|Home | About | Journals | Submit | Contact Us | Français|
Although migration from Mexico to the United States is more than a century old, until recently most other countries in Latin America did not send out significant numbers of migrants to foreign destinations. Over the past thirty years, however, emigration has emerged as an important demographic force throughout the region. This article outlines trends in the volume and composition of the migrant outflows emanating from various countries in Latin America, highlighting their diversity with respect to country of destination; multiplicity of destinations; legal auspices of entry; gender and class composition; racial, ethnic, and national origins; and the mode of insertion into the receiving society. The review underscores the broadening of international migration away from unidirectional flows toward the United States to new streams going to Europe, Canada, Australia, and Japan, as well as to other countries in Latin America itself.
Although it is often forgotten, migratory processes are reversible. Countries of immigration can become countries of emigration, and nations that traditionally have sent out large numbers of migrants can become net receivers. This transformation has, in fact, occurred in many countries of Europe and the Americas. After centuries of outflow, by 1990 all countries in Western Europe had become immigrant-receiving nations, including such historical nations of emigration as Ireland and Spain. Although it is less well appreciated, in the past half century, Latin America made the opposite transition; it ceased being a place of destination for immigrants from Europe, Asia, and the Middle East and became a sender of migrants to locations throughout the world.
At the beginning of the nineteenth century, Latin America participated in the global flow of migrants out of Europe and elsewhere, receiving some 29.5 million persons, or 15 percent of the 191 million people who moved globally during the period. By the 1920s, these migrants had come to represent 5.5 percent of the total Latin American population of 523 million. Later in the twentieth century, many Latin Americans themselves decided to become migrants. Most opted to stay within the hemisphere, and some 25.3 million currently live in the United States; 3.5 million reside elsewhere in the Americas. More recently, however, Latin Americans have begun emigrating to Europe, which now houses around 2 million Latin Americans, as well as to Japan, which contains just under half a million.1
The transformation of Latin America into a region of out-migration has been slow but finally came to involve most countries in the region. Whatever the pace of change, however, each country exhibits its own unique process of international migration. Some migratory processes are markedly unidirectional, with nearly all migrants going to one foreign destination; whereas others are multidirectional, with migrants going to a variety of different destinations internationally. Indeed, heterogeneity is a distinctive trait of Latin American migration. Nonetheless, it is possible to generalize about basic patterns and processes; to distinguish, in particular, certain common phases in the dynamics of out-migration; and to define the characteristic features of each phase and how they vary from country to country.
In the present analysis, we approach the issue of Latin American migration from a historical perspective, offering a brief overview of the immigrant-receiving phase during the postcolonial era and a fuller analysis of the sending phase that began during the 1950s. We then move on to analyze in greater detail three fundamental migratory processes that characterize movement in the region today: north–south migration to the United States and Canada; interregional migration within Latin America; and transoceanic migration to Europe, Japan, and Australia. We end by exploring in greater depth two distinctive migratory patterns that have emerged only recently and are unique to Latin America, phenomena we label relay migration and transgenerational migration.
Latin America carries in its name a certain ambiguity inherent in the definition of all regions. Depending on how one defines “Latin” and “America,” one may include or exclude a variety of countries. For our purposes, we chose a bounded definition of Latin America as those nations that were colonized by either Spain or Portugal and therefore speak Spanish or Portuguese. This definition excludes a variety of smaller Anglophone and Francophone countries in the hemisphere, including those that speak English- or French-based creoles, countries that exhibit very different colonial histories and migratory experiences. Our analysis focuses on three broad regions: Mesoamerica (Mexico and Central America except Belize), the Spanish Caribbean (Puerto Rico, the Dominican Republic, and Cuba), and South America (all countries on the continent except the three Guyanas).
The history of migration in Latin American can be divided into two broad phases. The first phase is, in Braudelian terms, one of long duration that begins in the early 1500s, when the region was first settled by European colonists and African slaves. Indigenous populations soon dwindled after exposure to European microbes for which they had no acquired immunity. Aside from the slave trade, which over the next three hundred years saw more than 12 million Africans arrive to populate certain countries in the New World, the pace of settlement was slow and steady. It was only after the turmoil of the wars of independence that large numbers of immigrants began to arrive in Latin America, first from Europe and then from the Middle East and Far East. From this multiplicity of flows arose a new mixing of races and cultures that is best described by the Spanish term mestizaje.
During the latter half of the nineteenth century and first half of the twentieth, European emigrants went primarily to five countries in Latin America: Argentina (around 4 million), Brazil (2 million), Cuba (600,000), Uruguay (600,000), and Chile (200,000) (Nugent 1996). After 1939, refugees from the Spanish Civil War arrived in Mexico, Chile, and the Dominican Republic (Gardiner 1979). The most recent country to receive European immigrants was Venezuela, which began with its first oil boom in the 1940s. From the mid-1940s to the mid-1970s, more than half a million immigrants arrived in Venezuela from Spain, Italy, and Portugal (Van Roy 1987; Vannini 1983).
Asian migrants to Latin America came mainly from China and Japan. Chinese immigrants have long been a presence throughout Latin America, but they formed notable concentrations in Peru, where they arrived to work in coastal plantations; Panama, where they helped to build the canal; and the Caribbean, where they worked on sugar plantations in Cuba, the Dominican Republic, and Costa Rica (Lausent 2000). Japanese immigrants came during the first half of the twentieth century as European immigration began to slow down, but just two countries attracted the vast majority of Japanese immigrants: Brazil and Peru. Whereas around 190,000 Japanese immigrants went to Brazil, some 20,000 settled in Peru (Lesser 2006; Morimoto 1999).
Flows of migrants from the Middle East were much less numerous, but unlike the streams of Chinese and Japanese migrants, they did not enter as workers but as entrepreneurs and commercial functionaries. Moreover, rather than settling in just one or two countries, they dispersed widely throughout Latin America and had an important influence on both politics and commerce, most notably in places such as Mexico, Argentina, Peru, Chile, Uruguay, Costa Rica, and Colombia (Lesser 2006; Díaz de Kuri and Macluf 1995). Carlos Slim, the richest person in Latin America, and Carlos Menem, the former president of Argentina, are both of Middle Eastern origin, as is the Colombian pop star Shakira and the Mexican actress and producer Salma Hayek.
During the first phase of Latin American migration, the policies of most countries offered an open door, and some even worked actively to promote immigration. A common view at the time was that the region was a “horn of abundance.”2 Celebrated thinkers argued that Latin America needed to promote immigration because it lacked sufficient workers to take advantage of so many riches and to exploit the region’s abundant resources. Although optimism about the future led to widespread agreement about the need for immigrants, there were disagreements about who should be encouraged to come.
Officials of the large companies that controlled the mines, plantations, and massive construction projects were not very particular—they only wanted labor and cared little about national origins as long as people were willing to work. Politicians and government officials, however, often sought not only to import labor but also to “improve the race.” In practical terms, this meant encouraging and facilitating the immigration of white Europeans who would, through generations of intermarriage, gradually bring about a genetic “improvement” of Amerindians and Africans through the process of mestizaje3 (Gardiner 1979; Massey et al. 1998; Masato 2002; Johansson 2006; Telles 2004).
The second phase in the history of Latin American migration begins around 1950 and continues to the present. It corresponds to a change in the direction of the migratory current from in- to out-migration. Although the shift happened at different times in different countries, with the exception of Mexico all the transitions occurred after World War II, and even in Mexico the postwar period saw a revival of flows that had been dead for more than a decade. New conditions in the global political economy were responsible for the shift. In Western Europe and Japan, war recovery triggered a long economic boom to create labor shortages that kept workers at home, whereas Eastern Europe, Russia, and China imposed restrictive controls to prevent the emigration of citizens. In Latin America, meanwhile, U.S. cold war policies supported and installed authoritarian regimes that limited population movement as a means of political control during a period when demographic growth was rapidly accelerating.
After World War II, migratory flows shifted throughout the world and were reconfigured along regional lines. In Europe two key trends emerged: recruitment of workers from selected third-world nations and immigration from colonies and former colonies (Martin and Zucher 2008). Having lost its colonies in World War I, Germany, in particular, embarked on an aggressive guest-worker program that brought in millions of Turkish guest workers. France also undertook guest-worker recruitment but also received large numbers of immigrants from former colonies in Africa and the Caribbean. Britain relied less on guest-worker recruitment and more on immigration from countries and colonies in the British Commonwealth. Countries such as the Netherlands, Belgium, Switzerland, and the Scandinavian nations adopted various mixtures of ex-colonial immigration and guest-worker recruitment depending on their particular histories and circumstances.
