Since the onset of the “Great Recession,” comparisons with previous economic crises have been frequent. Many observers have noted that the protectionist and autarkic policies of the Great Depression have failed to materialize, despite pessimistic forecasts and isolated instances of protectionism. Less remarked upon has been the weakness of other forms of “economic nationalism” in which new or firmer boundaries are drawn between citizens and noncitizens in determining access to the labor market or eligibility for social rights or benefits. To be sure, states have taken some initial measures to protect citizens from economic distress, such as awarding contracts to firms owned by or employing citizens, imposing new limits on immigration, negotiating deals with private companies to keep jobs onshore, and cutting back on certain benefits provided to noncitizens. But despite a few high-profile examples, policies enacted in response to worsening economic conditions and based on membership in the nation-state have to date remained modest.Footnote 1

This stands in stark contrast to the depression years of the 1930s. In the United States and France—the two states with the largest foreign-born populations per capita at the time—noncitizens experienced widespread restrictions on jobs and participation in public works projects, uncertain or discriminatory access to social assistance, and, perhaps most infamously, new and often irresistible pressures to “repatriate” to sending countries. In these ways, one’s citizenship status intruded into areas of life where it had previously held little or no relevance in both countries. This article draws upon and supplements theories on the relationship between state formation and the making of modern national communities to explain how the significance of citizenship status increased in both countries. Its core argument is that this was the result of the expansion of the power of receiving states to control entry, compel exit, and, in the context of economic crisis, fund social assistance programs through subsidiary levels of governance. Sending states, meanwhile, increasingly provided what I call “remote protections”Footnote 2 to their citizens abroad through bilateral labor treaties and expanded consular support. In both ways, receiving and sending states began to more comprehensively “cage” their populations—to use Mann’s apt metaphor (Mann 1993, pp. 20, 251)—and this enabled citizenship to have new impacts on the lives of foreign-born populations with the onset of the Great Depression.

The new salience of citizenship status was most dramatically seen in the unprecedented repatriation pressures experienced by noncitizens in both countries. According to US statistics, total emigrant departures jumped from 61,882 in 1931 to 103,295 in 1932 and 80,081 in 1933—the first years when net migration to the United States were negative. Populations of Mexican origin were most affected by repatriation pressures, and the Mexican Migration Service recorded that 457,023 nationals returned between 1929 and 1937. The majority of Mexican returnees were “self-propelled,” although approximately 32,000 were formally deported, 44,000 opted for “voluntary” deportation to avoid deportation proceedings, and it is estimated that up to 90,000 returned through organized repatriation by relief agencies acting with or without the cooperation of the Mexican government (Hoffman 1978, p. 232). In France, over the five-year period spanning 1931–1935, 357,842 left the country, with the more recent arrivals from Eastern Europe accounting for over half of this total.Footnote 3 We know much about the repatriation pressures captured by these statistics thanks to the contributions of an interdisciplinary group of scholars who have approached this history from the perspective of immigrant groups or a single mode of state activity, such as migration control or the provision of social assistance. But the existing literature also exhibits certain weaknesses. First, it has yet to give due attention to the numerous ways the expansion of the powers wielded by sending and receiving states interacted to elevate the importance of citizenship status and create new repatriation pressures. Second, given that the experience of single countries or specific immigrant populations has dominated research agendas, the existing literature has not attempted to link this experience systematically to the structural changes underway within and between nation states at this time. For this reason, there have been no attempts to uncover the theoretical implications of this process.

I aim here to fill these gaps by investigating the relationships among state expansion, citizenship, and repatriation pressures during the interwar period in the United States and France and by scrutinizing a neglected feature of the repatriation phenomenon. But before turning to the empirical evidence, I first discuss how my theoretical approach builds upon prior efforts to study processes of nation state formation and their impact on noncitizens. I then briefly turn to the main substantive stakes involved in my reinterpretation of the repatriation phenomenon by critically engaging a growing body of scholarship on the repatriation of individuals of Mexican origin from the United States during the Great Depression.

Theoretical framework

Scholars have approached the relationships among state expansion, migration control, and citizenship from four distinct, yet interrelated, angles. The first considers the development of the state’s “infrastructural power” (Mann 1993, p. 59) to “control the legitimate means of movement” (Fahrmeir, Faron, and Weil 2003; Torpey 2000) through the adoption of administrative innovations such as the usage of passports or other forms of documented identity (Noiriel and About 2007), consular “remote control”(Zolberg 2003) or border surveillance and law enforcement (Rosenberg 2006). Another focuses on the politics of immigration restriction (Freeman 1995; Tichenor 2002) and the evolution of immigration law and policy (Calavita 1984; Ngai 2004). A third literature approaches the state as an institution through which novel forms of social closure emerge and where struggles are waged over membership in the state and over the collective resources distributed by it (Brubaker 1992; Rothschild 1981; Wimmer 2002). As several scholars have noted, the emergence of welfare states in particular created new tensions between immigrants and native-born populations (Bommes and Geddes 2000; Fox 2012; Freeman 1986). A fourth approach shifts the focus away from the states that have historically been the targets of immigration to the policies of sending states (Choate 2007; Fitzgerald 2009; Martin 2008; Rodriguez 2010).

I draw selectively on all four approaches to explain the repatriation pressures that are the subject of the substantive sections of this paper. My main agenda theoretically, however, is to build upon theories that link the historic process of state-building to the boundaries drawn between “insiders” and “outsiders” based on citizenship status. A starting point is provided by Wimmer’s general theoretical model describing the making of modern national communities in terms of “the mechanisms that tie nationalist principles to the institutionalized practices of inclusion and exclusion in the modern state” (Wimmer 2002, p. 9). In its ideal type, there are five dimensions--legal, political, military, social, and cultural—across which national communities become closed in a reflexive process concurrent with the historical formation of modern states. By the end of the long twentieth century, the core features of the nation state model (and the social closure it brings with it) could be readily seen in a growing number of states: citizenship laws dividing the lines between citizens and aliens; democratic and nationalist movements displacing foreign rulers in favor of a government representative of the sovereign nation; conscription practices replacing mercenaries with citizen soldiers. But what is of interest here is what occurred after World War I, when, as Wimmer puts it, national closure “reaches its zenith with the emergence of welfare institutions and of state control over migration” (Wimmer 2002, p. 62). At this stage, improvements in state control of migration and the provision of social assistance, whatever their benefits to insiders, also make possible new discriminations between citizens and aliens. When and how states develop this capacity (to the extent that they develop it all) and its impact on processes of closure of course vary across time and space. But for my empirical purposes, what is of special interest is the fact that both France and the United States were approaching their respective “zeniths” during the interwar period. I show in this article that neither state approached the comprehensive closure along national lines as presented in Wimmer’s ideal-typical model, but were nevertheless experiencing homologous processes of state expansion that led to the deterioration of the status of noncitizens.

