Samuel Huntington's analysis of the “Hispanic challenge”—his claim that Mexicans are on their way to forming a separate nation within the U.S.—rests on a series of misconceptions that are not his alone.1
At the heart of his difficulties is a widely shared form of reasoning about racial and ethnic populations that has become increasingly problematic in the contemporary era of mass immigration: it anticipates a single, predominant outcome for group members, such as assimilation or racialized exclusion; instead, it is the diversity within groups of patterns of incorporation into American society that needs recognition today. This is all the more true of Mexican Americans because of the long history across which their immigration stretches and their presence in the Southwest and California before the arrival of European Americans.That long history is implicated, for instance, in what Huntington (and others) perceives as a stagnation in Mexican-American socioeconomic attainment after the second generation. In the Mexican-American case, the cross-sectional comparison of generations is misleading, as we will see below, in part because different generations originate in different periods of Mexican immigration and settlement and in part because the institutional discrimination of the pre-civil rights era thwarted mobility and interfered with past intergenerational advance. Direct evidence of diversity in the adaptation to U.S. society comes from patterns of linguistic assimilation, which do not support the notion of a cleavage into separate, language-based subsocieties. That is, while there is arguably more bilingualism in later generations of Mexican Americans than there was among European Americans, by the third generation the prevalent pattern is English dominance and even monolingualism; and that pattern does not appear to be weakening over time. Nor do Mexican Americans show any signs of forging separate institutional structures, a prerequisite, one would think, to the creation of a subsociety. The mobility of some descendants of Mexican immigrants from the Mexican-American category to that of non-Hispanics with some Mexican ancestry further speaks against the notion that a chasm is opening up.
Thus, rather than forming a group that is increasingly separating itself from the U.S. mainstream, Mexican Americans turn out to be a highly diverse population in many respects—socioeconomic attainment, linguistic assimilation, racial appearance, and legal status. A substantial fraction among them—the majority, in all likelihood—is pursuing the American Dream and clearly advancing beyond the humble status of the parental generation. A significant minority, however, because of a lack of legal documents, the absence of economic and educational opportunities, or racial and ethnic discrimination, is excluded from this attempt. These are very serious problems, but they can be addressed by enlightened social policy. Thus, even here, the outcome is not already determined.
The Paradox of Generations
Huntington presents data that appear to show very low levels of Mexican-American educational advancement beyond high school, regardless of generation.2
Huntington 2004b, 234.
The comparison and the interpretation that Huntington educes from it suffer from both deficiencies in the data and a lack of appreciation of the problematic nature of generational comparisons in the Mexican-American case. The data deficiencies are easier to explain. Huntington grossly understates Mexican-American educational attainment. For example, the data he presents for Mexican Americans collapse all post-secondary education that falls short of a degree into the high-school-diploma-only category. However, it is the “some college” category, absent from his presentation, where the majority of Mexican-American college goers finish. As the 2000 census data in Table 1 reveal, the percentage that has attended college in a recent cohort, individuals born in the U.S. between 1970 and 1975, is quite substantial—about 50 percent among Chicanas and more than 40 percent among their brothers. These data also show that the percentage with baccalaureate degrees is not as low as Huntington's data indicate. As a consequence of these corrections, the Mexican-American gap in college education falls from more than 4-to-1 to about 2.5-to-1. The corrections are also sizable for the lowest category, the percentage without a high school diploma, which the Latino survey data appear to inflate.