In North America, meanwhile, the United States adopted a dual policy of recruiting unskilled workers from adjacent Latin American locations while expanding or contracting refugee admissions in response to the ebbs and flows of the cold war. During the 1940s and 1950s, Puerto Rico and Mexico served as natural reserves of low-wage labor attuned to the needs of the U.S. economy, the former under the aegis of a colonial relationship and the latter under the auspices of neocolonial dependency. During this time, the vast majority of Mexican migrants entered with short-term labor contracts or without documents and were thus officially “disposable,” whereas Puerto Ricans entered as U.S. citizens and were thus more permanent and less vulnerable (Duany 2002; Durand, Massey, and Parrado 1999).
Whereas migrants from Mexico and Puerto Rico were recruited to serve U.S. labor needs, those from Cuba and the Dominican Republic were admitted to serve U.S. geopolitical strategies. After the Cuban Revolution of 1959 and the shift of the Castro regime toward socialism and an alliance with the Soviet Union, the United States opened its doors to Cuban refugees, accepting virtually anyone who managed to make it to the United States for asylum and a fast track to legal residence. Then in 1965, when political disorder overtook the Dominican Republic, the United States intervened militarily, and the embassy began offering legal immigrant visas to leftist students to get them out of the country and restore stability.
As the latter examples suggest, U.S. policy in Latin America after World War II focused mainly on hemispheric security rather than economics. Indeed, from 1945 to 1950 Belgium and Luxembourg alone received more direct aid than all of Latin America (Park 1995, 172). Instead of money, the United States provided arms, training, and materiel to conservative governments throughout the region. In 1947 it signed the Rio Pact, which offered military assistance to nations facing communist insurgencies and left-wing populist movements, and in 1948 it founded the Organization of American States as a means of achieving regional consensus and control in Latin America.
As a result of these machinations, for four decades from the 1950s to the 1990s most of Latin America remained mired in a nightmare of dictatorial governments and military juntas. The nightmare began in 1954 when the CIA helped overthrow the center-left government of Jacobo Arbenz in Guatemala after he sought to expropriate land owned by the United Fruit Company (Poitras 1990). In subsequent years, the United States systematically acted to support dozens of right-wing dictatorships in nations throughout the region, including those in Argentina, Brazil, Bolivia, and Nicaragua. Occasionally the United States helped stage military coups to prevent the left from exercising political power, as in Chile in 1973.
Eventually the Sandinista Revolution of 1979 succeeded in dislodging the U.S.-backed Nicaraguan dictatorship of Anastasio Somoza, which destabilized the neighboring countries of El Salvador, Honduras, and Guatemala, which were then led by right-wing governments but faced their own Soviet-backed insurgencies. During the 1980s, the Reagan administration responded by arming, training, and fielding an army of right-wing Contras to battle the Sandinistas while providing covert aid for counterinsurgency efforts among its neighbors, yielding a decade of violence, bloodshed, and civil war that destroyed local economies and generated large flows of refugees to the United States (Lundquist and Massey 2005).
Following the dictates of cold war policy, U.S. authorities welcomed Nicaraguan refugees and put them on an easy path to permanent residence through the Nicaraguan Adjustment and Central American Relief Act (NACARA) but refused to grant refugee status to most Guatemalans and Salvadorans, granting some temporary protected status (TPS) and pushing others into undocumented migration (Durand, Telles, and Flashman 2006). Throughout the period, Cuba remained the only country where U.S. cold war policy decisively and permanently failed. The Castro regime was influential regionally in launching numerous guerilla campaigns and supporting left-wing populist movements (Poitras 1990).
In all cases, once direct U.S. intervention in the Caribbean and Central America unleashed streams of refugees toward the United States, subsequent flows came to be sustained by the expansion of migrant networks and the ongoing demand for unskilled labor in the United States. In contrast, indirect U.S. interventions in South American nations such as Argentina, Brazil, and Chile did not generally produce streams of refugees northward. Indeed, as already noted, the military juntas supported by the United States tended to forestall mass emigration by imposing barriers to free movement and currying favor with the masses with paternalistic measures and populist economic policies. Those who did leave South America for political reasons were not workers or peasants but middle-class dissidents who preferred asylum in France, Canada, or Sweden to refuge in the United States (Wright and Oñate Zúñiga 2007).
Although recruitment in the case of Mexico and Puerto Rico, along with cold war politics in the rest of Latin America, indeed played key roles in determining migratory flows in the postwar period, economic and demographic factors were also relevant. From the 1950s through the 1970s birthrates in Latin America remained high and population growth was explosive. Whereas in 1950 the total fertility rate (TFR) in the region stood at 5.88 children per woman, by 1955 it had risen to 5.93, and it continued rising until 1960 when it reached a plateau of 5.97 before trending downward. Beginning in the 1970s, however, family planning programs were implemented throughout the region, and fertility rates dropped precipitously to reach a regional average of 2.52 children per woman in 2000 (United Nations 2007). In terms of migration, however, the reduced demographic pressure was not felt until the new century, as during the 1980s and 1990s Latin American baby boomers continued to enter the workforce in large numbers.
Finally we must consider as a permanent backdrop to migration the unfavorable economic situation of Latin America during the latter half of the twentieth century. Although the United States was primarily obsessed with security and cold war politics, the Kennedy administration realized that stability in Latin America required more than arms and military aid and in 1961 launched the Alliance for Progress to provide economic aid in the region. The program was never sufficiently funded, however. It came to be considered a failure and was terminated in 1970 by President Nixon (Park 1995).
Through the early 1960s, import substitution industrialization (ISI) policy delivered steady economic growth and social improvement across Latin America, but toward the end of this decade ISI had begun to reach its limits as a development model. Although a middle class tied to government spending had been created, private industry languished, markets stagnated, inequality increased, and public indebtedness grew.
As growth slowed and inequality increased in the late 1960s, emigration began to emerge as a response in three South American countries—Colombia, Ecuador, and Peru—which came to sponsor important migratory colonies in the United States (Jokisch 2007; Chaney 1980; Cardona 1983; Díaz-Briquets 1983; Altamirano 1992, 1996; Herrera, Carrillo, and Torres 2005). During the 1980s, mounting international debt exploded into a balance of payments crisis in Mexico and Brazil, leading to a prolonged period of economic retrenchment, successive devaluations, falling real incomes, and rising unemployment that came to be known as the “lost decade” (Kliksberg 2001).
The exhaustion of ISI led to the imposition of a new economic model based on neoliberal principles defined by the “Washington Consensus,” which involved privatizing state-owned firms, downsizing the public sector, lowering tariff barriers, relaxing import quotas, ending controls on foreign investment, and generally opening up the national economy to global market forces. In most Latin American nations, the structural adjustment policies of the World Bank, the International Monetary Fund (IMF), and the International Development Bank only added to the general economic misery, with only narrow segments of the population linked to foreign trade and investment clearly benefiting. Only in Chile was the new model successful in placing the country on a path to economic growth, social progress, and democracy. Between 1986 and 2000, Chile’s gross domestic product (GDP) nearly doubled (Sabatini and Wormald 2005).
Nevertheless, by the late 1990s economic recovery had begun in several Latin American nations and the region entered into a period of relative stability. With the end of the cold war, the United States moderated its interventionist approach and became instead a fervent promoter of democracy. After the turn of the century, the sun finally seemed to shine on Latin America as the economy of the region began to grow again, averaging 4.5 percent annually since 2000. Chile, Brazil, and Peru have led the way with high growth rates for at least five years, and although Mexico and Argentina have not yet made such sustained progress, there are clear signs of recovery. However, the new round of economic progress has less to do with U.S. policies than with the rapid growth of demand by China and India for raw materials produced in Latin America.4
Over the past six decades, patterns of Latin American migration have been closely tied to the structure and rhythm of economic growth. Between 1950 and 1980, the largest flows were internal and stemmed from ISI’s concentration of economic activities in large metropolitan areas, which led to high rates of rural-to-urban migration; rapid urbanization; and the emergence of gigantic megalopolises around Mexico City, Buenos Aires, São Paulo, Santiago, Caracas, and Lima (García Canclini 2004). During the lost decade of the 1980s, Latin America’s intertwined political and economic crises generated new pressures for out-migration, and international migration intensified from older sending nations such as Mexico and spread to new nations such as Argentina and Brazil. With the collapse of the Soviet Union, even Cuba was pushed toward a new migration regime. In the spring of 1980 it was forced to relax exit controls and allow the departure of some 120,000 people from the port of Mariel, definitively transforming the outflow from white, middle-class dissidents seeking refuge to dark-skinned, working-class breadwinners seeking opportunity (Pedraza 2007).