To explain how state expansions affected the stakes involved in citizenship status, and specifically new repatriation pressures, I also propose an amendment to Wimmer’s model. In his account, the national closure that results from the expansion of state power in the fields of social assistance and migration control considers only processes bounded by the reach of the state within its own borders. It is therefore inattentive to the activities of other states that are projected across state boundaries.Footnote 4 This is a significant omission, for as Noiriel (1996, p. 78) once suggested in broad terms, the two “decisive factors” that defined immigration in the early twentieth century was not just “the development of the welfare state” within receiving states but also “the intensification of international relations” beyond them. More specifically, receiving states were not the only ones expanding their powers in these fields. Sending states were expanding these powers too. And their activities, even when designed to ensure the protection of their citizens across state boundaries by, for example, negotiating labor treaties ensuring equal access to social assistance, nevertheless served to bolster nationality (rather than, for example, local residency) as the criterion through which social and legal rights and protections could be withheld or denied. To refer to Mann’s “caging” metaphor once again, the expansion of the infrastructural power of sending states began to add further bars to the national cages enclosing citizens, regardless of geographic location.

In my analysis of state expansion, citizenship, and repatriation during the interwar period, I therefore follow the work of Wimmer and others cited above who have demonstrated how changes in the state’s capacity to control migration and provide social assistance changed traditional patterns of inclusion and exclusion. But I also add to this list of state activities forms of “remote protection” provided by sending states. By remote protection, I mean any state activities that are ostensibly designed to provide “protection” to populations claimed as citizens of the state who are outside the state’s borders. What constitutes actual protection is open to debate (a point to which I return throughout this article) but it still captures the basic logic of the state activity. Such sending state activities may include negotiating international treaties, monitoring compliance with local, contractual, and international laws, or providing assistance to citizens to return “home.” To illustrate using one of my comparative cases, international treaties, supervised contracts, and consular support of various kinds became key features in the systems of migration control and the provision of social assistance in France during the interwar period. Thanks to recent scholarship, we know that far from being meaningless accords or a collection of empty promises, bilateral labor treaties and the protections they offered foreign workers were indeed respected, however grudgingly, at various levels of governance. Investigations of migrant rights and citizenship in Marseille and Lyon (Lewis 2007), the policing of foreigners in Paris (Rosenberg 2006), and the expansion of social assistance programs at the local level (Smith 2003) have all shown how the possession of citizenship evolved into a more concrete system of membership through which certain rights and protections flowed or were denied. Those who have studied the experience of individual immigrant groups (Marès 1989; Milza 1986; Noiriel 1986; Ponty 1988; Slaby 2005) have also noted, though not fully explored, this new feature in the landscape of international migration.

The upshot is that a theoretical framework attentive to the evolution of the structures of both receiving and sending states across multiple domains is needed to understand how state power and citizenship affected the lived experience of foreign-born populations during the interwar period and beyond. The activities of sending states, to be sure, have become an important subfield of scholarship on international migration. But their insights have so far neither been applied to the repatriation experience of the interwar period nor put into interaction with the expanding powers of receiving states, as I propose in this article. Indeed, the specializations dividing disciplines and scholars focused on migration control, citizenship, social assistance, and sending states has meant that their collective insights are seldom put in dialogue with one another in the analysis of concrete historical cases. Marshalling their collective expertise within the theoretical framework developed above shows how this chapter in American and French history took each country another step closer to making national citizenship what it remains to this day: a powerful and seemingly “natural” mechanism of social closure separating citizens from noncitizens both within and beyond the nation state’s borders. It also, as I argue below, casts the repatriation pressures experienced by populations of Mexican of origin in the United States in a new light.

Empirical stakes

Until recently, historians have dominated the scholarship on the repatriation pressures on populations of Mexican origin. The scope of their research has varied, attempting to cover either the entire United States (Balderrama 1995; Guerin-Gonzales 1994; Hoffman 1974; Reisler 1976a) or single states or localities (Arredondo 2008; Betten and Mohl 1973; Humphrey 1941; McKay 1982; Simon 1974; Valdés 1988). These contributions, the most comprehensive of which had the stated goal of “relating the story rather than exploring sociological theories” (Balderrama 1995), provide invaluable documentation of local conditions and chronicle the activities of diverse groups of individuals and organizations. But while most refer to at least some of the ways in which the central governments of the United States and Mexico were monitoring, claiming, categorizing, identifying, arresting, deporting, transporting, and disbursing jobs and social goods, none has attempted a systematic accounting of them or sought to place the Mexican immigration experience within a broader theoretical or comparative context. Such neglect may account for the tendency in some accounts to resort to a vague passive voice or characterize (misleadingly) Mexican repatriation as a “program of forced removal of half a million Mexican immigrants and Mexican Americans, sanctioned by the United States government” (Guerin-Gonzales 1994, pp. 74, 77). This interpretation becomes problematic with a clearer accounting not only of the numbers involved, but also of who was doing what within the expanding orbits of state power as discussed in the previous section.

More recently, M. Ngai (2004) and C. Fox (2012) have pushed beyond group-specific accounts by researching Mexican repatriation in a comparative context. Ngai investigates the “sociolegal” history of immigration restriction, situating the Mexican experience within a series of case studies of peoples originating from China, Japan and the Philippines. She argues that these groups shared a common status as “alien citizens” who may have possessed formal membership but for reasons of race were treated essentially as foreigners. For Ngai, the repatriation of Mexicans was a consequence of their “racialized foreignness” and was “a racial expulsion program exceeded in scale only by the Native American Indian removals of the nineteenth century” (Ngai 2004, pp. 71, 74). Fox compares the experiences of Mexicans, European immigrants, and African Americans in order “to clarify the relative influence of race, formal citizenship and legal status for the extension of social citizenship.” She proposes that each group occupied “three different worlds of relief” – worlds that were diverse in terms of regional location, labor market participation, political incorporation, and perceived foreignness on the part of social workers and relief officials. Her assessment of repatriation is more nuanced than Ngai in that she concludes that “there is probably no single reason why Mexicans were targeted for expulsion” (Fox 2012, p. 185). With regard to access to social assistance more generally, she found that populations of Mexican origin were “offered limited access to relief, strongly mediated by their citizenship and legal status (or perceived citizenship and legal status)” (Fox 2012, p. 290). The attitudes of local social workers, who were key gatekeepers of relief and came to see Mexicans as an unassimilable population, emerge as important variables in her account.

