By 1970, it was obvious that a separate question was needed to identify Hispanics—foreign birth, language, surname, and other proxies had become increasingly problematic, given the preponderance of U.S.-born Hispanics, increasing patterns of intermarriage (making surname a less reliable indicator), and cultural assimilation that included the transition from Spanish- to English-language dominance. The addition of a specific question about “Spanish-speaking people” on the long-form questionnaire in 1970 resulted from political pressure by the U.S. Inter-Agency Committee on Mexican American Affairs.24 In 1976, as noted previously, the Census Advisory Committee on the Spanish-Origin Population for the 1980 Census recommended the adoption of the panethnic label “Hispanic.” In this way, the current census policies about Latinos’ racial and ethnic status reflect both the state’s construction of race and ethnicity as well as the efforts of Latino power brokers to negotiate the spaces between white, off-white, and non-white racial categories.
Today the Census Bureau intentionally treats Latinos as an ethnic group, rather than as a racial group.25 Since 1980, people living in the United States have been asked to identify themselves as a member of a racial group and, in a separate question, as Hispanic or non-Hispanic.26 The Hispanic ethnicity question (which in 1990 followed the race question but in 2000 preceded it) asks, “Is this person Spanish/Hispanic/Latino?” and provides five possible choices: no, Mexican American, Puerto Rican, Cuban, and other Spanish/Hispanic/Latino (then asking for specific subgroup identification).27 In 2000, the race question directed respondents to choose one or more of the following options: white, black, American Indian or Alaskan Native, one of ten Asian categories, or “some other race.”28 Of those who identified themselves as Hispanic, 99 percent selected only one racial category, and the census officially reports that “among Hispanics, 92 percent were White.”29
The reality, however, is considerably more complicated and may well reflect Mexican Americans’ history as off-white. When answering the race question, Latinos select “some other race” in substantial proportions, rejecting white, black, Indian, or Asian racial identification. In 2000, 42 percent of Hispanics preferred “some other race,” while 48 percent selected white, 4 percent selected black, and 1 percent selected American Indian.30 Latinos’ preference for “some other race” is all the more striking because, combining all non-Hispanics, only 1 percent of people selected “some other race.”31 Moreover, this makes all the more perverse the Census Bureau’s disregard for this preference by folding those persons who chose “some other race” into those who selected “white” as their race.
To explore why substantial numbers of Latinos reject identifying as white, African American, American Indian, or Asian American—as the Census Bureau would prefer they do—we must return to a variation of the debate that preoccupied the nation in the mid-1800s: where do Mexican Americans—and now Hispanics—fit in the national racial order? Consciously or not, Latinos are engaging in a debate about this question when they identify themselves racially. The debate is both descriptive and normative—it is about how they see themselves and how they perceive others as seeing them, and it is very much about the desire to belong in the national community. In the context of debates over immigration and citizenship in Europe, sociologist Rogers Brubaker has said,
The politics of citizenship today is first and foremost a politics of nationhood. As such, it is a politics of identity, not a politics of interest (in the restricted, materialist sense). It pivots more on self-understanding than on self-interest. The “interests” informing the politics of citizenship are “ideal” rather than material. The central question is not “who gets what?” but rather “who is what?”32
Similarly for Mexican Americans today (and perhaps since 1848), the politics of identity are about who they are racially, compared to other Americans and in the context of a complex history in the United States.
The fact that many Hispanics self-identify racially as “other” has been observed since 1980—the year in which census questionnaires first included a so-called Hispanic ethnicity question as well as racial self-identification (as opposed to racial categorization by a census employee).33 In 1980, 40 percent of Hispanics selected “other” rather than black, white, or Indian as their race.34 In 1990, the number of Hispanics who selected “other” was about the same as those who selected “some other race” in 2000 (44 percent in 1990, 47 percent in 2000),35 suggesting that the different wording of the question and the different order of the race and Hispanic identity questions did not influence choices.36 These data show that, over time, the proportion of Hispanics who reject white racial identification is rising (during the same three decades, the proportion of Hispanics choosing “white” has decreased from 64 to 48 percent).37
What does it mean that Hispanics—60 percent of whom are Mexican Americans—divided their responses to the race question between white (48 percent) and “some other race” (42 percent)? Four explanations appear most consistently in the literature.38 One explanation implicitly validates the 50 percent of Latinos who selected “white” as choosing the “correct” answer and labels those Latinos who selected “some other race” as ignorant or confused. I include in this category those who are generally critical of the census approach to race, such as sociologist Nathan Glazer, a longtime critic of race-based government policy. “Do the American people in general know? If they do not—and it is hardly likely they are fully briefed on the legislation and regulations and politics and the pressures that have made the census form, with respect to race and Hispanicity, what it is today—what are they to conclude?”39 How would Glazer respond to the fact that virtually all of the non-Hispanics who responded to the race question in 2000 seemed to have no trouble selecting a category? A variant of this explanation targets Hispanics specifically: because they are recent immigrants to the United States, it implies, they must be confused about how race works here and so they erroneously select “some other race.”40 This explanation would seem to be the one adopted by the Census Bureau itself given how it collapses “some other race” into “white” responses.