Changes in the political economy of Latin America were accompanied by important shifts in immigration policies among key receiving nations. In 1985, Japan amended its restrictive legal code to create flexible terms of entry and residence for the descendants of earlier Japanese emigrants to Latin America, yielding an outflow of several hundred thousand Nikkei workers to Japan (Takenaka 2005; Lesser 2006). In 1986, the U.S. Congress passed the Immigration Reform and Control Act, which simultaneously increased border enforcement and criminalized undocumented hiring, but which also legalized some 2.6 million former undocumented migrants from Latin America. While legalization spurred additional migration through family reunification, the policy of border enforcement backfired by reducing rates of return migration among undocumented migrants and promoting a new wave of permanent settlement (Massey, Durand, and Malone 2002). As a result, the Latino population of the United States grew rapidly and in 2000 topped 35 million to become the nation’s largest minority group (U.S. Census Bureau 2000).
Although the United States has continued to pull in large numbers of immigrants from throughout the region, the turn of the century witnessed a new diversification of destinations. In particular, Europe began to emerge as a new and forceful pole of attraction for Latin American migrants; and within the region, Argentina, Chile, and Costa Rica became important destinations for those from neighboring countries. At the same time, emigration spread to countries that before had not participated in international migration as senders, such as Argentina and Brazil.
Although better data and improved estimation methods have improved the situation, estimates of emigration from Latin America still tend to understate its true size and scale. In general, however, we can distinguish two levels of participation: countries of mass emigration, where at least 10 percent of the national population resides abroad; and countries of migratory consolidation, where citizens abroad make up at least 7 percent but not yet 10 percent of the country’s population. In the first category we include Puerto Rico, El Salvador, Mexico, the Dominican Republic, Cuba, and Ecuador; in the second category we place Colombia, Nicaragua, Paraguay, and Peru.5
Before going on to analyze patterns and processes of Latin American migration, we must define what we mean by these terms. Procedurally, migration involves movement along three dimensions: social, temporal, and spatial. It is social because it is influenced not merely by market conditions and wage levels but by network formation, social capital accumulation, cumulative causation, and various processes of societal transformation that affect sending and receiving areas alike (Massey et al. 1998). It is temporal because it unfolds sequentially in time according to certain developmental stages: departure, where the analytic focus is on causes; arrival, where adaptation and integration are most salient; return, where the motivations of migrants come into play; and long-term consequences, which pertain not only to receiving nations (Alba and Nee 2003; Portes 2007) but to sending societies as well (Massey et al. 1998; Egea et al. 2005; Durand, Telles, and Flashman 2006).
Finally, migration is spatial because it always implies a change of geographic location, which may or may not carry geopolitical implications. Although the spatial referents of migration have traditionally been studied by focusing on places of origin, transit, and destination, more recent studies have sought to take into account what have been variously called migratory circuits (Durand 1986), transnational spaces or fields (Levitt and Schiller 2004), migratory flows (Anguiano and Trejo 2007), binational societies (Guarnizo 1994), and circulatory territories (Tarrius 2000). To undertake an analysis of migration processes, we must take into account this collection of themes, phases, and perspectives.
In terms of patterns, migration involves characteristics or modalities that define and distinguish various processes from one another. In a sociological sense, pattern refers to the type, model, or road taken by the migratory process in each case. A process may display various patterns that develop sequentially over time or express themselves simultaneously, depending on circumstances (Durand 1994; Durand and Massey 2003). The definition of a migration pattern, as with all typologies, is an abstraction that requires simplification and a delimitation of fundamental traits. As Portes (1999) argues, to create and elaborate typologies is the first step in the process of theorization.
At this juncture, we distinguish three basic processes of international migration that have been consolidated and routinely expressed in Latin America: intraregional migration, south–north migration, and transoceanic migration. For each process, we distinguish the particular combination of patterns that render it distinctive relative to the others; at the end of this section we close by describing two new patterns of migration that have only recently emerged within Latin America and are distinctively characteristic of that region, patterns we have labeled “relay” and “transgenerational” migration.
Intraregional migration involves movement within Latin America. Some refer to it as involving migration between neighboring countries (Balán 1988); others classify it as migration between bordering countries (Cerrutti and Maguid 2007), although strictly speaking, not all sending and receiving countries share a border, as with Peru and Argentina. Some define intraregional migration using specific geographic or political denominations, such as the Southern Cone, Mercosur, or the Pacto Andino, though this approach is problematic because member countries change over time (Maguid 2005; Sassone 2004b). For her part, Pellegrino (1989) simply refers to migration in the Americas, which includes flows to the United States and Canada, whereas Martínez (2004, 2005) uses the term “intraregional” but defines it as a pattern and not a process.
Here we opt for a broad definition of intraregional migration that includes all of Latin America, thus going beyond neighboring nations, but excludes migration to the United States or Canada. The process of intraregional migration is characterized by its long duration and wide diffusion as well as its moderate intensity. It began in the first decades of the nineteenth century when national borders were first defined. The early flows were predominantly between neighboring countries; and until the 1970s most movement occurred within Mesoamerica, the Caribbean, or South America, in large measure because of the high cost of transportation and communication but also because of the need for passports and visas. The major exceptions were Argentina and Brazil, which early on opened their doors more broadly to migration from within the region and closed them only partially during the military dictatorships of the 1970s and 1980s.
The shift to open markets and globalization in Latin America during the 1990s was accompanied by a liberalization of transit controls for trade, tourism, and labor, bringing about the emergence of three distinctive patterns of intraregional migration, the first of which is border migration. This pattern of movement is characterized by temporary moves of short distance that are tied to seasonal harvests. Examples include Bolivians migrating to work in the sugar and tobacco industries of northern Argentina (Dandler and Medeiros 1991); Paraguayans going to subtropical estates in northeast Argentina for horticultural work (Balán 1988); Peruvians moving back and forth to harvest bananas and mangos in Ecuador; Nicaraguan peasants and Panamanian Ngobe Indians traveling to the annual coffee harvest in Costa Rica (Alverenga Venutolo 2000; Rosero Bixby, Camacho, and Mok 2002); Guatemalans migrating seasonally to coffee farms in Chiapas, Mexico (Mosquera Aguilar 1990); Colombians working on farms in Venezuela’s Zulia and Andes provinces (Van Roy 1987); Dominicans going to harvest coffee and sugar cane in Puerto Rico (Pascual Morán and Figueroa 2000); and Haitians migrating to cut sugar cane and harvest coffee in the Dominican Republic (Catanese 1999; Grasmuck 1982).
The second pattern of intraregional movement is ethnic migration, which occurs when indigenous people have ancestral lands that straddle a national boundary that was imposed in the postcolonial era. Strictly speaking, this movement is not migration because people move within their own recognized territory; but given the supremacy of national boundaries over ethnic territories, it nonetheless constitutes a form of intraregional migration. In some cases, treaties authorize free circulation, as with certain tribes along the Canada–U.S. border (the Iroquois, Blackfoot, and Sioux), the Mexico–U.S. border (the Kikapoo, Yaqui, and Pima), and the Costa Rica–Panama border (Ngobes) (Reid 2007; Fabila 1945a, 1945b; Durand 1994). In other cases the circulation is de facto rather than de jure, as with Mayans in Mexico and Guatemala; Aymaras in Peru and Bolivia; Guaranis in Paraguay, Argentina, and Brazil; Tobas in the Chaco region of Argentina and Paraguay; Yanomami in the Brazilian and Venezuelan Amazon; and Guajiros in Colombia and Venezuela. Some of these groups are involved in seasonal agriculture, others tend lands that simply span the border, and others engage in cross-border trade and contraband (Gordillo 1996).
Finally, the last kind of intraregional movement is city-directed migration, which comes in two varieties: professional and unskilled. The distinction is necessary because the two types of migrants rarely interact, even when they come from the same country. Migrants with technical or professional training typically locate in capital cities and generally travel as individuals in search of opportunities for education, work, or professional development. In some cases, they move in response to network connections, old family links, or intermarriage; and more and more cases of migration are associated with transnational firms that have plants and offices in multiple countries.
Two countries, Venezuela and Mexico, have been the principal receivers of professional migrants in Latin America. In the Venezuelan case, the oil boom from 1950 to 1980 generated a high demand for professional skills. Data from a 1980 legalization program revealed that 12 percent of Bolivians had a university education, as did 10 percent of Peruvians, 8 percent of Chileans, and 9 percent of Argentines (Van Roy 1987). High salaries and living standards were the primary magnets attracting professional migrants. Latin American immigrants currently constitute 4.4 percent of Venezuela’s population.