Both scholars have advanced our understanding of this pivotal period between the wars when the United States took its first steps in creating regimes of immigration restriction and social assistance now familiar today. However, Ngai posits that race was the key driver of the repatriation pressures affecting Mexican populations—a view that is puzzling given that she signals in the introduction of her book the need for American immigration history to be “rethought” in the context of “global developments and structures,” with specific reference to the “hypernationalist” regimes that followed World War I. Below, I follow this promising line of inquiry while also challenging her exclusively racial interpretation. I supplement Fox’s comparative analysis of the uneven expansions of social citizenship in the United States by exploring in more depth an important feature of the “world” in which populations of Mexican origin (not to mention the social workers who worked with them) navigated: the activism of the Mexican state and engagement of consular officials with relief authorities. Although not entirely absent in her account, Fox does not fully develop the multiple ways in which the latter contributed to Mexicans’ real or perceived “foreignness,” discouraged naturalization, and acted as a spur, rather than a hindrance to repatriation pressures. Supplementing her account with the remote protections provided by the Mexican state also helps explain better how citizenship status became reinforced as an important social category at the local level where it previously held little significance. Indeed, sending state remote protections in France, though taking a quite different form, had similar affects in French localities during this period.

In short, my main critique of the extant US literature is that the discriminations felt by noncitizen populations and especially the repatriation pressures experienced by Mexican citizens in the United States during the depression does not give due attention to the new ways state power was being exercised and citizenship being used by sending and receiving states.Footnote 5 Moreover, there is no attempt to link this historical episode to the more generic processes of nation state formation to which the United States, as the following section explores in detail, was not immune.

The United States

In fact, after World War I the United States was becoming “a nation like the others” (Zolberg 2006) by expanding its power with respect to restricting entry, policing borders, and monitoring and expelling noncitizen populations. Yet it diverged in one key respect. In contrast with the intergovernmental bargaining and treaties that were structuring labor migration in Europe, the United States concluded no international immigrant labor agreements until World War II.Footnote 6 Combined with the Foran Act of 1885 forbidding the importation of labor under contract, the US international labor market was comparatively “free,” at least once American borders were crossed. Instead, controlling the borders themselves became the focus of the federal government’s immigration policies during the interwar years. Denying entry through quotas and restrictive legislation, relying on consular officers to maintain “remote control” from abroad (Zolberg 2003), policing borders with the newly established border patrol, initiating new systems of documenting individual identity (Torpey 2000), and deporting those who breached these controls became the federal government’s means of governing international migration during the 1920s. The effects of this pattern of migration control proved to be unevenly felt among the diverse foreign-born populations already in the United States or actively seeking entry to it.

In brief, war and restrictive quotas meant that populations from southern and eastern Europe—those who had made up the largest share of new arrivals in previous decades and were the current bêtes noires of restrictionists—returned to their countries of origin less often, naturalized more, and, with the exception of a brief surge in 1921, came over in less numbers. Mexican-born populations, by contrast, came (and went) under a de facto guest worker system “governed” by a sending state unable to control emigration but active in providing forms of remote protection to its citizens abroad and a receiving state that allowed limited entry through legal means but lacked the capacity or interest to prevent illegal entry. Return migration was frequent, naturalization low, and the proportion of undocumented entries high. As distinctions between aliens and citizens became more consequential in everyday life and as the power exercised by sending states to “protect” and receiving states to restrict and expel grew stronger, Mexican citizens in the United States were uniquely vulnerable to new mechanisms of social closure based on citizenship status. And in the context of depression, they experienced unprecedented and often irresistible pressures to repatriate. This section provides evidence for these claims, structured in terms of the three expansions of state power emphasized in this article: migration control, provision of social assistance, and remote protection. The analysis follows Mexican-born populations and the activities of the Mexican state in the greatest detail, with ad hoc references to other noncitizen populations and sending states to provide context and bring their particularity into starker contrast.

Restriction, exclusion, expulsion

A seldom discussed legacy of the Great War and the immigration acts of the 1920s was the disruption of patterns of return migration, especially among individuals from countries most affected by quota restrictions (Piore 1979; Wyman 1993). Between 1908 and 1910, approximately one in three who crossed America’s borders from Europe eventually returned to their country of origin. For southern and eastern European immigrants, return migration regularly exceeded fifty percent (Piore 1979, p. 151). The war disrupted this migration pattern, with the number of entries from European countries dropping from just over 1 million in 1914 to under 200,000 the following year (Zolberg 2006, p. 238). Would-be emigrants were effectively caged by warring states concerned about manpower shortages, while legislation restricting entry made its first breakthrough in the United States in 1917 in anticipation of the surge that would follow the peace. When this proved inadequate in the years after the war (the number of emigrants from Europe climbed to 652,364 in 1921) the Immigration Acts of 1921 and 1924 reversed the trend. By the end of the 1920s, the so-called “birds of passage” crossing and re-crossing the Atlantic were coming to permanent rest in American cages. Restriction, coupled with aggressive federally-supported Americanization campaigns (Zolberg 2006, p. 263) ironically had benefits for the groups in the United States most targeted by them. Stigmas that had historically been associated with sojourner populations (Shumsky 1992) eroded and naturalizations increased.

Migration from Mexico followed the opposite trajectory during the 1920s. According to government statistics, the number of entries from Mexico had already doubled during the previous decade, driven northward by the “push” of political unrest, violence, and land reforms in Mexico and the “pull” of demand for cheap labor in the north. During the 1920s, legal migration from Mexico, free from the 1921 and 1924 quotas, grew from 50,000 per year in 1920 to nearly 90,000 in 1924 before falling to a low of 40,000 in 1929. Undocumented migration also surged during this period and return migration was frequent. Indeed, a statistic that is often lost in scholarly accounts of the repatriation pressures encountered by Mexican populations during the 1930s is that between 1926 and 1929, 4 years of relative prosperity and high labor demands, the Mexican Migration Service counted 218,000 returnees, which compares with 306,268 recorded over the following 4 years of depression between 1930 and 1933 (Hoffman 1978, p. 233). Despite the growing share of migration from Mexico, measures taken by the United States government took the form of “ad hoc policy pronouncements” (Corwin 1978a). Caught between growers, railroad companies, and industrialists claiming chronic labor shortages and restrictionists calling for the closure of America’s “backdoor,” workers from Mexico were thrust into a system in which the appearance of control rather than its reality was the most prominent characteristic. US authorities possessed neither the organizational capacity to control the mobility of recruited laborers nor a willingness to collaborate with their Mexican counterparts (Cardoso 1980; Mapes 2009, pp. 122–142). Most importantly, the employers of Mexican labor had little use for government interference on either side of the border.