Educational Attainment, 1971–1975 Birth Cohort
Still, the disparity between Anglos and Mexican Americans is large and calls out for some interpretation. Others have made the same comparison that Huntington has and drawn the conclusion that a stagnation sets in with the Mexican-American third generation.3
This appears to support the characterization of Mexican Americans as a group at risk of what Portes and Zhou have called “segmented” or “downward” assimilation: a failure to advance beyond the humble status of immigrant or ethnic parents, which is then transformed into a negative self-evaluation because of a more thorough absorption of mainstream standards of social ranking.4 However, an important question concerns how Mexican Americans do when compared to their parents; this is the concern at the heart of an intergenerational comparison, for one assumes that in comparing, say, the third to the second generation, one is comparing (roughly) children to their parents.As some have noted, the long history of Mexican immigration to the U.S., which goes back more than a century, and the institutional discrimination from which Mexican Americans suffered up through the civil-rights era threaten the validity of conclusions from intergenerational comparisons of the sort used by Huntington, which are typically based on cross-sectional data.5
Perlmann 2005; Smith, 2003.
The deeper, more revealing problem, however, is that of the institutional discrimination faced by Mexican Americans in areas where they were concentrated, which limited their ability to move ahead educationally and economically before the civil-rights era. In states like Texas, Mexicans were confined to separate and underfunded schools into the second half of the twentieth century. The negative impact of this discrimination is quite visible in a comparison of the educational attainments of different birth cohorts of Mexican Americans: the cohorts born in the U.S. before mid-century have substantially lower years of education than those born later.6
See Alba et al., 2004.
A more suitable way of examining the Mexican-American educational trajectory is therefore to compare children to their parents. James Smith has contrived an indirect way of doing this, by simultaneously taking birth cohort and generation into account.8
He find that when cross-generational comparisons also involve birth cohorts separated by 30 years, a rough gauge of the age gap between parents and their children, then a pattern of steady intergenerational progress is revealed. Even better for diagnostic reasons is a direct comparison, which is possible with a limited number of data sets, such as the National Longitudinal Study of Youth (NLSY) of 1979.9Alba et al., 2004.
Yet a gap remains and shows no clear sign of closing. If anything, it may be growing because of improvements in the educational attainments of recent cohorts of non-Hispanic whites. As revealed by census data by birth cohort, the high-school dropout rates of second- and third-generation Mexican Americans continue to be more than twice as high as those of their Anglo counterparts, and the discrepancy in college attendance and graduation rates has grown somewhat. There is something of a paradox in these results, one that calls out for finer analysis in order to resolve it. Each generation of Mexican Americans makes a greater advance beyond the education of its parents than do their Anglo peers and yet fails to close the educational gap. How can this be possible? I will return to this question at the end.
Language Assimilation
Much of the threat that Huntington perceives in the Hispanic challenge is linked to the notion that Hispanics, and especially Mexicans, could eventually establish a Spanish-language-based subsociety on U.S. soil. This subsociety could come to monopolize power and economic opportunity in some regions of the U.S. and thus create disadvantages for ordinary, English-monolingual Americans; it could even, in those regions taken from Mexico in the mid-nineteenth century, serve as the demographic springboard for revanchist claims. Huntington writes, for instance, that “Mexican immigration is leading toward the demographic reconquista of areas Americans took from Mexico by force in the 1830s and 1840s,” and describes them as being “Mexicanized.”10
Huntington 2004b, 221.
Such assertions overlook the strength of language assimilation among Hispanics, Mexicans included; and they fail to provide a sociological analysis of the conditions necessary to create the outcomes he fears.
It does appear to be the case that the language assimilation of Hispanics does not match the pattern of mother-tongue extinction within three generations that was evident among most European-American groups, though even there important exceptions existed, such as Germans in the Midwest, who created bilingual public school systems that functioned up to World War I, and French Canadians in New England. Hispanic groups show much higher rates of bilingualism among second-generation adults than the European groups of the last great wave (1880–1925) generally did, and bilingualism persists among a minority in the third and later generations. Nevertheless, the abundant data about language practices among Hispanics demonstrate unequivocally that 1) with rare exceptions, U.S.-born Hispanics speak English well, as do the majority of immigrants who have lived in the U.S. for 10 years; 2) about half of the second generation is English dominant; and 3) by the third generation English dominance, if not monolingualism, is the prevalent pattern. The seemingly high rates of Spanish use among Hispanics today are due mainly to very high rates of recent immigration: in 2000, the foreign-born made up 40 percent of the entire Hispanic population. These facts do not lay the basis for a separate Spanish-language subsociety.