On a smaller scale, Mexico, Chile, and Argentina also attract professional migrants by offering comparatively high salaries. The professional category also includes persons leaving for political reasons, mostly leftist dissidents who left during the 1970s and 1980s, notably Chileans, Argentines, Uruguayans, Bolivians, and Central Americans. It is estimated that during the Pinochet dictatorship, some two hundred thousand persons, or around 2 percent of the population, left Chile for exile in Mexico and Venezuela, not to mention Canada, France, Sweden, and various socialist nations (Angell and Carstairs 1987). The Cuban exile, perhaps the most important in the region, is directed principally to the United States. Nevertheless, Cuban migration within Latin America has been growing in recent years, with new streams going to Puerto Rico, Mexico, Brazil, and Venezuela (Duany 2002; Migration Policy Institute 2008a).
Military dictatorships and authoritarian regimes tend to be as fearful of the departure of nationals as the arrival of foreigners. Authoritarian regimes justify restrictive mobility policies with arguments about national security. In the paradigmatic case of the Dominican Republic under Rafael Trujillo, exits from the country were severely limited by the national police, who tightly controlled the issuance of passports. Upon coming to power, a common practice among military regimes in Latin America has been to deport dissidents and lock the doors to outsiders in order to stabilize the political situation. In the case of Chile, an estimated 200,000 natives left after the 1973 coup, and the foreign population fell from 90,441 in 1970 to 84,345 in 1982 (Mármora 1997; Gardiner 1979).
The pattern of city-directed migration also includes workers of both rural and urban background. Unlike border movers, however, these migrants move over longer distances and settle in urban locations for year-round employment in the secondary labor market as domestic servants, elder care workers, cleaning personnel, construction workers, garment operatives, personal service providers, and low-level sales personnel. A distinctive feature of working-class city-directed migration is the emergence of ethnic economies housed in zones that come to acquire a particular national identity and ethnic niches in particular segments of the labor market. Examples include the Peruvian “nanas” who go into domestic service in Santiago, the Bolivian and Paraguayan men who do construction in Buenos Aires, Colombians who work informally in Caracas, Nicaraguans who take the dirty jobs in San Jose, and Dominicans who perform low-end services in San Juan (Duany, Angueira, and Rey 1995; Grasmuck and Pessar 1991; Cardona 1983; Rosero Bixby, Camacho, and Mok 2002; Sassone, Owen, and Hughes 2004).
The case of city-directed migration to Argentina is perhaps most relevant because of its relative age and diversity. Chileans, Paraguayans, Bolivians, and most recently Peruvians have long been an important presence in several Argentine cities, but most importantly in Buenos Aires, where they have taken over entire neighborhoods to form national ethnic enclaves (Vior 2006; Bertone de Daguerre 2003; Vargas 2005; Sassone 2004a). In Caracas, this kind of migration was important during the 1980s, especially among Colombians and Ecuadorians, but it ceased being a pole of attraction at the end of the twentieth century and shifted to become a zone of emigration for political reasons.
During the first decade of the twenty-first century, intraregional migration in Latin America was limited to just a few countries. Perhaps the most extreme case is that of Costa Rica, where the population born in Nicaragua represents around 7 percent of the total population and comprises 70 percent of all foreigners. In Argentina the foreign-born population represents 4.2 percent of the population; and those born in Chile, Bolivia, Paraguay, and Peru together constitute only 2.8 percent of the total population. In Chile, immigration is a very recent development, and foreigners currently represent only a little more than 1 percent of the population, of whom 26 percent come from Argentina, 21 percent from Peru, 6 percent from Bolivia, 5 percent from Ecuador, and 42 percent from other countries (Migration Policy Institute 2008a).
Intraregional migration in Latin America has been notably facilitated by the liberalization of migration requirements as a direct consequence of economic integration under Mercosur; the Andean Community; and free trade agreements negotiated in Central and South America, most recently UNASUR, the Union of South American Nations, launched by Brazil in 2008. South Americans can now travel throughout the region without a visa and in some cases without having to show a passport, just a national identity document. In Central America, the CA4 program permits free movement between Nicaragua, Honduras, El Salvador, and Guatemala. Only three Latin American countries have restrictive policies: Puerto Rico (because it falls under the laws of the United States), Mexico, and Costa Rica. In the case of Mexico, visas are required from most Latin Americans as a way of preventing the entry of migrants seeking to enter the United States. Costa Rica seeks mainly to control undocumented migration from Nicaragua and Panama and generally permits the free entry of other Latin Americans.
The literature generally refers to south-north migration in global economic rather than geographic terms and highlights the asymmetries between developed industrialized societies, which are generally in the north, and poor developing societies located in the south (Zolberg 1989; Portes 2007; Martínez 2005). In Latin America, a key feature of south–north migration is that it unfolds in the context of dependency, domination, disparity, and attraction exercised by one country in particular: the United States. Emigration to the United States from Latin America is the quintessential south–north flow: a generalized migratory process with deep historical roots, a long tradition, and massive scale that distinguishes it from more recent flows directed to Europe, Japan, and other industrialized nations, which we prefer to classify as transoceanic.
Since the early twentieth century, Latin America—but especially Mexico and the Caribbean—has served as a reserve of unskilled labor for the U.S. economy. In these regions, migration to the United States has long been a daily reality. Los Angeles, New York, Chicago, and Miami have consistently functioned as widely recognized referents for Mexicans, Cubans, Puerto Ricans, and Dominicans. Although south–north migration involves Canada as well as the United States, differences between the two are notable. Whereas Latin Americans make up almost half of the foreign-born population of the United States (49 percent), in Canada Latinos represent less than 3 percent of all foreigners. Mexicans and Salvadorans each constitute just 0.8 percent of the foreign-born in Canada, followed by Chileans at 0.4 percent (Migration Policy Institute 2008a, 2008b; Pew Hispanic Center 2008c; García 2006). Whereas Latin Americans constitute the largest and most visible immigrant population in the United States, with a huge social, economic, and cultural weight, in Canada they are a small minority with little influence on any dimension.
Although considerable data are available on Latin American migrants in the United States, the information must be analyzed carefully given the diversity of definitions and classifications. The immigrant population itself constitutes the first generation. It can be divided into documented and undocumented components, and documented migrants can themselves be subdivided into legal permanent residents and naturalized citizens. Another subgroup includes the children of Latin American immigrants who were born in the United States, who constitute the second generation. The grandchildren of immigrants are the third generation, and so on. Collectively these generations compose the Hispanic or Latino population of the United States.
As of 2006, the first generation of Latin American immigrants comprised around 23.4 million persons, with Mexicans in first place with 11.5 million, followed by Puerto Ricans at 3.9 million, Salvadorans at 1 million, Cubans at 932,000, Dominicans at 764,000, and Colombians at 589,000. The case of El Salvador is noteworthy because although migration from that country began only in the 1980s, its immigrant population has now surpassed that of Cuba or the Dominican Republic, each of which began sending migrants in the 1960s. Nonetheless, as Table 1 confirms, the general order of importance is Mexico, followed by the Caribbean, Central America, and finally South America.
The Latin American immigrant population is currently the most rapidly growing in the United States, not only through immigration but also through childbirth. Whereas the general fertility rate (the birth rate of women of childbearing age or GFR) for non-Hispanic women stood at 63 per 1,000 women aged fifteen to forty-four in 2006, the rate for Hispanics was 84, with a figure of 73 for native-born Hispanics and 96 for those of foreign birth. In contrast, non-Hispanic whites had a birthrate of 61, with figures of 69 for African Americans and 63 for Asians (Pew Hispanic Center 2008a). The most prolific of all were women of Mexican origin, who reported a birthrate of 106 per 1,000.
When combined with rapid immigration, high fertility levels have produced vertiginous growth in the Latino population over the past four decades. Between 1960 and 2000, the population of Latin American origin multiplied by a factor of five, going from 6.9 to 35.3 million persons. Apart from this notable increase in absolute numbers, perhaps the most important symbolic change was the overtaking of African Americans by Hispanics as the nation’s largest minority in 2000. Six years later, the distance between the two groups had only widened, with the Hispanic population reaching a total of 44.3 million persons, or 15 percent of the U.S. population, compared with just 12.2 percent for African Americans (Pew Hispanic Center 2008b).
The Latino population of the United States is currently growing as much from natural increase as from immigration, and by 2050 the Latino population is projected to reach 102.5 million persons (U.S. Census Bureau 2004). With respect to undocumented migrants, estimates suggest that Latin Americans constitute the vast majority of a population that currently exceeds 12 million. According to Passel (2005), 81 percent of all undocumented migrants are from this region, with the majority coming from Mexico (57 percent) and the rest (24 percent) coming from other Latin American nations.