The state’s new capacity to restrict during the 1920s was freely used in the context of economic depression. Thanks to the expansion of consular remote control, legal immigration to the United States from Mexico, already reduced in 1929 to 40,000, was brought down to as few as 2171 by 1932. At the point of entry, the border patrol’s initial budget allocation of $1 million dollars was doubled in 1928 (Reisler 1976a, p. 59) and in March 1929 legislation was passed making illegal entry a misdemeanor and illegal re-entry after deportation proceedings a felony. Beyond the border, the aggressive deportation activities of the Bureau of Immigration peaked during the Hoover administration. The Bureau’s raids of private residences and public places throughout the country were rhetorically justified as a means of alleviating unemployment and tended to be either “solicited” by local officials who welcomed their tactics or “unsolicited” by less cooperative local authorities. Most noteworthy was the cooperation between local relief authorities and Bureau officials, which grew especially close in southwestern states. The Bureau’s aggressive tactics and frequent disregard for due process continued until 1933, when the Roosevelt Administration took steps to rein in bureau agents. The number of annual deportations then declined. But between the years 1930 and 1932, over 54,000 were formally deported, of which 44 % were Mexicans (Fox 2012, p. 127).Footnote 7 Overall, Hoffman (1974, p. 126) reports that between 1929 and 1935, “the federal government removed from the United States, either under warrant proceedings or as voluntary departures, a total number of Mexicans in the neighborhood of 82,400.”

Even noncitizens able to prove their legal right to remain in the country were not spared from other formal exclusions based on citizenship status. Much had changed since the last prolonged depression in the 1890s, when Higham (1955, p. 73) found that “discriminatory legislation was far from common,” and that “a depression environment exacerbated anti-foreign sentiment in much broader, less obvious ways.” Since then, formal, citizenship-based distinctions had become specific, punitive, and pervasive at all levels of governance. During World War I, naturalization and citizenship status became the focus of Americanization campaigns, conflicts over draft eligibility, and federal legislation in 1918 that doubled the income tax requirements for a vaguely defined category of “non-resident aliens” (Higham 1955, p. 248). State legislatures throughout the country also passed laws formally barring access to certain professions and public works jobs to noncitizens; a trend that accelerated during the depression.Footnote 8 By the eve of the depression, four out of five labor unions and three out of five jobs in large and small automotive, transportation, manufacturing, and construction corporations demanded citizenship as a prerequisite for membership or employment according to one survey (Fields 1935, p. 534).

Such restrictions within the United States prompted one Mexican consul stationed in Chicago to complain that jobs were being “conserved in preference for North American citizens” (Arredondo 2008, p. 98). The consul had reason for concern since Mexicans lagged Europeans in naturalization rates.Footnote 9 Commonly cited reasons for this have included the migratory labor market they filled, geographic proximity and fluidity of borders, racial hostility encountered from native-born and immigrant populations, the fact that party machine politics were less prevalent in the regions where Mexican citizens lived and worked, and the expensive and burdensome requirements that applying for citizenship entailed, especially after 1929 (Rich 1940). On the question of naturalization, several researchers (Arredondo 2008; Reisler 1976b) have also noted the hostile stance of some consuls. Regardless of the precise reasons, the fact that Mexican nationals were not subject to migration control during the 1920s and so infrequently sought naturalization were an important basis for formal discrimination and widespread views, even among “progressive” social workers administering social assistance programs, that Mexicans were an undesirable and inassimilable population (Arredondo 2008, pp. 90–93; Fox 2012).

Social assistance and public works

The denial of social assistance benefits by public and private charities or welfare bureaus throughout the country and their decision to organize the transportation of Mexican citizens, and in many cases their American-born children, has become most associated with repatriation during the depression. Statistical evidence is fragmentary, but Hoffman estimates that between 1930 and 1937 “perhaps over 90,000 Mexican nationals received some kind of assistance” from these sources.Footnote 10 Los Angeles County represented the most organized case, with 13,332 returned to Mexico between 1931 and 1934 with county funds (Hoffman 1978, pp. 233–234). Past research has provided ample documentation of how localities denied benefits and the particular circumstances that led to repatriation pressures. My aim here is to clarify certain aspects of this phenomenon that have been deemphasized or ignored altogether due to the insufficient attention that has been directed to the expansions of state power highlighted in this article. The first concerns the timing, scale, and level of coercion used in the organization of repatriation and how the expansion of federal relief efforts during the New Deal affected it. The second concerns the shifting criteria used to determine eligibility for certain benefits and the role played by the activities of central state authorities in either encouraging or discouraging the extension of citizenship status into this domain. The third concerns the central role of Mexican consuls in cooperating with local authorities and assisting the repatriation of their citizens.

As Hoffman documented, the earliest and heaviest periods of repatriation were, for the most part, “self-propelled” by the most mobile segments of the population who were responding to worsening economic conditions. The intensification of deportation activities no doubt provided additional pressure to repatriate. The first cases of organized repatriation of the more destitute elements of Mexican communities by local public and private welfare authorities did not begin until 1931. From Los Angeles County to Detroit, these early programs were local in organization and were reportedly voluntary. But between 1932 and 1933, there appears to have been a shift not only in the number of individuals willing to participate in the programs as word of conditions in Mexico spread, but also in the level of coercion that was used. It was in this respect that the experience of European aliens, who were also repatriated by relief authorities during this period, differed from Mexican. Fox (2012) found that European repatriate cases were pursued on a “casework” basis, involved little coercion, and did not take on the mass character of the programs affecting Mexican repatriates.

The passage of the Federal Emergency Relief Act (FERA) in 1933, which provided approximately $500,000,000 in funds to state-run relief projects over the next 2 years, changed local circumstances drastically. The federal government explicitly required that citizenship not be used as a criterion for eligibility, and further it forbade the usage of its funds for the purpose of repatriation. However, Fox (2012) has documented that other federally funded relief programs—the Civilian Conservation Corps, Public Works Administration, Civil Works Administration, and Works Projects Administration (WPA)—all used citizenship or the possession of first papers to determine eligibility or preference. In the case of the WPA, legislation was explicitly passed in 1937 and 1939 that began to impose formal restrictions along civic lines where previously none existed, reflecting a growing mood of civic nationalism as economic distress lingered. The federal government therefore used citizenship selectively in determining eligibility for unemployment relief and public works projects (officially rejecting it in the former, largely accepting it on the latter). The effect of citizenship status on determining eligibility for the different programs of the Social Security Act was similarly uneven. Citizenship did not determine eligibility for Social Security Insurance. State-level discretion was allowed, however, for means-tested programs like Old Age Assistance, where citizenship restrictions were the norm in states with a high percentage of foreigners.

Most significantly, the usage of citizenship status to withhold or grant certain protections at the state and local level became increasingly widespread during the interwar period. As already noted, this could be seen in state laws barring noncitizens from professions and participation in public works projects. Traditionally, access to unemployment assistance and poor relief organized through public and private organizations rested on the duration of residency in the community. But residency appears to have been abandoned in many states and localities, as shown by the turn to repatriation programs already mentioned. Indeed, a fascinating glimpse into the contested significance of residency versus citizenship can be found in a portion of a letter of complaint written by the Mexican consul of Chicago to the trustee of the Gary, Indiana welfare authorities in 1932. Although the main purpose of the letter was to admonish the latter for failing to notify the consulate promptly of “fifteen Mexican families” who reputedly desired to return to Mexico through county assistance, he also called “attention to the fact that the Mexican Government is in no way obliged to take care of those people, who regardless of their nationality are members of your community” (emphasis added).Footnote 11 That the consul found it necessary to remind this local authority that nationality should have no bearing on assistance may be seen as one indicator of the extent to which authorities were engaging in just such a practice. And as I show in the next section, there was also some irony in this consul’s stance given that the Mexican state over the previous decade had also asserted its own obligation “to take care of those people.”

