February 1999 Volume 5 Number 4
The Mexican-American Second Generation: Yesterday, Today...and Tomorrow -- David E. Lopez
DRAFT DO NOT CITE WITHOUT PERMISSION OF AUTHORS DRAFT
Yesterday, Today...and Tomorrow
David E. Lopez
University of California at Los Angeles
Ricardo D. Stanton-Salazar
University of California at San Diego
To Appear in , Edited by Ruben Rumbaut and Alejandro Portes, University of California Press, forthcoming
The Mexican American "new" second generation shares many points in common with the other cases included in this volume. They display much the same complex pattern of partial acculturation and ambivalent ethnic identity that is typical of other new second generation youth. They are socially defined as "non-white," though race is for them a source of confusion and ambivalence. They feel caught between, on the one hand, the demands of their parents, who are struggling to build new lives in the United States, and on the other hand, their own struggles to combat what they perceive as a hostile environment and their need to construct a new identity that will allow them to face that environment with confidence.
But in other ways the Mexican-American case stands apart. They and their parents lack many of the resources that have allowed other recent groups of newcomers to the United States to thrive. The majority of Mexican immigrants come from the humblest sectors of Mexican society On average they have only a few years of schooling, limited urban job skills and little or no knowledge of English. Immigrant Mexican communities are by no means "disorganized" in the old-fashioned sociological sense of the term; two-parent households and high levels of labor force participation are the norm. But in comparison to many other immigrant communities, they lack the web of organizations and practices that have allowed those groups to utilize traditional culture to help children achieve. The children of Mexican immigrants generally do poorly in school and their occupational prospects are bleak. Their socioeconomic disadvantages and dismal school performance are particularly striking in California, where most other contemporary immigrant groups are notable for just the opposite.
If Mexican American youth were just another in the vast array of new second generation groups, there would be only modest cause for concern regarding their below average achievement and future prospects. It would be nice if all of America's ethnic groups and all its children were, as in Lake Wobegon, above average; in the real world some have to fall below the mean. But in California and the Southwest, Mexican Americans are not "just another" immigrant-origin ethnic group. They are instead by far the largest "minority" and are rapidly becoming the single largest ethnic group, destined to outnumber whites sometime in the next century. Mexican Americans and other Latinos today constitute the single largest ethnic group among California's school children, and four out of five are the children of immigrants. Mexican American youth are also much more likely to be in blue collar/working poor families than whites, Asians or even African Americans. It would be no exaggeration, then, to say that the Mexican American second generation is on its way to becoming the single largest segment of California's population, when it is broken down along the three dimensions of ethnicity, generation and class.
In this chapter we tell the story of the "new" Mexican American second generation, drawing from Census data and other research as well as from the CILS surveys of Mexican American youth in San Diego. As concerned as we are with understanding the challenges facing adolescents growing up second generation and the vagaries of their ethnic identity, our primary goal is to understand the trajectory of socioeconomic adaptation and advancement for this rapidly-growing segment of the population. Our focus is on California, since the Mexican American CILS sample comes from San Diego, and because immigration and the children that are its result are more immediately consequential for California than for the Southwest or the rest of the United States. We begin with a discussion of the unique importance of the Mexican case and some of the difficulties involved in comparing it with other "new" second generation groups. We then provide a description of the San Diego Mexican-American community and the sample of Mexican American school children in San Diego that provides much of our data. In Part III we turn to the thorny question of race and felt discrimination among Mexican Americans, which we consider to be a key issue that is usually over-simplified. Part IV reviews the socioeconomic status and accomplishments of second generation Mexican Americans, from the school and labor force accomplishments of an earlier second generation, through the school performance of second generation youth in San Diego and California today.
I. The Unique Importance of the Mexican Case
In addition to poverty and sheer numbers, two additional important characteristics set the Mexican-American case apart: historical depth and racial ambiguity and negative stereotypes. Today's is by no means the first Mexican-American second generation. The very roots of California and the Southwest are of course Mexican, and Mexicans have continuously inhabited the region for hundreds of years. In the early part of this century these tiny traditional Mexican/Hispanic populations were overwhelmed by migration from Mexico, migration that produced its own "second generation" that came of age forty years ago. Indeed, a discernable third generation was coming of age when large-scale immigration began again in the 1960s.
It is instructive to compare the age/generation profiles of California's Mexican American population in 1960 and 1996 (Table I). 1960 was the low point in the proportion foreign-born, and the high point in the proportion "native of native." The compositional differences are equally striking: in 1960 only 20 percent of the state's Latino population was foreign-born, and the median age of the foreign-born was 42; only 5 percent were school-age children and immigrants constituted two-thirds of the aged. The second generation was one-third of the total Latino population, but nearly half of those in their prime working years, 25-44; it was the core of working-age adults. Nearly half of all California Latinos were native-of-native, overwhelmingly third generation in California; but they were mostly children, and constituted two-thirds of all Latinos under the age of 16. By 1996 these patterns had been reversed. Immigrants are a much larger proportion of the total, but the sharpest contrasts come when the data are organized by age band: fully 69 percent of those 25-44 are immigrants, and two-thirds of the children are second generation, which has an average age of 10. In 1960 it made some sense to think of the generational categories as actual generations, as demonstrated by the median ages of each group. In 1996 California's Latino population is actually an overlay of two quite different groups: a numerically predominant recent and continuing immigrant population, which provides most of the workers and most of the school children, and an aging population whose immigrant origins tracing back 70 years into the past.
The pre-existing Mexican-American community serves as an important part of the "context of reception" for new immigrants, but beyond this it is important in understanding today's second generation for two reasons. First, their social and economic attainments provide a rough guide to the fate of today's second generation, and perhaps tomorrow's third. The Mexican socioeconomic origins and context of reception for Mexican immigrants today are not all that dissimilar to what immigrants from Mexico faced 80 years ago. Economic conditions and attitudes about race have certainly changed, but these changes are complex, and there is no reason to believe that the net effect of these changes is to make social assimilation and success easier today for Mexican Americans than in the past. Second, the status of the pre-existing Mexican American population inevitably determines the expectations and stereotypes that others apply to this new generation of Mexicans in the United States. The largely negative and highly racialized stereotypes associated with Mexican Americans provide an additional and heavy burden for the new second generation.
