In the first week of October 1932, an International Conference on Migration Statistics was held in Geneva. Over the course of five days, some thirty statisticians from twenty-six countries discussed how to produce more reliable international migration statistics. This kind of methodological discussion about statistical standardization was not at all unusual in the new world of international organization. Since 1920, the standardization of statistics had become an ordinary activity in the “Palace” of the International Labour Organization and the League of Nations in the hills above Lake Geneva.
The International Conference on Migration Statistics offers particularly interesting insights into the historical attempt by international organizations to measure the world. On the one hand, “international migration” was not yet a category in scholarship and policy making. It was an international invention intended to bring together the existing categories of “emigration” and “immigration.” Before this time, these last two categories were perceived as two fundamentally separate phenomena. Perhaps more plainly than other objects targeted by statistical analysis, “international migration” was connected to the effort to construct a new international understanding of the world after the Great War.
On the other hand, the discussions among the statisticians at the conference also revealed barriers to and complications for the creation of an international consensus on the subject under consideration. The biggest problem was that the statistical object “international migration” touched on delicate political questions concerning the significance of territories, borders, and belonging in a world organized substantially along colonial lines whose legitimacy faced increasingly serious opposition. Focusing not only on the statistical knowledge itself but also on the context in which it was produced, the discussion of statistical problems in Geneva in 1932 manifested contemporary international political tensions and genuine ambivalence about the role of international organizations and their new international point of view.
Statistics and the International Point of View
The special character of the statistical object discussed at the Migration Statistics Conference was depicted in the January 1933 issue of International Labour Review as follows:
Most of the subjects (wages, prices, unemployment, etc.) dealt with by previous conferences have a primarily national character; migration, on the contrary, is an international phenomenon. Every emigrant who leaves one country becomes also an immigrant in another. Neither the statistics of emigration nor the statistics of immigration are in themselves sufficiently adequate records of the total streams of migration in the world.1
Nowadays, the existence of phenomena of international character is not disputed. Migration in particular is widely seen as an international or global issue, but this was not the case in the early 1930s, when the phenomenon was still viewed almost exclusively from the perspective of individual sovereign states2. Whereas “emigration” and “immigration” were meaningful categories in political discourses, “international migration” existed principally as a statistical category, one invented by the International Labour Office in the 1920s. The novelty of the category was evident in the conference description, which assumed that readers still had to be informed about why “migration” was an “international phenomenon.”3
This reasoning shows how the term “international migration” was used to demonstrate that the international standpoint was objective. Only from an external, international position was it deemed possible to grasp the entirety of the situation that immigration and emigration comprised. Whereas the “national perspective” was depicted as naturally partial (but not wrong), the “international perspective” was held to open up a view on the “total streams of migration in the world.”4 Consequently, this undertaking in the production of international knowledge was about not only making visible a new issue but also constructing and legitimizing a new viewpoint.
Statistics were the medium through which the international point of view was made visible.5 Since 1922, statisticians in the International Labor Office produced regular reports on “Migration Movements.” They collected statistics from member countries and developed criteria to use in the transfer of these numbers into international tables. Contemporary reports, however, clearly showed that the varying definitions of “emigration” and “immigration” used by the reporting offices made the data difficult to compare across state boundaries. Nonetheless, the transformability of numbers made it possible to convert them into the new order of international tables. Consequently, the statisticians treated the numbers as if they were based on internationally agreed-upon definitions. In turn, this translation of statistical knowledge made the resultant international statistical tables the basis for discovering the world from an international standpoint.6
By interpreting the international tables, statisticians uncovered purportedly wider connections and deeper truths. In 1923, in their first report on “Migration Movements throughout the World,” which drew on statistical tables summarizing total “emigration in the world” before and after the Great War, experts in the Labour Office concluded,
An examination of the part played by each country in these movements cannot fail to bring out the fact that this decline in migration has been universal. It seems to result from causes which go very deep, and which are greater than any particular event or tendency. The phenomenon is world-wide.7
The statistics here were comparable with maps in that they were used as a medium to depict the world as a coherent space with drastically reduced complexities and variations. In its most reduced form, the world here was reduced to a single number representing “the total volume of international migration” in a specific time period.8 As with map making, however, the significance of statistical representations of the world were burdened with historical preconditions.9
The Problem of Colonial Borders
Compared to the evident faith in the possibility of uniform migration statistics at the beginning of the 1920s, the mood among the statisticians at the 1932 conference was rather sober:
In view of the imperfect state of migration statistics and the differing needs and practices in the various States, the Conference did not consider it possible at the present time to lay down any precise or limited definition of migration with any prospect of its being universally applied.10
Inconsistency between national and international statistics did not appear to be the main problem, however. In the view of the statisticians represented in Geneva, the problems of international migration statistics were mainly due to the varied meanings of borders and categories of belonging within empires.
