These similari
ties are not coincidental. The frontiers of Manchuria and the rice fields
and rubber plantations of Southeast Asia were as much a part of the
industrial processes transforming the world as the factories of Man
chester and the wheat fields of North America. Power and capital were
centered in the North Atlantic, but massive migration flows often took
place beyond the direct influence of Europe.
From a global perspective, the usual periodization in which the age
of mass migrations ended in 1914 is not appropriate. World migration

* I am indebted to suggestions and assistance from Jerry Bentley, Jeffrey Burds, Adam
Kosto, Greg Mann, Patrick Manning, David Northrup, Pablo Piccato, Qiu Liben, Sam Rob
erts, Elizabeth Sinn, Anand Yang, Bin Yang, an anonymous reviewer for the Journal of World
History, and audiences at the World History Center of Northeastern University, June 2002,
the history department at the University of Canterbury in Christchurch, New Zealand, July
2002, and the World History Association Conference in Seoul, Korea, August 2002. All
remaining errors and poor conceptualizations are solely the result of my own obstinacy in
the face of good advice.

Journal of World History, Vol. 15, No. 2

? 2004 by University of Hawai'i Press

155

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 i56 JOURNAL OF WORLD HISTORY, JUNE 2OO4

reached new peaks in the 1920s, and the immigration restrictions of

the 1920s were also part of much longer trends of regulation, border
control, and nationalism that had grown concurrently with migration
since the middle of the nineteenth century. In fact, the segmentation
of global migration into different systems, which has facilitated the
ability to focus only on transatlantic migration and ignore the rest of
the world, was as much a consequence of political intervention into
migration as of economic processes. Thus, a global perspective on
migration provides insight not only into the global reaches of an
expanding industrial economy, but also into how this integrative econ
omy grew concurrently with political and cultural forces that favored
fragmentation into nations, races, and perceptions of distinct cultural
regions.

Overview of World Migration

Table 1 offers an overview of the three main circuits of long-distance

migration from 1846 to 1940. It is based on immigration, emigration,
and customs statistics from around the world (the appendix contains
a detailed review of sources). They count mostly ship passengers who
traveled in third class or steerage, or people such as migrants from India
who were categorized under bureaucratic definitions of "emigrants" or
"laborers," or migrants who registered under officially sponsored colo
nization schemes such as those from Russia to Siberia and central Asia.
Most of the available statistics can be classified according to three main
destinations: the Americas, the broad expanse of North Asia stretch

Table 1. Major long-distance migration flows, 1846-1940

Destination Origins Number Auxiliary origins

Americas Europe 55-58 million 2.5 million from
India, China,
Japan, Africa
Southeast Asia, India, southern 48-52 million 4 million from
Indian Ocean China Africa, Europe,
Rim, South northeastern
Pacific Asia, Middle East
Manchuria, Siberia, Northeastern 46-51 million
central Asia, Asia, Russia
Japan
Sources: See Appendix.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
McKeown: Global Migration, 1846-1940 157

ing from the Russian steppes to Siberia and Manchuria, and a region
centered on Southeast Asia but extending across the rims of the Indian
Ocean and the South Pacific. The majority of migrants to each desti
nation also came from similar origins, although the column of "auxil
iary origins" shows migrant flows from other places.
These classifications are intended to suggest the larger trends and
broad comparability of long distance migration around the globe.
Nonetheless, the table is highly selective. It does not account for much
migration through Africa, western Asia, or within each of the sending
and receiving regions. Much of this was short-distance or overland
migration for which statistics are not always readily available (although
some flows, such as those from Europe to North Africa, are well docu
mented), and will be discussed in the next section.
The transatlantic migrations to the Americas are the best known
of these migrations. Over 65 percent of these migrants went to the
United States, with the bulk of the remainder divided between
Canada, Argentina (which had the largest proportion of foreign-born
residents), Brazil, and, to a lesser extent, Cuba. Over half of the emi
gration before the 1870s was from the British Isles, with much of the
remainder from northwestern Europe. As migration increased along
with new transportation technologies in the 1880s, regions of intensive
emigration spread south and east as far as Portugal, Russia, and Syria.
Up to 2.5 million migrants from South and East Asia also traveled to
the Americas, mostly to the frontiers of western North America or the
plantations of the Caribbean, Peru, and Brazil. Half of this migration
took place before 1885, after which the decline of indentured labor
recruitment and the rise of anti-Asian immigration laws began to take
effect.
Migration to Southeast Asia and lands around the Indian Ocean
and South Pacific consisted of over 29 million Indians and over 19 mil
lion Chinese. Most migration from India was to colonies throughout
the British empire. Less than 10 percent of this migration was inden
tured, although much of it was undertaken with assistance from colo
nial authorities, or under some form of debt obligation under kangani
labor recruitment systems.1 Over 2 million Indians also migrated as
merchants or other travelers not intending to work as laborers.2 Migra

1 David Northrup, Indentured Eabor in the Age of Imperialism, 1834-1922 (Cambridge,

U.K.: Cambridge University Press, 1995), p. 65.
2 Claude Markovits, "Indian Merchant Networks Outside India in the Nineteenth and
Twentieth Centuries: A Preliminary Survey," Modern Asian Studies 33 (1999): 895, esti
mates 1.5 million commercial emigrants, but his numbers do not include all Indians that

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 i58 JOURNAL OF WORLD HISTORY, JUNE 2OO4

tion expanded with the increasing restriction of indenture contracts

after 1908 and the abolishment of indenture in 1920. Nearly 4 million
Indians traveled to Malaysia, over 8 million to Ceylon, over 15 million
to Burma, and about 1 million to Africa, other parts of Southeast Asia,
and islands throughout the Indian and Pacific Oceans.
The vast majority of Chinese migrants came from the southern
provinces of Guangdong and Fujian. Less than 750,000 Chinese
migrants signed indenture contracts with European employers, includ
ing 250,000 to Latin America and the Caribbean before 1874, 250,000
to Sumatra from the 1880s to the 1910s, and a smaller number to
mines, plantations, and islands scattered throughout the Pacific and
Indian Oceans (indentured laborers to South Africa from 1904 to 1908
and to Europe during World War I were mostly from north China).
Many more Chinese worked for Chinese employers under various
forms of contract and debt obligation, wage labor, and profit sharing.
Up to 11 million Chinese traveled from China to the Straits Settle
ments, although more than a third of these transshipped to the Dutch
Indies, Borneo, Burma, and places farther west. Nearly 4 million trav
eled directly from China to Thailand, between 2 and 3 million to
French Indochina, over 1 million to the Dutch Indies (for a total of
over 4 million if transshipments from Singapore are included), less
than 1 million to the Philippines, and over 500,000 to Australia, New
Zealand, Hawai'i, and other islands in the Pacific and Indian Oceans.
Migration into the broad expanse of North Asia is the least well
studied of these systems. Small trickles of migrants had moved into
central Asia, Siberia, and Manchuria for hundreds of years, but the
Qing government's gradual relaxation of restrictions against move
ment into Manchuria after i860 and the emancipation of serfs in
Russia in 1861 set the stage for more massive migration. Both govern
ments actively encouraged settlement with homesteading policies in
the i88os, each partly inspired by the desire to forestall territorial
encroachment by the other. Railroad construction in the 1890s further
strengthened the migrant flows.3 Between 28 and 33 million Chinese

may have been included under immigration categories such as "other than laborers." See
also Markovits, The Global World of Indian Merchants, 1750-1947 (Cambridge, U.K.: Cam
bridge University Press, 2000), p. 17; and Kernial Singh Sandhu, Indians in Malaya: Some
Aspects of Their Immigration and Settlement (1786-195-/) (Cambridge, U.K.: Cambridge Uni
versity Press, 1969), pp. 373-380.
3 Robert H. G. Lee, The Manchurian Frontier in CWing History (Cambridge, Mass.: Har
vard University Press, 1970); Donald Treadgold, The Great Siberian Migration: Government
and Peasant in Resettlement from Emancipation to the First World War (Princeton, N.J.: Prince
ton University Press, 1957).

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
McKeown: Global Migration, 1846-1940 159

Table 2. World population growth (millions) by regions, 1850-1950

1850 population 1950 population Average annual growth (%)

Receiving
Americas 59 325 1.72
North Asia 22 104 1.57
Southeast Asia 42 177 1.45
Sending
Europe 265 515 0.67
South Asia 230 445 0.66
China 420 520 0.21
Africa 81 205 0.93
World 1200 2500 0.74
Sources: Colin McEvedy and Richard Jones, Atlas of World Population History (L
Penguin, 1978).

migrated into Manchuria and Siberia (most of whom em

short sea voyage from Shandong to the Liaodong peni
with nearly 2 million Koreans and over 500,000 Japanese.
million Koreans migrated to Japan, especially in the 1930
million Russians moved into central Asia and Siberia over
In addition, up to 1 million northern Chinese, Koreans, a
migrated to a diverse range of destinations, including
Americas, Hawai'i, Southeast Asia, South Africa, and E
Global migrations caused a significant shift in the distr
the world's population. All three destinations experien
population growth, with their populations increasing by f
5.5 from 1850 to 1950 (see Table 2). Growth rates in these
over twice that for world population as a whole, and abou
greater than in Africa, a region of small net immigration
son, growth rates in the sending regions were lower than
lation growth, and less than half of those in the recei
Taken together, the three main destination regions acc
percent of the world's population in 1850 and 24 perc
Southeast Asia grew more slowly than the other two dest
that growth took place within a much more restricte
much more entrenched native population. From 1870 to
imately 35 million migrants moved into the 4.08 million
meters of Southeast Asia, compared to the 39 million m
moved into the 9.8 million square kilometers of the Unite
Emigration rates tend to be uneven within particular r
some villages or counties sending numerous migrants

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 16o JOURNAL OF WORLD HISTORY, JUNE 2OO4

send hardly any at all. Nonetheless, average emigration rates from dif
ferent sending regions are broadly comparable. At first glance 19 mil
lion overseas emigrants from China or 29 million from India seems
like a drop in the bucket compared to the several millions from much
smaller countries like Italy, Norway, Ireland, and England. But if we
look at regions of comparable size, the rates are very similar. Some of
the peak emigration rates ever recorded were an annual average of 22
emigrants per 1,000 population in Ireland during the famine of 1845 to
1855, or 18 per 1,000 from Iceland in the 1880s.4 Some South Pacific
and Caribbean islands probably experienced similar rates. More typi
cal rates in periods of high overseas emigration are 10.8 per 1,000 from
Italy, 8.3 per 1,000 from Norway, and 7 per 1,000 from Ireland in the
first decade of the twentieth century.5 In comparison, the average
annual overseas emigration rate from Guangdong province in south
China, which had an area slightly larger and a population slightly
smaller than Italy, was at least 9.6 per 1,000 in the peak years of the
1920s. Hebei and Shandong provinces (sources of migration to
Manchuria) had a rate of 10 per 1,000 during that same decade.6

Other Flows

Transoceanic migration accounts for only a portion of global migration.