Another important difference from other immigrant groups today is racial ambiguity and the negative stereotypes attached to being Mexican in America: by official statistics Mexicans and other Latinos are white, black, Asian or Native American, but in practice "Latino" and especially "Mexican" serve as quasi-racial terms. However ambiguous race may be on the individual level for Mexican Americans, on the group level Mexicans are very much a stigmatized and despised group in California and the Southwest. Just as Mexican immigrants early in this century inherited the traditional status of Mexicans in the United States, so today's immigrants and their children are inheriting the inner and outer burdens of Mexican "color" as they developed throughout the twentieth century. We believe that these four distinctive characteristics__poverty, group size, historical depth and racial stereotypes__interact to create special barriers to upward mobility for Mexican American youth, in such a way that their school performance and socioeconomic trajectories cannot be simply analyzed by individual-level quantitative measures.
Thirty years ago, in their massive study The Mexican-American People, Grebler, Moore and Guzman (1970) argued that the future of Mexican Americans might go in either of two directions: they might proceed along a somewhat delayed path of assimilation and increased economic equity, like Italians and other Euro-American ethnic groups; or they might share the caste-like fate of African Americans. The answer was to be found in the future, as third and fourth generation Mexican Americans became the majority of the group's population. What Grebler and his associates did not foresee was the great revival of immigration that has taken place in the past 35 years. The Mexican-American People appeared just as the "Old" second generation was hitting its peak earning years, and their children were moving into young adulthood. As we shall see below, their mixed success, under economic conditions generally considered better than today, does not support optimism about the economic futures of today's emerging "New" second Generation.
Much of our description and analysis revolves around the question of low academic achievement and its long-run consequences for Mexican Americans and for California. We believe that this is one of the gravest policy questions facing the state and the Mexican American/Latino community at the dawn of the millennium. But in the historical context of immigration to the United States it is not a particularly interesting sociological question. We are not shocked and amazed about dysmal academic performance by children of Polish peasants and Italian farm laborers a century ago. Why expect the children of Mexican campesinos to do any differently? They arrive with the same meager resources as Polish peasants and Italian agricultural laborers, as well as the added burden of being socially non-white and identified with oppressed and low-achieving pre-existing Mexican American population.
Most immigrant groups in the past took at least three generations to reach socioeconomic parity with older ethnic stocks. Low second generation achievement was the norm; it is the rare high achieving second generations that require explanation. The success of groups like Jews has been explained in terms of the interaction between their backgrounds and their specific "context of reception." Conversely, groups like Poles and Italians lacked those special characteristics and conditions, but this was the normal state of affairs for most immigrants; there was no reason for special theories to explain why they and their children had to work their way up slowly, generation by generation. As was the case with Jewish second generation success, there is also a lively debate about the causal significance of specific cultural traits above and beyond the clear socioeconomic advantages of today's Asian high-achieving second generation groups. This debate is fascinating, but not really relevant to understanding low achievement among the Mexican second generation, which appears to be following much the same pattern as European groups in the past.
To reiterate, the Mexican example reflects the normal run of things in American immigration history: the children of peasants who migrate get the minimal education offered by their new homeland, and find jobs somewhere between the dirty work their parents did and the "careers" of more privileged groups. Their parents found work, hard and ill-paid, to be sure, but better-paid than what they left in Mexico; and if their children's education and occupational prospects may be modest by U.S. standards, they are vastly superior to what they could have expected in Mexico. In relative terms, there may be more upward mobility among the Mexican American second generation than in the case of a Taiwanese engineer whose son goes to Stanford, or a Korean college graduate whose daughter goes to Harvard. There is, therefore, no reason to invoke cultural values to explain the modest success of the current Mexican-American second generation in schools. If the burden of enduring racial stereotypes is added to their socioeconomic disadvantages, then low Mexican-American achievement levels become all the more comprehensible.
Why, then, is there such an obsession about "explaining" Mexican American "low achievement," an obsession that we admittedly share? We see three reasons:
first, today's Mexican-American second generation happens to be growing up alongside the truly unusual Asian American second generation groups, among whom even many poor children do well in school, and they are inevitably compared unfavorably with them. Even if achievement differences between Asians and Latinos could be explained by statistically controlling socioeconomic factors (and it appears that they cannot), the general public cares little for multivariate analysis, and focuses on the group differences, not their causes;
second, the previous conditions that faciliated second and then third generation mobility in the past, particularly the availability of good, unionized manual jobs that do not require high levels of formal education, simply do not exist today. In contrast to the increasing industrial productivity of the 1940s and 1950s, we live today in an increasingly service-oriented economy, with an increasingly bifurcated job market. These are sociological platitudes, of course, but what is not always made clear is that this structural change is really a problem only for those groups entering the U.S. job market at the bottom, and Mexicans are the only large immigrant group that does so today. The emerging Mexican American second generation will not be satisfied with jobs on the bottom, but will have difficulty competing for jobs on the top, and it is not clear that they will be satisfied with whatever remains in-between;
third, there is substantial evidence that the older Mexican American third generation did not experience the same levels of upward mobility and social assimilation as that experienced by Italians and other European ethnics, but rather seems stuck near the bottom, along with African Americans (Grebler, et al., 1970; Ortiz, 1996). As with African Americans, a small middle-class segment is doing just fine, but a substantial portion seems to be left behind, albeit not in quite the dire straits of the black "urban underclass." This suggests that today's Mexican-American "new" second generation may not serve as a transition between their parents and their fully integrated and assimilated children, but rather represent the transition from a permanently disadvantaged minority to a permanently disadvantaged majority in California and the Southwest.
