Racial Formation in Theory and Practice: The Case of Mexicans in the United States

Abstract

Mechanisms of social stratification require the categorical definition of an out-group to that can be excluded and exploited. Historically, in the United States, African Americans have been the subject of a systematic process of racial formation to define socially in this fashion. Beginning in the 1970s, however, and accelerating in the 1980s and 1990s, Mexicans were increasingly subject to processes of racialization that have rendered them more exploitable and excludable than ever before. Over the past decade, Mexican Americans moved steadily away from their middle position in the socioeconomic hierarchy and gravitated toward the bottom. This paper describes the basic mechanisms of stratification in the United States and how Mexicans have steadily been racialized to label them socially as a dehumanized and vulnerable out-group.

Introduction

Social stratification occurs because human societies are characterized by social structures that divides people into categories based on a combination of achieved and ascribed traits. Achieved characteristics are those acquired in the course of living, whereas ascribed characteristics are set at birth. Stratification involves the unequal distribution of people across social categories characterized by differential access to scarce resources. The resources may be material, such as income and wealth; they may be symbolic, such as prestige and social standing; or they may be emotional, such as love, affection, and sex. Stratification systems order people vertically in a social structure characterized by a distinct top and bottom. Although the number and range of social categories have risen dramatically in recent centuries, the basic means by which people are granted more or less access to scarce material, emotional, and symbolic resources have remained remarkably similar through the ages.

Given socially defined categories and people distributed among them, inequality is generated and perpetuated by two basic mechanisms: exploitation and exclusion (Tilly 1998). Exploitation occurs when people in one social group expropriate a resource produced by members of another social group and prevent them from realizing the full value of their effort in producing it. Exclusion occurs when one social group restricts access to a scarce resource, either through outright denial or by exercising monopoly control that requires out-group members to pay rent in return for access. Either way, stratification is enabled through a socially defined process of exclusion.

Both mechanisms are social in origin and follow from the pursuit of core social motives common to all human beings (Fiske 2003). In a very real way, stratification begins psychologically with the creation of cognitive boundaries that allocate people to different social categories. Before categorical inequality can be implemented socially, categories must be created cognitively to classify people based on some combination of achieved and ascribed characteristics. This categorization of people follows its own logic.

Racial Stratification in Theory

The roots of social stratification ultimately lie in the construction of boundaries to enable social distinctions, a task that comes naturally to human beings, who are mentally predisposed to engage in categorical thought (Fiske 2003). We construct general categories about the world in which we live and use them to classify and evaluate the people and things we encounter. These conceptual categories collectively are known as schemas. They represent cognitive structures that serve to interconnect a set of stimuli, their various attributes, and the relationships between them (Fiske 2003).

People use schemas to evaluate themselves and the social roles, social groups, social events, and social actors they encounter, in a process known as social cognition (Fiske 2003). The categories into which they divide up the social world may change over time and evolve with experience, but among mature human beings they always exist and people always fall back on them when they interpret objects, events, people, and situations (Fiske 2003). Human beings are programmed psychologically to categorize the people they encounter and to use these categorizations to make social judgments.

Social schemas do not exist simply as neutral mental representations, however. They are typically associated with emotional valences. The human brain is composed of two parallel processors that, while interconnected, function independently (Panksepp 1998; Konner 2002). The emotional brain is rooted in a set of neural structures that are common to all mammals and are known collectively as the limbic system, whereas the rational brain is centered in the noecortex, especially the prefrontal cortex, which is most developed in humans (Damasio 1999). The two portions of the brain are neurally interconnected, but the number and speed of connections running from the limbic system to the neocortex is greater than the reverse, so that emotional memories stored in the limbic system, which are typically unconscious, greatly affect how human beings make use of categories that exist within the rational, conscious brain (LeDoux 1996; Zajonc 1998).

Emotions stored in the limbic system may be positive or negative but when they are associated with particular classes of people or objects they contribute to prejudice, which is a predetermined emotional orientation toward individuals or objects (Fiske 2003). A prejudicial orientation for or against some social group contains both conscious and unconscious components (Bargh 1996). All human beings, whether they think of themselves as prejudiced or not, hold mental schemas that classify people into categories based on age, gender, race, and ethnicity (Taylor et al. 1978; Stangor et al. 1992). They cannot help it. It is part of the human condition; and these schemas generally include implicit memories that yield subconscious dispositions toward people and objects, leading to bias and stereotyping (Fiske 1998). Stereotypical notions are always present, but people are most likely to fall back on them in making judgments when they feel challenged and face threat or uncertainty (Bodenhausen and Wyer 1985; Bodenhausen and Lichtenstein 1987).

In making social judgments about others, human beings appear to evaluate people along two basic psychological dimensions: warmth and competence (Fiske et al. 2002). Warmth is how likeable and approachable a person is. We are attracted to people we view as high on the warmth dimension and seek to interact and spend time with them. We find people who are low on the warmth dimension to be off-putting and generally avoid them and seek to minimize the number and range of social contacts; we do not like them and find them “cold.” In addition to these subjective feelings of attraction and liking, we also evaluate people in terms of competence and efficacy—their ability to act in a purposeful manner to get things done. We may or may not like people who are highly competent, but we generally respect them and appreciate their ability to get things done in an effective way.

These two dimensions of social perception come together in the stereotype content model, which holds that human social cognition involves the cognitive placement of groups and individuals in a two-dimensional social space defined by the intersection of independent axes of warmth and competence (Fiske et al. 2002). As shown in Fig. 1, the social space of stereotyping has four basic quadrants. The top right quadrant contains people within the person’s own group, along with members of groups perceived to be similar to one’s own. Naturally, we think of members of our own group as warm and competent and, hence, approachable and worthy of respect. The relevant emotion associated with perception of one’s in-group is esteem or pride.

Fig. 1

Graphic representation of the stereotype content model

The intersection of the two dimensions yields three distinct kinds of out-groups, however, which vary in terms of approachability and respect. The bottom right quadrant contains those groups that are viewed socially as competent but not warm. They are respected but not liked, and the relevant emotion that people feel toward them is envy. This quadrant includes the classic middlemen minorities, such as Jews in Medieval Europe, Chinese in Malaysia, Tutsi in Rwanda, and Indians in East Africa. In a stable social structure, people show public respect and defer to members of envied out-groups, but if the social order breaks down they may become targets of communal hatred and violence, even genocide, because they are not liked and are not perceived as people “like us.”