For instance, according to the 2002 survey of the Pew Hispanic Center, nearly half of the second generation is English dominant, and nearly half is bilingual, when the definitions of language proficiency entail both speaking and reading.11
(Speaking ability alone is not a good indicator of language proficiency because many individuals have oral competencies developed in family and private contexts that, because of restricted vocabularies and range of expression, do not give them facility with a language in other settings.) Only a small percentage of the second generation (7 percent) is scored as Spanish dominant. In the third and later generations, more than three-quarters is English dominant, and less than a quarter bilingual; Spanish dominance is no longer a significant pattern.In fact, much of the third and later generations has at best marginal competence in Spanish, learned now more in school than at home. Two studies in census data of parents' reports of their children's home languages reveal that the majority of Mexican-American children in these generations speak only English: in 1990, this was true of nearly two-thirds; a decade later, it was true of 70 percent.12
This increase in assimilation, remarkably, occurred despite a surge of Mexican immigration during the 1990s, which might have added to the viability of Spanish in later generations. It seems likely, moreover, that parental reports overstate bilingualism, crediting this to children who speak snatches of a mother tongue with parents and other adult relatives.13Lopez, 1996.
These language shifts are revealed in a variety of ways. For instance, a Washington Post/Kaiser Foundation survey of Hispanics at the end of the 1990s found that two-thirds of the second generation watched mainly English-language television programs (compared to only a quarter of the immigrant generation); in the third and later generation, the fraction rose to about 90 percent.14
Goldstein and Suro, 2000.
Mexican-American Hegemony?
Huntington raises the prospect that Mexican Americans will gradually, “from below,” take over the economic and political structures of the Southwest, California, and Texas, to form a “Republica del Norte.” The prototype he has in mind is the “Hispanization of Miami,” his term for the ascendance of Cuban immigrants and Cuban Americans to create a city that fuses North American and Latin American societies and cultures. The consequences of such a Mexican-American ascendancy would include disadvantages for Anglo Americans, who could taste the sourness of the minority experience in their own country, and the creation of what amounts to a quasi-separate nation, a Quebec, in the United States.
Huntington recognizes that the processes of Mexican-American ascent to domination, if they occur, will have to look quite different from those of the Cubans, whose initial immigrant waves contained the elite of the society overthrown by the revolution. These elites entered a stagnating American city, brought some financial capital and the knowledge of how to get things done, formed a coherent social stratum based in part on preexisting ties, and benefited from a refugee resettlement program that was unusually generous by U.S. standards.15
Portes and Stepick, 1993.
How, then, would Mexican Americans achieve dominance in states like Arizona, California, and Texas? Huntington rests his claim on the notion of demography as destiny: “In the long run, however, numbers are power, particularly in a multicultural society, a political democracy, and a consumer economy.”16
Huntington, 2004b: 253.
Alba and Nee argue that much assimilation is incidental to the everyday decisions of individuals and families to improve their lives materially and socially: assimilation occurs while people are making other plans, so to speak.17
Alba and Nee 2003; see also Bean and Stevens, 2003; Clark, 2003.