While we stress the uniqueness of the Mexican American case, we believe that it fits neatly within the general "segmented assimilation" theoretical framework of the CILS project that frames the potential for success in terms of background resources, labor market conditions and fit, and the general political/social context of reception (Portes and Rumbaut, Legacies, ch. 3). The Mexican-American case demonstrates the importance of cultural and material capital, or rather their absence. Today, as throughout the twentieth century, labor market conditions have played an important role in both attracting Mexican labor, and also in determining the social fates of Mexican Americans raised in the United States. Even more so does it underline the importance of the cultural as well as material dimensions of the "context of reception," into which immigrants arrive, which in turn sets the stage for the world in which their children grow up. In the case of Mexican Americans this includes expectations based on historically-rooted "racial" stereotypes that are more complex than white attitudes towards African Americans or Asians, but no less consequential.
II. San Diego: Setting and Sample
San Diego County has the third largest concentration of Latinos in California, following Greater Los Angeles and the San Francisco Bay Area. According to the 1990 Census, one in five residents in the County were of Latino origin, totalling a bit more than a half a million people. About 86% of these Latino residents are of Mexican-origin. While San Diego has only half the Latino proportion of Los Angeles, and only five percent of the state's total Latino population, its Mexican American population is similar to that of the entire state in both history and socioeconomic makeup.
San Diego County has historically served as the principal gateway for migration from Mexico and Latin America into the United States. In fact, the San Diego-Tijuana border is recognized as the busiest international border in the world. The cities of San Diego and Tijuana, whose downtown centers are only 19 miles from each other, together constitute one of the major trading and industrial centers of the Pacific Rim. The San Diego/Tijuana region is also one of the principal population and industrial centers along the U.S.-Mexican border.
Like California as a whole, San Diego's Mexican-origin population grew rapidly as a result of emigrants fleeing the economic and political turmoil of the Mexican Revolution. The city and smaller municipalities stretching to the border have well-established Mexican neighborhoods, most low and modest income, but including some traditional middle-class areas established by better-off segments of the refugees fleeing the Revolution.
Despite, or perhaps because of its proximity to Mexico, Mexican Americans have never felt particularly welcome in San Diego, and the white elite and electorate have managed to keep them marginalized economically and politically throughout the twentieth century. The resurgence of immigration from Mexico has increased white/Mexican tensions. It is along the border just south of San Diego that homeowners groups flooded crossing areas with their headlights, not to aid migrants but rather to force the border patrol to arrest them. And before becoming the most notoriously anti-immigrant Governor in California's recent history, Pete Wilson was the mayor of San Diego.
As in other major California cities, San Diego's economy is increasingly dependent upon Latino workers. Latino labor participation in the county jumped from 13 percent in 1980 to 20 percent in 1990 and will probably surpass 25 percent by the year 2000. While the 1980s saw Latinos heavily concentrated in the manufactoring sector, in the 1990s Latinos are more likely to be involved in the construction, trade, and services industries. As in the state as a whole, the Latino workforce is increasingly composed of immigrants, most of whom have only a few years of schooling and poor English skills. The Latino population in the county is also substantially younger than the region's total population. While the median age for the general population was 30.9 in 1990, for Latinos it was 24.3 years. Latinos also accounted for 29.3% of the regions population under 5 years of age, and 29.1% of the school-age population (age 5-17).
The growing presence of Latino youth is most marked within the public school system, where demographic shifts have been remarkable. In 1977, white students constituted 71 percent of the county's student enrollment (ages 5-17); by 1989, this percentage had decreased to 54 percent, and by the time of the CILS last survey in 1996, whites constituted less than 50% of the K-12 population. Latino students, representing by far the largest segment of the new majority, constituted about 27 percent of the county's student body. This is somewhat less than the statewide Latino presence (40 percent), and much less than the overwhelming Latino presence in Los Angeles and other highly impacted municipalities.
Latinos are a large and rapidly-growing part of San Diego's public schools, but they are still a numerical minority. Forty-three percent of the Latino student body in San Diego in 1990 was classified as Limited English Proficient (LEP) with a bit less than a third classified as a bilingual, yet Fluent English Proficient (FEP). Another 28% are classified as English dominant. The school system has had to respond to this growing presence; for example, in 1990, San Diego County employed 875 teachers with bilingual credentials; yet, this number only accounted for a bit less than half of the regions need for bilingual and credentialed educators (San Diego Association of Goverments, 1991).
The youth in our sample have many characteristics in common, in part in order to meet the sampling criteria. They are hardly all peas in a pod; there is some diversity on a number of dimensions: parental socioeconomic status, birthplace, languange preference, choice of identity and school achievement. Internal diversity allows us to assess the interrelations among these variables within the sample, with the goal of de-constructing the "ethnic" effect of being Mexican American. Internal diversity is especially important when making comparisons with other ethnic groups, and trying to assess the degree to which ethnic differences are due to socioeconomic factors. Unfortunately our sample is a comparatively homogeneous one, and the diversity and level of correlation among variables is modest...far from sufficient to allow for the explanation of low achievement by socioeconomic and demographic factors. At least by the conventional sociological measures we use here, homogeneity, not diversity, is the dominant theme. Still, there is some diversity, and it does correlate with achievement to some degree.
Table II provides a demographic sketch of the sample. Seven hundred and twenty-seven Mexican-American youth were interviewed in 1992, and researchers were able to re-interview 578 in 1995-96. This 80 percent re-interview rate is quite good by survey research standards, though it is the lowest of all the ethnic group samples in San Diego. All statistics given here apply only to the remaining 578 cases.
At first glance the sample appears to have considerable generational diversity, with 61 percent U.S. born and 39 percent born in Mexico. However in reality there is little sociological diversity here: 79 percent have been in the U.S. all their lives or at least since the age of 5, and most of the remaining fifth arrived between the ages of 6 and 12. Sampling criteria of necessity limit the degree to which respondents might legitimately be considered immigrants. Only those who had already been in the U.S. for several years before 1992 were eligible, so even most of the 11 percent who did arrive after the age of 10 were only 11 or 12 when they entered the United States. This is, after all, a study of the second generation, children who are the children of immigrants but are growing up in the United States. Certainly it makes little sense to distinguish between the native-born and those who arrived before the age of 5. Those who arrive between the ages of 6 and 12 are progressively more subject to the set of advantages and disadvantages of growing up in Mexico instead of growing up Mexican in San Diego. But, except for obvious matters such as a firmer grip on Spanish, and a greater probability of accented English, among this sample it is difficult to attribute any lasting effect to age at immigration, especially since our data come from adolescents, not adults. In this sample of adolescents age at immigration per se is particularly difficult to interpret since, of course, it is so correlated with the length of time that one's parents have been in the United States, which is a major determinant of economic advancement for adult immigrants.