Remote protection

Indeed, just as the United States was adopting novel forms of migration control, the Mexican state was extending its reach among migrant populations both at home and abroad. Their activity represented something new during the interwar period, changing relationships not only with their fellow nationals living and working in the United States but also public and private authorities at the state and local level. Compared to other sending states during this period, Mexico was unique. For instance, the Italian government did indeed seek to maintain links with their populations scattered throughout Europe and the Americas, especially under fascist leadership. But in contrast to the Mexican government’s stance explicitly discouraging naturalization and warning that naturalized Mexicans would relinquish any “protection” they could provide, the Italian government did not provide much protection, encouraged their citizens to naturalize in the United States, and generally abstained from direct intervention in favor of religious and organizational proxies (Choate 2007, p. 739; Smith 2006, p. 742). Similarly, the Polish government that emerged from World War I mainly funded educational and cultural activities on such a scale that its impact was minimal (Pula 2004, pp. 48–50).

What explains the Mexican government’s expanded presence? It was in part in response to their frustrated attempts to control the terms and volume of migration as well as the very real mistreatment often experienced by their nationals inside US borders. As the Secretary of Foreign Relations put it in 1917, “since it is not possible for the government to prevent emigration, it must take every measure to reduce the hardships of our fellow citizens while they reside abroad” (Corwin 1978b, p. 187). These “measures” were embodied by the active engagement of the 55 Mexican consuls stationed throughout the United States, whose activities have been researched mainly in terms of labor advocacy and disputes (Balderrama 1982; Cardoso 1980; Gonzalez 1999). Like the consular affairs officers of the United States who made it possible to carry out government restrictions at a distance by remote control, Mexican consuls were charged with establishing ties with and organizing Mexican colonias abroad and providing remote protections as needed.

Need first arose during the United States government’s own mobilization efforts upon its entry into World War I. After Selective Service personnel began to enforce the law that all male citizens and aliens with their first papers between the ages of 21 and 31 must register with local draft boards, Mexican consuls began to provide documentary evidence attesting to an individual’s Mexican nationality, including, in some cases, those who were actual American citizens (Cardoso 1980, p. 50). Such greater engagement with Mexican citizens in the United States was reinforced during the depression of 1920-1921 when the government reportedly assisted 23,354 workers left stranded by unscrupulous employers or destitute by worsening economic prospects (Hall 1992, p. 764). This role was eventually institutionalized during the 1920s with the establishment of a special “protection” division in the Foreign Service within the Ministry of Foreign Relations (Corwin 1978b, p. 189). While the nationalist sentiment of government officials and the very real hardships experienced by their citizens were no doubt significant motivators, the state also had an interest in building bridges with a potentially valuable segment of their citizenry and maintaining good public relations domestically with vocal elements of the urban nationalist middle classes. Often invoked by the latter was the Mexican constitution of 1917, which called for the state to ensure the social rights of all its citizen workers, defined both within and beyond state borders on the basis of descent.

By the 1930s, Mexican consuls had become a familiar and active presence in many cities and localities throughout the country, a fact that proved decisive in the repatriation efforts that took place. Indeed, two scholars have been particularly explicit in acknowledging, if not systematically investigating, this relationship. Gonzalez (1999, p. 31) has argued that “the U.S.-sponsored repatriation campaign would never have succeeded to the extent that it did without the full participation and acknowledgement of the Mexican government,” while Cardoso (1980, p. 148) has noted that “consuls cooperated to the fullest with charity bureaus which carried out repatriation drives.” Individual case studies of Los Angeles County, Detroit, Chicago, Gary, and various cities in Texas have further corroborated consuls’ direct cooperation with local public and private authorities in helping to assist their citizens back to Mexico. To be sure, in some cases consuls were simply reacting to circumstances in which local authorities would have attempted to move forward without their involvement. The reverse was also true in that in some localities, like Detroit, it was the consul who was responsible for the original idea (Fox 2012, p. 184). In emphasizing here the role of consuls and the Mexican state, I do not lay claim to an original discovery, but rather wish to signal an aspect of the repatriation pressures that pushes against their characterization as “racial expulsion programs” and highlights the ambiguous effects of new sending state interventions in the lives of citizens. This was especially relevant for the expanded role of Mexican consuls during the interwar period, who desired to “protect” their citizens abroad but also maintain ties to populations that were understood as a state asset. Such protections were no doubt welcome in some cases, particularly where employers and members of the local community were all too willing to exploit or discriminate against Mexican workers in good times and cut them loose when they were no longer needed or desired. However, the consuls’ nationalistic appeals may have also stunted the rate of naturalization and their central role in encouraging Mexican nationals to repatriate (often to Mexican colonization projects of dubious viability) calls into question whether they truly put the interests of their fellow citizens first.

The remote protections provided by sending state governments took on a different form in France. They were, however, no less significant for noncitizens and resulted in a hardening of citizenship-based boundaries far more comprehensive than in the United States.

France

During the interwar period France became a country of mass labor migration. Between 1921 and 1930, the foreign-born population, coming mainly from Europe and North Africa, had grown from approximately 1.53 million to 3 million, representing a shift from 4 to 7 % of the total population (Mauco 1935, p. 185; Slaby 2005, p. 105). But despite becoming a receiving country on a scale close to the United States, France diverged in several ways with regards to the terms of migration and the expansion of state power. Whereas bilateral labor treaties that the United States concluded with China and Japan were exceptions that proved the rule, in France it was migrant workers from non-treaty countries that proved to be most exceptional during the interwar period. While the United States continued to outlaw contract labor, France made the possession of a formal contract from a French employer one of the primary conditions for entry and legal settlement. Finally, whereas relatively few constraints on geographic and occupational mobility existed for foreign workers in the United States, the labor market in France during the 1920s and 1930s was, at least in theory, a highly regulated regime of contracts, identity cards, work permits, state-administered foreign labor services and departmental job placement centers. The result, as described by one labor scholar of the period, was a segmented labor force “divided between the French citizen and the marginalized alien,” with the former increasingly identifying “with its nationality and the power of the French state to improve its life chances” (Cross 1983, p. 165). When the prosperity of the 1920s turned to economic crisis during the 1930s, the state power that had already been established to control entry to the country and access to jobs was used to compel exit and determine eligibility for social assistance.