Much migration was temporary or permanent movement to nearby
cities, towns, factories, mines, and plantations. Other migration took
place within the main sending and receiving regions and through
places such as Africa and the Middle East, which were at the interstices
of the main long-distance systems. Most of these migrations were
closely linked to processes that shaped the major long-distance systems,
and the three systems delineated in Table 1 quickly blur into a spec
trum of overlapping migrations.

4 William Smyth, "Irish Emigration, 1700-1920," in European Expansion and Migration:

Essays on the Intercontinental Migration from Africa, Asia, and Europe, ed. Pieter Emmer and
Magnus M?rner (New York: Berg, 1992), pp. 51-52; Helgi Sk?li Kjartansson, "Icelandic
Emigration," in European Expansion, pp. 105-106.
5 Walter Nugent, Crossings: The Great Transatlantic Migrations, 1870-1914 (Blooming
ton: Indiana University Press, 1992), p. 43.
6 Population data for China is from Thomas Gottschang and Dana Lary, Swallows and
Settiers: The Great Migration from North China to Manchuria (Ann Arbor: University of
Michigan, Center for Chinese Studies, 2000), pp. 172-173; and Robert Marks, Tigers, Rice,
Silk, and Silt: Environment and Economy in Late Imperial South China (Cambridge, U.K.: Cam
bridge University Press, 1998), p. 280.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
McKeown: Global Migration, 1846-1940 161

The transatlantic migrations could be extended to include over 10

million people who moved to the western frontiers of North America,
first primarily across the United States and eventually into the western
plains of Canada. This process also spurred the relocation of great num
bers of Native Americans and the migration of over 2.5 million Mex
icans to the agricultural areas of the southwestern United States in the
early twentieth century. The industrial centers of the northeastern
United States also attracted over 2.5 million Canadians, and then over
1 million African Americans and Mexicans in the early twentieth cen
tury.7 In other parts of the Americas, great numbers of Andean peoples
moved to coastal plantations and cities, and over 300,000 Caribbean
peoples migrated to plantations in Central America and Cuba, to the
Panama Canal Zone, and to the United States.8
Massive internal migration also took place within the major send
ing regions of long-distance migration. In Europe, migrants from Ire
land traveled to England for work, and from eastern and southern
Europe to industrial areas in northern Europe, especially France and
Germany. In Russia, migrants moved into the growing cities and south
ern agricultural areas.9 Within India they moved to tea plantations in
the south and northeast, to the mines and textile-producing regions of
Bengal, and to newly irrigated lands and urban areas throughout the
subcontinent.10 In China, they migrated to growing coastal cities, to
areas of the Yangtze basin left underpopulated by the Taiping rebel
lion, and to borderland areas of the northwest and southwest, includ

7 Dirk Hoerder, Cultures in Contact: World Migrations in the Second Millennium (Dur
ham, N.C.: Duke University Press, 2002), pp. 217, 355-356; Daniel Johnson and Rex
Campbell, Black Migration in America: A Social Demography (Durham, N.C.: Duke Univer
sity Press, 1981 ); Jean Meyer, "Les migrations mexicaines vers les Etats-Unis au Xx?me si?
cle," in Les Migrations Internationales: De la fin du XVll??me si?cle ? nos jours, ?d. CIDMSS
(Paris: Editions du CNS, 1980), pp. 681-696; Bruno Ramirez, Crossing the 49th Parallel:
Migration from Canada to the United States, 1900-1930 (Ithaca, N.Y.: Cornell University
Press, 2001).
8 Orlando Patterson, "Migration in Caribbean Societies: Socioeconomic and Symbolic
Resource," in Human Migration: Patterns and Policies, pp. 106-145.
9 Jeffrey Burds, Peasant Dreams and Market Politics: Labor Migration and the Russian Vil
lage, 1861-1905 (Pittsburgh: University of Pittsburgh Press, 1998); David Moon, "Peasant
Migration, the Abolition of Serfdom, and the Internal Passport System in the Russian
Empire, c. 1800-1914," in Coerced and Free Migration: Global Perspectives, ed. David Eltis
(Stanford, Calif.: Stanford University Press, 2002), pp. 324-360.
10 Kingsley Davis, The Population of India and Pakistan (New York: Russell and Russell,
1951 ), pp. 107-123; Arjan de Haan, "Migration on the Border of Free and Unfree Labour:
Workers in Calcutta's Jute Industry, 1900-1990," in Migration, Migration History, History:
Old Paradigms and New Perspectives, ed. Jan and Leo Lucassen (Bern: Peter Lang, 1999), pp.
197-222; Hoerder, Cultures in Contact, pp. 380-383.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 I?2 JOURNAL OF WORLD HISTORY, JUNE 2OO4

ing overland migration to Burma.11 Each of these systems involved at

least 20 million journeys. Southeast Asia and the South Pacific were
also sites of migration, including up to 500,000 Javanese traveling to
plantations in Sumatra and the Southeast Asian mainland and over
400,000 Melanesians and Micronesians working on plantations and as
seamen throughout the region.12
Africa experienced net transoceanic immigration, but at much
smaller numbers than other main destinations and from a wider vari
ety of origins. The immigrants included over 3 million French and Ital
ians into North Africa and up to 1 million other Europeans, Syrians,
Lebanese, Arabs, Indians, and Chinese throughout the continent.13
The end of the transatlantic slave trade led to increased movement of
slaves into the western Sudan, the Middle East, and areas bordering the
Indian Ocean in the late nineteenth century. Labor migration to plan
tations and mines in southern and central Africa increased through
the late nineteenth and twentieth centuries, as did movement to agri
cultural areas and coastal cities in western and eastern Africa. Millions
of people took part in these movements, some of whom were coerced
and many of whom went to work for European enterprises, but many
of whom also found independent occupations.14
The Middle East and ex-Ottoman lands were also at the interstices
of the main long-distance flows described above. Much of the move
ment in this region was the kind of labor migration that predominated
in much of the rest of the world. Projects such as the Suez Canal and
development of an infrastructure for cotton cultivation in Egypt
attracted large amounts of local migration, while Lebanon and Syria
experienced some of the highest overseas emigration rates in the
world.15 Over 3 million people also took part in the hajj to Mecca from

11 Ge Jianxiong, Cao Shuji, and Wu Songdi, Jianming Zhongguo yimin [Concise history
of Chinese migration] (Fuzhou: Fujian Renmin Chubanshe, 1993), pp. 460-492.
12 Lydia Potts, The World Labour Market: A History of Migration (London: Zed Books,
1990), p. 71.
13 Imre Ferenczi and Walter Willcox, eds., International Migrations, vol. 1, Statistics
(New York: National Bureau of Economic Research, 1929), p. 1028.
14 Dennis Cordell, Joel Gregory, and Victor Pich?, Hoe and Wage: A Social History of
a Circular Migration System in West Africa (Boulder, Colo.: Westview, 1996); Philip Curtin,
Why People Move: Migration in African History (Baylor, Texas: Markham Press Fund, Baylor
University Press), pp. 33-39; Patrick Harries, Work, Culture, and Identity: Migrant Laborers
in Mozambique and South Africa, c. 1860-1910 (Portsmouth, N.H.: Heinemann, 1994);
Fran?ois Manchuelle, Willing Migrants: Soninke Labor Diasporas, 1848-1960 (Athens: Ohio
University Press, 1997); Pat Manning, Slavery and African Life: Occidental, Orientai, and
African Slave Trades (Cambridge, U.K.: Cambridge University Press, 1990), pp. 171-178.
15 Akram Fouad Khater, Inventing Home: Emigration, Gender, and the Middle Class in
Lebanon, 1870-1920 (Berkeley: University of California Press, 2001).

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
McKeown: Global Migration, 1846-1940 163

1879 to 1938.16 This was also an area of massive migration caused by

war and politics, a harbinger of the kinds of migration that would
become increasingly prominent over the twentieth century. The disso
lution of the Ottoman empire and wars with Russia led to an exchange
of 4 to 6 million people, with Muslims moving south from the Balkans,
Greece, and Russia into Turkey, and Christians moving in the other
direction. Around 1 million Armenians were expelled from Turkey to
points around the world, and nearly 400,000 Jews moved to Palestine
in the early twentieth century.17 The massive movement of refugees
would extend to other parts of Europe in the wake of World War I and
the Russian revolution, including the movement of 3 million Russians,
Poles, and Germans out of the Soviet Union.18
In addition to the migration of settlers and workers, some of the tra
ditional merchant diasporas continued to flourish. For centuries before
the i8oos, these ethnic networks had been some of the most promi
nent exemplars of long-distance migration. Their importance dimin
ished under the economic transformations and new labor migrations
impelled by industrialization, but many old and new diasporas contin
ued to play prominent roles at the frontiers of the expanding global
economy. Jewish merchant networks were incorporated into the oper
ations of European capital and Armenian merchant networks were
decimated by the traumas of genocide. But other diasporas increased
in number and spatial extent while maintaining and adapting tradi
tional forms of commercial organization. Chinese merchant networks
helped channel Chinese labor throughout Southeast Asia and later
established dense networks of shops and services in places as distant as
the Amazon rubber groves, South Pacific atolls, and upriver Borneo,
not to mention restaurants, corner stores, and laundries that served
other migrants in plantations and urban neighborhoods throughout
the world. Merchants from India expanded trade networks into cen
tral Asia, Africa, and Southeast Asia. Chettiars from southern India
followed British expansion into Burma, and Parsis facilitated the India

16 Nearly 700,000 pilgrims were from the Dutch Indies. Centraal Kantoor voor de Sta
tistiek in Nederlandsch Indie, Statistisch Jaaroverzicht van Nederlandsch Indie 1938 (Bata
via), p. 140.
17 Calvin Goldscheider, "Israel," in Handbook on International Migration, ed. William
Serow, Charles Nam, David Sly, and Robert Weller (New York: Greenwood Press, 1990),
p. 132-135; G?lten Kazgan, "Migratory Movements in the Ottoman Empire and the Turk
ish Republic from the End of the 18th Century to the Present Day," in Les Migrations Inter
nationales, pp. 212-213; Justin McCarthy, Death and Exile: The Ethnic Cleansing of Ottoman
Muslims 1821-1922 (Princeton, N.J.: Darwin Press, 1995).
18 Aaron Segal, An Atlas of International Migration (London: Hans Zeil, 1993), p. 56.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 164 JOURNAL OF WORLD HISTORY, JUNE 2OO4