The upper left quadrant includes out-groups that are viewed as warm and thus likeable but not competent. Those falling into this category include people who have experienced some misfortune but who are otherwise seen as “people like me,” such as the disabled, the elderly, the blind, and the mentally retarded. One could imagine being in their shoes but for some accident of fate, and so the relevant emotion is pity. We like them but, recognizing their lack of competence, we also feel sorry for them and do not respect them. In a stable social structure, members of pitied out-groups tend to be looked after and cared for; but in times of social disorder they may suffer from neglect (as seen in the aftermath of Hurricane Katrina in New Orleans), though they generally do not become targets of intentional hatred or communal violence.

Finally, social groups occupying the lower left quadrant are perceived simultaneously as low in warmth and low in competence. Being neither likeable nor capable, people within these out-groups are socially despised and the dominant emotion is disgust or contempt. This quadrant contains social outcasts such as drug dealers, welfare cheats, sex offenders, derelicts, and bums. It also includes members of groups that have been subject to an ideological process of group formation and boundary definition that questions their humanity. African Americans in the Jim Crow South were perceived by whites to be neither competent nor warm. They were socially labeled as inferior, even subhuman; and because they were perceived as less than fully human, they could be exploited, segregated, humiliated, and killed without remorse.

Recent work in neuroscience has implicated a particular region of the brain as central to the process of social cognition (see Harris and Fiske 2006). Whenever individuals perceive a stimulus as human and therefore a potential social actor, an area of the brain known as the medial prefrontal cortex lights up when observed under functional magnetic resonance imagery (fMRI). Harris and Fiske (2006) pretested a number of photographic images of social actors to establish the quadrant into which they fell; then they showed these images to experimental subjects such that each person saw a total of eighty images—20 of ingroup members, 20 of envied out-groups, 20 of pitied out-groups, and 20 of despised out-groups.

As they viewed the various social images, the brains of subjects were scanned under fMRI and centers of activity recorded. As expected, the investigators found that images of people representing in-groups, envied out-groups, and pitied out-groups triggered clear reactions in the medial prefrontal cortex. Startlingly, however, images of despised out-groups did not (Harris and Fiske 2006). Whereas out-groups triggering feelings of pity and envy were instantly perceived as human beings and social actors, those that were despised were not seen in social terms at all—at the most fundamental level of cognition. Despised out-groups thus become dehumanized at the neural level, giving human beings license, in their own minds, to treat them as if they were animals or objects.

This basic feature of human social cognition provides the psychological foundations for exploitation and exclusion in the real world. The position of a group within the social space defined by warmth and competence is not fixed, but malleable, varying across time, space, and culture (Leslie et al. 2006). Although social categories may be constructed and maintained by individuals within their own minds, the process by which boundaries are expressed is social. Group identities and boundaries are negotiated through repeated interactions that establish working definitions of the categories in question, including both objective and subjective content, a process that sociologists have labeled boundary work (see Gieryn 1983; Lamont and Molnar 2002). When social actors succeed in establishing the limits and content of various social categories in the minds of others, psychologists refer to the process as framing (Kahneman and Tversky 2000). In essence, boundary work involves defining categories in the social structure and framing involves defining them in human cognition.

People naturally favor boundaries and framings that grant them greater access to material, symbolic, and emotional resources and they seek to convince others to accept their favored version of social reality (see Lakoff 2002; Lakoff and Johnson 2002). In general, social actors who control more resources in society—those toward the top of the stratification system—have the upper hand in framing and boundary work. Whites historically have perpetuated negative stereotypes of African Americans as unintelligent, violent, hypersexual, and shiftless; and rich people likewise have promoted a view of the poor as lazy, unmotivated, undisciplined, and undeserving. To the extent that such stereotypes become a part of everyday social cognition, individual members of the stereotyped out-group will tend to experience discrimination and exclusion in society.

Of course, exclusionary social distinctions and demeaning framings are always contested by people on the receiving end (Barth 1969). Those subject to exploitation by a particular framing of social reality work to oppose it and to substitute an alternative framing more amenable to their interests. Likewise, when they encounter categorical boundaries that prevent them from accessing a desired resource, people work actively to resist and subvert the social definitions as best they can. Members of subjugated groups have their own expectations about how they should be perceived and treated, and even if they outwardly adapt to the social preconceptions of more powerful others, they generally work inwardly to undermine the dominant conceptual and social order in small and large ways.

Though such two-way interactions, however asymmetric they may be, people on both sides of a stratified social divide actively participate in the construction of the boundaries and identities that define a system of stratification. No matter what their position in the system, people seek to define for themselves the content and meaning of social categories, embracing some elements ascribed to them by the dominant society and rejecting others, simultaneously accepting and resisting the constraints and opportunities associated with their particular social status. Through daily interactions with people and institutions, people construct an understanding of the lines between specific social groups (Barth 1981).

Racialization of Mexicans in the Past

In the sweep of American history, many groups have been targets of prejudice and discrimination (Perlmutter 1999; Jacobson 1999). African Americans have traditionally been the objects of the harshest and most durable processes of exploitation and exclusion (Foner 1988; Massey and Denton 1993; Higginbotham 1996; Packard 2002; Katznelson 2005; Massey 2007). Nonetheless, successive waves of European immigrants and their descendants had to struggle long and hard to be accepted as “white” (Ignatiev 1996; Brodkin 1999; Jacobson 1999) and some groups did not achieve this until the 1970s (Alba 1990). Although presently there is little framing or boundary work to create important social distinctions among European origin groups, mechanisms of exploitation and exclusion continue to operate against African Americans (see Bertrand and Mullainathan 2004; Darity and Mason 1998; Fiske 1998; Fischer and Massey 2004; Massey and Blank 2006) and Mexicans (Telles and Ortiz 2008). Indeed, in many ways mechanisms of exclusion and exploitation directed at Mexicans have grown harsher in recent years (Massey 2007).

By themselves, Mexican Americans constitute the nation’s second largest minority group, with around 28 million people, compared to 38 million African Americans, and when they are combined with others of Latin American origin, they constitute the nation’s largest single minority group—Latinos. Not only do Mexicans constitute two-thirds of the entire Latino population; but it is doubtful whether the average Anglo-American can distinguish between mestizos of Mexican, Guatemalan, Salvadoran, Peruvian, or Colombian origin. To the extent that boundary work and framing are carried out within American society to position Mexicans as a racialized “other,” these mechanisms are bound to have spillover effects on other Latinos.