Remote protection

France signed labor treaties with Italy as early as 1904 and with Belgium in 1906. But the treaties concluded during and after World War I were more comprehensive in the context of mass labor migration and recruitment (Sauvy and Gazzera 1949). Their proliferation resulted from a conflict between French private firms and trade associations in the agricultural, mining, and metallurgical sectors wanting unfettered access to foreign labor, and sending state governments committed to controlling the terms of labor migration to protect their own interests as well as those of “their” workers. French firms and recruiters, at least prior to the war, were reluctant to involve their own government and attempted to negotiate directly with foreign authorities (Kerel 1991). Only when the governments of labor-supplying countries began to demand more comprehensive protections and inspections on French soil, as Italy did in 1913, did employers and recruiters turn to their own state for “protection” (Cross 1983, p. 27). The outbreak of war and ensuing manpower shortages compelled the French government to act, not only by negotiating and ultimately agreeing to the terms extracted by the Italian (and later Greek and Portuguese) governments, but also by recruiting over 222,000 workers directly from French colonies in North Africa and East Asia. The wartime precedent of state-to-state negotiated labor migration endured during the interwar period, with France concluding treaties with new states (Poland, 1919; Czechoslovakia, 1920; Austria, 1928; Yugoslavia, 1929) as well as old (Italy, 1919; Luxembourg, 1923; Belgium, 1923; Greece, 1929; Romania, 1930; Spain, 1932).

These treaties included similar language guaranteeing equal pay with French nationals, required the possession of a standardized work contract, and extended the same social protections granted to French nationals. For example, the treaty concluded between France and Poland in 1919, which was subsequently amended to include conventions on social assistance in 1920, ensured that Poles would be eligible for protection by “all laws of social insurance against various risks such as illness, disability, and unemployment, currently in force or which may be subsequently established.” More specifically, Article 5 of this treaty stated that “mutual assistance subsidies against unemployment, relief funds for unemployment, and public assistance work provided by public institutions will be awarded in each of the Contracting States to nationals of the other State.”Footnote 12

The expansion of sending-state activities also went beyond negotiating and signing bilateral labor treaties. Within their own borders, sending states began channeling workers into desired sectors of employment or simply inhibiting them from emigrating through bureaucratic procedures and red tape (Martin 2008). The Italian Commissariato General Dell’emigrazione (later renamed under fascist leadership the Directorate of Italians Abroad) was especially active “as a kind of trade union for emigrants” (Cross 1983, p. 115). It demanded higher wages in work contracts and only allowed entry into those trades, such as construction or railroad work, which offered the opportunity to acquire some savings and bring valuable skills back to Italy. In addition to intervening in the emigration process, foreign governments encouraged and funded educational and cultural activities. Poles were noted during this period for having a vibrant associational life, the nationalistic and religious elements of which were criticized from both the French republican left and nationalist right. Indeed, authorities became so concerned with sending state involvement in the education of the children of foreign-born workers that new prohibitions were enacted throughout the country, affecting Poles as far north as Pas-de-Calais (Slaby 2005) and Italians as far south as Nice (Schor 1986).

Despite such expansions of consular support for citizens abroad and the existence of treaty constraints, there were also limits to the efficacy of remote protections enjoyed by noncitizens. Crucially, the entire administrative system that emerged during the interwar period was fundamentally designed to provide a series of legal measures with which to deny rights and jobs to noncitizens, especially during times when they were no longer needed or wanted. When residency or work permits expired or when one intentionally entered the realm of the “irregular” or the undocumented to seek better employment opportunities, the rights guaranteed by remote protection were suspended. And even in cases where French authorities were legally bound to uphold their end of the “social contract,” the lack of systematic oversight made noncompliance probable.Footnote 13 Also, given the bureaucratic and legalistic procedures that such a system brought with it, workers were often uncertain of the rights and the obligations that were owed to them.Footnote 14 Finally, remote protections provided scant actual protection to those engaged in political activities--the so-called violators of French “hospitality.” Expulsion (or the threat of it) was frequently used as a means of social control, just as it came to be used in the United States against labor organizers and members of leftist organizations.Footnote 15 No one expressed this attitude better than the prefect of Pas-de-Calais. In response to an investigation made by the Ministry of Labor in 1934 into whether illegal expulsions had been carried out against certain Poles, the prefect defended himself by writing that “the measure of expulsion … is not because he has been laid off from the mine; it is not at all at the instigation of the company, but because he has abused the hospitality that our country has freely given him and because his pernicious action has been recognized by the police services…. [F]oreign workers stay in France who want to, [but] on the condition, of course, they keep a correct attitude” (Slaby 2005, p. 310). Such attitudes held by French officials are potent reminders that the interwar expansions of state power in international migration control and regulation of the labor market brought new threats as well as new rights, with the threat of expulsion being foremost among them.

Restriction, exclusion, expulsion

New administrative controls and law enforcement measures expanded during the interwar period, further codifying the boundaries drawn between French citizens and noncitizens. Several policies toward foreigners enacted under the extraordinary circumstances of war became permanent during the period of ensuing peace. In addition to the restoration of passport restrictions that had long been neglected prior to the war (Torpey 2000, p. 112), two decrees passed in April 1917 required foreigners above 15 years of age to carry documents bearing their citizenship status, occupation, photograph, and signature. The requirement that all noncitizens establish legal residence proved to be a powerful means for departmental prefects within the Ministry of the Interior to grant, refuse, or revoke residency by confiscating identity papers through the administrative procedure known as refoulement (literally, “pushing back”). This course of action rendered continued presence in the country illegal, but did not entail physical expulsion. The power to expel was also reinforced during the 1920s through the expansion of police powers within the Ministry of the Interior (Rosenberg 2006). That such power existed not only in theory, but also in practice can be seen in the nationwide statistic that, between 1920 and 1932, 93,130 foreigners were expelled out of an immigrant population of 2.7 million (Cross 1983, p. 181). In addition to enmeshing foreign workers in systems of administration and surveillance, new governmental organizations, including the Foreign Labor Service under the jurisdiction of the Ministry of Labor and Departmental Placement Offices under the Ministry of the Interior, were established during the war to help to channel workers to geographic areas and occupational sectors where labor was needed most. These organizations endured during the peace, with the Foreign Labor Service reportedly placing 1,345,093 aliens into jobs between 1921 and 1930 (Cross 1983, p. 151).

By the time the global depression spread to France in 1931, French authorities had already developed an administrative pattern based on previous recessions in 1921, 1924, and 1927. Legal entries were reduced on the frontiers (especially for those seeking work in industrial sectors) while the Ministries of Labor and the Interior established guidelines for noncitizen populations already in the country. The Foreign Labor Service, which had the power to legalize the status of noncitizens in the country without the proper papers, was ordered to suspend all further legalizations. The Ministry of the Interior issued a circular to all prefects on February 7, 1931 requesting information on unemployed foreigners under their jurisdiction and asking that they encourage employers to terminate the contracts of foreign workers (Perry 2004, p. 343). In the following years, such rulings by administrative decree shifted in response to changing economic, social, and political conditions.