Mexican Americans, because of their low starting position in American society, are likely to prove a case in point. Thus, the increasing dominance of English over Spanish in the U.S.-born generations testifies partly to their perceptions of where the superior opportunities lie, in the English-speaking mainstream versus the Spanish-dominated part of the labor market. Mexican-American families, like the ethnics of other immigrations but unlike African Americans, who are restricted by more severe racial discrimination, tend to leave Mexican-dominated residential areas as their economic position improves and they acculturate linguistically.18
There is a racial element here, however: the residential situations of Mexican Americans who call themselves “white” on the U.S. Census are somewhat better overall than are those of their co-ethnics who view themselves as non-white.Perhaps the bottom line on separation is delivered by Gregory Rodriguez (2004), who points out that Mexican Americans simply have not attempted to build a parallel, ethnic institutional structure:
Nor have Mexican Americans ever shown much interest in distancing themselves from the mainstream by building parallel ethnic institutions. For example, in Los Angeles, home to more Mexicans than any other city in the U.S., there is not one ethnic Mexican hospital, college, cemetery or broad-based charity.19
Without a separate set of institutions, Mexican Americans of necessity depend on institutions shared with other Americans to satisfy a wide range of everyday needs.
Two Assumptions
Huntington's analysis shares two assumptions with the bulk of the literature on contemporary immigration that are questionable at best. The first is that there will be no significant abatement in immigration from Mexico for several decades at least; the other, that the social cleavage between Mexicans and the mainstream is so deep and difficult to cross that Mexican Americans remain permanently a part of the group and identifiable as such.
In reflecting on the assumption about the continuing supply of immigrants, one should remember that the number of legal and illegal Mexican immigrants has risen sharply in recent decades and that many of the social and cultural phenomena associated today with Mexican immigrants and Mexican Americans reflect this surge. In 1980, the number of Mexican-born residents of the U.S. was just 2.2 million; by 1990, it had doubled to 4.3 million; and by 2000, it had doubled again, to 8.7 million. The accelerating pace of Mexican immigration cannot be kept up for much longer, and the demographic picture of Mexican society raises the possibility that there could even be some fall-off within the coming two decades. The youthful population is the key to migration trends because labor-force migration is almost entirely a matter of young adults, who can anticipate working long enough to compensate for the initial costs and dislocations of moving across national borders. The 10–14 age group in Mexico is a good harbinger because it is too young to have been directly reduced much by migration but it is within a decade of the ages when migration rates are high. According to the Mexican population projections for 2000–2050 by Consejo Nacional de Población (CONAPO),20
the size of this age group is currently at an all-time high (11.4 million strong in 2005), but will start to slip later in the decade and continue to slide for the foreseeable future. Forecast at 10.6 million by 2010, it is projected to decline to 9.1 million 10 years later. Should the full magnitude of these declines be realized, they would mean substantially fewer labor-market entrants to be absorbed by the Mexican economy. By itself, the demographic changes do not guarantee a decline in migration to the U.S., but they set up the preconditions for a decline, which can then occur if economic opportunities for young adults improve in Mexico. They make, however, any assumption that the recent pattern of steadily rising immigration can continue utterly implausible. In the Mexican-origin population, the immigrant fraction is destined to decline in the future, and thus the language and other characteristics of immigrants will become less prevalent.Another assumption that runs throughout the literature on Hispanics is that there is great consistency in self-identification as Hispanic and as Mexican: thus, we can meaningfully compare groups so designated over time because they are consistently defined and affected only by the core demographic processes of fertility, mortality, and migration. Intergenerational comparisons in effect assume this and are not meaningful if there is, say, selective departure from the group in the third and fourth generations.
There is, in fact, departure from the group, i.e., a significant number of individuals who apparently have ceased to call themselves Mexican or Hispanic. The evidence for this claim comes from comparisons of birth cohorts of U.S.-born Mexican Americans across U.S. censuses. Thus, if one compares such cohorts in 1980, as counted by the Hispanic origin question, to their equivalents in 2000 (e.g., the 10–14 age group in 1980 translates into the 30–34 year-old group in 2000), the numbers typically decline by 10–12 percent compared to what should be expected based on patterns for all U.S.-born Americans, which factor in mortality differentials by age. This reduction comes about because of identity shifts, reflected in changes in the way people answer the Hispanic-origin question on the census: insofar as the changes can be traced, they appear to be complex, with some people migrating into pan-Hispanic categories, while others embrace more mainstream identities.21
Alba and Islam 2005; see also Jiménez 2004.