Mexicans first came into the United States in 1848, when the Treaty of Guadalupe Hidalgo ended the Mexican-American War and ceded to the United States the present states of Texas, Arizona, New Mexico, and California, along with parts of Utah, Colorado, and Nevada. With its signing, more than 50,000 Mexicans suddenly became U.S. citizens (Jaffe et al. 1980); and with the stroke of a pen they were transformed from being a majority in their own country to a minority in an alien land (MacLachlan and Beezley 2003).

The bulk of the new citizens lived in New Mexico and Texas, and the latter, being a slave state, quickly relegated Mexicans to the wrong side of the color line—not enslaved like African Americans, but certainly not accorded the rights and privileges of white Europeans (Gutierrez 1995). Through a variety of categorical mechanisms—some legal and some only quasi-legal—Mexicans were systematically disenfranchised of their property and liberties and turned into landless laborers for white land and business owners (Zamora 1993; De Leon 1993). Although Mexicans were also disenfranchised and subordinated outside of Texas, the degree of subjugation elsewhere was not as severe initially, but grew progressively worse over time as Mexican populations grew (Gutierrez 1995).

By the end of the nineteenth century, Mexicans had been transformed socially and economically into a subordinate stratum subject to widespread discrimination and systematic exclusion (Grebler et al. 1970). The degree to which Mexicans had been transmuted from masters of their own domain into a racialized source of cheap labor for whites is indicated by the Senate Dillingham Commission report of 1911, which described Mexicans as “notoriously indolent and unprogressive in all matters of education and culture” doing dirty jobs fit only for “the lowest grade of nonassimilable native-born races” though their “usefulness is, however, much impaired by [their] lack of ambition and [their] proneness to the constant use of intoxicating liquor” (U.S. Commission on Immigration 1911, p. 59, 94, 110).

The roughly 50,000 Mexicans absorbed into the United States in 1848 expanded mainly through natural increase until the twentieth century. Although fertility rates were high, so were levels of mortality, and the rate of demographic expansion was modest. By 1900, the Mexican origin population of the United States was only around 150,000, and in that year just 237 Mexicans arrived as immigrants. Although racism against Mexicans was real and ongoing, their small numbers and geographic isolation meant that categorical mechanisms of inequality established to ensure their subordinate status had little effect on inequality within the nation as a whole.

This status quo was upset in 1907. In that year, the United States and Japan concluded a “Gentlemen’s Agreement” under which Japanese authorities agreed to prevent the departure of its citizens for the United States, in return for which the United States agreed not to inflict upon Japan the indignity of prohibiting their entry (Zolberg 2006). This action caused a serious disruption of western labor markets, for Japanese immigrants had come to comprise the backbone of the region’s unskilled workforce. In response, western employers looked southward and began to recruit Mexicans, first to work on the railroads, then in mines and farm fields, and finally in factories (Cardoso 1980; Durand and Arias 2000).

As the Dillingham Commission put it, “Mexican immigration may increase for some time as this race offers a source of labor to substitute for the Asiatics in the most undesirable seasonal occupations.” Indeed, the Commission continued, “in the two southern California districts where the force of field workers is predominantly Mexican, the Mexican is preferred to the Japanese. He is alleged to be more tractable and to be a better workman in one case. In the other he is said to be a quicker and better workman than the Japanese….” (U.S. Commission on Immigration 1911, p. 50, 110). Few Mexicans entered the United States before the Gentlemen’s Agreement with Japan; but thereafter the flow of legal immigrants accelerated rapidly from near-zero in 1907 to reach 16,000 in 1909.

With the outbreak of the First World War in 1914, U.S. employers were cut off from supplies of immigrant labor in Europe as well as Asia, and as demand for American products grew, employers increased their recruitment of workers from south of the border (Cardoso 1980). When the United States entered the First World War in 1917, labor demand spiked and worker shortages became chronic as white factory workers were mobilized for military duty. In response, the U.S. government established an official labor recruitment program to assist growers in the southwest (Reisler 1976; Morales 1982) and factory owners throughout the Midwest redoubled their private recruitment efforts (Durand and Arias 2000).

Thus, the number of contract workers entering the United States grew to 17,000 in 1920 and legal immigration reached an unprecedented 51,000 persons. American insecurities about all things foreign came to a head during the recession that followed the war, and the Red Scare of 1918–1921 was accompanied by a wave of anti-immigrant hysteria. Congress passed the first quota law in 1921 to curtail immigration from southern and eastern Europe and enacted an even stricter version in 1924. During this period, both legal immigration and contract migration by Mexicans fell, reaching lows in 1922 of 18,000 and 12,000, respectively.

Economic recovery, however, led to a sustained economic boom known as the “roaring twenties,” and given the new restrictions on European immigration and steadily tightening labor markets, recruitment of Mexicans soon resumed. Legal immigration surged again to peak at 88,000 in 1924 and the entry of contract laborers climbed to 18,000. Until this time, the Mexico–U.S. border was little more than a line on a map and was mostly unmarked (Massey et al. 2002). In response to growing immigration from Mexico, however, in 1924 Congress created the U.S. Border Patrol and for the first time, the border with Mexico became a tangible reality.

The creation of the Border Patrol brought into existence a new category of Mexican in the United States—the illegal migrant. During its first year of operation, the U.S. Border Patrol apprehended around 4,600 Mexicans attempting to enter the country without documents (Massey et al. 2002). Legal entries rebounded in the middle part of the decade, however, to reach 68,000 documented immigrants and 17,000 contract workers in 1927. As these two streams declined in the late 1920s, illegal immigration correspondingly rose, with apprehensions reaching 18,000 in 1930.

From the Gentlemen’s agreement in 1907 to 1930, the number of Mexican-born persons in the United States more than quadrupled, going from 178,000 to 739,000 in a little over two decades and making Mexicans a visible minority in cities throughout the Southwest, Midwest, and Pacific regions (Durand and Arias 2000). With the crash of the American stock market, however, the surge in Mexican immigration ended as quickly as it had begun. In keeping with percepts laid out by the Dillingham Commission, Mexicans were considered expendable as workers and unassimilable as citizens, and in an era of rising austerity whites framed them as taking jobs that rightfully belonged to “real” Americans and burdening taxpayers with relief payments that rewarded their natural “indolence” (Hoffman 1974). In keeping with these perceptions, federal authorities joined with state and local officials to organize a series of deportation campaigns that, over the course of a few years, cut the Mexican population of the United States in half (Jaffe et al. 1980).