The various migratory patterns that produce this large and growing Latino community are quite heterogeneous and diverse; and each country has its own history, unique characteristics, and distinct rhythm of growth. Since the end of the nineteenth century, Mexico has been the dominant national origin group among Latinos in the United States, which is hardly surprising given its special historical and geographical relationship to its northern neighbor. The 2000 Census enumerated some 20.6 million persons of Mexican origin in the country, comprising 59 percent of all Latinos. A notable feature of the Mexican population is its divided legal character, with people of three very different statuses—citizen, legal resident, and undocumented—often within the same family (Massey, Durand, and Malone 2002).
Persons of Caribbean origin occupy second place and in 2000 comprised 15 percent of the total Latino population. Emigration from the Caribbean developed in distinct phases. It began with the migration of Puerto Rican workers after World War II in response to aggressive labor recruitment and was followed by the arrival of Cuban refugees in the 1960s, Dominicans in the 1970s, Central Americans in the 1980s, and South Americans in the 1990s. Each new flow entered under a very different legal regime, with Puerto Ricans arriving as citizens, Cubans as refugees, Dominicans as legal immigrants, and Central Americans as undocumented migrants or asylum seekers (Duany 1995; Grasmuck and Pessar 1991; Lundquist and Massey 2005; Pedraza 2007).
Immigration from Central America began during the 1980s and 1990s and by 2000 accounted for some 4.8 percent of all Latinos in the United States. The most important factor instigating emigration from the region was the civil war in Nicaragua (1976–1979) and its spillover effects in El Salvador (1979–1991) and Guatemala (1980–1996). More recently the devastation wrought by Hurricane Mitch in Honduras (1998) sparked a new outflow of environmental refugees from the region. The contribution of these nations to Central American immigration is quite unequal, however, with El Salvador and Guatemala being most important, followed by Honduras and Nicaragua, and finally at much lower levels by Panama and Costa Rica (Hamilton and Stoltz Chinchilla 2001; Menjívar 2000).
Finally, persons of South American origin make up just 3.8 percent of all Latinos in the United States and generally are the most recent arrivals. Even if the original emigrants from some countries began arriving in the United States as early as the 1950s, the surge of migrants acquired particular momentum toward the end of the twentieth century. It was only during the 1990s that persons of Colombian origin crossed the half-million mark, and Ecuador and Peru tripled their U.S. populations between 1980 and 2000. When emigration from South America first began before 1970, it was relatively easy to obtain a residence visa because there were no numerical limits on immigration from the Western Hemisphere. In 1968 a hemispheric cap was imposed, and in 1976 country quotas were initiated; since then legal immigration has occurred mainly through family reunification provisions (Reimers 1992). Under certain circumstances, where there is a demand for particular skills, Latin Americans have been able to take advantage of temporary labor visas offered under the H2 program to gain access to the United States. Some eighty thousand Mexicans enter the country each year to work in agriculture and services, and around three thousand Peruvians of Amerindian origin enter to work as shepherds in the mountains of the American West (Paerregaard 2005). Compared with Mexicans or Central Americans, South Americans are more likely to enter with tourist visas and then violate their terms by staying too long or taking paid employment (Altamirano 1992, 1996; Cardona and Rubiano de Velásquez 1980).
Table 2 summarizes the social and demographic characteristics of immigrants from different regions. The average age ranges from thirty-three to forty-three years, with Mexicans being the youngest and Caribbean migrants the oldest. The gender composition has changed over the years to reflect a higher share of females, especially among migrants from the Caribbean and South America, where women outnumber men. With respect to education, the Latin American immigrant population is very heterogeneous. Mexicans and Central Americans generally report lower levels compared with Caribbeans and South Americans. Among the Caribbeans, 29 percent had a secondary school education and 11 percent a professional education; among South Americans, the respective figures were 29 and 19 percent. In contrast, just 21 percent of Mexican reported a secondary education and only 3 percent a professional education, compared with 25 and 7 percent, respectively, for Central Americans. These educational disparities suggest that emigration from Mexico and Central America is selective of peasants and unskilled workers, whereas that from South America and the Caribbean is drawn more from the urban middle classes.
In synthesis, south–north migration without a doubt continues to be the most important international population movement in the Americas, not only because of its large volume but because of its age; its diverse effects on sending and receiving societies; and the large quantity of remittances it generates, around $180 billion according to the Interamerican Development Bank (2007). Within the region, northward flows are especially important in Mexico and Central America, whose migratory outflows are directed overwhelmingly to the United States; whereas South America and the Caribbean, with the exception of Puerto Rico, evince a wider range of destinations.
Three cases of south–north migration are extreme in their own ways. Mexican emigration is exceptional for its remarkable scale. With more than 6 million undocumented Mexican migrants living in the United States, 11 million persons of Mexican birth, and 21 million of Mexican origin, the potential effects of Mexican immigration on U.S. society are enormous. At the same time, the effect on Mexico of losing 10.5 percent of its national population and the annual receipt of $24 billion in migrant remittances is equally large, if not larger. Indeed, it is difficult to compare the Mexican case directly with others in the region given its closeness to the United States, the age of the flows, the large numbers involved, and the unique historical relationship between the two countries (Durand and Massey 2003; Massey, Durand, and Malone 2002).
The case of Puerto Rico is perhaps even more exceptional given its status as a U.S. commonwealth in “free association” with the United States, an arrangement that is difficult to understand and explain sociologically except as a historical artifact of colonialism. Its exceptional pattern of migration embraces three fundamental facts: Puerto Ricans have U.S. passports and are citizens by birth; Puerto Rico is poorer than the poorest state in the American union; and more than half (50.5 percent) of all persons of Puerto Rican origin now reside in the continental United States (Duany 2002). Although the last figure is impressive, it must be placed in context. Although Puerto Rican migration may be considered international, in that it emanates in Latin America, it is also domestic because Puerto Rico is a U.S. territory. When considered as a special kind of internal migration, the case is less exceptional, given that many U.S. states have more than 50 percent of their natives living in other states.
El Salvador also constitutes an extreme case given that such a large share of its population (14.5 percent) lives outside the country, mostly in the United States (Pellegrino 2001). As a result, its economy is disproportionately dependent on remittances from the United States (Massey et al. 1998). From our viewpoint, we consider migration massive whenever emigrants exceed 10 percent of the national population. According to this criterion, migration from El Salvador is more massive than from anywhere else in Latin America, except the special case of Puerto Rico. It is also more heavily dependent on remittances than most, with money sent by migrants conservatively estimated at around 4 percent of GDP and around a third of national exports (Massey et al. 1998). In a very real way, migrants are El Salvador’s biggest money-earning export. The total volume of remittances estimated for all of Latin America and the Caribbean was $234 billion in 2006, of which $180 billion came from the United States and $54 billion from the rest of the world (Interamerican Development Bank 2007). The export of labor is obviously a very big business for Latin America generally.
Transoceanic migratory processes include all those directed outside of the Americas. In a global sense, many form part of the broader south–north flow, but unlike movements to the United States, asymmetry between origin and destination is not a critical factor. Instead, other factors come into play, such as colonial ties, migration histories, rights accorded to descendants of earlier emigrants, bilateral accords, and race- or ethnic-conscious migration policies. In the Latin American case, two relatively new destinations stand out: Europe and Japan.
Latin American emigration to Europe arose primarily in the last decade of the twentieth century and continued to develop in the first decade of the twenty-first century. It is not a generalized phenomenon throughout Latin America, however. On the contrary, emigration to Europe derives from a select set of countries: mainly Ecuador, Colombia, the Dominican Republic, Argentina, and Peru and, to a lesser extent, Bolivia, Cuba, and Brazil (Ponce Leiva 2005). In terms of destinations, three countries are most relevant, for both migratory and colonial reasons: Spain, Italy, and Portugal. Table 3 shows the regional origins of legally registered immigrants in selected European nations to reveal the panorama of destinations. As can be seen, Latin American immigrants generally have a limited presence in the European Union. They constitute less than 5 percent of registered immigrants in Germany, France, and Britain. The percentage is higher in Italy (9 percent) and Portugal (15 percent), but the great exception, not surprisingly, is Spain, where Latin Americans make up 35 percent of the total immigrant population.
By omitting unregistered migrants, these figures understate the true size of the immigrant population in each country, and the number and share of undocumented immigrants probably differ from place to place. Nonetheless, Spain, Italy, and Portugal have all carried out regularization programs in recent years to bring immigrants out of the shadows, thus rendering them countable (Padilla and Peixoto 2007). At the same time, however, many Latin Americans have dual citizenship and thus escape official tabulations of foreigners. In Spain, Valls and Martínez (2006) derived an approximate count of undocumented migrants by taking the difference between the number of foreign residence permits issued and the number of foreigners listed in the national population registry. (Even though they may be in the country illegally, immigrants are eligible for a variety of services and benefits if they sign up with the registry and thus have a strong incentive to do so.) Their calculations indicated that taking account of undocumented migrants would add another 50 percent to the regularized population, at least in the year they undertook their study.