Wherever they are born, most children of immigrants to the United States begin life speaking their ethnic mother tongue, but then progressively adopt English and loose their propensity and ability to speak the language of their parents. It has been well documented that this linguistic shift is slower for Mexican American youth today than it is for the children of Asian immigrants, but the difference is one of degree, and it is difficult to say if Spanish is being maintained more tenaciously than European immigrant languages early in this century. That said, the vitality of Spanish as a used language in San Diego and Southern California generally is incontrovertable: immigrants, who make up half the workforce in many industries, often speak Spanish as their only language; their children are mostly bilingual, with varying degrees of competence in each language, and a significant minority of the Mexican-American/Latino third generation continues to use Spanish, if only as an interactional symbol of ethnic solidarity (Portes and Hao, 1998; Lopez, 1996).
As high school students in San Diego, all of our sample speak English to some degree, and over half report that they speak it very well. As the children of Mexican immigrants, almost all speak Spanish to some degree (the principal exceptions being those with one or more parents who do not speak Spanish regularly) and two-thirds say that they speak it very well. However in 1995, 78 percent of the U.S.-born and 63 percent of the Mexico-born declare that they now prefer to speak English, substantial increases from 1992 for both groups. High as these rates are, they are lower than the Asian subgroups also included in the San Diego survey.
All reports on language preference, use and ability are intrinsically fuzzy measures, but there is no question that this is a population undergoing linguistic change.
Only time will tell if this second generation will choose to pass Spanish on to their children, but the results of this study, which are well supported by other research, strongly suggest that this is a population undergoing transitional bilingualism. On the other hand a solid majority does report that they can speak Spanish very well, suggesting that at least the potential for language maintenance is present. Unlike other states in the region with high concentrations of Mexican-origin youth, the Latino youth scene in California, and specifically southern California, is largely defined by children and youth growing up in the U.S. with immigrant parents. They, as well as many "native of native" Latino children grow up in barrios where the cultural and linguistic milieu is characterized by the clash and constant negotiation between youth subcultures and Mexican-Latino immigrant culture, between the forces of youth identity formation and the nostalgic re-creations of immigrant adults. Over the last 15 years, however, the apparent Mexicanization of barrios across the state has served to infuse a new Mexicanist ethnic identity into contemporary Mexican American youth culture, which we believe will continue to be supportive of the Spanish language.
As with generation and language, this sample of Mexican American youth is also quite homogeneous with respect to their parents economic status. Data about their parents education and jobs that are filtered through the eyes of adolescents is certain to be only approximate, but it does seem to roughly correspond what we know of the educational and occupational profile of Mexican immigrants. Only 10-15 percent of the parents are reported to have any college education at all, and the median years of schooling is about eight. A miniscule 2 percent of their fathers have better white collar jobs (managerial/professional/ technical) and a bare 10 percent have lower white collar jobs (sales and clerical). The vast majority are classified as "wage workers" or blue collar workers. The mothers who are in the workforce follow much the same pattern. But of course immigrants have traditionally done the dirty work and, as emphasized above, Mexican immigrants follow the "classic" immigrant pattern much more than Asian immigrants today.
As Hayes-Bautista (1988) and others have emphasized, economic success and assimilation for Mexican immigrants is found not through education, the professions or even extraordinary rates of entrepreneurship, but rather through stable families acting collectively to achieve economic goals. In this sample this is reflected in rates of 2-parent households and home ownership that are high in comparison to African Americans, or Puerto Ricans in the East, though lower than average for the other immigrant groups. Family size, here represented by the percent in smaller families, is close to the norm for other groups included in the study, as are labor force participation rates, which are somewhat below average. Perhaps the most significant measure is the remarkable increase in home ownership in the brief three years separating the two surveys, up 20 percent for households of U.S.-born children and a remarkable 50 percent for the households of immigrant children.
In sum, data from the San Diego survey reflect other studies of Mexican immigrant households: in comparison to African Americans and Puerto Ricans, levels of two-parent families, home ownership, evidence of familial cooperation and other "positive" characteristics are comparatively high, but they do not stand out when compared with other immigrant groups. Put another way, there is no evidence that the social and cultural characteristics of Mexican immigrant households give them an advantage when compared to other immigrants, and therefore there is no reason to believe that they can compensate for the economic and educational disadvantages typical of Mexican immigrant households.
Adolescence is typically a time of identity confusion and shift, along the dimensions of ethnicity, gender and social status. Understandably, ethnic identity is particularly salient to these children of immigrants. In 1992 the most popular terms of identity were panethnic (Latino, Hispanic, Chicano): 45 percent for the U.S. born, 51 percent for those born in Mexico. In 1992 AND 1995 40 percent of the U.S. born labeled themselves as Mexican-American, but only 12-15 percent of those born in Mexico used this term at either point in time; THEIR preferred identity was Latino/Hispanic in 1992, but in 1995 they largely abandoned these pan-ethnic terms in favor of some varient of the straight-forward term "Mexican." Only 2 percent of the U.S.-born and none of the immigrants had adopted non-ethnic terms like "American" in 1992 or 1995.
When identity in 1992 is cross-classified with identity in 1995, the picture gets even more complex. In Table III we have combined the two nativity groups, and separated the term "Chicano" from the pan-ethnic terms, since we believe it is better seen as a nationalistic political identity. Table III shows that the principal shift has been from pan-ethnic terms to "Mexican" and, to a lesser degree, to "Mexican American." Those previously using the term "Chicano" have also shifted substantially to these categories. In contrast, those who identified as "Mexican" or "Mexican-American" in 1992 tended to stick with these terms.