During the period 1929–1937, some 458,000 Mexicans were arrested and expelled from the United States without due process, including many legal resident aliens and their U.S. citizen children. By 1940, only 377,000 Mexican immigrants were left in the country (Jaffe et al. 1980). Those who remained were pushed to the margins of society, segregated in dilapidated barrios where they attended segregated schools and received inferior services (Grebler et al. 1970). In these enclaves, Mexicans were transformed from aspiring immigrants into a self-conscious domestic minority, increasingly calling themselves not Mexicans but Chicanos (Sanchez 1995; Gutierrez 1995). The degree to which Mexicans were racialized during this period is indicated by the fact in 1930 the U.S. Census Bureau, for the first and only time in its history, enumerated Mexicans as a separate race, alongside blacks (Bean and Tienda 1987).

With the entry of the United States into the Second World War, however, American industry once again mobilized and full employment resumed. In combination with renewed military conscription, the war created new labor shortages, especially in the American southwest and federal authorities quickly forgot about the deportations and quietly turned southward to negotiate a binational treaty for the “temporary” importation of farm workers from Mexico, who became known as braceros. The resulting “Bracero Program” was operated by the U.S. Departments of State, Labor, and Justice in cooperation with the Mexican government and in September of 1942, the first braceros arrived for agricultural work in Stockton, California (Calavita 1992).

The Bracero Program was instrumental in restarting a migratory flow that had been dormant for more than a decade. In the years leading up to 1942, Mexican immigration to the United States was virtually nil; and although labor flows were revived by the Bracero Program, the number of contract workers remained rather small throughout the war. From 1942 through 1945, a total of only 168,000 braceros were recruited into the United States. Within urban areas, the children of earlier Mexican immigrants took advantage of the return to full employment and moved upward economically, working at unionized jobs in war industries and translating their newfound affluence into a flashy style known as “pachuco,” whose emblem was a baggy ensemble known as the “zoot suit” (Mazon 1984).

As with the Japanese earlier in the century, white Californians resented racial inferiors rising above their assigned station, and in the charged atmosphere of wartime Los Angeles, anti-Mexican rioting broke out. On June 3, 1943, a group of servicemen on leave complained that they had been assaulted by a gang of pachucos wearing zoot suits (Obregon Pagan 2006). In response, an angry mob of white soldiers and civilians headed into the Mexican barrio of East Los Angeles where they attacked all males wearing zoot suits, beat them severely while ripping off the offensive garments and burning them on the spot.

Rather than protecting U.S. citizens of Mexican origin, the Los Angeles police swept into the barrio and arrested hundreds of already beat-up pachucos for “disturbing the peace,” several of whom died in jail for want of medical treatment. Although nine white sailors were arrested over the next few days, eight were released without charge and one was let go after paying a small fine. The attacks on Mexicans only ceased when military authorities declared Los Angeles to be off-limits to service personnel. For Mexican Americans, however, a strong message had been sent: even in progressive California, people of Mexican origin were not going to be accepted as equals, no matter where they were born, how much they earned, or how stylishly they dressed (Mazon 1984).

Although originally envisioned as a “temporary” wartime measure, the booming postwar economy perpetuated growers’ fears of a labor shortage, and under pressure from the Texas and California congressional delegations, the House and Senate extended the Bracero Program on a year-to-year basis through the late 1940s. Despite the extensions, however, the number of bracero visas remained insufficient to meet rising grower demand and so employers increasingly took matters into their own hands by recruiting illegal migrants directly, especially after 1950 when reinstatement of the draft during the Korean War markedly tightened U.S. labor markets. As Mexicans crossed the border in larger numbers on their way to farms and fields where they knew they would be hired, the annual number of apprehensions went from around 7,000 in 1942 to reach 544,000 in 1952.

With the end of the Korean War, a brief economic recession combined with another surge of anti-foreign hysteria during the McCarthy era to make illegal migration a hot political issue. In 1953–1954, the U.S. Immigration and Naturalization Service (INS) responded to the rising clamor by launching “Operation Wetback” (Calavita 1992). In cooperation with state and local authorities, the INS militarized the Mexico–U.S. border and organized a mass arrest of Mexicans—or, more accurately, Mexican-looking people. During 1954, the number of Mexicans apprehended by the INS swelled to over a million for the first time in U.S. history.

A key difference compared with the deportation campaign of the 1930s, however, was that this time Congress simultaneously acted to expand the number of temporary work visas, roughly doubling the annual number of braceros admitted per year. From 1955 through 1959, between 400,000 and 450,000 braceros were annually imported into the United States. Legal immigration also surged, going from 9,600 in 1952 to 65,000 in 1956 before leveling off (Massey et al. 2002). It was this increase in access to legal avenues for entry, more than stepped-up border enforcement that reduced illegal migration to a trickle during the late 1950s. From a figure of 1.1 million in 1954, the number of apprehensions fell to just 30,000 in 1959, where it remained well into the 1960s.

For a time, Americans seemed content in having a disposable workforce that seasonally traveled to the United States for difficult and demanding work but which regularly returned to Mexico once that work was done. In the wake of Operation Wetback and the expansion of the Bracero Program, illegal immigration disappeared as a political issue; but as the civil rights movement picked up steam immigration became controversial in a different way. In addition to overturning Jim Crow and banning discrimination from U.S. markets, civil rights activists sought to purge the nation’s immigration system of its racist legacy. As a result, during the 1960s both the Bracero Program and the national origins quotas both came under attack.

The Bracero Program was viewed by civil rights advocates as a corrupt, coercive, and exploitive labor system, roughly on a par with black sharecropping in the South. Under intense pressure from religious groups, unions, and civil rights organizations, Congress downsized the Bracero Program in the early 1960s—reducing the annual number of work visas from 438,000 in 1959 to 178,000 in 1964—before voting in the following year to end the program altogether. In the same year, 1965, Congress passed amendments to the Immigration and Nationality Act that finally abolished discriminatory national origins quotas and lifted the ban on immigration from Asia and Africa.