Colonial empires claiming sovereignty over large parts of the world in the 1930s undermined the meaning of the category “international migration” and the credibility of the international point of view. The category “international migration” presupposed national borders and their function of transforming people who crossed them from “nationals” or “citizens” into “aliens” or “foreigners.” In a methodological paper written as a basis for discussions at the conference, Labour Office statisticians described the problem with colonial statistics this way:
The classification of migrants according to nationality presents special difficulties in the case of colonial migration. Generally speaking, it is possible to distinguish:
(1) migration from the home country to its colonies, or vice versa;
(2) migration from one colony to another of the same nation;
(3) migration between colonies of different States.
This distinction may be important from the point of view both of the colonial Power and of the colony. From the international point of view, it is important to find criteria which will make it possible, in the first place, to distinguish between nationals and aliens. The information at present recorded in the statistics of colonial Powers and their colonies makes this distinction difficult or even impossible.11
Neither the imperial organization of territories and their borders nor the legal meanings of belonging in colonial situations corresponded to international criteria for ordering people in relation to nation-states.
We can see that “methodological nationalism” was inherent not only in the scientific knowledge production of individual countries but also in the knowledge production of international organizations.12 At the same time, the problems faced by international statisticians show that the nation-state was not yet a global norm in the interwar period.
Tensions between the points of view of colonial powers and the fledgling international point of view also show how statistical and political questions could be connected. This overlap became especially evident in the critical stance of the British government toward the international production of global migration statistics. Already in its first official statement in 1922, the British government rejected the plans of the Labour Office. Neither the development of international classificatory schemes nor the production of uniformly organized sets of statistical migration data was deemed possible in the administration of the British Empire.
If the international categorization of people on the move did not make sense from a British perspective, the very category of “international migration” raised questions about the significance of borders—in terms of both belonging and territoriality—within the Empire. Viewed in the context of nationalistic movements and their use of the category “nation” to raise claims against imperial rule after the Great War, international definitions of national borders implied a potential danger for conceptions of a British imperial world order.
When the invitation to the Migration Statistics Conference in Geneva arrived in London on May 28, 1932, the decision within the British administration to reject it was undisputed. The responsible administrators in the Board of Trade deemed the conference “a waste of time.”13 Moreover, British functionaries tried to convince statisticians in the employ of other states that the conference was senseless. This attempt to sabotage the conference failed, however, and it took place according to the Labour Office’s plan.
Although the British absence at the conference wasn’t mentioned in the Labour Office’s conference report, the participation of a statistician from the United States was emphasized: “Twenty-six States and the Secretariat of the League of Nations accepted invitations to the Conference; it was particularly gratifying that the United States—the greatest immigration country of the world—sent a representative.”14 Indeed it was remarkable that the U.S. government sent a representative to Geneva at all given its notably ambivalent relationship with the Genevan world of international organizations in the interwar period.
In contrast to the British case, however, the international point of view and the category of international migration became knowledge of national interest in the United States during the 1920s. On the initiative of the Social Science Research Council in New York, the International Labour Office was even asked to undertake a historical study on international migration statistics. The study, published in 1929, was the largest statistical publication on “international migrations” in the interwar period.15 In this case, however, the international ordering of people offered a potential source of knowledge for the construction of “national origins” in the United States.
It is well known that statistical knowledge in general and U.S. Census Bureau statisticians in particular played an important role in the conceptualization and implementation of the so-called National Origins Act of 1924 and its drastic restrictions on immigration based on racial and national categories.16 A deeper look into the specific circulation of this knowledge in the U.S. Census Bureau could bring new findings on the extent to which international statistics were used to legitimize the new U.S. immigration policy.