Overall, there is a clear shift from pan-ethnic to national "Mexican" identity among immigrants in the sample, and a secondary shift from Mexican-American to Mexican, largely among the native-born. There is also a substantial, and somewhat puzzling decline in the use of the term "Chicano," a term that expressed the political aspirations of second and third generation youth in the 1970s. Those abandoning the term moved in all directions, to Mexican, Mexican-American and Hispanic, suggesting a plain decline in the term's popularity rather than any meaningful shift.
It is probably dangerous to attribute too much meaning to these shifts. They are, afterall, only changes over a three-year period in answers given by adolescents; those who live with them know that teenagers are apt to change their identities and appearances overnight! On the other hand, the particular three years are probably more significant historically than developmentally. It was precisely during this time that anti-immigrant sentiments bubbled and boiled over in California, as reflected in the bitter campaign over Proposition 187, which, though framed as a denial of rights to illegal immigrants, was widely interpreted as an attack on all Mexicans in the United States, legal or illegal. Locally in San Diego County school boards and city councils debated what to do about "the children of illegals" in schools and hospitals, and citizen groups wore down their car batteries shining headlights along the border. In this context, a nationalistic reaffirmation of identity is hardly surprising.
Will these identity changes have lasting consequences? We have no way of knowing to what degree these findings might mirror similar changes in other parts of California and the Southwest, or the degree to which they reflect developmental as opposed to historical changes. Portes and MacLeod (1996), analyzing the 1992 CILS survey results, found that pan-ethnic identity (including Chicano as well as Hispanic and Latino) was actually identified with lower SES and lower levels of assimilation and self-esteem. Given the changes from 1992 to 1995, it is not surprising that we did not find a similar pattern within the Mexican-American sample in 1995. "Mexican" is the clear favorite among the Mexico-born children, but, after all, they are Mexicans. Mexico is a fervently patriotic and nationalistic society, and Mexicans are apparently increasingly willing to express these feelings even when they live in Southern California, as demonstrated on Mexican national holidays and when Mexican soccer teams play in the U.S. But on the other hand among the U.S. born, Mexican-American continues to be the single most popular term, followed by Latino/Hispanic. This suggests that specifically Mexican identity wanes among the second generation, and it is logical to expect it to be even weaker among the third. On the other hand, it is worth pointing out that over 98 percent of the sample chose SOME form of Mexican/Latino identity; at least among second generation adolescents there is no evidence that ethnic distinctiveness is being obliterated by assimilation.
In sum, the San Diego sample would seem to well-represent the Mexican-American second generation in general, both in terms of social and cultural makeup. There is relatively little socioeconomic or linguistic diversity, and a strong identity with the terms "Mexican" and "Mexican-American." The clear tendency over time and "generationally" from immigrants to the native-born is an increasing preference for English, and a tendency to express their Mexican identity in "hyphenated-American" form. We turn now to the confusing but essential question of race and Mexican Americans, which we believe is at the crux of the dilemma of understanding the present and future place of the Mexican-American New second generationand, indeed, all Mexican Americans.
III. The Mexican "Race:" Ambivalence, Ambiguity and a Sense of Discrimination
The major students of European assimilation and its discontents in American life__such as Milton Gordon (1964), Richard Alba (1985) and Stephen Steinberg (1989)__all emphasized the disjunction between the typical Euro-American experience and that of racially-distinct groups, by which they principally meant African Americans, but with less certainty included Asians and Latinos as well. It would not be inapposite to say that African Americans have faced the ultimate in negative "contexts of reception," generation after generation. In their recent work Alba and Nee (1997) are skeptical about the significance of non-whiteness per se as a barrier to assimilation, as indeed they would have to be in the face of the evidence: Asian immigrant groups today demonstrate cultural and economic assimilation more rapid than Europeans a century ago, and their rates of social assimilation more closely approximate the European experience than the exclusion that faced all non-whites fifty years ago, and that still is the norm for African Americans. Middle-class Black anglophone Caribbeans seem to outpace native African Americans, though poor black Puerto Ricans are among the most marginalized "immigrants" in the continental United States. Clearly not color per se, but rather the background characteristics of specific populations as well as the specific stereotypes and discrimination that they face are decisive, as segmented assimilation theory emphasizes.
In addition to the complexities of class, among Mexicans race is particularly variable and ambiguous. In the popular press Mexicans are sometimes depicted as just another ethnic group ("like Italians"), on the same path towards assimilation. Chicano activists, often twisting the meaning of the Spanish term "la raza," are equally certain that Mexican Americans are uniformly an oppressed racial group. The Census Bureau has never known what to make of Mexicans. Variously identified by mother tongue and surname in the past, in the last three decennial censuses they have been identified through a distinct "Hispanic Origin" question, separate from the question on race on which Latinos must identify themselves as white, black or Asian. If they insist on writing in something like "Mexicano" race it is re-coded as white. This results in two separate definitions of the nation's white population: all classified as white on the race question and the real white population, which the Census Bureau awkwardly but revealingly refers to as "non-Hispanic whites." One can write in "Mexican" on the ancestry question, but only Europeans and miscellaneous other whites are included in publications based on this ethnic definer; Mexicans and other Latinos are relegated to separate publications and tables, just like Blacks and Asians.
It is perhaps unfair to chide the Census Bureau, since race among Mexican Americans is confusing. Race is both a variable and a constant for Mexican Americans. 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. The sizeable minority that looks essentially Euro-American has at least the potential to "pass" as individuals, but to the degree that they continue to be identified as Mexicans they are subject to much the same stigma as their darker brothers and sisters. The complexity of this stigma has two sources: U.S. attitudes towards Mexicans as a "race" nearly as despised as Africans, and ideas about race brought from Mexico itself.