China trade using some of the capital they earned to establish textile
mills in India. Of particular interest are the Sindworkies from the town
of Hyderabad in what is now Pakistan. After the 1860s, they spread
from Japan to the Panama Canal and Tierra del Fuego, establishing
upscale tourist shops that sold "curios" from around the world and
becoming prominent carriers of Japanese trade in the Dutch Indies.
Other merchant diasporas such as the Hadhramis (from Yemen), Hausa,
and Lebanese Christians joined the Chinese and Indians at this inter
face between expanding industrial enterprises and dispersed individual
producers and consumers around the world.19

Historical Trends

The broad historical trends of global long-distance migration are traced

in Figures i and 2. The year 1846 is a somewhat arbitrary starting point
for this history, chosen for the availability of systematic migration sta
tistics from Europe, India, and Hong Kong. Relatively complete sta
tistics for Siberia and southern China ports other than Hong Kong
become available in the late 1870s, and for Chinese migration into
Manchuria in 1891. Long-distance and transoceanic migration had
been increasing gradually around the world since at least the 1820s.20
Prior to the nineteenth century, long-distance migration was under
taken primarily by merchants, African slaves, and a relatively small
trickle of settlers, agriculturalists, and miners to frontiers throughout
the world. Much migration was to nearby frontiers such as hills, forests,
and swamps. Many of these settlements were undertaken by indepen
dent settlers, but some were massive movements impelled by violence
and crisis or sponsored by governments and private companies, such as
the repopulation of Sichuan Province in western China after the Ming
to Qing transition, migrations into eastern Europe and the Ukraine,

19 Linda Boxberger, On the Edge of Empire: Hadhramawt, Emigration, and the Indian
Ocean, 1880S-1930S (Albany: State University of New York Press, 2002); Philip Curtin,
Cross-Cultural Trade inWorld History (Cambridge, U.K.: Cambridge University Press, 1984);
Christine Dobbin, Asian Entrepreneurial Minorities: Conjoint Communities in the Making of
the World-Economy 1570-1940 (Richmond: Curzon, 1996); Markovits, The Global World of
Indian Merchants; Adam McKeown, "From Opium Farmer to Astronaut: A Global History
of Diasporic Chinese Business," Diaspora 9 (2000): 317-360.
20 Leslie Page Moch, "Dividing Time: An Analytical Framework for Migration His
tory Periodization," in Migration, Migration History, History, pp. 41-56, and "The European
Perspective: Changing Conditions and Multiple Migrations, 1750-1914," in European
Migrants: Global and Local Perspectives, ed. Dirk Hoerder and Leslie Page Moch (Boston:
Northeastern University Press, 1996), pp. 115-140.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 18 !

1846-1851-1856- 1861- 1866-1871- 1876- 1881-1886- 1891-1896- 1901-1906- 1911-1916- 1920-1926- 1931-1936
50 55 60 65 70 75 80 85 90 95 1900 05 10 15 20 25 30 35 40

figure i. Aggregated global migration (five-year totals).

1800 -,

1600 -

figure 2. Global migrations (five-year averages).

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 166 JOURNAL OF WORLD HISTORY, JUNE 2OO4

and the extensive population movements sponsored by the Ottoman

government.21 In addition, merchants, craftsmen, laborers, and trans
portation workers made circuits throughout markets and towns in
much of Asia, Europe, Africa, and even in a Russia bound by serfdom.
Entire villages specialized in skills such as banking, stonemasonry, let
ter writing, or trade in particular products and sent migrants to ply their
trade across the region.22
The rise of a global economy centered on European, North Ameri
can, and Japanese industrialization was the context for increased long
distance migration of settlers and workers. Foodstuffs and resources
from frontiers near and far helped supply growing industrial centers,
and economic transformations disrupted old migration patterns. Chi
nese and Indian migration to Southeast Asia increased along with
transatlantic migration, took advantage of steamship lines after the
1860s, and opened up mines and agricultural areas in of the jungles of
Southeast Asia. The decline of the transatlantic slave trade after the
1820s led to the rise of indentured Asian migration in the 1840s.
Indentured migration would reach its zenith in the 1880s, declining
just as Asian migration began to boom. Similarly, exiles and prisoners
to Siberia declined just as Siberian migration began to increase in the
i88os (dropping from 75 percent in the 1870s to nearly nothing by the
1890s).23
Migration rates increased dramatically around the world in the last
quarter of the nineteenth century. After the depression of the early
1870s, transatlantic migration boomed and clearly surpassed Asian
migration for the first time in the late 1870s, although migration to
Southeast Asia soon picked up in the 1880s. Migration to North Asia
followed suit in the 1890s. Developments in transportation technol
ogy such as steamships and railways in all of these areas facilitated the
growth in migration. In turn, increased migration facilitated more
industrial expansion, which encouraged more migration. As migration

21 Ge Jianxiong, Jianming Zhongguo yimin, pp. 457-458; Richard Hellie, "Migration in

Early Modern Russia, 1480S-1780S," in Coerced and Free Migration, pp. 292-323; Hoerder,
Cultures in Contact, pp. 110-114, 281-287; Leong Sow-Theng, Migration and Ethnicity in
Chinese History: Hakkas, Pengmin, and Their Neighbors, ed. Tim Wright (Stanford, Calif.:
Stanford University Press, 1997).
22 Hoerder, Cultures in Contact, 288-300; James H. Jackson and Leslie Page Moch,
"Migration and the Social History of Modern Europe," Historical Methods 22 (1989): 27-36;
Leslie Page Moch, Moving Europeans: Migration in Western Europe since 1650 (Bloomington:
Indiana University Press, 1992), pp. 25-89; G. William Skinner, "Mobility Strategies in
Late Imperial China: A Regional Systems Analysis," in Regional Analysis, vol. 1, Economic
Systems, ed. Carol A. Smith (New York: Academic Press, 1976), pp. 327-364.
23 Hoerder, Cultures in Contact, p. 322.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
McKeown: Global Migration, 1846-1940 167

grew, larger portions of migrants to the Americas migrated to industrial

occupations and towns rather than frontier homesteading, a pattern
that would be followed in North Asia at about a fifteen- to twenty-year
delay. Rural populations continued to grow more rapidly than urban
ones in Southeast Asia, but the plantations, mines, and rice-growing
areas of the region were as much a part of the global economy as North
American factories and fields.24
Migration in each region ebbed and flowed along with business
cycles, climaxing in the years before World War I. Transatlantic migra
tion reached a spectacular peak of over 2.1 million in 1913, and migra
tion to Southeast and North Asia also reached unprecedented peaks of
nearly 1.1 million per year from 1911 to 1913. Transatlantic migration
was hit hardest by World War I, but recovered to 1.2 million migrants
in 1924, after which immigration quotas in the United States severely
curtailed immigration from southern and eastern Europe. Asian migra
tion also reached new peaks in the 1920s, with 1.25 million migrants
to Southeast Asia in 1927 and 1.5 million to North Asia in 1929. The
Great Depression put a stop to much migration, with the significant
exception of the command economies of Japan and the Soviet Union
where coercion, government promotion, and relatively strong econ
omies produced rates of up to 1.8 million migrants per year into North
Asia by the late 1930s.
Migration increased more quickly than world population. Trans
oceanic (and Siberian) migration in the 1850s amounted for 0.36 per
cent of the world's population. It amounted to 0.96 percent in the
1880s and 1.67 percent in the 1900s, then declined to 1.58 percent in
the 1920s. Concurrent growth around the world was not coincidental,
but linked through an increasingly integrated global economy. It was a
world on the move, flowing into factories, construction projects, mines,
plantations, agricultural frontiers, and commercial networks across the
globe. Coercion and violence played a role in this migration, especially
in the mid-nineteenth century and in the 1930s, but the bulk of it was
channeled through independent networks of friends, family, and vil
lagers. Ultimately, European, North American, and Japanese industri
alization, capital, and military power generated and dominated much
of this movement. But migrants around the world were not just carried
about by Europeans. They embodied the expanding global political
economy.

24 Anthony Reid, "South-East Asian Population History and the Colonial Impact," in
Asian Population History, ed. Ts'ui-jung Liu (Oxford, U.K.: Oxford University Press, 2001),
pp. 45-62.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 i68 JOURNAL OF WORLD HISTORY, JUNE 2OO4

Beyond the Atlantic

Rudolph Vecoli introduced his edited volume A Century of European

Migrations, 1830-IQ30 with the statement "[w]e need to move beyond
the framework of the 'Atlantic Migration'... It [has] blinkered us to
the global nature of [migration]."25 He went no further than to articu
late this agenda and offered no suggestions about how migration would
look different without those blinkers. At the most obvious level, atten
tion to global migration will provide more grist for comparative micro
and macrostudies about the operation of migration networks, the
causes of migration, migration's relationship to economic and demo
graphic change, the role of gender, and integration into local societies.
But serious work in this direction will be undertaken only once it is
clear that global migrations are broadly comparable, an assumption
that is still not well established in migration history. Even scholars who
are aware of migration beyond the Atlantic tend to characterize it as
directly subject to European expansion and not generated by the same
impulses that shaped transatlantic migrations.
It is easy to harvest numerous quotes from state-of-the-field vol
umes that privilege the uniqueness of the transatlantic system. Many
of these assertions are simply statements of quantity, such as Virginia
Yans-McLaughlin's assertion that "[b]y chance or choice, almost half
of these world-travelers [world migrants] settled in the United
States."26 Some of these statements go further and assert that trans
atlantic migration was categorically different in quality, such as Ira
Glazier and Luigi De Rosa's claim that "[i]t is North and South Amer
ica in the nineteenth and twentieth centuries that provide the great
stage for the migration drama, where migration assumes extraordinary
dimensions. While for the other continents migration was a means of
relieving demographic pressure by moving surplus population to
regions of lower density, in North and South America the problem was
one of providing a labor force to work the vast areas of open land wait
ing to be brought under cultivation."27 Douglas Massey asserted the