Besides raising questions about the specific functions of international statistical knowledge, an examination of the statistical category of international migration and of the logic of statistical standardization points to how both contributed in the long run to the legitimization of efforts to sort people along national lines in the resulting international world order. Considering such effects from a history of knowledge perspective suggests that changes in the movements of people do not by themselves explain the global spread of the category “international migration” in the second half of the twentieth century. Instead, we must also consider the role that the distinct international point of view—reinforced by the migration statistics compiled in interwar Geneva—played in this period of imperial decline.17
Yann Stricker is pursuing a PhD in history at the University of Lucerne, Switzerland.
- International Labour Organization, “An International Conference on Migration Statistics,” International Labour Review 27, no. 1 (January 1933): 1. ↩︎
- On the history of the statistical perspective of individual sovereign states, see Léa Renard Nationalité, migration et ‘race’ au prisme de la statistique (1880–1920): Une tentative d’histoire croisée franco-allemande” (PhD diss., University of Potsdam, forthcoming). ↩︎
- Ibid. ↩︎
- Ibid. ↩︎
- On the mediality of statistical numbers, see Bettina Heintz, “Numerische Differenz: Überlegungen zu einer Soziologie des (quantitativen) Vergleichs,” Zeitschrift für Soziologie 39, no. 3 (2010): 162–81. ↩︎
- On the connections between international statistics and representations of the world, see Bettina Heintz, “Welterzeugung durch Zahlen: Modelle politischer Differenzierung in internationalen Statistiken, 1948–2011,” Soziale Systeme 18, nos. 1–2 (2013): 7–39. ↩︎
- International Labour Organization, “Migration Movements throughout the World in 1913, 1920, and 1921,” International Labour Review 7, no. 4 (1923): 539. ↩︎
- International Labour Organization, “Migration Movements throughout the World in 1913, 1920, and 1921,” International Labour Review 7, no. 4 (1923): 538. On the effectiveness (and absurdity) of statistical numbers as “world representations,” see Daniel Speich, Die Erfindung des Bruttosozialprodukts: Globale Ungleichheit in der Wissensgeschichte der Ökonomie (Göttingen: Vandenhoeck und Ruprecht, 2013); and Patricia Hongler “Erzählweisen der Dekolonisierung: Berichte aus dem Archiv der OECD” (PhD diss., University of Lucerne, forthcoming). ↩︎
- On the historical construction of national maps, see David Guggerli and Daniel Speich, Topografien der Nation: Politik, kartografische Ordnung und Landschaft im 19. Jahrhundert (Zürich: Chronos, 2002). ↩︎
- International Labour Organization, “An International Conference on Migration Statistics,” International Labour Review 27, no. 1 (January 1933): 5. ↩︎
- International Labour Office, “Statistics of Migration: Definitions—Methods—Classifications,” Studies and Reports, ser. N (1922): 101–2. ↩︎
- Andreas Wimmer and Nina Glick Schiller, “Methodological Nationalism and Beyond: Nation-State Building, Migration and the Social Sciences,” Global Networks 2, no. 4 (2002): 301–34. On the entanglement of “nationalism” and “internationalism,” see Glenda Sluga, Internationalism in the Age of Nationalism (Philadelphia: University of Pennsylvania Press, 2013). For an example of how migrant knowledge from the Labour Office was used by academics, see Kijan Espahangizi, “The Granddaughter’s Dissertation: Some Thoughts on Knowledge about Migration in 1960s Switzerland,” History of Knowledge, August 10, 2017, https://historyofknowledge.net/2017/08/10/the-granddaughters-dissertation/. ↩︎
- Internal report by H. Leak, June 29, 1932, International Conference of Migration Statisticians, Board of Trade, BT 70/35, British National Archives. ↩︎
- International Labour Organization, “An International Conference on Migration Statistics,” International Labour Review 27, no. 1 (January 1933): 3. ↩︎
- Walter F. Willcox and Imre Ferenczi International Migrations (New York: National Bureau of Economic Research, 1929). ↩︎
- Margo J. Anderson, The American Census: A Social History (New Haven, CT: Yale University Press, 1988); and Mae M. Ngai, “The Architecture of Race in American Immigration Law: A Reexamination of the Immigration Act of 1924,” The Journal of American History 86, no. 1 (1999): 67–92. ↩︎
- This approach is inspired by our discussions in Daniel Speich’s research group, Statistics between Empire and Nation: Generalized Knowledge in European-African relations, University of Lucerne, Switzerland. ↩︎