In nineteenth century California and the Southwest, Mexicans formed something of a middle group between whites and non-whites...Asians, Africans and Indians...and internal color/class distinctions within the Mexican population were explicitly recognized. White Mexican property owners mixed socially with Yankees, while "greasers," working-class mestizos, were considered little better than Negroes or Chinamen. As the old Mexican elites declined, the acceptance of "High-class Mexicans" declined as well, but never totally disappeared. Mexicans were defined as non-white and restricted by miscegination laws, laws that tended to not be enforced against whiter and middle-class Mexicans. With the onset of mass immigration from Mexico early in the twentieth century, social prejudice against Mexicans was reinforced by white working-class fears of competition. Social prejudice and discrimination against Mexicans in early twentieth century California was somewhere in-between the situation of blacks and the lastest Southern European arrivals in Northeastern cities. But to say that it was "in-between" hardly means that it was minor: Italians and other Southern Europeans were themselves despised in quasi-racialist terms, and suffered social prejudice the remnants of which are still with us today ( Almaguer, 1994)
However varied the actual experience of individual Mexican Americans may have been, negative stereotypes among the general population were well-crystalized by the 1930s. The poverty and limited prospects for upward mobility through education were reinforced by strong anti-Mexican attitudes. During the Depression, job competition heightened these feelings, and led to mass deportation of Mexicans, including some who were legal citizens. This antipathy reached its head during World War II in the "Zoot Suit Riots," in which white servicemen attacked "cholos" who affected the baggy dance garb of the time, and the attackers were applauded by the press who saw the cholos as draft-dodgers. The fact that Mexican Americans, both naturalized citizens and the vanguard of the first second generation, volunteered and died during World War II in record numbers was either unknown or of no consequence.
Anti-Mexican attitudes attenuated somewhat in the Postwar period, but by no means disappeared. The term "Mexican" continued to be associated in the public mind with a host of negative characteristics of status and of character. Residential and social segregation were not as uncompromising as that faced by Blacks, at least in the context of California. Data from the 1960s suggested a marked upturn in social assimilation as measured by intermarriage rates. Though difficult to prove, we believe that this partial decline in discrimination was the result of two processes: a modest general decline in anti-Mexican attitudes, and the fact that as Mexican Americans became more middle-class they were more accepted, as they had been in the past(Grebler, et al., 1970).
The very acceptance of "white" Mexicans holds within it the key to continuing resentment about racism and the internalization of self-deprecation, even among those Mexican Americans who have personally experienced little direct racism: the experience of being told that one does not "look Mexican" or having well-meaning friends say things like "I thought you were Spanish" might lead an individual lacking in ethnic pride to hide his Mexicanness, but it also has the effect of reinforcing negative stereotypes. When even a well-meaning individual uses the term "Mexican" or "Latino" in opposition to "white," she is explicitly categorizing them as non-white, with all the negative baggage that that implies in American culture.
Mexicans like to accuse North Americans of racism, with some justice, but Mexican culture itself is as racist any: the white/mestizo/Indian caste-like divisions on which Mexican society was built for four hundred years still broadly describe the physiognomy of the upper, middle and lower classes, in both fact and cultural representations. The mestizo/Indian distinction is as much cultural as it is based on phenotype, making it possible for individuals to move "up" through education and intermarriage; but they do so within this firm racialist structure. On this has been grafted a twentieth century nationalist ideology emanating from the Mexican Revolution that glorifies the Indian past and centers "being Mexican" as being mestizo. Mexican social scientists have long noted the contradiction between the cultural idolization of indigenas in Mexico and the inexcusable ways in which they are actually treated. They have been much more hesitant to publically discuss racism as it affects mestizos in Mexico.
Mexican immigrants bring with them a set of conflicting racial attitudes about themselves that reflect the contradictory role of race in their homeland. On the one hand they are imbued with a strong nationalistic pride in "la Raza," which is really a cultural term referring to all Mexicans and Mexican culture; on the other hand, as mestizos they cannot escape the color prejudice inherent in Mexican culture, even if that prejudice be directed against themselves. Spanish language television serves to remind them of the color hierarchy, since news and entertainment show hosts are white, as are the middle-class and elite characters portrayed in telenovelas, the Spanish-language soap operas that continue to be the most popular form of Spanish-language programming. Mexican color prejudices, we believe, reinforce and support anti-Mexican attitudes encountered in the United States. Like African Americans, Mexican Americans bear internal as well as external burdens of color, but their internal burdens are brought at least in part from Mexico.
The racial stigma of the Mexican-American population provides a particular challenge to understanding the role of race in the process of segmented assimilation. We believe that today race and racism is more consequential for this racially-ambiguous population than it is for the more racially-straightforward East Asian immigrant populations. On the one hand race among Mexicans and other Latinos is more ambiguous at the level of individuals; on the other hand the stereotypical asssociation between low socioeconomic status and being of Mexican "race" is more ingrained and enduring. For at least the last forty years there has been no clear association in the American mind between being Asian American and being poor and poorly-educated; indeed the reverse has increasingly been the case. We do not mean to imply that Asians do not experience racially-based discrimination and prejudice; of course they do, and they are the special victims of the sort of racial lumping that ignores distinctions among Asian subgroups. But they do not struggle under stereotypes that relegate them to the status of "greasers" or "niggers." We believe that anti-Mexican racism, in both its external and internalized forms, is an important factor in explaining the continued low status of second and third generation Mexican Americans in the United States.
We argue that a special burden of the new Mexican American second generation is that they inherit this pre-existing stigma, and it is reinforced by their own life experiences. Their parents, struggling to "ganar la vida," are probably less affected by U.S. anti-Mexican attitudes, though we believe that they do convey the contradictory message about race that they bring from Mexico. Portes and Bach (1985, pp. 282, 333) argued that, in a sense, a heightened awareness of discrimination can be seen as an indicator of increased participation in and understanding of American society. In other words, when children learn what it means to be "Mexican" in California they are undergoing precisely what Gordon meant by "acculturation," learning, though not necessarily internalizing, and certainly not benefitting from, the norms of the dominant society. The severe poverty of their parents reinforces the caste-like stereotype of being Mexican, both in their eyes and in the eyes of others. We believe that it helps explain otherwise puzzling findings regarding identity and felt discrimination, to which we now turn.