Instead of racist quotas, the new legislation set a neutral cap of 20,000 immigrant visas per year for each country outside the Western Hemisphere. These visas were allocated to people using a “preference system” that took into account national employment and humanitarian needs (Zolberg 2006). Although viewed as a landmark achievement by the civil rights movement, the 1965 Immigration Act also launched a new trend of restrictive immigration policies toward Mexico by imposing the first-ever numerical limits on immigration from the Western Hemisphere. Subsequent amendments successively put nations in the Western Hemisphere under the 20,000 per-country visa limit, abolished separate hemispheric quotas, and finally established a single worldwide ceiling that was reduced to just 270,000 visas. Whereas in 1965 Mexicans seeking to become legal immigrants faced no numerical limit, by 1980 they found themselves competing with aspiring immigrants from all over the world for a limited number of visas.

Together with the termination of the Bracero Program, these legislative changes dramatically reduced the possibilities for legal entry and virtually guaranteed a rise in undocumented migration. Whereas around 480,000 Mexicans entered the United States each year as braceros or legal immigrants during the late 1950s, the total inflow of Mexicans through legal channels fell to just 62,000 per year from 1965 to 1985, 13% of its former level. The gap between the demand for visas on the part of employers and workers and the paltry number offered by the government was increasingly made up through undocumented migration, and annual apprehensions along the border climbed steadily from 55,000 in 1965 to 1.6 million in 1985.

In essence, the shift in U.S. immigration policy after 1965 transformed Mexican migration from a de jure guest worker program based on the circulation of braceros into a de facto guest worker program based on the circulation of undocumented migrants (Durand and Massey 2003). Until 1985, this flow remained overwhelmingly circular, composed primarily of young men moving back and forth for seasonal work in agriculture, construction, manufacturing, and services (Massey et al. 2002). Under the unofficial temporary worker program that prevailed between 1965 to 1985, Mexicans remained disposable as workers and unwanted as citizens, as in the Bracero Program, but labor flows were regulated informally as a product of border enforcement rather than under the terms of a formal binational agreement.

Although the presence of undocumented Mexicans in the United Sates was technically illegal, until the mid-1980s the consequences of this illegality were relatively benign. The size and budget of the Border Patrol rose modestly each year to keep pace with a gradual increase in the underlying volume of undocumented migration and the probability of apprehending a undocumented border crosser remained constant at around 33% (Massey and Singer 1995). The political economy of Mexican migration before 1985 was such that an undocumented migrant could reasonably expect to arrive at the border and achieve entry after a few tries at small personal risk and with modest financial investment.

The rise of undocumented migration after 1965 was accompanied by new demonizaton of Mexicans, however. Undocumented Mexicans were increasingly framed as a threat to the nation’s security, workers, culture, and way of life. Chavez (2001) studied U.S. magazine covers devoted to immigration between 1965 and 2000 and classified them as affirmative, alarmist, or neutral in their portrayal of immigrants. Covers coded as “affirmative” used text and images to celebrate immigration; “alarmist” covers used text and images to convey problems, fears, or dangers associated with immigration; and “neutral” covers were accompanied by articles that offered balanced and factual coverage of immigration issues that was neither affirmative nor alarmist.

Chavez found that alarmist themes overwhelmingly predominated in coverage of immigration after 1965, characterizing two-thirds of all covers devoted to the topic from 1965 through 1999, compared with just 9% classified as neutral and 19% as affirmative. The frequency of alarmist covers also increased markedly over time. Whereas 18% of the alarmist covers appeared in the 1970s, 38% were published in the 1980s and 45% appeared in the 1990s. Upsurges in alarmist text and imagery also coincided with recessionary periods in the United States (Chavez 2001, pp. 21–24).

The words printed in the texts that accompanied the images generally reinforced the sense of alarm and urgency communicated by the pictures (Chavez 2001). In time-honored fashion, editors made heavy use of marine metaphors, depicting immigration as a “tidal wave” that was “flooding” the United States and threatening to “inundate” its culture. During the 1970s and 1980s, however, a new metaphor appeared with growing frequency as immigrants and immigration were framed increasingly in martial terms. The Mexico–U.S. border was portrayed as a “battleground” that was “under attack” from “alien invaders” who constituted a “time bomb” waiting to explode and destroy American culture and values. In this militarized portrayal, Border Patrol Officers became “defenders” who were “outgunned” as they tried to “hold the line” against attacking “hoards” (Dunn 1996; Andreas 2000).

Whether the metaphorical language was martial or marine, however, it always portrayed immigration from Mexico as a “crisis.” Going back to the earliest days of the republic, politicians have found it convenient to demonize immigrants during periods of social upheaval and economic insecurity (Higham 1955) and during the 1980s the symbolic portrayal of immigrants as a threat reached new heights (Zolberg 2006). Ronald Reagan, in particular, framed immigration as a question of “national security” linked to his ongoing prosecution of the Cold War. As a result of communist insurgencies in Central America, he predicted “a tidal wave of refugees—and this time they’ll be ‘feet people’ and not boat people—swarming into our country seeking safe haven from communist repression to the south” (Washington Post, June 21, 1983). In a 1986 speech, he reminded Americans that “terrorists and subversives are just two days driving time from [the border crossing at] Harlingen, Texas” (quoted in Kamen 1990) and his 1987 Task Force on Terrorism reported that immigrants constituted a potential fifth column in the United States because extremists would “feed on the anger and frustration of recent Central and South American immigrants who will not realize their own version of the American dream” (quoted in Dunn 1996).

The labeling of immigration as a national security threat, the symbolic portrayal of the Mexico–U.S. border as a defensive bulwark, and the demonization of Latinos as subversives reached a crescendo in the middle 1980s (Massey et al. 2002). The year 1986 was pivotal. Late in that year congress passed the Immigration Reform and Control Act, or IRCA, which contained three far-reaching provisions that drastically reshaped the political economy of migration and transformed the position of Mexicans in the United States (Durand et al. 1999). First, IRCA sharply increased funding for border enforcement, initiating an unprecedented expansion in the Border Patrol that continued unabated for two decades. Second, it authorized two legalization programs: one for long-term residents who had been in the country for at least 5 years and another for short-term agricultural workers. Finally, it criminalized the hiring of undocumented workers and applied sanctions against employers who knowingly did so.