25 Rudolph Vecoli, "Introduction," in A Century of European Migrations, 1830-1930,

ed. Rudolph Vecoli and Suzanne Sinke (Urbana: University of Illinois Press, 1992), p. 13.
26 Virginia Yans-McLaughlin, "Introduction," in Immigration Reconsidered: History,
Sociology, and Politics, ed. Virginia Yans-McLaughlin (New York: Oxford University Press,
1990), p. 3.
27 Ira Glazier and Luigi De Rosa, "Introduction," in Migrations across Time and Nations:
Population Mobility in Historical Context, ed. Ira Glazier and Luigi De Rosa (New York:
Holmes and Meier, 1986), p. 5.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
McKeown: Global Migration, 1846-1940 169

historical predominance of European migrations as a way to highlight

what he believed was the increasingly global quality of contemporary
migration: "Although international migrants were not exclusively
European, the overwhelming majority came from that continent. . . .
Before 1925, 85 percent of all international migrants originated in
Europe, but since i960 Europe has contributed an increasingly small
fraction of emigrants to world migration flows."28
Most of these works do not engage directly with non-Atlantic
migrations, and such statements can potentially be dismissed as lack of
direct knowledge. But the decision to disregard the rest of the world is
still shaped by assumptions about that world, as can be seen more
clearly in Pieter Emmer's comparative work. Emmer categorizes long
distance migration as being intercontinental or not. Thus, he includes
Russians who crossed the Ural mountains and French who went to
Algeria as intercontinental, but not Chinese who went to Singapore.
After counting 5 to 6 million intercontinental African and Asian
migrants from 1800 to i960, he concludes that "[t]he study of migra
tion as part of the process of European expansion and contraction
clearly shows that Europeans have participated much more extensively
in intercontinental migrations than Africans and Asians."29 This dis
tinction then becomes the basis for qualitative judgments. In his essay
"Was Migration Beneficial?" he divides global migration into temper
ate and tropical plantation systems. The tropical system is made up
almost entirely of intercontinental Asian and African migration to
plantations under conditions of indenture and to white settler nations
filled with anti-Asian sentiment, making it easy to conclude that
Asians benefited much less from migration than Europeans. He even
suggests that it does not make much sense to ask if Asian migration
was beneficial "given the relatively small volume of both internal and
external migration."30
Emmer's distinctions may seem crude, but the general assumption
that Asian migration consisted almost entirely of indentured labor
dominated by Europeans pervades much of the literature. This assump
tion is somewhat understandable for Indian migration given its rela

28 Douglas Massey, "Why Does Immigration Occur? A Theoretical Synthesis," in

Charles Hirschman, Philip Kasinitz, and Josh De Wind, eds., The Handbook of International
Migration: The American Experience (New York: Russell Sage Foundation, 1999), p. 35.
29 P. C. Emmer, "European Expansion and Migration: The European Colonial Past and
Intercontinental Migration; An Overview," in European Expansion, pp. 3, 10-11.
30 Pieter Emmer, "Was Migration Beneficial?" in Migration, Migration History, History,
p. 113.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 170 JOURNAL OF WORLD HISTORY, JUNE 2004

tively tight organization within the British empire and the relatively
high level of assisted migration to European-owned plantations. Much
more perplexing is the characterization of Chinese migration as inden
tured and the consistently low estimations of Chinese migration in
Western language scholarship ranging from 2 to 8 million.31 Even the
incomplete Chinese numbers in the most widely used source of migra
tion statistics, Ferenczi and Willcox's International Migrations, account
for nearly 8 million Chinese.32 Most citations, however, tend to draw
from other secondary works, eventually leading to studies by Chen Ta,
Arnold Meagher, and Chen Zexuan, all of whom were counting only
contract labor migration.33 The projection of these numbers as a com
plete accounting of Chinese migration is laden with assumptions about
the nature of Asian migration.34
Dirk Hoerder's recent attempt at a comprehensive global history of
migration, Cultures in Contact, also explains Asian migration primarily
in terms of European intervention. His chapter on Asian migration in
the industrial age is titled "Asian Contract Labor System (1830s to

31 Lynn Pann, Encyclopedia of Chinese Overseas (Cambridge, Mass.: Harvard University

Press, 1999), p. 62; Potts, World Labour Market, p. 70; and Walton Look Lai, "Asian Con
tract and Free Migrations to the Americas," in Coerced and Free Migration, p. 230. Hoerder,
Cultures in Contact, pp. 366-367, 377, and 389, has inconsistent numbers but notes that free
migrants are not included in most estimates. Donna Gabaccia has done exemplary work that
includes Asians in her comparative analyses of migration to the Americas. In "The 'Yellow
Peril' and the 'Chinese of Europe': Global Perspectives on Race and Labor," in Migration,
Migration History, History, p. 180, she did take a look at Ferenczi and Willcox and offers an
estimate of 2 to 10 million, suggesting that less than a third of them were coolies. A figure
of 20 million migrants does circulate in contemporary Chinese language scholarship.
32 Tables in Ferenczi and Willcox count 5.5 million immigrants to the Straits Settle
ments from 1881 to 1915, 3.7 million departing Chinese ports from 1876 to 1901, and 2.4
million leaving Hong Kong from 1900 to 1924.
33 Chen Ta, Chinese Migrations, with Special Reference to Labor Conditions (Washington,
D.C.: Government Printing Office, 1923), gives only scattered numbers. Chen Zexuan,
"Shijiu shiji cheng xing de tiaoyue huagong zhi" [The nineteenth century Chinese contract
labor system], Lishi Yanjiu (1963), no. 1, counts 6 million contract laborers (tiaoyue hua
gong), but his definition is expansive, including all forms of assisted and debt migration,
including migration to California. David Northrup, Indentured Labor, p. 56, counts 386,901
indentured Chinese and 1,336,030 Indians. He relies heavily on Arnold Meagher, "The
Introduction of Chinese Laborers to Latin America: The 'Coolie Trade,' 1847-1874" (Ph.D.
dissertation, University of California at Davis, 1975), which stands as the most exhaustive
archive-based estimate of Chinese indentured labor migration before 1874.
34 Jan and Leo Lucassen, "Migration, Migration History, History: Old Paradigms and
New Perspectives," in Migration, Migration History, History, p. n, notes that historians of
non-European migration tend to focus on coerced migration and suggests that the distinc
tion with free European migration should be broken down. Some of the essays in this vol
ume make valuable steps in this direction, but, like the articles in Eltis, Coerced and Free
Migration: Global Perspectives, consider Asian migrations mostly in the context of move
ment to the Americas.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
McKeown: Global Migration, 1846-1940 171

1920s) and Transpacific Migration." It begins with the assertion that

"[p]arallel to the voluntary and self-bound migrations in the Atlantic
system, the migrations of [Asians] involved a minority of free migrants,
large numbers of self-bound migrants, and forced moves," and over
two-thirds of the chapter focuses on indentured migration and Euro
pean-run plantations.35 He cites statistics that show high numbers of
migration into Southeast Asia and Manchuria, yet he still provides
maps in which flows of Indian migrants to the Caribbean are depicted
as twice as large as those to Southeast Asia, and twice as large from
North China to overseas destinations than to Manchuria (whereas the
Manchurian migrations were at least 250 times larger).36 Hoerder's
explanations of Asian migration are also categorically different than
those for European migration. The causes of Chinese overseas emigra
tion include "[(Imperial Chinese maladministration and revolts, over
population and natural disasters as well as colonial penetration."37
European research (as Hoerder is aware) has consistently argued that
the first three causes can rarely be used to explain the establishment of
steady migration patterns, and "colonial penetration" is, of course,
irrelevant for Europe. This leaves the impression that Chinese do not
migrate within the same kinds of networks and by making the same
kinds of decisions as Europeans. Similarly, the Manchurian migrations
are treated as peasant resettlement driven by overpopulation, with lit
tle discussion of the industrial transformations of northeastern Asia.
The overall effect is to render Asian migrations of little significance
without European presence or despotic Asian regimes to impel them.
Detailed comparative analyses between different migrant flows
around the world will certainly provide continued insights into the
processes of migration. But such comparisons will not reach their full
potential without a revision of the larger assumptions of world history
used to contextualize migration. The nearly contemporaneous rise of
global migration suggests that non-Europeans were very much involved
in the expansion and integration of the world economy, well beyond
the direct intervention of Europe. The division of migration into dis
tinct systems often obscures the links between these global processes.
As we shall see, a global history of migration must also understand the
processes of regulation and intervention by which global migration has
come to be segmented into distinct systems, both in perception and in
practice.

35 Hoerder, Cultures in Contact, p. 366.

36 Ibid., pp. 368, 372, 389, 494-495
37 Ibid., Cultures in Contact, p. 12.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 172 JOURNAL OF WORLD HISTORY, JUNE 2004

Periodization

The late 1920s was one of the high points of global mass migrations.
But studies that privilege the transatlantic migrations generally take
1914 to be the end of the age of mass migration. This date makes some
sense from the perspective of the North Atlantic, where migration pat
terns were severely disrupted by World War I and its aftermath. The
war produced rigid passport controls and massive refugee streams that
were unfamiliar to western Europe (although not to Russia, south
eastern Europe, and Asia Minor). Germany severely restricted immi
gration after the war, the United States erected quotas in 1921 and
1924, and Italy restricted emigration in 1927. It appeared to be a new
era of migration control, rigid borders, and nationalistic prejudices.38
But it is far from clear that movement was severely diminished
within the larger transatlantic system beyond those three (admittedly
important) nodes of Germany, Italy, and the United States. Over 2
million workers traveled to France in the 1920s, and migration to
Australia, New Zealand, and American destinations other than the
United States (and the United States before 1924) resumed levels
equivalent to the first decade of twentieth century. Emigration from
eastern Europe and Portugal?areas at the expanding periphery of the
transatlantic system?even increased during the 1920s.39 Immigration
to the United States from other parts of the Americas also increased
to over 150,000 a year, the highest rates ever.40 It is also unclear that
immigration policy was categorically more restrictive. Immigration
policy actually grew less restrictive in Canada over the 1920s, while
remaining fairly constant in Argentina and Australia.41
What seems to be at stake in insisting that 1914 was the end of mass
migration is not just the number of migrants, but the belief that a new
era of nationalism, rigid borders, and government regulation had
taken root. The postwar years are depicted as a break from the regime