As Rumbaut (1966, p. 123) has pointed out, coming to grips with discrimination and prejudice can be much more psychologically damaging to adolescents than for adults. The degree of damage depends on individual factors, as well as on the strength of family and community networks and when the available support comes from peer groups rather than families, the result can be a general rejection of schooling and other "white" institutions and goals (Fordham and Ogbu, 1986; Sanchez and Fernandez, 1993). The resulting oppositional culture and sub-society is an important, if ultimately destructive form of segmented assimilation (Suarez-Orozco, 1991; Portes and Zhou, 1993). As Rumbaut points out, there can be a great diversity of adaptive responses within the same community, depending on individual as well as social heterogeneity. Putting aside the individual factors, which do not concern us here, it follows that the comparatively low level of social heterogeneity in the Mexican American San Diego sample would lead us to expect comparatively little heterogeneity in adaptive responses among them.
Table III presents the rates of perceived discrimination among San Diego Mexican American youth, divided according to language dominance. We chose this dimension of internal variation because it came closest to yielding distinct groups. Perhaps the most striking finding is the degree to which perceptions of discrimination and racial conflict are pronounced across all three groups. Even among the Spanish dominant, who are generally the most recently-arrived from Mexico, 88 percent agreed that there is racial discrimination in America, and 67 percent report having personally experienced discrimination. As Ruggiero and Taylor (1997) have pointed out, it is common for individuals of all groups to report higher levels of discrimination directed against their group than against themselves. The figures on these two measures are similar for the other two groups, with a slight tendency for the more bilingual to have slightly lower pereceptions and experiences of discrimination. All three groups also agree that there is considerable racial conflict in America with, again, a slight tendency for bilinguals to agree at a lower rate. When the sample is divided by gender (omitted from Table III) there is a very slight tendency for boys to preceive higher levels of racial conflict and discrimination.
As impressive as these rates of perceived discrimination are, they are not much different than the rates among other second generation youth in the San Diego surveys. Eighty percent or more of ALL groups agree that discrimination and racial conflict are common, and two-thirds of each group report personal experiences of discrimination.
What does it mean when minority adolescents report discrimination? There is a well-established body of scholarship on the phenomenon that minority students tend to attribute any negative academic feedback to discrimination, and use this explanation as an shield for their self-esteem (Gurin and Epps; 1975; Fordham and Ogbu, 1986; Crocker et al., 1993). Persuasive though this literature may be, we do not find strong evidence in support of it in the San Diego data. For example, all sub-groups, including the Mexican-American, were split more-or-less evenly on the question of whether or not non-whites had an equal opportunity to get ahead in the United States. We suspect that much of the research tradition on "minority youth," which traditionally has been carried out among African Americans, with some studies including Chicanos and Puerto Ricans as well, has only limited applicability to understanding today's second generation youth, particularly Asians and others who do not come with a heavy burden of negative racist stereotypes.
Of all the groups in question, we continue to believe that these negative "vicious circle" effects are most likely to impact Mexican-American youth, but we do not find compelling evidence for (or against) this hypothesis in the San Diego data. However, in a separate study carried out by the second author in a large urban high school in Southern California, strong evidence that Mexican American students feel discriminated against did emerge. Two-thirds agreed that Latino students suffer prejudice and discrimination in schools, at the hands of teachers and staff, and only 14 percent disagreed. On the other hand, the same sample generally agreed that Latino students did have an equal opportunity with others to achieve at school. How might this seeming contradiction be explained? We believe that a strong ethic of individualism and self-reliance pervades inner city high schools, just as it does American society generally. Mexican-American students are willing to acknowledge prejudice and discrimination, but apparently are not willing to consciously blame poor school performance on it. Stanton-Salazar (forthcoming) found that both students and school counselors had a very narrow conception of the range of support students should receive, and stressed unrealistic norms of self-reliance over more communitarian and holistic approaches that employed social support networks.
Qualitative interview data from this study provide further insight into how discrimination is perceived by Latino students. Although students generally acknowledged the existence of societal racism, and were usually able to cite examples in their lives or their familys lives where they encountered prejudice and discrimination, such incriminations were often tempered with accounts of how Mexicans often create their own problems. One young woman, Rosa, in her last year in high school, tells of having a job interview over the phone, then coming in and feeling that the manager looked surprised to see that see was Mexican:
"I guess they heard me over the phone, and I guess I sounded kind of white. Once I got to the store, I saw there was only white girls working there. Well, they never called me back."
Later, Rosa states that differential access to information about educational and occupational opportunities is something Mexicans in the U.S. must frequently contend with:
"...we don't get enough information about things."
Yet, her analysis ultimately reverts to a victim blaming perspective:
"We don't know what's going on, and we don't bother to find out. And a lot of us are too lazy. We're not trying hard enough. We're not all lazy, but we just don't try hard enough."
Others were similarly able to cite illustrations of discrimination in their environment, but ambivalent about how it personally affecting them. Jacqueline, an English dominant senior, states:
"Yeah, I've seen it all the time, it's everywhere, but I don't really notice it. I've never experienced it."
Although minimizing its influence in her own life, she is quite cognizant about how discrimination has affected her father:
"My dad, he works as a machinist. And at the place where he works, like its's most white people, and he's been on a the job for a long time. They get him to train other guys, but then they turn around and give one of these new guys the position that my dad was asking for."
Clearly, perceptions of individual and institutional racism are bound up in complex ways with the experience of being the teenage child of immigrants in America today. While acknowledging discrimination, Mexican-American second generation youth seem less willing to attribute all their problems to racism, and seem to evince a strong ethic of self-reliance and rugged individualism, no matter how inappropriate it may be to the world in which they live.
IV. Educational Aspirations and Achievement
Educational aspirations and expectations, as well as the role of significant others in shaping these orientations, are a principal link between socioeconomic background and eventual attainment in adulthood (Duncan, Featherman, and Duncan, 1972; Wilson and Portes, 1976). Since aspirations have been theoretically tied to the encouragement and moral support of parents and other significant members of a young persons social network, aspirations have played an important role in accounting for those minority youth from low-SES communities who do experience academic success and educational mobility (Velez, 1989). Kao and Tienda (1995), examining a large national sample of eighth graders, found that relative to the children of U.S.-born parents, the children of immigrants had higher educational aspirations. Most telling was the pronounced hopes of immigrant parents, expressing aspirations for their children that even exceeded those of the children themselves.