Conclusion
That Mexican Americans are not isolating themselves from the mainstream society and striving to establish a separate nation within the U.S. does not mean that their incorporation is problem free; in this respect, Huntington's account broaches critical issues. At least three prominent difficulties confront many Mexican immigrants and U.S.-born Mexican Americans and imply that assimilation is not the only pattern of incorporation relevant to the group's experience. Many Mexican Americans have very humble origins because of the marginal economic situation of immigrant parents. They are entering an economy where the “good” jobs in the middle, i.e., stable, well-paying jobs requiring less education than a college degree, are getting harder to find, and where therefore the need for educational credentials beyond the high school diploma is intensifying.23
At the same time, a sizable minority fail even to attain that diploma. There can be little doubt, then, that many Mexican Americans, but not the majority of the U.S.-born, are condemned to the low-wage sectors of the economy.Racial discrimination also poses problems for a substantial part of the Mexican-American population. Mexicans are of course a racially diverse group, containing a large proportion of racially mixed individuals, whose appearance shows traces of Indian and European ancestries with perhaps some African thrown in, and also a large number of individuals of Indian phenotype. The racial discrimination to which Mexicans with Indian features and skin coloring are exposed is not as well documented as that confronting African Americans.24
Perhaps it is less rigid; we cannot say because the research has not been done. But there is no reason to think it does not exist: as David Lopez and Ricardo Stanton-Salazar note, “those who fit the mestizo/Indian phenotype, who ‘look Mexican,’ cannot escape racial stereotyping any more than African Americans, though the stigma is usually not so severe.”25Finally, the issue of legal status is of great moment for many Mexican immigrants and even for Mexican Americans. Undocumented status is very widespread among recent immigrants from Mexico, shared by perhaps as many as half of them. Lack of legal status drives immigrant parents into a social and economic underground, where they are fearful to insist on the rights that legal residents and citizens see as their due, and forces many of them to work at jobs that are exploitative in terms of pay and benefits, security of employment, and working conditions. We have no systematic evidence yet about how lack of legal status intrudes, directly or indirectly, on the second generation, whose members are U.S. citizens by virtue of birth but grow up in households where parents and perhaps older siblings must live with the associated uncertainties; but an impact must be suspected until proven otherwise. Moreover, because American immigration policies have induced a settlement process, even for the undocumented, many undocumented Mexican children are now growing up in the U.S.26
They are being educated in American schools and placed in a social limbo as a consequence: raised as Americans, they cannot easily go back to Mexico; but without legal status, their U.S. educations are of no significance in the U.S. labor market. Their situation is one of the quiet tragedies of the contemporary immigration scene. It is one that cries out for a solution through enlightened policy. The proposed Dream Act, which addresses the situations of college students and is still awaiting passage by Congress, is not sufficient: why should any young person who has attended primary and secondary schools in the U.S. be denied the right to live and work here?Instead of these systemic blockages lodged in poor chances for educational mobility, ethnic and racial discrimination, and legal status, Huntington's argument identifies cultural and social isolation as the key impediment preventing some Mexican Americans from striving to enter the mainstream. The danger of this argument is that it will stimulate a nativist insistence on the need to Americanize new immigrant populations.27
Etzioni, 2005.
A wise policy of assimilation, like a wise educational policy, does not seek to destroy the attitudes and memories that are there, but to build on them. There is a current opinion in America, of the “ordering and forbidding” type, demanding from the immigrant a quick and complete Americanization through the suppression and repudiation of all the signs that distinguish him from us.28
Thomas, Park, and Miller 1971, 280–1.
Over the long haul, the best inducement to assimilation has been the lure of the opportunities available in the mainstream. Of course, that lure is effective only if immigrants and their descendants perceive that opportunities are available to them. Thus, reducing the barriers to opportunities for Mexican Americans is still the wise policy.