Immigrants from all Latin American countries have some representation in Europe, but those from South America clearly stand out. Table 4 shows the regional origins of registered immigrants living in Spain, Portugal, and Italy, the three most important destinations. Across these three countries, 88 percent of all Latin Americans came from South America, 9 percent from the Caribbean, and 2 percent from Mexico or Central America. Thus, migration to Europe is substantially a South American rather than a Central American or Caribbean process.
This statement is certainly true of Spain, where 89 percent of Latin Americans came from a South American nation. According to the estimates of Padilla and Peixoto (2007), the largest contributor was Ecuador (35 percent), followed by Colombia (21 percent), Peru (9 percent), Argentina (8 percent), the Dominican Republic (6 percent), and Bolivia (5 percent). Emigration to Spain stands out for its explosive growth in the 1990s, especially from countries in which a visa was not required for entry, such as Ecuador and Argentina. Nonetheless, immigrants for whom a visa was required often found other routes to Spain. Migrants from Peru typically entered through the Netherlands, where no visa was required, and then made their way to Spain without a passport once inside the EU.
A common argument that has been wielded to explain the rapid growth in emigration from South America is that emigration from this region was “favored” by policies implemented in 1996 to create a process of “ethnic substitution” in which European-origin migrants were favored over those from Africa, either northern or sub-Saharan (Valls and Martínez 2006). Others have argued that successive legalizations have simply encouraged still more migration, in effect issuing a “call” for undocumented Latin Americans to leave for Spain (Padilla and Peixoto 2007). Even if an illegal entrant missed the prior legalization, that person would simply have to wait a few years for the next one to be offered.
Italy is in second place after Spain as a destination for Latin American migrants, with 82 percent coming from South America, 13 percent from the Caribbean, and 6 percent from Central America. In the Caribbean, the Dominican Republic is the leading origin country, comprising 7 percent of all Latin American immigrants, just ahead of Cuba at 6 percent. Among South American nations sending migrants to Italy, five stand out: Peru (24 percent), Ecuador (24 percent), Brazil (13 percent), Colombia (8 percent), and Argentina (7 percent).
Argentina comes in only fifth as a source of immigrants to Italy, but the size of the flow is probably understated by these figures. During the first decades of the twentieth century, emigration from Italy to Argentina was massive. A large fraction of Argentines can therefore claim dual nationality through an Italian grandparent and enter with an Italian passport, thereby escaping registration as foreigners. In addition, many Argentine emigrants of Italian origin use their Italian passports to enter Italy but then move immediately to Spain, given their greater facility with Spanish. Thus according to the latest data, 86,921 Argentines live in Spain compared with only 14,360 in Italy (Padilla and Peixoto 2007; Bonifazi and Ferruzza 2006). Italy generally lags behind Spain in the preferences of Latin Americans, though it comes before France and Britain.
Portugal comes in third place as a destination for Latin American emigrants, and nearly all (98 percent) are from South America, overwhelmingly from just two countries: Brazil (88 percent) and Venezuela (6 percent). As Table 4 indicates, trivial numbers of migrants enter Portugal from Central America and the Caribbean. The dominance of Brazilians obviously reflects historical ties of language, culture, and colonization (Padilla and Peixoto 2007). Venezuela figures as a significant source because thousands of Portuguese immigrated there as labor migrants during the years of the oil boom (Van Roy 1987).
The gender distribution of Latin American emigrants to Europe is generally even, except for the Dominican Republic and Brazil, which have unusually high proportions of women (69 and 70 percent, respectively). Insertion into the labor market across all destinations generally follows the traditional pattern of men going into construction and agriculture and women into domestic service and hotel work (Valls and Martínez 2006).
Apart from Europe, Brazil also sends significant numbers of migrants to Japan, as does Peru and to a lesser extent Bolivia. These flows are composed primarily of second- and third-generation descendants of Japanese emigrants who entered these countries as labor migrants early in the twentieth century. During the final decades of the twentieth century, Japan found itself with a high standard of living, a low birthrate, and a rapidly aging population that yielded a shortage of workers for unskilled and semiskilled jobs and, hence, a demand for immigrants. Given the reluctance of the Japanese to import foreigners, officials turned to the Japanese diaspora to find workers, arranging special visas and easy terms of entry for Latin Americans of Japanese descent. This policy ultimately led to the emigration of several hundred thousand workers and their families from Brazil and Peru. This phenomenon will be considered in more detail in the next section under the rubric of transgenerational migration.
Upon shifting from immigration to emigration, Latin America developed a new set of migratory patterns with distinct traits and characteristics that require more detailed elaboration, for in many ways they are unique to the Latin American context.
Relay migration has developed over the course of the past several decades and stems from the disequilibrium that is created in labor markets whenever there is a significant departure of workers. It reflects the articulation of various migratory processes with one another in response to the needs of global capital and the shocks to local markets that are generated when workers are recruited to sites of capital investment, both national and international.6 This movement might also be called “chain migration,” highlighting the concatenation of labor markets in time and space, or “step migration,” underscoring the different salary levels in labor markets as one moves from poorest to wealthiest nations.
In Latin America, the best example of relay migration is the movement of workers among Haiti, the Dominican Republic, Puerto Rico, and the United States. In this case, the process began at the upper part of the labor ladder, in the United States, and resulted from deliberate labor recruitment. After World War II, labor markets in the northeastern United States required new workers to replace aging European immigrants who had arrived earlier in the century. Europe was beginning its own postwar economic boom and was no longer sending out migrants. In response, beginning in 1948 employers established recruitment programs to attract workers from Puerto Rico, going so far as to subsidize air travel between that island and New York City (Friedlander 1965). Once begun through recruitment, however, the outflow became self-sustaining in the 1950s owing to the operation of migrant networks, which led to large-scale displacements from the Puerto Rican countryside (Fleisher 1963).
The departure of workers to the United States from rural areas created job openings that were then filled by immigrants from the Dominican Republic, a short boat ride away, who arrived in significant numbers to cut sugar cane and harvest coffee. In this case there was no active labor recruitment because news about the availability of seasonal jobs filtered back to the neighboring island, where salaries were lower, initiating voluntary undocumented migration. In turn, rural areas of the Dominican Republic began to experience a shortage of workers owing to the departure of workers for Puerto Rico as well as the capital city and, after 1965, the United States. The shortage of workers in the coffee and sugar industries was then met by recruiting workers from Haiti, the poorest country in the Western Hemisphere, whose inhabitants viewed seasonal labor in the Dominican Republic as an economic opportunity.
Other authors have also commented on the same migratory pattern. Grasmuck (1982) and Martínez (1995) both examined the pattern from the viewpoint of dependency theory as a movement of peripheral workers and hence failed to see the process as an integrated whole. Instead, they focused on emigration from Haiti and the Dominican Republic, on one hand, to the United States, on the other, and left aside the critical link of Puerto Rico. Grasmuck, in particular, focused primarily on migration between periphery and core, leading her to identify different gradations of exploitation, or “heat,” that different groups could tolerate in the labor inferno. Haitians were willing to accept the worst jobs in the Dominican Republic, whereas Dominicans were willing to take the worst jobs in the United States. Subsequent ethnographic work has shown that Dominicans going to the United States are not those who worked in the cane or coffee fields but were urban in origin (Georges 1990; Grasmuck and Pessar 1991). Recent ethnographic work reveals that it was Dominicans, and not Haitians, who went to Puerto Rico to work in the sugar and coffee industries (Duany, Angueira, and Rey 1995).
In the end, dependency theory offers a rather narrow and somewhat mechanical way of analyzing labor markets. Grasmuck thus saw the movements as separate, overlooking the replacement features of the ongoing relay, stating explicitly that “the argument here is not that Haitian migrants take the place of Dominican emigrants” (1982, 374). In reality the various labor markets are interconnected and operate more dynamically, like interconnected cups in which water leaking from one leads to inflows from others. Dominicans went to Puerto Rico because wages were higher than in the Dominican Republic, not because they would have to work at home in jobs that were appropriate for Haitians.
As Piore (1979) has noted, in places of destination, jobs held by immigrants invariably lose prestige and become unattractive to natives. Thus, Dominicans are cane workers in Puerto Rico but not in the Dominican Republic, where the work is done by Haitians.7 At the same time, Puerto Ricans who no longer were willing to cut cane at home were nonetheless willing to do so in Florida and Hawaii (Natal 2001). The migration relay extends even further, because it was African Americans’ no longer wishing to cut cane in Florida that caused the demand for Caribbean workers in the first place.