38 Kulischer, Europe on the Move; Moch, Moving Europeans, pp. 155-170; Saskia Sas
sen, Guests and Aliens (New York: New Press, 1999), pp. 85-90.
39 Kulischer, Europe on the Move, p. 152. Kulischer also notes that Siberian migration
reversed in the immediate post-revolution years, as local polices were erected against the
interests of Russian settlers, but migration resumed familiar eastward patterns by the mid
19208, facilitated by central directives overriding local policies (pp. 74-84).
40 Mauice Davie, World Immigration: With Special Reference to the United States (New
York: McMillan, 1936), p. 208.
41 Ashley Timmer and Jeffrey Williamson, "Immigration Policy Prior to the 1930s:
Labor Markets, Policy Interactions, and Globalization Backlash," Population and Develop
ment Review 24 (1998): 743.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
McKeown: C3lobal Migration, 1846-1940 173

of free migration in place before the 1920s. But rather than posing a
dichotomy of before and after 1914, it would be better to understand
regimes of regulation as part of a cumulative process that had been
taking place since at least the 1870s. The concurrent growth of migra
tion since the mid-nineteenth century was part and parcel of the
expansion of borders and regulation, including numerous projects to
encourage, restrict, select, protect, distribute, and monitor migration.
Extending the era of mass migration into the 1920s acknowledges
both the global scale and the long-term relationship of migration and
politics since the early nineteenth century.42 This also helps provide
insight into how different migration systems became segmented across
the globe.
"Free" migration was itself a product of government regulation in
many instances. The abolishment of the slave trade required intensive
government intervention against the activities of private traders and
recruiters. The regulation of the Chinese coolie trade in the 1860s and
18 70s entailed the collaboration of Chinese and European officials to
circumscribe the activities of private recruiters and establish regula
tory institutions that assured that each migrant had entered into his
contract "freely" and "voluntarily."43 The expansion of Indian migra
tion throughout the British Empire was a direct result of this type of
intervention. All migration moves through social and political land
scapes. The idea of a free migration must always be contextualized in
its political context in which certain other processes and interests
must also be constrained.
Colonial and national policies toward migrants and natives have

42 Jane Caplan and John Torpey, eds., Documenting Individual Identity: The Development
of State Practices in the Modern World (Princeton, N.J.: Princeton University Press, 2001);
Cindy Hahamovitch, "Creating Perfect Immigrants: Guestworkers of the World in Histor
ical Perspective," Labor History 44 (2003): 70-94; John Torpey, The Invention of the Passport:
Surveillance, Citizenship and the State (Cambridge, U.K.: Cambridge University Press, 2000);
Aristide Zolberg, "The Great Wall Against China: Responses to the First Immigration Cri
sis, 1885-1925," in Migration, Migration History, History, pp. 291-315, and "International
Migration Policies in a Changing World System," in Human Migration: Patterns and Policies,
ed. William McNeill and Ruth Adams (Bloomington: Indiana University Press, 1978), pp.
241-286. For summaries of migration laws, see International Labour Office, Emigration and
Immigration (Geneva: 1922), and Ashley Trimmer and Jeffrey Williamson, "Racism, Xeno
phobia or Markets: The Political Economy of Immigration Policy Prior to the Thirties,"
NBER Working Paper 5867, 1996, pp. xix-xxxi.
43 Robert Irick, Ch'ing Policy Toward the Coolie Trade 1847-1878 (Taipei: Chinese
Materials Center, 1982); Anthony Reid, "Early Chinese Migration into North Sumatra,"
in Studies in the Social History of China and South-east Asia: Essays in Memory of Victor Pur
cell, ed. Jerome Ch'en and Nicholas Tarling (Cambridge, U.K.: Cambridge University Press,
1970), pp. 289-320.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 174 JOURNAL OF WORLD HISTORY, JUNE 2004

also had a significant effect on the size and nature of migrant flows.
Chinese migration to colonial Southeast Asia and Siam in the middle
of the nineteenth century was inseparable from the enormous revenue
farms auctioned out to Chinese entrepreneurs, which were nexi for
labor recruitment and settlement of the interior. Later policies that
framed natives as rural dwellers with rights to the land and in need of
protection from exploitative "alien" Chinese or Indians were crucial
in the development of extensive "middleman" networks by Chinese
and Indians. In turn, these networks channeled further migration and
expanded well beyond direct European influence to create a world
densely enmeshed in an expanding global economy.44 These policies
were much different than those in the Americas that treated many
(but not all) immigrants as potential citizens, and developed policies
of selection, recruitment, and protection to facilitate the realization of
this ideal.
Nationalism and the policing of territorial boundaries were also
part of the same historical processes that produced migration. Many
immigration restrictions and border controls were erected in reaction
to immigration, building on the fear of cultural contamination, declin
ing living standards, or political subversion. Leo Lucassen has argued
that the rise of mass political participation and the welfare state was
most responsible for the perceived need to restrict the entry of migrants
who may compete for political resources.45 But migration also has a
positive relation to nationalism and the control of territorial bound
aries. The opening and encouragement of migration into Siberia and
Manchuria were both partially impelled by the perceptions of the
Russian and Chinese governments that these borderlands must be
claimed and fortified. Japan and Italy both encouraged overseas migra
tion as a way to expand national interests. The Italian emigrant pass
port requirements of 1901 were intended to facilitate high-quality

44 John Butcher and Howard Dick, The Rise and Fall of Revenue Farming: Business Elites
and the Emergence of the Modern State in Southeast Asia (New York: St. Martin's, 1993); Gary
Hamilton and Tony Waters, "Ethnicity and Capitalist Development: The Changing Role
of the Chinese in Thailand," in Essential Outsiders: Chinese and Jews in the Modern Trans
formation of Southeast Asia and Central Europe, ed. Daniel Chirot and Anthony Reid (Seat
tle: University of Washington Press, 1997), 258-284; McKeown, "From Opium Farmer to
Astronaut"; Carl Trocki, Opium and Empire: Chinese Society in Colonial Singapore, 1800
1910 (Ithaca, N.Y.: Cornell University Press, 1990).
45 Leo Lucassen, "The Great War and the Origins of Migration Control in Western
Europe and the United States (1880-1920)," in Regulation of Migration: International Expe
riences, ed., Anita Buocker, Kees Groenendijk, Tetty Havinga, and Paul Minderhoud (Ams
terdam: 1998), pp. 45-72.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
McKeown: Global Migration, 1846-1940 175

migration that was not refused entry at foreign ports, as well as to main
tain tabs on young males who were required to perform military ser
vice.46 Emigrant communities have also been important sites for learn
ing and propagating ideals of nationalism and for the reification of a
homeland that must be nurtured, fortified, and protected.47
Perhaps one of the earliest and most significant examples of migra
tion control has been the cutting off of Asian migration to the Amer
icas and Australia in the last two decades of the nineteenth century.
Nearly 50,000 Chinese migrated to the California gold rush from 1851
to 1855. Total migration to California and Australia amounted to
135,000 in the 1850s, perhaps amounting to a quarter of all Chinese
emigration. All these migrants were funded and arranged by Chinese
capital (albeit transported on European ships) and depended on Chi
nese mining skills. Chinese were flowing into the same paths as Euro
pean migrants until international intervention in the coolie trade in
the 1870s and anti-Asian immigration laws in the Americas and Aus
tralia after the 1880s severely reduced migration to these destinations.
Chinese migrants to the Americas declined from 20-35 percent of all
Chinese migrants in the 1850s through the early 1870s to 2-5 percent
per year after 1890. Half of all Asian migration to the Americas had
already taken place by the mid-i88os and was cut off just when Asian
migration in general began to boom. The resulting segmentation of
migration systems has since framed understandings of Asian migrations
as categorically different and less significant than transatlantic migra
tions. But this segmentation was precisely an example of the kinds of
government intervention that continued to accumulate through the
1920s and 1930s.
Some of the larger implications of taking 1914 as a break can be
seen in work by Jeffrey Williamson and his collaborators, especially The
Age of Mass Migrations with Timothy Hatton and Globalization and
History with Kevin O'Rourke. They argue that migration was a critical
aspect of globalization in the North Atlantic, contributing to a con
vergence of wages, living standards, and industrial production across
nations. This argument is linked to a periodization in which the late
nineteenth century until 1914 was a period of globalization, the period
from 1914 to 1950 was one of deglobalization, and the period after

46 Torpey, Invention of the Passport, pp. 103-105.

47 Adam McKeown, "Conceptualizing Chinese Diasporas, 1842 to 1949," Journal of
Asian Studies 58 (1999): 322-326; Nina Glick Schiller, "Transmigrants and Nation-States:
Something Old and Something New in the U.S. Immigrant Experience," in Handbook of
International Migration, pp. 94-119.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 176 JOURNAL OF WORLD HISTORY, JUNE 2OO4

1950 was one of renewed globalization.48 O'Rourke and Williamson

provide evidence that wages and production converged across nations
before 1914, but diverged or remained stagnant in the interwar period.
They argue that domestic inequalities caused by globalization led to
backlash and the erection of barriers against globalization, such as high
tariffs and anti-immigration laws, which caused economic divergence
and, ultimately, the Great Depression.49 But they offer little evidence
that the flows of globalization actually receded after World War I.
Generalizations across the period from 1914 to 1946 can easily show a
decline in international trade, capital flows, and migration. A focus on
the 1920s, on the other hand, shows flows of trade and capital higher
than ever before, and tariff barriers that were little changed from the
1890s.50 The same is broadly true for migration and its controls.
For Williamson, the lesson based on taking 1914 as an end of glob
alization is that inappropriate government intervention disrupts the
positive effects of globalization, which will tend toward global eco
nomic equilibrium if left to its natural course. Backlash against domes
tic inequalities provokes this inappropriate intervention, and appro
priate policies are necessary to remove the factors that provoke this
backlash. Of course, this conclusion is based on the analysis of "glob
alization" in a small segment of the world centered on the North
Atlantic. Hatton and Williamson justify their lack of attention to
migration in the rest of the world by arguing that labor markets in
"what is now known as the Third World . . . were segmented by dis
crimination, language, and custom. They were segmented by long dis
tance and high migration cost. They were segmented by the poverty
of the Third World labor-surplus areas, areas so poor that potential
emigrants would have found it impossible to finance the move to the
booming OECD labor markets anyway."51 Segmentation due to dis
crimination certainly seems accurate. Appealing to language and cus
tom is more problematic, given that Hatton and Williamson disregard
language and custom in their economically driven analysis of transat

48 In a World Bank publication, Williamson goes so far as to call the period before
1914 a "belle ?poque," the period from 1914 to 1950 the "dark middle years," and the sub
sequent era a "renaissance." See Jeffrey Williamson, "Globalization and Inequality, Past
and Present," The World Bank Research Observer 12 (1997): 118.
49 Kevin O'Rourke and Jeffrey Williamson, Globalization and History: The Evolution of
a Nineteenth-Century Atlantic Economy (Cambridge, Mass.: MIT Press, 1996), p. 286.
50 Paul Bairoch, Economics and World History (New York: Harvester Wheatsheaf, 1993).
51 Timothy Hatton and Jeffrey Williamson, The Age of Mass Migration: Causes and
Economic Impact (New York: Oxford University Press, 1998), p. 249.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
McKeown: Global Migration, 1846-1940 177

lantic migration. The large mid-nineteenth-century migrations from

Asia to the Americas also question the assumption that long distance
and poverty were factors.
The assertion that migration and trade outside of the Europe
America axis is largely irrelevant to North Atlantic "globalization"
implies that the rest of the world could not take part in the positive
effects of globalization because of factors that had little to do with the
North Atlantic economy or with active government intervention into
world migration systems. This reduces Williamson's thesis to a tautol
ogy: globalization had beneficial effects on economies because it took
place in areas where economies were strong. But, as we have seen, even
if the actual overlap in migrants travelling between major migration
systems was small and grew increasingly smaller over time, migration
around the world still followed the same broad cycles. We should not
be satisfied with a priori assumptions of segmentation and categorical
difference across the globe. Segmentation was not a product of natural
conditions that divided the world into discrete units, but a structural
aspect of the larger global political economy and political intervention.
This, of course, is the argument of dependency and world-systems
theories, that global inequalities are the result of positive structural
interactions between cores and peripheries. Yet dependency and world
system theorists also highlight economic processes over political inter
vention in understanding global relationships. They dismiss the rele
vance of Asian and African migrations in a different way than do
more orthodox economists, seeing them as little more than a by-prod
uct of the expansion and intervention of Europe, but still categorically
different than the transatlantic migrations.52 They are able to explain
only those small portions of non-European migrations that were
indentured or otherwise directly subject to Europeans, and leave little
means of understanding the broad similarities between global migra
tion patterns.