Administrative repatriation pressures peaked between 1934 and 1936. This was a time when the country sank deeper into depression, when the obvious initial targets for repatriation (noncitizens of irregular status or minimal ties to the country such as recent arrivals and single men) had already left the country, and when a government of the right came to power.Footnote 16 The latter’s response was to call for a review of all existing identity cards, issue stricter guidelines for their renewal, and restore restrictions on occupational and geographic mobility not seen since the war (Bonnet 1976, pp. 296–313).Footnote 17 Since the beginning of the depression in 1931, the Ministry of the Interior also called upon the prefectures of police to intensify their surveillance of foreigners, particularly those engaged in political activity, and to use their powers to expel or jail those of irregular status. In Paris, the number of prosecutions for those who did not possess a valid identity card increased from 511 for the first 11 months of 1930 to 557 for December alone (Cross 1983, p. 269). Out of a foreign population of approximately 370,000 living in Paris, over 50,000 people received notices of expulsion or refoulement during the 1930s as a whole (Rosenberg 2006, p. 105). Given that the immigration service in this city was reputedly concerned about provoking international incidents with foreign governments and handled certain categories of noncitizens with relative concern for due process, the number probably would have been far greater were it not for such constraints.

While employer hiring and firing practices remained relatively unfettered from government interference, the Law for the Protection of National Labor, passed in August 1932, set a quota of 10 % on the portion of noncitizens that could be engaged in public sector employment. It also established that similar quotas could be applied in private industries if labor or employer organizations petitioned that one be set. Such formal, legalistic means of regulating the workforce were initially resisted by employers and only requested by unions representing small groups of skilled workers in specialty or service trades that employed proportionally few foreign workers (Cross 1983, p. 194). By the mid-1930s, just as the administrative procedures used to encourage repatriation were intensified, the government pressured employers and unions to submit petitions for more quotas. The result was an increase from 72 quotas between 1932 and 1934 to 472 for the years 1934 to 1936. In addition to quota legislation, new laws excluded noncitizens outright from practicing medicine in 1933 and government-paid jobs in 1934 (Noiriel 1996, pp. 65, 220). In such an atmosphere of growing unemployment and new restrictions on employment opportunities, it is hardly surprising that access to social assistance also became a source of contention between citizens and noncitizens.

Social assistance and public works

Early in 1931, the Ministry of Labor issued a circular to departmental prefects reminding them of their obligations with respect to noncitizen workers already within the country. It divided the population into three categories: citizens from countries that had signed bilateral labor treaties (Italy, Poland, Belgium, and Czech Republic), citizens from countries where a treaty had been signed but not yet ratified (Romania, Austria, and Yugoslavia), and “everyone else.” For citizens in the first category, equal access to social assistance or public works was guaranteed as long as their documents were in order. For those in the second, giving aid was recommended but not obligatory, while the fate of “everyone else” was left entirely to the discretion of local authorities (Bonnet 1976, p. 266). For noncitizens who had managed to resist the repatriation pressures described above, the treaties that had been agreed to by central governments and forced upon otherwise reluctant departmental and municipal authorities offered critical protection from the economic distress of the 1930s.

Even before the onset of the depression, local relief authorities and municipal hospitals had already begun to resent the promises that had been made without their consultation and for which they received little compensation from the central government. In the mid-1920s, departmental officials in Pas-de-Calais suggested to the Ministries of Labor and the Interior that the social provisions included in the bilateral treaties be abandoned or, at the very least, that the employers of immigrant labor be required to pay additional taxes to cover the increased costs of providing healthcare and education for noncitizens (Slaby 2005). By the early 1930s, Paris’s municipal council responded by trying to establish a separate hospital for European noncitizens (similar to one established for North Africans). This proposal was eventually abandoned due to the treaty guarantees for equal treatment and the diplomatic complications such an initiative would cause (Rosenberg 2006, p. 185).

Not surprisingly, those to whom assistance was not legally required became the most vulnerable as social expenditures skyrocketed throughout the country. For instance, until 1937 Russians were officially ineligible to receive unemployment assistance in Marseille, and in Lyon separate soup kitchens were set up for Armenians instead of the standard meal vouchers given to citizens and noncitizens from reciprocal treaty countries (Lewis 2007, pp. 8, 72). As another example of the extent to which loopholes in treaty protections were assiduously exploited, Spanish workers in the commune of Drancy (a suburb of Paris that would later become infamous during World War II) continued to be excluded from eligibility on the grounds that the recently concluded reciprocal treaty with Spain had not yet entered into force (Perry 2004, p. 353).

North African workers, who occupied an ambiguous legal category between colonial subject and full citizen and lacked protections beyond those offered by the French state, were especially vulnerable. Rosenberg (2006, p. 11) offers that this was due to the fact that “the most important constraints that protected foreign, mostly European, migrants did not apply to North Africans.” Their “special” status was reinforced by the separate organizations established for surveillance and the administration of aid, particularly the North African Indigenous Affairs Services (SAINAs) in Marseille, Lyon, and other large cities. The Paris Prefecture also created a special North African Brigade, whose officers and spies did not share the fear of provoking diplomatic incidents that other police divisions dealing with foreigners felt. Nor did they have to go through the same procedures for purposes of expulsion. Instead, “the police could simply arrest colonial migrants for vagabondage and put them on a train to Marseille, where they would be forced onto a ferry crossing the Mediterranean” (Rosenberg 2006, p. 164). Similarly, Lewis (2007, p. 211) found in Marseille and Lyon “repatriation drives” reminiscent of those already discussed in the United States, which resulted “from decisions made by local authorities regarding the provision of social welfare within their jurisdictions.” Although systematic statistics on the numbers of North Africans repatriated in this way are not, to my knowledge, available, Lewis does note that by the late 1930s in Lyon, just over 1000 North Africans had been repatriated with the assistance of local SAINA authorities.

As a whole, the number of noncitizens that were “assisted” or coerced into leaving the country during the 1930s is also difficult to quantify. Aggregate numbers of repatriations that occurred during this period are unfortunately of little use in this respect, as they do not differentiate in terms of the means that were employed to repatriate workers and their families. There has, to my knowledge, been no systematic attempt to disaggregate these numbers on a statewide basis beyond scattered references to public authorities and private employers organizing train shipments or handing out en masse free or reduced-price train tickets.