Racialization of Mexicans in the Present

After 1986, the size and budget of the Border Patrol grew rapidly in a way that was disconnected from the underlying volume of migration (Massey et al. 2002). Despite vast expenditures on border enforcement, however, the rate of illegal entry remained steady (Donato et al. 1992). By 1990, it was clear that the legislation was not working and with illegal migration once again growing, Congress passed a new series of restrictive amendments to the Immigration and Naturalization Act, imposing new limits on legal family migration, authorizing funding for thousands of additional Border Patrol officers, tightening employer sanctions, streamlining deportation procedures, increasing penalties for immigration violations, and stepping up internal enforcement. From a figure of just 11,000 in 1985, the annual number of Mexicans arrested in the interior of the United States and forcibly deported grew steadily to reach 150,000 in 2005, the largest number since the deportation campaigns of the Great Depression.

The militarization of the border and the escalation of internal enforcement helped to break the traditional pattern of circular migration that had prevailed historically (Massey et al. 2002). Whereas the probability that a Mexican would initiate undocumented migration remained constant in the years following IRCA, the likelihood of returning to Mexico once entry had been achieved plummeted after 1986. Rather that deciding not to leave for the United States in the first place, undocumented migrants instead chose to stay longer once they had run the gauntlet at the border and successfully made it into the country (Durand and Massey 2003).

The falling rate of out-migration and the steady rate of in-migration combined to increase the rate of net undocumented migration and cause an unprecedented acceleration in the number of Mexicans living north of the border. In essence, restrictive U.S. immigration and border policies backfired. Instead of reducing the net annual inflow of Mexican migrants, they doubled it. U.S. immigration and border policies transformed Mexican immigration from a circular flow of male workers into a settled population of families. By the year 2008, the total number of undocumented Mexicans present north of the border had reached 7 million and the total number of foreign-born Mexicans had climbed to around 12 million.

These figures imply that one of every 10 people born in Mexico now lives in the United States, and around 55% are present in illegal status. At the same time, immigrants have come to comprise a growing fraction of all Mexican Americans, roughly 40% as of 2005; and the large number of undocumented among them mean that midway in the first decade of the twentieth century, more than half of all Mexican-born persons—and more than a fifth of all persons of Mexican origin—lack any social, political, or economic rights in the United States. Mexicans are now more exploitable than at any time since the 1850s.

Rather than blaming the deterioration of wages and working conditions on structural changes in the political economy, however, conservative economists such as Borjas (1995) have attributed the trend to a “declining quality of immigrants,” singling out Mexicans in a way that would be familiar to members of the Dillingham Commission. Although the demonization of Latino immigrants as “invaders” and “terrorists” slackened somewhat during economic boom of the 1990s, these framings returned with a vengeance after September 11, 2001, both inside and outside of academia.

Within the academy, intellectuals such as Harvard political scientist Huntington (2004) offered a reprise of the Dillingham Commission’s assertions about the unassimilability of Latin American immigrants:

The persistent inflow of Hispanic immigrants threatens to divide the United States into two peoples, two cultures, and two languages. Unlike past immigrant groups, Mexicans and other Latinos have not assimilated into mainstream U.S. culture, forming instead their own political and linguistic enclaves-from Los Angeles to Miami-and rejecting the Anglo-Protestant values that built the American dream. The United States ignores this challenge at its peril.

Outside of the ivory tower, former Nixon speech writer and conservative pundit Buchanan (2006), has warned of an “Aztlan Plot” fomented by Mexican conspirators to recapture lands lost under the Treaty of Guadalupe Hidalgo, thus effecting a “reconquista” of the American southwest. Comparing Mexicans to the barbarians invading ancient Rome, he referred to “the Third World invasion and conquest of America” as a “state of emergency.” In an interview with Time magazine, he warned:

If we do not get control of our borders and stop this greatest invasion in history, I see the dissolution of the U.S. and the loss of the American southwest—culturally and linguistically, if not politically—to Mexico. It could become a part of Mexico in the way that Kosovo is now a part of Albania” (August 28, 2006, p. 6).

Even more hyperbole was forthcoming from Chris Simcox of the Minutemen Civil Defense Corps, a vigilante group he founded to patrol the Mexico–U.S. border. On the organization’s website, he rhetorically asks “Are terrorists exploiting our porous borders?” and then supplies the following answer: “We know drug dealers, gang bangers and way too many criminal foreign nationals are creating havoc in our communities and threatening our public safety.”

The legal foundations for the criminalization not just of undocumented hiring, but of undocumented migrants themselves were laid by the 1996 Antiterrorism and Effective Death Penalty Act, which gave the federal government new police powers for the “expedited exclusion” of any alien who had ever crossed the border without documents (no matter what his or her current legal status) or who had ever committed a felony (no matter how long ago). These provisions—coming on the heals of a decade of draconian drug laws and three strikes legislation—instantly rendered thousands of legal resident aliens deportable, many of whom had entered as infants and spent their entire lives in the United States.

The law also delegated to the State Department absolute authority to designate any organization as “terrorist,” thereby making all members of groups so-designated immediately excludable and deportable. It also narrowed the grounds for asylum and added alien smuggling to the list of crimes covered by the Racketeer Influenced Corrupt Organizations (RICOs) statute, while severely limiting the possibilities for judicial review of deportations. According to Legomsky (2000, p. 1616), this legislation constitutes “the most ferocious assault on the judicial review of immigration decisions” ever launched “by creating new removal courts that allow secret procedures to be used to remove suspected alien terrorists; by shifting the authority to make ‘expedited removals’ to immigration inspectors at ports of entry; and by setting unprecedented limits on judicial review of immigration decisions.”

The events of September 11 thus occurred against a background of rising animus toward immigrants and a growing assault on their civil liberties and social rights. In response to the terrorist attacks Congress on October 26, 2001 passed the USA PATRIOT Act, which granted the executive branch expansive new powers to deport, without any hearing or presentation of evidence, all aliens—legal and illegal—that the Attorney General had “reason to believe” might commit, further, or facilitate acts of terrorism. For the first time since the Alien and Sedition Act of 1798 Congress voted to permit the arrest, imprisonment, and deportation of non-citizens upon the orders of the Attorney General without judicial review.