The Mexican-American youth in our San Diego sample reported uniformly high educational aspirations for themselves, and reported that their parents had even higher aspirations (Table V). For example, 67 percent of the U.S.-born aspired to complete college, and reported that 84 percent of their parents wanted them to at least complete college. Obviously these expectations are highly optimistic, and in fact when asked what level of schooling they realistically expected, rates were markedly lower, though still unrealistic, given that only 10-20 percent of this sample will probably actually finish college. Occupational aspirations were equally unrealistic (60 percent hoped to have professional or managerial jobs). Aspirations varied little by gender, though girls had modestly higher educational aspirations, and multivariate analysis yielded no clear pattern of determinants of aspirations in this sample.
Rather than interpret these high aspirations causally, we prefer to view them as a recognition among Mexican-American youth of the importance and desirability of educational achievement as an essential step towards occupational success. Other studies (e.g. Kao and Tienda, 1995) have reported similar high aspirations among Mexican-American youth and their parents, with recent immigrants often having even higher aspirations. In the San Diego study other ethnic groups reported even higher aspirations. The irony is that the higher aspirations of the other (mostly Asian) youth might actually be more realistic: in California today Asians are attaining truly spectatular rates of college attendance and graduation (see Table VII).
The importance Mexican immigrant parents place on higher education is also reflected in the survey and qualitative research by the second author mentioned above. He found that adolescents had a strong sense of the unequivocal value their parents place on education, but it was also clear that immigrant parents typically had only vague ideas about what higher education actually was or how to help their children actually attain it. As one girl put it:
"I think that most parents of Mexican origin, they push you along, they tell you for the millionth time that you've gotta work hard so that you can do something they never did; or, you know, just climb the social ladder like they never did and never will."
Parents repeatedly told tales of hardship and sacrifice, including their own lack of educational opportunities and their determination that their children would succeed. They were relentless in their exortations. For example:
"I am someone who washes toilets, who cleans houses, who takes the most strenuous and difficult jobs, all because I didn't study...I wanted to be successful, hoping that God would help me...but for their (her children) benefit so that they wouldn't have to kill themselves like me...I want them to study."
Well-meaning as such exortations are, in themselves they are not likely to be effective motivators. In both the San Diego survey and in Stanton-Salazar's field work, parents overwhelmingly reported that they spoke with their children about school work and the importance of studying. Stanton-Salazar found that, at the same time that they talked about the importance of education, parents also put various duties and pressures on their adolescent children that militated in the opposite direction. Their attention was often shifting to their younger children, and they expected their teenage children to help out with childcare, housework, dealing with the outside world, and often contribute to the family income. They were often simply unable to help their teenage children with homework, and the children often spoke of the divide between the world of school and the world of home. As one girl put it:
"Since I was young, they really haven't had much involvement with my school work...they are apart from my schoolwork."
In sum, educational success is valued by Mexican-American children and their immigrant parents, but the scarce material and educational resources parents bring to the table means that in many cases they are not able to translate those values into effective educational support for their children, especially as they confront the difficult years of adolescence. Though we lack the data to support it, we are certain that Mexican immigrant parents stress education much more than immigrants from Poland or Italy 80 years ago, if only because higher education was much less pervasive in those days. The tragedy is that exhortation alone will not move today's Mexican-American second generation into the middle class.
These family dynamics help us understand low achievement among the Mexican-American second generation but, as we emphasized in Part I, in reality there is no compelling reason to explain their modest success in school, as this is the typical pattern among poor immigrants for the past 150 years. It is also, alas, the pattern established by earlier generations of Mexican-Americans in the United States, including the third generation that was coming of age when the research was being done for The Mexican American People (1970). The evidence is compelling that the previous second and third generations were not faring well in school or the workplace, even before the return of large-scale immigration. Grebler and Moore (1970), utilizing 1960 Census data and the surveys from Los Angeles and San Antonio that they had commissioned about 1965, concluded that second generation Mexican Americans were doing better occupationally than their parents but not nearly so well as European Americans of the same generation. Other studies that included but did not focus on Latinos, including Blau and Duncan's pathbreaking The American Occupational Structure (1967), reached much the same conclusions. So did subsequent studies focused on Latinos, such as Bean and Tienda's (1986) analysis of the 1980 Census, and Ortiz' (1996) analysis of the 1990 Census. All these studies document disparities in education, in occupational status net of education, and in income attainment net of occupation and education. The continuing effect of educational disadvantage is by far the greatest factor explaining low occupational and income attainment among Mexican Americans. What is even more striking is the small difference between second and third generation Latino achievements in education and subsequently in the studies referred to above. Whereas historically for European ethnic groups the step up from second to third generation was as significant as the step up from first to second, for Mexican Americans the change was insignificant.
One of the most historically-valuable components of the Mexican American People is the study of children in the Los Angeles Unified School District in 1966 (Grebler, et al., 1970, pp.142-180; 609-617). Though not conceived specifically in generational terms, we know that roughly two-thirds of the Mexican-American children that were the focus of the study were third generation, with most of the balance being second generation. In comparison to their anglo fellow students, both groups of Mexican Americans fared poorly in grades and test scores, though the third generation children, particularly those from English-dominant and more assimilated homes did better, whether measured by grades or test scores. The researchers attempted to disentangle SES and generational effects; they did expect SES to be the predominant explanation for low achievement among Mexican-American youth. They examined three measures: reading comprehension, IQ test scores and grades. For all, the effect of family income and occupational level largely disappeared when parental education was separated out, supporting a general finding in educational research that parental education is the single most important component of parental SES. The more tangible aspects of SES did have stronger associations with IQ scores, however. All these effects were observable among both Mexican-American and Anglo children. T