But the main point is that with the conclusion of bilateral treaties and expansions of state power specified above, distinctions based on citizenship status became foundational for interwar labor migration in France. By possessing French citizenship, French nationals were spared the new and intrusive administrative constraints involved in finding (or changing) a job and establishing residency, less exposed to the surveillance of police authorities that regarded immigrants as disproportionately likely to commit crimes, and shielded from the threat of expulsion or refoulement not just for criminal offenses but for participation in radical political activities. Noncitizenship in this sense was therefore a category defined negatively in comparison to French nationals. All noncitizens, regardless of country of origin, family status, or occupation, were required, at least in theory, to carry identity cards, obtain work permits, register with local authorities, and live in an apolitical manner euphemistically referred to in official correspondence as respecting the French state’s “hospitality.” But beyond the French/non-French dichotomy, citizenship was the basis for further distinctions among noncitizens themselves. In one category were citizens from sending countries that had signed the bilateral labor treaties and provided “remote protection” through consular or diplomatic means. In another were citizens who lacked such formal state protection: colonial subjects, “stateless” Armenians and Russian émigrés, and, increasingly during the 1930s, political refugees. In summary, distinctions among noncitizens on the basis of the citizenships they did possess became a means with which to grant as well as withhold rights and protections during the depression.

Conclusion

Following World War I, state expansions across the domains researched in this article elevated the stakes attached to citizenship in both the United States and France. However, the route each country took and the precise ways in which citizenship intruded on the social lives of noncitizens diverged as well as converged in many respects. By way of conclusion, I discuss some of the most noteworthy features in terms of the three expansions of state power investigated in this article.

Both states experienced similar expansions in their capacity to control migration and expel noncitizens during the interwar period. Both also deployed their deportation powers with special vigor, albeit sporadically, in response to worsening economic distress. How they differed was in terms of the criteria for granting entry and the role of surveillance within their own borders. In the United States, politicized and racialized quota restrictions based on national origin became institutionalized while in France the possession of a labor contract became the most widespread means to gain legal entry. After entry, there was little government surveillance in the United States. The number of formal and informal restrictions on occupation was indeed increasing at the state and local level, but this was nothing compared to the formalized regime of police surveillance and administrative controls that came to prevail in France.

With the expanded power to control migration also came the expanded power to protect. In the United States, the federal government unevenly upheld the principle of nondiscrimination on the basis of citizenship status, one that appears to have been abandoned in many state and local governments facing limited budgets and unprecedented demands. Overall, however, the material stakes involved in citizenship were clearly heightened in the context of prolonged economic crisis and expansions in relief and social assistance. Further indicators of this, beyond the formal legal restrictions, are numerous contemporary reports of naturalization offices being swamped with new applications in response to citizenship-based relief programs, the fact that 1.7 million applications for naturalization were filed during the 1930s (Fox 2012, p. 288) and that the percentage of the foreign-born population who were naturalized or with first papers increased from 55 % in 1920 to 75 % by 1940.Footnote 18 Cases of naturalization fraud also suggestively reached unprecedented levels during this decade (Weil 2013, p. 44-52). For its part, the French central state also inhibited localities from withholding relief from noncitizens, a power that it could exercise thanks to the expanded financial support it was providing at the local level and the sole discretion it claimed for negotiating and ratifying international labor treaties. But, as in the United States, the central government’s role was uneven in protecting noncitizens, as shown by legislation passed to protect “national labor” and the first appearance of citizenship-based quotas on private industry. Not surprisingly, naturalization rates in many French localities also experienced noteworthy surges during the 1930s (Noiriel 1986, p. 626).

Where the United States and France differed most was in terms of sending state remote protection. Sending states were mostly a non-factor in the United States thanks to the latter’s unwillingness to participate in international labor treaties. The one exception was the Mexican state, which stood alone in terms of its activism and role in encouraging the repatriation of “their” citizens. In France, sending state remote protection via bilateral labor treaties meant that the citizenship held by foreigners could constrain what municipal and departmental authorities could legally do. Localities did their best to circumvent the accords that they had never agreed to, but for the most part they honored them.

Despite variations in the effects of sending state remote protections, and despite intermittent attempts by central state authorities to ensure the extension of social protections to noncitizens at the local level, repatriation pressures intensified in both countries during this period, following similar chronological trajectories. First, the initial economic downturn led to an immediate halt to entries, and the most mobile or vulnerable segments (i.e., irregular or undocumented workers and single men) began to leave in large numbers on their own accord or as a result of heightened government surveillance. These made up the largest percentage of the repatriated population and represent the most continuity from previous economic downturns, when the logic of labor supply and demand was as consequential, if not more, than state attempts to discourage entry and encourage exit. It was during the second phase—when the contradictions inherent in “temporary” noncitizen workforces were exposed, when it became clear that individual choices had led many workers to stay and start families (to evolve, in other words, from migrants into immigrants), and when the political community was summoned to come to their assistance—that we found contestation and the kinds of new repatriation pressures that have become infamous in certain localities.

In France, those left unprotected by reciprocal treaties thanks to the citizenship they possessed or those who entered the realm of the “undocumented” or “irregular” proved to be most vulnerable. The less protected foreigners from non-treaty countries and subjects from North Africa became special targets of discrimination. Sending-state remote protection was therefore a key variable determining how noncitizens were treated in the receiving state, particularly at the local level. In the United States, populations of Mexican origin proved to be most vulnerable thanks in part to the particular ways state power had expanded in the United States and Mexico. Restriction disrupted migration patterns from southern and eastern Europe and was coupled with aggressive Americanization campaigns prior to the Depression. Meanwhile, migration from (and often back to) Mexico of both the legal and illegal variety surged. Naturalization lagged, and so new citizenship-based discriminations affected Mexican nationals as a group the most. Overall, it is also worth emphasizing that in both countries, those with liminal statuses in relation to citizenship (e.g., short- and long-term residents) confronted new incentives (and in many cases pressures) to naturalize or leave.

Looking beyond the experience of immigrants to the more generic process of state formation and the institutionalization of citizenship, this article has provided a sketch of each country’s step toward achieving greater control of migration, extending greater social protections, and coming to terms with the assertion of sending state remote protections. Far from being an aberration of the unique circumstances of the interwar period, these expansions have become common features not just in France and the United States, but around the world that are ever-present but seldom remarked upon, much like the “unwaved flag” of nationalism that Billig (1995) once observed. Migration control has ensured that approximately 97 % of the members of the world’s population remain caged in their country of birth—a fact that no doubt enables most countries that currently offer social assistance to be even more liberal than France and the United States were during the interwar period in extending benefits to noncitizens and upholding the historic role of local residency. Sending states, for their part, have today grown even more sophisticated in managing the terms of labor migration, further solidifying the stakes involved in citizenship status with equally ambiguous results for workers on the move. The Philippines today perhaps represents the ideal typical model of the modern “labor brokerage state” (Rodriguez 2010). This article has provided some historical and theoretical context for its activities and that of its ancestors.

To return to the observation that began this article, the fact that our own Great Recession has not (so far) produced new discriminations or repatriation pressures against noncitizens should not be read as a diminution of the power of nation states or a decline in the importance of citizenship. To the contrary, it signals the triumph of the caging powers of the former and the naturalization of the rights and exclusions associated with the latter.