Over the course of U.S. history attacks on immigrants have waxed and waned. What distinguishes the current wave of anti-immigrant hysteria from its predecessors is not its demonizing of foreigners or its harsh treatment of non-citizens per se, but its use of the fear of foreigners to launch a broader assault on the civil liberties not just of immigrants, but of all Americans, for the PATRIOT Act also permits unprecedented surveillance and incarceration of U.S. citizens, again at the discretion of the executive branch and without review. As Zolberg (2006) notes “while the challenges posed by international migration are real and warrant a worldwide reconsideration of prevailing regimes, the resurgence of nativist responses constitutes a more immediate threat to liberal democracy than immigration itself.” It is no coincidence that the only U.S. citizen now held in indefinite detention without charge and without a hearing is a dark-skinned Latino, Jose Padilla.

The Consequences of Racialization

All of the boundary work being done by academics, pundits, and politicians to frame Latin American immigrants as a threat and categorize them socially as undesirable has affected public opinion, turning it steadily against Latinos. According to polls conducted by the Pew Charitable Trusts, as late as 2000 only 38% of Americans agreed that “immigrants today are a burden on our country because they take our jobs, housing, and health care.” Five years later, the percentage had risen to 44%; and as the drumbeat of anti-immigrant rhetoric reached a crescendo in 2006 it became a majority viewpoint at 52%. In keeping with this shift, the percentage of Americans who rated immigration as a moderately big or very big national problem rose from 69% in 2002 to 74% in 2006 (Kohut and Suro 2006).

As of the year 2006, almost half of all Americans (48%) opined that “newcomers from other countries threaten traditional American values and customs” and 54% said that the United States needed to be “protected against foreign influence.” Not surprisingly, given these views, 49% said they believed that “immigrants kept to themselves and do not try to fit in;” 56% said they “don’t pay their fair share of taxes;” 58% believed that immigrants “do not learn English in a reasonable amount of time;” and 60% of those who had heard of the Minutemen approved of their activities (Kohut and Suro 2006).

Recent studies by Lee and Fiske (2006) applied the stereotype content model to various immigrant groups. Based on respondent and subject ratings, they plotted the position of different groups in the two-dimensional space defined by the intersection of warmth and competence, and the results of this exercise are reproduced in Fig. 2. As expected, the social space generally occupied by esteemed in-group members (high warmth, high competence) includes groups such as Canadians, Europeans, documented immigrants, and third generation immigrants. Likewise, the space generally occupied by envied out-groups (high competence, low warmth) is occupied by classic middlemen minorities such as the Koreans, Chinese, Japanese, and Asians generally. Eastern Europeans, first generation immigrants, Russians, Germans, French, and Middle Eastern immigrants occupy a middle position of moderate warmth and moderate competence.

Fig. 2

Stereotype content model applied to immigrants (Source: Lee and Fiske 2006)

Note, however, which groups occupy the space of low warmth and low competence that corresponds to the most despised out-groups: South Americans, Latinos, Mexicans, farm workers, and Africans; and the most despised immigrant group of all is undocumented migrants, who rate lowest of all on both warmth and competence, such that they approach the portion of the graph usually reserved only for the most detested and socially stigmatized groups, such as criminals and drug dealers. In societal terms this is dangerous territory, since it implies that undocumented migrants are not perceived as fully human at the most fundamental neural level of cognition, thus opening a door to the harshest, most exploitive, and cruelest treatment that human beings are capable of inflicting on one another.

In this context, the U.S. immigration system has assumed a new centrality in the exploitation and exclusion of Latinos. The implementation of employer sanctions increased discrimination against Hispanics in U.S. labor markets, lowering their wages, depressing the returns to human capital, and closing off long-established pathways of upward mobility. At the same time, IRCA promoted a wholesale shift to subcontracting in the unskilled labor market. The militarization of the Mexico–U.S. border, meanwhile, raised the rate of undocumented population growth to increase the number of people in exploitable, powerless categories. Finally, as private discrimination increased and larger shares of the population were being exploited economically, Congress increased the social penalties for being poor, Hispanic, foreign, and undocumented, cutting even legal immigrants off from public services for which they had heretofore qualified.

As a result of these deliberate policy actions, the political economy facing Hispanics is now vastly harsher and more punitive than the one prevailing before 1986. Historically, Hispanics have occupied a middle position between blacks and whites in the American stratification system, but with the restructuring of the political economy of immigration in the late 1980s and early 1990s, the relative standing of Hispanics declined and they came to replace African Americans at the bottom of the class hierarchy. Figure 3 illustrates this change by showing the ratio of minority to white income for blacks and Hispanics from 1972 to 2002.

Fig. 3

Hispanic and black personal income as a ratio of white income (Source: U.S. Bureau of the Census)

The bottom two lines show income ratios for Hispanic and black men. In the early 1970s, the average black male earned roughly 60% of what the average white male earned, while Hispanic males earned around 70% of what the white male level. These relative income ratios prevailed through the early 1980s, but in the middle of that decade IRCA transformed the structure of the low wage labor market and the bargaining position of Hispanic men deteriorated. From 1983 to 1986, the ratio of Hispanic to white income fell from .70 to around .60, where it hovered until 1991 and then dropped below .60 for the first time ever. At about the same time, black male incomes began rising relative to those of white miles and in 1993 the income ratio crossed over the Hispanic–white ratio and from that point onward Hispanics replaced blacks at the bottom of the male earnings hierarchy.

The top two lines show trends in the relative earnings of Hispanic and black women to reveal a similar switching of positions, one that occurred even earlier than among minority men. In 1972 Hispanic women earned the same income as white women, whereas black women earned around 92% of their white counterparts. During the 1970s and 1980s, however, the earnings of women in both minority groups deteriorated relative to those of white women; but the decline was more rapid among Hispanic women and in 1981 the two lines crossed and from then on Hispanic women replaced black women at the bottom of the female earnings hierarchy.

Despite this early crossover, however, the Hispanic–white and black–white income ratios remained quite close to one another until IRCA passed in 1986. After this date, the deterioration in black female income slowed down and then in the early 1990s reversed and began to move upward. In contrast, the deterioration in the Hispanic women’s earnings accelerated and the two income ratios began pull apart at a rapid pace, yielding a widening gap. From 1987 to 2002, the ratio of black female income to white female income rose from .80 to around .96 whereas the income ratio for Hispanic females fell from .80 to around .70 before coming back up to end the decade at around .77.