Migrant Networks

Much recent work on the sociology and history of migration has placed
migrants in the context of extensive networks that cross nations and

52 In addition to Hoerder, Cultures in Contact, which offers a modified world-systems

approach, see Edna Bonacich and Lucie Cheng, eds., Labor Immigration under Capitalism
(Berkeley: University of California Press, 1984).

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 178 JOURNAL OF WORLD HISTORY, JUNE 2OO4

regions.53 It is at the level of these transnational networks that expand

beyond the boundaries of nation-states that we can best perceive social
formations that are directly embedded in social processes at a global
scale. This is also the level at which global economic patterns are
embodied in specific institutions and human activities. Attention to
these networks, rather than taking distinct Atlantic and Asian systems
or national histories for granted as a fundamental organizing rubric,
can provide keys to understanding broad global patterns of migration.
Recent migration scholarship has emphasized the family as funda
mental arena of decision making. Families can often be understood as
an investment portfolio, sending children and affines to a variety of
diversified places and occupations in the hope that one will provide a
return on the investment. The specifics of this family economy change
in accordance with local family structures, inheritance customs, pat
terns of land tenure, and changing opportunities to acquire property
and status at home and abroad. The experience of migration may also
alter the original intentions of the migrant and his family, but few
migrants ever make choices only as isolated individuals. Decisions are
made in the context of information and assistance obtained from rel
atives and village members around the world. These networks are also
institutionalized as mutual aid societies, labor recruitment enterprises,
and dense commercial linkages. These geographically dispersed com

53 Historical and theoretical work on migration networks around the world include
Te?filo Altamirano, Presencia Andina en Lima Metropolitana (Lima: Pontificia Universidad
Cat?lica del Per?, Fondo Editorial, 1984); John Bodnar, The Transplanted: A History of Immi
grants in Urban America (Bloomington: Indiana University Press, 1985); Monica Boyd,
"Family and Personal Networks in International Migration: Recent Developments and New
Agendas," International Migration Review 23 (1989): 638-670; Madeline Hsu, Dreaming of
Gold, Dreaming of Home: Transnationalism and Migration between the United States and South
China, 1882-1943 (Stanford, Calif.: Stanford University Press, 2000); James H. Jackson
and Leslie Page Moch, "Migration and the Social History of Modem Europe"; Ivan Light,
Parminder Bachu, and Stavros Karageoris, "Migration Networks and Immigrant Entrepre
neurship," in Immigration and Entrepreneurship, ed. Ivan Light and Parminder Bachu (New
Brunswick, N.J.: Transaction Publishers, 1993), pp. 25-49; Jan and Leo Lucassen, eds.,
Migration, Migration History, History; Fran?ois Manchuelle, Willing Migrants: Soninke Labor
Diaspora; Adam McKeown, Chinese Migrant Networks and Cultural Change: Peru, Chicago
and Hawaii, 1900-1936 (Chicago: University of Chicago Press, 2001); Alejandro Portes and
J?zsef B?r?cz, "Contemporary Immigration: Theoretical Perspectives on Its Determinants
and Modes of Incorporation," International Migration Review 23 (1989): 606-630; Oded
Stark, The Migration of Labor (Cambridge, U.K.: Basil Blackwell, 1992); and Charles Tilly,
"Migration in Modern European History," in Human Migration, pp. 48-72, and "Trans
planted Networks," in Immigration Reconsidered, pp. 79-95. For a general review of migra
tion theories, see Douglas Massey, Joaqu?n Arango, Grame Hugo, Ali Kouaouci, Adela
Pellegrino, and J. Edward Taylor, "Theories of International Migration: A Review and
Appraisal," Population and Development Review 19 (1993): 431-466.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
McKeown: Global Migration, 1846-1940 179

munities are the social space within which many migrant decisions and
activities take place. They produce a social geography that is not con
gruent with physical geography. A migrant may have more knowledge
of distant Penang, Omsk, or Chicago than the nearby market town,
because that is where all his uncles and cousins are.54
Such networks both create and constrain opportunities. Migration
would generally be too expensive and risky without the information,
assistance, and opportunities provided by friends, relatives, and other
resources available through these networks. It would also be a point
less endeavor without the links to home and family maintained within
these networks, through which money, support, and the other fruits of
migration could be channeled. But the very strength of these channels
also restricts opportunities. It is very difficult for an established net
work to access new job opportunities and new destinations, even if only
rudimentary new skills are necessary and the economic benefits are
high. Many migrant elite even had an interest in controlling and lim
iting information and opportunities, so as to further profit from their
position at the nexus of opportunities.
Studies of aggregate nation-to-nation movements can be quite suc
cessful at explaining the ebbs and flows of particular streams, but they
provide few tools to understand the mechanisms of migration. For
example, one village may have a population so devoted to emigration
that even the fields are left barren or leased to outsiders, while another
village a few kilometers down the road with almost identical condi
tions may produce no migrants at all. Aggregate models and economic
approaches are also hard pressed to explain why migrants may choose
to travel to one location when wages and opportunities are clearly
superior elsewhere. Why would some Italians choose to work in Mon
treal when the local French Canadians were looking for better-paying
work in New England?55 In other words, individuals making decisions
(the "free" migrants) are the basic unit of most economic and aggre
gate models, yet those studies have little understanding of the social
conditions under which those decisions are made (although scholars

54 J. D. Gould, "European Inter-Continental Emigration: The Role of 'Diffusion' and

'Feedback,'" European Journal of Economic History g: 294-295; McKeown, Chinese Migrant
Networks, pp. 70-76.
55 Bruno Ramirez, "The Crossroad Province: Quebec's Place in International Migra
tions, 1870-1915," in A Century of European Migrations, pp. 243-260. It may also be noted
that criticisms of economic models as based exclusively on wage differentials may be attack
ing straw men. Hatton and Williamson, Age of Mass Migration, for example, develops a com
plex model that takes networks, population, wage differentials, and commercialization into
account.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 i8o JOURNAL OF WORLD HISTORY, JUNE 2OO4

who emphasize social networks over economics have difficulty explain

ing the decline of migration flows).
A map of the world drawn from these geographically dispersed
spaces and networks would look much different than the familiar
mosaic of geographically discrete territories. They make up a world of
complex and overlapping flows and nodes, none of which can be
entirely captured within a single national or regional history.56 Under
standing migration simply as relocation from one nation to another
tends to create a simplified historical narrative whereby home is con
flated with the past, and historical change is identified with integra
tion and settlement in the new land. Each network occupies a distinct
niche that articulates with many local histories but has its own specific
patterns. To be sure, states, international borders, and local social struc
tures are important aspects of migration history, but they are also just
some of many interwoven factors. The means of recruitment, the insti
tutions established by previous migrants, and the ways in which travel,
information, money, and opportunities are channeled and controlled
are all aspects of migration not reducible to a single location. These
processes and institutions are also inseparable from regulation, legal
rights, and boundaries. These boundaries erect potential barriers, but
they are also a source of opportunity because many migrants make their
living precisely by their skills in negotiating these boundaries, manag
ing multiple regulatory regimes, crossing customs and migration obsta
cles, and operating in spaces that are ultimately subject to no single
regulatory regime. Migration itself becomes a source of sustenance and
a self-reproducing goal, above and beyond the sending and receiving
societies.
Donna Gabaccia's work on female migration suggests how migrant
networks can be understood on their own terms and also at a con
juncture of local and global historical processes.57 She starts from the
oft-noted phenomenon that the proportion of male migrants and the

56 For critiques of national perspectives, see Hasia Diner, "History and the Study of
Immigration: Narratives of the Particular," in Migration Theory: Talking Across Disciplines,
ed. Caroline Brettell and James Hollifield (New York: Routledge, 2000), pp. 27-42; Donna
Gabaccia, "Is Everywhere Nowhere? Nomads, Nations, and the Immigrant Paradigm of
United States History," Journal of American History 86 (1999): 1115-1134; and Nancy
Green, "The Comparative Method and Poststructural Structuralism: New Perspectives for
Migration Studies," in Migration, Migration History, History, pp. 57-72.
57 Donna Gabaccia, "Women of the Mass Migrations: From Minority to Majority,
1820-1930," in European Migrants, pp. 90-111. See also J. D. Gould, "European Inter
Continental Emigration: The Road Home: Return Migration from the U.S.A.," European
Journal of Economic History 9 (1980): 41-112; Ewa Morawska, "Return Migrations: Theo
retical and Research Agenda," in A Century of European Migrations, pp. 277-292; and Portes
and B?r?cz, "Contemporary Immigration."