Relay migration does not imply an automatic or mechanical process, but a series of discrete labor market adjustments that are made slowly over time and take advantage of connections that exist between neighboring countries among people doing the same kind of work. The chain or relay extends from the top to the bottom rung of the development ladder. In this example, the top is the United States and the bottom is Haiti. As the poorest country in the region, Haiti receives no immigrants. The sugar industry, however, is historically complex, and the direction of flows has shifted in the past. There have been periods when Puerto Ricans migrated to cut cane in the Dominican Republic and Haitians to cut cane in Cuba (Álvarez 1988; Duany 2002; Martínez 1995). Nevertheless, at the end of the twentieth century, the flow of workers had stabilized as a relay system linking Haiti, the Dominican Republic, Puerto Rico, and the United States in a progressive chain of jobs.
Relay migration always involves an asymmetry between two social and economic perspectives: that of the migrants looking up to the next higher step that offers better wages and that of the employers looking down in search of “others” who are willing to undertake the work that they themselves, and others like them, no longer wish to do. Whereas the rungs of the labor ladder ascend for the former, they descend for the latter, indicating the fundamental asymmetry that drives the process of relay migration. For example, although Puerto Ricans perceive themselves to be white (Loveman and Muñiz 2007), or at least nonblack (Duany 2002), and have all the privileges of U.S. citizenship, they are nonetheless routinely stigmatized by white European Americans as “black” and stereotypically considered lazy, ignorant, criminal, poor, sexually obsessed, and culturally unassimilable (Duany 2004).
For their part, of course, Puerto Ricans look down on Dominicans and compare them to groups toward which they display great intolerance—homosexuals, ex-convicts, and the homeless—reflecting the fact that many are undocumented, a majority are women, most are darker-skinned, and they tend to be of rural origin, leading to stereotyping of them as dangerous, strange, dirty, ignorant, and violent (Duany 2004). Of course, these are the same traits and stereotypes that the Dominicans, in turn, assign to Haitians. As Grasmuck (1982) notes, the work of cutting cane in the Dominican Republic is not only considered to be poorly paid, but is work fit only for Haitians, blacks, and slaves. Even if Dominicans themselves often have African phenotypes, they classify themselves racially as “Indians” of different types and complexions, though in reality they also realize they are “black behind the ears,” as one popular saying highlighted by Candelario (2007) has it. For the Dominican government, the “Haitian question” is decidedly racial and marked historically by murder, persecution, and expulsion, including mass deportations in 1937 and 1991 (Gardiner 1979; Martínez 1995).
Once imported from a lower rung of the labor ladder, relay migrants typically encounter racial and social stigmatization, and the jobs they hold become stigmatized and exclusively identified with the racialized out-group. Although economic activities change and societies evolve over time, the migratory flows and stigmatizations persist. In Puerto Rico, the cutting of sugar cane at this point is practically all mechanized and no longer requires much labor. Nonetheless, a strong demand for Dominican labor continues in industries such as coffee, construction, and services, and these jobs are now reserved for Dominicans.
Disequilibria in labor markets owing to emigration have generated the conditions for relay migration elsewhere in Latin America as well. In Mexico, a relay circuit has been established that brings migrants from Guatemala to take the place of migrants from Chiapas who have gone to Jalisco because people there have left for the United States. In El Salvador, out-migration has been so great that the country now imports Hondurans and Nicaraguans to work in the sugar cane fields. In the Argentine region of Patagonia, meanwhile, older European immigrants who once worked on farms and ranches have abandoned them for the city and are being replaced by Bolivians (Sassone, Owen, and Hughes 2004).
Transgenerational migration is another migratory pattern born in Latin America. It refers to the migration of the children, grandchildren, and great-grandchildren of emigrants who came to Latin America from Europe or Japan during the nineteenth and first half of the twentieth centuries. It is one of the fundamental characteristics of the new phase of migration now unfolding in Latin America and has been detected by various authors (see Pellegrino 2001; Martínez Pizarro 2004, 2005; Takenaka 2005; Takeyuki 1999; Masato 2002; Valls and Martínez 2006). The key to understanding transgenerational migration is that it is not a form of return migration, as Takeyuki (1999) has labeled Brazilian migration to Japan and as Martínez Pizarro (2004) has labeled Latin American emigration to Europe, calling it “return migration generationally deferred.” First-generation immigrants may return, but second- or third-generation immigrants necessarily emigrate.
Some authors emphasize the ethnic character of transgenerational migration, referring to policies that privilege ancestral ties as a means of selectively recruiting certain ethnic or racial groups and excluding others (Takenaka 2005; Valls and Martínez 2006). Despite this privileging of blood ties, however, recent ethnographic work suggests that second- and third-generation descendants of Japanese and European immigrants to Latin America often carry very different cultures, such that adaptation and integration within ancestral lands is often more complex and contradictory than expected, either by the receiving society or by the immigrants themselves (Takenaka 2005; Takeyuki 1999).
Various demographic and legal factors interact to produce transgenerational migration: aging populations, low fertility rates, and slow labor force growth, combined with new framings of citizenship and nationality; the implementation of legalization and amnesty; the allocation of visas on the basis of ancestry; and selective systems of recruitment that pay attention to ties of ethnicity, language, culture, and religion.8 Transgenerational migration occurs through the manipulation of identities, nationalities, genealogies, and surnames by migrants and authorities in receiving nations alike. Migrants draw on ties of common descent to achieve legal entry, exchanging what might be called “ethnic capital” for access to high-wage labor markets. Nations seek to avoid ethnic conflict by selecting immigrants who are viewed as ethnically “similar” to the native population.
In some ways, transgenerational migration stems from the failure of prior waves of immigrants to integrate, which led some policy-makers to conclude that common descent might facilitate the process of immigrant assimilation, or at least forestall rejection. For this reason, several countries have turned to ties of blood and ancestry to select, regularize, and recruit migrants from abroad. Obviously such procedures have racial implications, even though they are typically cast in positive terms of “affinity” and “compatibility” rather than racial undesirability, as might have been done in earlier eras.
Nonetheless, the decision of certain receiving nations to privilege common descent in allocating entry visas opened a new window of migratory opportunity for many Latin Americans, pushing them to reevaluate and appreciate individual and group characteristics that they had previously disregarded. In essence, immigration policies linked to blood and descent turn relationships to distant relatives and long-departed ancestors into a source of migration capital. A great-aunt living in some godforsaken town in southern Italy suddenly becomes a valuable resource enabling an Argentine to acquire an Italian visa and claim a nationality. A long-forgotten relative hardly anyone remembers can be contacted by Internet or telephone and become a valuable resource. The old passport of a deceased grandparent can provide the documentary proof needed to qualify for citizenship or a visa.
Old immigrant traditions of “not looking back” and “burning bridges” are no longer operative for second- and third-generation descendants of Spanish, Italian, and Japanese immigrants living in Latin America. Owing to the privileging of ancestral ties by certain receiving countries, they discover that they do, in fact, have a strong interest in looking back and maintaining bridges, something that is now quite possible owing to developments in transportation and communication, which have extended opportunities to travel, establish relationships, obtain information, and carry out administrative tasks over long distances. As a result, contact with family members—whether near, latent, or far—in the lands of one’s ancestors offers new pathways to emigration.
Transgenerational migration has always existed, though only in certain isolated cases. What is new is the scale of the phenomenon today and the intentional manipulation of linguistic, phenotypic, genealogic, and ethnic markers by aspiring migrants and governments. Although its dimensions are difficult to estimate precisely, trans-generational migration does appear in recent data gathered from both regularizing and naturalizing migrants. This new context transforms ethnicity into a fundamental source of capital accessible to potential migrants, along with whatever human and social capital they might possess. In the present global economy, these elements of ethnic capital may prove more valuable than human capital traits such as skills or education, at least for the purpose of gaining entry.
A key element in the emergence of transgenerational migration is the nature of the legal regime that defines citizenship and nationality. Traditionally, nationality has been defined according to one of two principles—jus sanguinis (citizenship by blood) and jus soli (citizenship by birth on national soil). In the current era, both principles have been subject to their own particular casuistry under the laws of various nations, but birthright citizenship is generally less manipulable than bloodline citizenship. Birthright citizenship is inclusive and makes no distinctions by ethnicity, race, or culture. It is a clear and transparent right that does not permit interpretations. In contrast, citizenship by blood is a social construction and subject to interpretation and ambiguity. It is exclusive rather than inclusive and privileges purity of blood in ways that cannot be anything but racial.