The shifting fortunes of Hispanics and African Americans in U.S. labor market are clearly reflected also in American poverty statistics, as shown in Fig. 4. Historically, rates of Hispanic poverty were much below those of blacks, but over the course of the 1980s and 1990s the differential disappeared and the two groups ended the twentieth century at rough parity in terms of material deprivation. Figure 4 shows the ratio of Hispanic to black poverty from 1972 to 2002. Through the 1970s and early 1980s, Hispanic poverty fluctuated at around 70–80% of the black level, but during the late 1980s and early 1990s Hispanic poverty rates rose and came to range between 80% and 90% of black rates. With the increase in settlement by undocumented migrants and the shift to family migration following the border build-up that began with operation blockade in El Paso, Hispanic rates of poverty came to equal or exceed those of blacks and the ratio pushed above 1.0 for the first time since poverty statistics had begun being collected.

Fig. 4

Ratio of Hispanic to black poverty rates (Source: U.S. Bureau of the Census)

The deterioration in the labor market position of Hispanics relative to blacks has been accompanied by a parallel shift within housing markets, for the 2000 Housing Discrimination Study revealed a significant increase in discrimination against Hispanics. Whereas in 1989 Hispanics were 19% less likely than blacks to experience adverse treatment in America’s rental markets, in 2000 they were 8% more likely suffer discrimination. Moreover, although the incidence of discriminatory treatment fell for both groups in the sales market, the decline for Hispanics was much smaller. As a result, whereas blacks in 1989 were twice as likely as Hispanics to experience discrimination in home sales, by 2000 Hispanics were were 18% more likely than blacks to experience it (Turner et al. 2002). Consistent with these data, in their audit of rental housing in the San Francisco Bay area, Purnell et al. (1999) documented extensive “linguistic profiling” that excluded speakers of Chicano English as well as black English from access to housing.

As discrimination against Latinos in housing markets increased so did levels of Hispanic residential segregation. Whereas the overall level of black segregation fell by 10 points over the past decade and black neighborhood isolation dropped by 12 points, Hispanic segregation rose by six points and isolation increased by ten points (Charles 2003); and whereas Hispanics did not satisfy the criteria for hypersegregation in any metropolitan areas during 1980 or 1990, by 2000 both New York and Los Angeles had earned the dubious distinction of becoming hypersegregated for Latino residents (Wilkes and Iceland 2004).

In the social realm, researchers have also documented the “chilling effect” of the 1996 immigration and welfare legislation on the use of public services by immigrants (Fix and Zimmermann 2004). Among undocumented migrants, the use of social services, always quite low, fell even further, so that after 1996 fewer than 5% reported receiving food stamps, welfare, or unemployment insurance while in the United States, and just 7% reported putting their children in public schools. More surprising was the decline in services consumed by legal immigrants. After 1996, usage rates for welfare, unemployment insurance, and food stamps all fell sharply to 10% or less Donato et al. (2006) and according to estimates by Borjas (2004), every 10% cut in the fraction of the public on public assistance raises the relative number of food-insecure households by five percentage points.

It thus seems that in 1996 Congress sent a clear signal to legal as well as illegal immigrants that they were unwelcome in what remained of the American welfare state, and both sets of immigrants got the message loud and clear. In addition to restricting the eligibility of legal immigrants for federal means-tested benefits, however, Congress in 1996 also raised the income threshold required to obtain an Affidavit of Support. This document is required of all persons seeking to sponsor the legal entry of a family member and requires a household to prove it has enough resources to support the immigrant should he or she become indigent. Congress, of course, sought to curtail family immigration. Rather than standing by and allowing the government to declare them ineligible to bring in relatives, immigrant households fought back by sending more family members into the workforce to bolster collective earnings so they could meet the higher income threshold. After 1996, older children in Mexican families increasingly dropped out of school and went to work, thus depressing already low levels of education among Latino children and permanently undermining their economic prospects (Donato et al. 2006).

Although Hispanics may not have been swept into the prison industrial complex to the same extent as African Americans, they nonetheless comprise a sizeable share of inmates that is disproportionate to their share of the U.S. population. Around 5% of Hispanic males aged 20–40 were in prison or jail in 2000, compared with 12% of blacks and just 2% of whites (Western 2006). Moreover, in the spring of 2006 the U.S. House of Representatives passed legislation (HR4437) sponsored by Republican Representative James Sensenbrenner to make “unlawful presence” in the United States a felony. It defined unlawful presence so broadly that almost every immigration violation, no matter how minor, technical, or unintentional, became a federal crime subject to imprisonment and deportation. If this act passes the Senate to become law, it will render 12 million people instantly subject to incarceration and represent the largest expansion of the prison industrial complex ever, potentially tripling the size of America’s prison population, already the largest on earth.

Conclusion

In American Apartheid, Massey and Denton (1993) offered a blueprint of how to build an underclass:

To begin, choose a minority group whose members are somehow identifiably different from the majority. Once the group has been selected, the next step in creating an underclass is to confine its members to a small number of continuous residential areas and then to impose on them stringent barriers to residential mobility… Once a group’s segregation in society has been ensured, the next step in building an underclass is to drive up its rate of poverty… The interaction of poverty and segregation acts to concentrate a variety of deleterious social and economic characteristics… Through prolonged exposure to life in a racially isolated and intensely poor neighborhood, poverty will quite likely be passed to children in the next generation. When this point is reached, a well-functioning and efficient structure for the construction and maintenance of an urban underclass will have been created.

The evidence reviewed here suggests that U.S. policies are moving Mexican Americans steadily away from their middle position in the economic hierarchy and toward the formation of a racialized underclass. Segregation levels are rising, discrimination is increasing, poverty is deepening, educational levels are stagnating, and the social safety net has been deliberately poked full of holes to allow immigrants to fall through. Whether or not Mexicans become a new urban underclass remains to be seen; but it is already clear that after occupying a middle socioeconomic position between whites and blacks for generations, the economic fortunes of Mexicans have now fallen to levels at or below those of African Americans.

In one critical way, moreover, Mexicans are much worse off than black Americans. Whatever discriminatory barriers African Americans still face, they at least have the legal right to live and work in the United States. In contrast, one-fifth of all Mexican Americans lack any legal claim on American society because they are present without authorization, and this fraction is rising rapidly. If the share of Latinos in undocumented status continues to rise, the resulting underclass will be even “better” than the one that emerged in black inner cities during the 1980s. Not only will its members be exploited and excluded; they will be outside the law itself, deportable at a moment’s notice and perhaps even at serious risk of incarceration for the felonious crime of living and working in the United States without permission.