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
McKeown: Global Migration, 1846-1940 181

return migration from the United States both rose over the course of
the twentieth century as migration moved away from frontier settle
ment to wage labor in cities and factories. While recognizing the broad
applicability of this generalization, she also points out that female
migration for all ethnic groups follows the same cycles?linked to busi
ness cycles?as male migration. Higher proportions of women also
migrated to industrial New England than to the rural West. That is to
say, the usual linkage between female migration, family settlement on
frontiers, female and return migration may not be accurate. Many
female migrants were also wage laborers, just like the men. More spe
cific explanations must be sought.
Gabaccia suggests that the difference in migration demographics
over the course of the nineteenth century can be understood also by
looking at family structure. Migrants from northern Europe who dom
inated migration to the United States in the mid-nineteenth century
were from areas where nuclear families predominated. Husbands and
wives were dependent on each other for economic sustenance, and
many women had experience in market activities. Southern and east
ern Europeans, who dominated in later migrations, tended to have
ideal families that were larger multigenerational units. Women could
remain at home as part of a larger economic unit and were also more
likely to be involved in subsistence than market activities.
The history of female migration from Ireland offers an even more
particular situation, based on historical circumstances. Irish migration
was unique in that it maintained nearly equal proportions of male and
female migration and very low return rates throughout the nineteenth
and early twentieth centuries. The potato famine encouraged the
migration of unusually high numbers of women in the mid-nineteenth
century, who got a lock on household service jobs in the United States.
This established a pattern of high rates of female migration and per
manent settlement in North America that would continue for nearly
a century (and was mirrored in short distance migration of women to
European cities to work in shops and as household servants).
These specific histories can explain some of the shifts in female
migration and return rates, but even northern Europeans tended to
produce increasing numbers of male sojourners over the late nine
teenth and early twentieth centuries. No single explanation can
account for everything, but multiple explanations highlight the ways
in which a single network was shaped by multiple forces, including his
torical contingencies, patterns set by previous migrants, structural eco
nomic relations between different parts of the world, and local cultural
forms like family structure.
No single explanation can account for everything, but multiple

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 l82 JOURNAL OF WORLD HISTORY, JUNE 2OO4

explanations highlight the ways in which a single network was shaped

by multiple forces, including historical contingencies, patterns set by
previous migrants, structural economic relations between different
parts of the world, and local cultural forms such as family structure.
Attention to the global contextualization of particular networks
can also provide insight into larger historical questions. For example,
Qiu Liben has insisted on a global perspective on migration, although
his main aim is to critique the Sinocentrism of his Chinese colleagues'
accounts of Chinese migration.58 He asserts that compared to Euro
peans, Chinese migration had relatively little effect on the world. He
points out that the 30 million Chinese and descendants of Chinese
that could be found around the world in the early 1990s amount to
nothing near the number of descendants of European migrants around
the world?indeed, they amount to little more than the European
descended population of Canada alone. Chinese and other Asians did
travel the world, he argues, but the story of modern world history is still
the story of the Europeans and Western capitalism.
Qiu is unwilling to include migrations to Manchuria into his calcu
lations of total Chinese migration, arguing that this is merely domestic
movement that is irrelevant to global comparison. But the current
Manchurian population of over 100 million suggests a demographic
transition similar to those of European settler colonies in temperate
zones. Without these migrations, Manchuria might not be a part of
China at all today, but part of Russia or Japan, or an independent
nation like Mongolia. Manchuria is best understood as part of the
global expansion into temperate frontiers. In contrast, most southern
Chinese and Indians migrated to tropical areas within or near well
established native states. The establishment of small family farms on
tropical frontiers remains difficult to this day, and most opportunities
for migrants have been as laborers and merchants?typical sojourner
occupations. When we turn the tables, perhaps only Costa Rica, Cuba,
Venezuela, and Colombia are tropical areas with populations of Euro
pean descent at all comparable to the large populations of Chinese,
Indian, and African descent in Southeast Asia and the Caribbean.
Of course, Qiu is talking not only about numbers but also about
power. Chinese and Indian merchants, miners, and agriculturalists in
Southeast Asia were ultimately subjected to financial, political, and
military power concentrated in the hands of Europeans. Chinese did
not settle the temperate frontiers of the Americas and Australia in

58 Qiu Liben, Cong shijie kan huaren [Looking at Chinese from a world perspective]
(Hong Kong: Nandao, 2000). He acknowledges that many of his colleagues estimate 20
million overseas Chinese migrants, but prefers an estimate of 10 million (pp. 19, 30).

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
McKeown: Global Migration, 1846-1940 183

great numbers because they did not have the power to influence gov
ernment regulations and social opportunities in their favor. Chinese,
however, were able to mobilize long-standing commercial and labor
recruiting networks to develop a niche as middlemen and business
men. No broad patterns of migration and settlement should be taken
for granted. They are all the product of particular social and economic
histories.

Contemporary Migration

After a decline in migration over the mid-twentieth century, the world

is once again experiencing a surge in migration. This surge is often
claimed to be an unprecedented expansion of mobility, subverting the
territorial nation-state and creating bold new transnational spaces.59
To what extent can these assertions be maintained? Is this an unprece
dented wave of migration that challenges national sovereignty in a
way that is different from earlier migrations?
Counting people of "migrant stock" is the preferred way of mea
suring migration by the United Nations and International Organiza
tion of Migration (see table 3). This number is obtained by counting
foreign-born peoples in national censuses. It is a very imperfect form of
measurement, because some censuses count foreign birth, while others
count only foreign residents who have not become citizens, and others
merely note racial or ethnic distinctions. This measurement may also
count people who have never moved all their lives while international
borders have moved around them. Thus, political disturbances and the
drawing of borders in South Asia has caused South Asian "migrant
stock" to include 20 million individuals over much of the late twen
tieth century. This accounts for nearly 15 to 25 percent of the world
"migrant stock" over this period, even though South Asia is more
important as a migrant-producing than as a receiving region.
Alternatively, we can attempt a crude estimate of annual migration
flows in the 1990s.60 Annual migration to the European Union
amounted to 1.2 million legal migrants and 400,000 to 500,000 ille

59 The literature on this topic is vast. Let me suggest only two of the more sustained
arguments, Stephen Castles and Mark Miller, The Age of Migration: International Population
Movements in the Modern World (New York: Guilford Press, 1993), and Alejandro Portes,
Luis E. Guarnizo, and Patricia Landolt, "The Study of Transnational ism: Pitfalls and Prom
ise of an Emerging Research Field," Ethnic and Racial Studies 22 (1999): 218-237.
60 Susan E Martin, ed., World Migration Report 2000 (Geneva: United Nations and
World Organization for Migration, 2000); Hania Zlotnik, "International Migration 1965
96: An Overview," Population and Development Review 24 (1998): 429-468.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 184 JOURNAL OF WORLD HISTORY, JUNE 2OO4

Table 3. Migrant stock as portion of world population

Year Foreign-born popuation (millions) Percentage of world population

1910 35.7 2.0

1930 42.9 2.1
1965 75.2 2.3
1975 84.5 2.1
1985 105.2 2.2
1990 119.8 2.3
2003 175 2.8

Sources: International Labour Office, World Statistics of Aliens: A Comparative Study of Census
Returns, 1910-1920-1930 (Geneva: 1936), p. 56; World Organization of Migration, World
Migration Report 2003; Zlotnik, "International Migration," p. 431. Segal, Atlas of World
Population, counts 90 million migrants in 1910, which amounts to 5.5 percent of the world's
population, but gives no source for this number.

gals. Migration to the United States averaged 860,000 legal migrants

a year and perhaps another 300,000 illegals (still less than the highest
numbers of 1912-1913). Migration into Canada, Australia, and New
Zealand accounts for another 300,000 each. Over 1 million migrants
have gone each year to the Persian Gulf states and Israel. Over 500,000
asylum applications were made each year around the world, often not
counted in migration statistics. Other major destinations include
Argentina, Venezuela, South Africa, and Japan, and large flows have
moved between countries in Africa, Southeast Asia, and the ex-Soviet
republics. A generous estimate of 2 to 3 million migrants a year for
these other destinations would give an annual migration of 7 to 8 mil
lion a year. With a hypothetical return rate of 40 to 45 percent, this
could account for the difference of 55 million migrants found in
migrant stock estimates from 1990 to 2003.
Peak transoceanic (and Siberian) migration rates around 1912 and
again in the late 1920s reached over 3.3 million a year. Contemporary
migration is double to triple that number, but a total of 80 million
migrants in the 1990s would account for only 1.5 percent of the pop
ulation, compared to the 1.7 percent of world population attained by
total migration in the decade after 1910. To match early twentieth
century proportions, migration in the 1990s would have to average at
least 9 million per year?a number that is unlikely but not impossible.
Of course the two numbers are not comparable. The numbers I cal
culated for 1840 to 1940 mostly count steerage class ship passengers.
The numbers for the 1990s include people defined by a bewildering
variety of administrative categories, including temporary labor
migrants, people planning to stay for more than a year, refugees, and

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
McKeown: Global Migration, 1846-1940 185

people who may have already resided abroad for decades but only now
applied for permanent residence (as is the case in United States
statistics). My numbers also count people who would not normally be
defined as "international migrants," such as those who traveled to
Manchuria, Siberia, Burma (within the colonial jurisdiction of India),
or Kazhakistan (which would now be counted as international migra
tion). But they do not count many migrants within Europe, the Amer
icas, Asia, and Africa who did cross international borders. Moreover,
increased immigration restrictions around the world means that many
contemporary long-distance migrants are uncounted because they are
illegal.
Despite these difficulties, it seems not unreasonable to assert that
the mobility of people intending to stay for a relatively long period of
time (more than a year) reached levels in proportion to world popula
tion in the 1990s that are comparable to the early twentieth century.
But, even if accurate, this still may not be a meaningful comparison.
For example, 454 million annual tourist arrivals were counted in 1990,
reminding us that temporary global mobility has surely reached unpre
cedented proportions.61 In fact, the shifts in categories used to measure
migration may be of more significance than the actual numbers them
selves. We cannot understand migration without simultaneously
understanding the social processes by which they have been produced,
recorded, and processed as "refugees," "guest laborers," or "illegal
aliens." The new categories (like the high tourist numbers) remind us
that the structures of global mobility are constantly changing in the
context of both political and economic processes. Volume alone is
meaningless if we do not understand the broad patterns and distinc
tions that shape migration, and have been used to understand, calcu
late and control migration. Those very borders that define interna
tional migration, and the populations inside them which are now seen
to be challenged by that migration are themselves the products of pre
vious migration and regulation, much of which has been quickly for
gotten.