In theory, of course, one’s parents confer the right to citizenship through direct inheritance. Although this principle is simple and direct, ambiguity accumulates as one moves from parents to grandparents and great-grandparents, creating possibilities for manipulation in determining what people in which generations carry citizenship rights. In Germany, for example, blood rights are preserved up to the third generation. In Spain, discussion now centers on whether grandchildren of earlier Spanish emigrants remain eligible for Spanish citizenship. In Japan, claims on citizenship are recognized through the fifth generation, with each generation having its own specific term in Japanese, though under current legislation each succeeding generation has fewer rights than its predecessor. According to one official interviewed by Takenaka (2005), “Japanese blood thins out over the generations.”
Moreover, even though phenotype may be preserved if strict endogamy is practiced, culture is not so easily preserved, and in fact, the transmission of cultural knowledge drops off rapidly beyond the first generation. In Peru, the first generation of Japanese immigrants organized sumo wrestling tourneys and endeavored to teach their children and grandchildren Japanese customs. Nonetheless, over the decades the language was lost, cultural knowledge dissipated, and sumo wrestling matches are no longer held in Lima or anywhere else in Peru.9
Whereas birthright citizenship is conducive to the formation of truly multiethnic, multicultural, and multiracial societies, bloodline citizenship lends itself more to racially and ethnically homogeneous societies that privilege certain origins over others. France and Germany are paradigmatic of the two extremes in Europe. France offers citizenship on broad and relatively generous terms, recognizing as nationals the children of immigrants who are legally present. In contrast, Germany is quite restrictive in extending citizenship, going back to a 1935 law that defined two types of citizenships: those pertaining to the German state and those pertaining to the German Empire, or Reich. The second is granted by right of blood and is certified by a document proving racial purity, in essence confirming Aryan origin, something that eluded children of mixed marriages, though this was subject to interpretation. In the range of possibilities, the children of intermarried Jews were always excluded, whereas children of mixed Nordic origins were often accepted (Garner 1936).
Despite the two different systems, the resulting patterns of integration or, more precisely, nonintegration, are not that different in practice. Although French citizens, second- and third-generation immigrants from North and sub-Saharan Africa continue to experience discrimination and have not been able to integrate, making them feel resentful and excluded (Brubaker 1992; Mestries 2007). In Germany, meanwhile, second- and third-generation descendants of Turkish guest workers who made the postwar economic boom possible are still not considered German, though they may have been born on German soil and speak German fluently, and they too experience discrimination and exclusion.
Clearly no model of citizenship or nationality can guarantee assimilation. Nonetheless, in countries that have experienced mass migration, discussions about nationality, culture, race, and ethnicity are probably inevitable, though the ultimate issue is always integration. For some groups, integration may be slow but occurs nonetheless, whereas for others assimilation may be segmented along different lines ranging from complete integration to no integration at all (Portes, Fernández Kelly, and Haller 2006; Alba and Nee 2003). Ethnographic evidence among trans-generational migrants in Japan and Europe reveals many problems and conflicts (Takenaka 2005; Takeyuki 1999), suggesting that segmented assimilation may occur even when immigrants share a nationality, surname, and phenotype.
In Latin America we can identify key cases of transgenerational migration implicated in the turn of the migratory tide back toward Europe and Japan. In Europe, the most important cases are Spain, Italy, and Portugal, three countries that sent out waves of immigrants to Latin America during colonization, independence, and industrialization. The other relevant case is Japan, which sent out migrants to Peru and Brazil in the first half of the twentieth century (Takenaka 2005; Pellegrino 2001; Lesser 2006). The phenomenon of transgenerational migration is not limited to Latin America, of course, and has begun to be observed in other locations, such as Ireland, England, Greece, and the special case of Israel.
With the exception of Mexico, emigration from most nations in Latin America is a relatively new phenomenon that has not yet reached maturity. In the future the various processes we have identified will likely intensify to incorporate new origins and new destinations. Some processes of emigration are old and very well established, such as those in Mexico and Puerto Rico. Others have erupted with unusual force and assumed a mass scale only in recent years, as with those in Ecuador and El Salvador. Some migratory processes are markedly unidirectional, such as those originating in Mexico and Central America, which go overwhelmingly to one single destination. Others are multidirectional with a diversity of destinations, as in the Dominican Republic and Ecuador, which send migrants to Europe as well as to the United States, and especially Peru, which sends migrants to some two dozen different countries.
The legal auspices of emigration from Latin America are diverse. Many migrate through irregular channels to work illegally in countries of destination. Others are recruited with legal but temporary work visas. Others obtain residence visas through legal provisions that authorize family reunification or skilled migration, and still others achieve legal entry by drawing on transgenerational ties to ancestors who departed as international migrants in earlier decades. Once they arrive in countries of destination, processes of adaptation and assimilation are strongly conditioned by legal status and other structural factors in society, and outcomes are often segmented.
With the exception of Chile and Costa Rica, most countries in Latin America have alternated between periods of boom and bust and experienced a prolonged economic crisis during the 1980s. Poverty is a burden throughout the region, and a lack of economic opportunity influences the outflow of migrants from all classes. In some cases, such as Mexico, Central America, and the Caribbean, those who emigrate are drawn principally from the lower sectors of society: workers or peasants. In other cases, as in many South American nations, migrants are drawn mainly from the middle and professional classes. In general, Latin American migration incorporates both men and women, although in certain cases, such as Peru, Brazil, and the Dominican Republic, female migrants dominate.
If migration is indeed a phenomenon that has increasingly spread throughout the region, there are nonetheless a few nations where it has not gathered much force. Venezuela, which until the 1980s received large numbers of immigrants from abroad, has recently begun to send out migrants, primarily from the middle and upper classes and mainly for political reasons. Meanwhile, Brazil, with 200 million inhabitants Latin America’s largest country, has just recently begun to send emigrants to the United States, Portugal, and Japan.
Although Latin American migration to the United States continues to be most important in numerical terms and is still the most dynamic, given the mounting restrictions on migration to the United States we may in the future see a moderation of south–north migration and an expansion of transoceanic migration. The latter is especially likely given the weakening of the U.S. dollar; economic decline in the United States; and the rise of transgenerational migration, a new modality made possible by the preferential treatment accorded to descendants of earlier waves of European immigrants to Latin America. Finally, the relaxing of transit regulations within South and Central America in concert with free trade agreements and common market accords can be expected to increase intraregional migration in the medium term as regional economies expand and income growth proceeds.
Jorge Durand is a professor and senior investigator at the Universidad de Guadalajara and an adjunct professor at the Universidad Iberoamericana in Mexico City. He has been elected as a foreign associate of the National Academy of Sciences, the American Academy of Arts and Sciences, and the American Philosophical Society. His most recent books are Detras de la Trama: Politicas Migratorias Entre Mexico y Estados Unidos (with Douglas Massey and Nolan Malone; Porrúa 2009) and Mexicanos en Chicago: Diario de campo de Robert Redfield 1924–1925 (with Patricia Arias; Porrúa 2008).
Douglas S. Massey is the Henry G. Bryant Professor of Sociology and Public Affairs at Princeton University and the president of the American Academy of Political and Social Science. His research focuses on international migration, race and housing, discrimination, education, urban poverty, and Latin America, especially Mexico. His most recent books are New Faces in New Places: The Changing Geography of American Immigration (Russell Sage Foundation 2008) and Categorically Unequal: The American Stratification System (Russell Sage Foundation 2007).
2The “horn of abundance” appears in various national emblems, including those of Peru, Venezuela, Honduras, Costa Rica, Panama, and Colombia.
3In Mexico and Peru the concern was an Indian genetic heritage, whereas in the Caribbean and Brazil the worry was over African genes.
4According to the newspaper El País (May 6, 2008), while the economies of Chile, Peru, Argentina, and Brazil, linked to European and Asian markets, grew by annual rates in the vicinity of 7 percent between 2000 and 2007, Mexico, being narrowly linked to the U.S. economy, grew by an annual rate of just 2 percent.
5These calculations are based on data from the United Nations, the U.S. Census of Population, and census enumerations of migrants in various European nations.
6The term “relay migration” (migración por relevos) was used earlier by Arizpe (1980) to define and explain the migration relay between fathers and sons as a survival strategy to diversify peasant household economies. It refers in this case to different phases of the family life cycle, when fathers who have worked as migrants are replaced by their sons.
7Dominicans who work in the sugar industry at home do not cut cane but have other jobs such as foremen, craftsmen, machine operators, and administrators (Grasmuck 1982).
8In Spain, for example, Law 36/2002, which came into force in 2003, facilitates access to dual nationality among descendants of Spanish immigrants to Argentina, as a special bilateral arrangement rather than a law that applies to all Latin Americans.
9Interview with Peruvian Nikkei, professor of medicine at the Universidad Peruana Cayetano Heredia, Lima, Peru, 2002.