Appendix: Sources for Migration Statistics

Most of the statistics used in the tables and figures, except for overseas
migration from South China, are from published sources, which are
mostly based on port and immigration statistics. The availability of sta

61 Zlotnik, "International Migration," p. 432.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 i86 JOURNAL OF WORLD HISTORY, JUNE 2OO4

tistics does not mean the numbers are necessarily reliable. ]. D. Gould,
"European Inter-Continental Emigration 1815-1914: Patterns and
Causes," European Journal of Economic History 8 (1979): 598-605, and
several articles in Imre Ferenczi and Walter Willcox, eds., International
Migrations, v. 2, Interpretations (New York: National Bureau of Eco
nomic Research, 1931), discuss the sometimes serious discrepancies
between departure and arrival statistics in the transatlantic migra
tions. I have encountered similar problems with Chinese and South
east Asian sources. The difficulties of accounting for return and repeat
migration further complicate the significance and reliability of any
quantitative estimate.
For the transatlantic numbers, Ferenczi and Willcox, eds., Interna
tional Migrations, v. 1, Statistics (New York: National Bureau of Eco
nomic Research, 1929), p. 172, is used for the years up to 1880, and
B. R. Mitchell, International Historical Statistics: Europe 1750-1993, 4th
ed. (New York: Macmillan Reference, 1998), pp. 129-140, is used for
subsequent years. Over 200,000 migrants from Turkey, Lebanon, and
Syria are not included in Figures 1 and 2, but discussion of this emi
gration can be found in Robin Cohen, Global Diasporas: An Introduc
tion (Seattle: University of Washington Press, 1997), p. 95, and Akram
Fouad Khater, Inventing Home: Emigration, Gender and the Middle Class
inLebanon, 1870-1920 (Berkeley: University of California Press, 2001).
Numbers for migration from Russia into Siberia and central Asia
before 1914 are from Donald Treadgold, The Great Siberian Migration:
Government and Peasant in Resettlement from Emancipation to the First
World War (Princeton, N.J.: Princeton University Press, 1957), pp.
33~35- Numbers for the period from 1914 to 1927 are more specula
tive. Estimates can be found in Eugene Kulischer, Europe on the Move:
War and Population Changes, 1917-47 (New York: Columbia Univer
sity Press), p. 83; V. V Obolensky-Ossinsky, "Emigration from and
Immigration into Russia," in Ferenczi and Willcox, International Migra
tions, v. 1, p. 576; and Treadgold, pp. 236-238. Kulischer (p. 112)
offers an estimate of 4.7 to 5 million migrants for the period from 1927
to 1939. This estimate has circulated widely, but archival research may
lead to a revision.
Nearly 25 million Chinese to Manchuria after 1891 are counted in
Thomas Gottschang and Diana Lary, Swallows and Settlers: The Great
Migration from North China to Manchuria (Ann Arbor: University of
Michigan, Center for Chinese Studies, 2000), p. 171, but much over
land migration may have passed uncounted. Gottschang and Lary (p.
64) also suggest that up to half a million Chinese moved into Siberia
before 1917. Numbers for migration to Manchuria before 1891 are

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
McKeown: Global Migration, 1846-1940 187

highly uncertain. Migration controls were loosened during the period

1876-1879 because of a famine in North China and 900,000 migrants
were estimated in 1876 alone (Gottschang and Lary, p. 47). Anecdotal
accounts also mention constant streams of new settlers after i860. Ge
Jianxiong, Cao Shuji, and Wu Songdi, Jianming Zhongguo yimin [Con
cise history of Chinese migration] (Fuzhou: Fujian Renmin Chuban
she, 1993), p. 478, estimates that 14 million migrants moved to Man
churia from i860 to 1907, based on population growth. This same
method gives them a serious underestimate of 4 million migrants from
1912 to 1930. Migration before the turn of the century likely contained
a larger proportion of families and settlers, and a smaller flow could
have produced higher population growth. I have used a rather low
estimate of about 3.5 million migrants before 1891 for the charts. The
actual number could have been as high as 6 to 7 million.
Numbers for Korean and Japanese migration are assembled from
Hideo Totsuka, "Korean Immigration in Pre-War Japan," in Commis
sion Internationale D'histoire des Mouvements Sociaux et des Struc
tures Sociales, ed., Les Migrations Internationales: De la fin du XVIII?me
si?cle ? nos jours (Paris: Editions du CNS, 1980), pp. 263-279;
Kulischer, Europe on the Move, p. 86; and Ferenczi and Willcox, Inter
national Migrations, v. 1, pp. 934-935.
Figures for Indian migration are from Kingsley Davis, The Popula
tion of India and Pakistan (New York: Russell and Russell, 1951), pp.
99-100. To estimate distribution between countries, I used Nalini
Ranjan Chakravarty, The Indian Minority in Burma (London: Oxford
University Press, 1971), pp. 21, 188; Ferenczi and Willcox, Interna
tional Migrations, v. 1, pp. 900-907, 915; and Kernial Singh Sandhu,
Indians in Malaya: Some Aspects of Their Immigration and Settlement
{1786-1957) (Cambridge, U.K.: Cambridge University Press, 1969),
pp. 373-380. Davis counted only those Indians who left as labor emi
grants. In years for which I could assemble fairly complete estimates
based on immigration statistics from different countries the numbers
were 15 to 30 percent higher than Davis's estimates. Claude Marko
vits, "Indian Merchant Networks Outside India in the Nineteenth and
Twentieth Centuries: A Preliminary Survey," Modern Asian Studies 33
(1999): 895, suggests that at least 1.5 million commercial migrants
could be added to Davis's numbers. I have adhered to Davis's estimates
in preparing the charts because they offer the most complete and con
sistent data set, and because so much of the migration to Ceylon and
Burma was repeat migration that I thought a low estimate of gross emi
gration would make the net significance of Indian migration more
comparable to the other migration flows.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
 i88 JOURNAL OF WORLD HISTORY, JUNE 2OO4

For emigration from South China ports other than Hong Kong and
Macao before 1876, I made estimates based on G.William Skinner,
Chinese Society in Thailand: An Analytical History (Ithaca, N.Y: Cornell
University Press, 1957), pp. 58-59, and on anecdotal reports found in
Meagher, pp. 143-145, and Lee Poh Ping, Chinese Society in Nineteenth
Century Singapore (Singapore: Oxford University Press, 1978), p. 86,
that suggest immigration rates of 10,000 a year to Singapore in the
1840s and 1850s and 20,000 a year in the early 1870s.
Statistics on indentured labor migration from China until 1874
and emigration from Hong Kong from 1846 to 1855 are from Arnold
Meagher, "The Introduction of Chinese Laborers to Latin America:
The 'Coolie Trade,' 1847-1874" (Ph.D. dissertation, University of Cal
ifornia at Davis, 1975), pp. 105A, 108A. Hong Kong emigration data
from 1856 to 1939 were generously provided to me by Elizabeth Sinn,
from her Hong Kong Research Grants Council funded project, "The
Impact of Chinese Emigration on Hong Kong's Economic Develop
ment, 1842-1941." A summary of some of this data is in Elizabeth
Sinn, "Emigration from Hong Kong before 1941: General Trends," in
Ronald Skeldon, ed., Emigration from Hong Kong: Tendencies and Impacts
(Hong Kong: Chinese University Press, 1995), pp. n-34. Data for
other ports in South China are from the Returns of Trade and Trade
Reports of the Chinese Maritime Customs from 1876 to 1928, Lian Xin
hao, "Jindai haigang jianji yu Dongnanya huaqiao yimin" [Modern
port hygiene and Chinese migrants to Southeast Asia], Huaqiao Hua
ren Lishi Yanjiu (1997): 10th anniversary: 50-51, for Xiamen (Amoy)
from 1931 to 1940 and for Shantou (Swatow) from 1920 to 1934; and
Li Zhiya and Huang Yinying, eds., Huaqiao yu Qiaowu Shiliao Xuanji
Bian [Selected documents on overseas Chinese and overseas Chinese
affairs] (Guangzhou: Guangdong Renmin Chubanshe, 1991), pp. 39,
133-134, for Hainan from 1937 to 1938 and for Shantou from 1937 to
1939. I made estimates based on colonial immigration reports for the
remaining missing years in the 1930s.
I used emigration data from Hong Kong and Chinese customs
reports as the basis for preparing Figures 1 and 2 because they offer the
most complete statistical series. They are generally considered to be an
undercount because they often did not count emigrants who traveled
in Chinese junks and many Chinese tried to avoid medical exams
given by customs officers, and I have made additions based on colonial
statistics. The reports of the protector of Chinese in Singapore for 1881
to 1939 generally corroborate quite well with Chinese emigration sta
tistics, with the exception of migration from Hainan in the years up to
1902. Skinner's Chinese Society in Thailand, pp. 61 and 173, gives esti

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms
McKeown: Global Migration, 1846-1940 189

mates based on Thai customs reports after 1882 that are up to 30 per
cent higher than emigration reports. Immigration statistics in the
Annual Report of the Philippine Commission and the Annual Report of the
Bureau of Customs in Manila from 1899 to 1939 are also up to 30 per
cent higher than emigration numbers.
Statistics for the Dutch Indies from 1900 to 1938 in Statistisch Jam
overzicht van Nederlandsch Indi? (Batavia) and Departement van Econ
omische Zaken, Volkstelling IQ30, v. 7, Chineezen en Andere Vreemde
Oosterlingen in Nederlandsch-Indi? (Batavia: 1935) are divided accord
ing to a variety of shifting and unexplained administrative categories.
I have found it impossible to make them correspond with each other
or with migration statistics from China and Singapore. I have thus
assumed that most migration to the Dutch Indies has been accounted
for by figures from China and Singapore. I have made the same assump
tion for seaborne migration to Burma.
Numbers for French Indochina are more perplexing. Immigration
statistics for Cochinchina (southern Vietnam) from 1879 to 1883 and
Hanoi from 1920 to 1924 are in Ferenczi and Willcox, International
Migrations, v. 1, pp. 154 and 920. Numbers for all of French Indochina
from 1923 to 1940 are in Annuaire Statistique de VIndochine (Hanoi). In
only one year (1926) do emigration statistics account for more than 30
percent of immigration statistics. Most migrants probably arrived by
junk, as suggested in Charles Robequain, The Economic Development of
French Indo-China, trans. Isabel Ward (London: Oxford University
Press, 1944): "The rich and well-to-do Chinese travel on big liners
like Europeans; but most of the immigrants are crammed together in
picturesque, ill-smelling groups on the decks of small boats." Assum
ing that many migrants to French Indochina were not counted as emi
grants by Chinese customs, I have constructed estimates of Chinese
migration to French Indochina based on available immigration statis
tics, population estimates, and anecdotal evidence that migration rates
decreased temporarily after the imposition of immigration restrictions
in 1906. See Wang Wen-Yuan, Les relations entre l'Indochine Fran?aise
et la Chine (Paris: Editions Pierre Bossuet, 1937), pp. 16-17.
The trends of Chinese migration recorded in Figures 1 and 2
account for 18.9 million Chinese emigrants to Southeast Asia and the
South Pacific. I think this estimate is conservative, especially before
1881 and after 1928, and the actual number could be as high as 22
million.

This content downloaded from

193.6.61.16 on Thu, 01 Dec 2022 12:46:29 UTC
All use subject to https://about.jstor.org/terms