A second explanation can be considered a hybrid of the “they’re confused” explanation. According to this explanation, the fact that the majority of Hispanics selected the “white” race category shows that Hispanics are, collectively, white in American society.41 Under this explanation, appropriately little weight is given to the fact that almost half of Hispanics selected “some other race” (since it reflects their racial confusion, it need not be taken seriously). However, the ubiquity of the phrase “non-Hispanic whites” in both the media and social science to distinguish “real whites” from Latinos suggests that who is and is not white is considerably more contested than this explanation would lead one to believe. A third theory posits that the census responses accurately reflect a Latino population that is racially bifurcated, roughly half (truly) white and the other half (truly) non-white.42
Both the second and third explanations, then, take Hispanics’ census responses at face value—that is, they assume that respondents intended their responses as literal reflections of their perception of their racial status in the United States. Several researchers, however, have instead posited a fourth interpretation—that Hispanics’ racial self-identification is highly symbolic of other dynamics, including their experiences of racial and economic discrimination in the United States, their general level of inclusion in or exclusion from the nation, and their different view of race as influenced by the Spanish-Mexican racial legacy described in Chapter 2.43 In a two-year study convened by the National Academies, sociologists Marta Tienda and Faith Mitchell found strong support for this explanation, concluding that Latinos see themselves as a separate race:
[Hispanics’] choice of “some other race” on the census forms reflects more than four centuries of mestizaje, or racial miscegenation, in Latin America and the Caribbean, as well as the differing conceptions of race noted above. Hispanics may also mark “some
other race” simply because they do not see themselves as fitting under any of the categories provided by the Census Bureau. . . . Thus, their rejection of the [Office of Management and Budget] racial classification by checking “some other race” on the census questionnaire reflected their lived experience rather than a statistical artifact or measurement error.44
If responses to the racial identity question reflect genuine differences in how Latinos see themselves in the nation, we should expect to see significant differences among Latino groups according to national origin. As noted, 90 percent of Hispanics overall divide their racial identity between “white” (48 percent) and “some other race” (42 percent).45 One finds interesting patterns when comparing racial identification with national origin. Of Hispanics, 90 percent identify as Mexican origin (60 percent) or as from one of eight other Latin American countries: Puerto Rico, Cuba, the Dominican Republic, El Salvador, Guatemala, Colombia, Peru, and Ecuador.46 Among these groups, self-identification as white versus “some other race” varies widely, ranging from a low of 28 percent for Dominicans to a high of 88 percent for Cuban Americans.47 It is not the case that large numbers of national-origin subgroups identify as black—those numbers range from 11 percent for Dominicans (the highest of any Hispanic subgroup) to 7 percent for Puerto Ricans and 4 percent for Cubans, to mention just three subgroups where we might expect to see a larger black identification.48
We can distinguish two tiers of Hispanics, in terms of their racial self-identification.49 The first, relatively small group embraces white racial identity at rates approaching 90 percent. Cubans and those who trace their origins to Argentina fall into this category, identifying as white 88 and 89 percent of the time, respectively. Hispanics who immigrated from or whose family origins are in Chile or Venezuela come close: 77 and 76 percent of them, respectively, identify as white. The second tier includes those subgroups whose members self-identify as white at rates between 28 and 52 percent. The two largest Latino groups fall into this category—Mexican Americans (50 percent) and Puerto Ricans (52 percent), together 70 percent of all Latinos. Central Americans (Guatemalans, Hondurans, and Salvadorans) identify as white in percentages ranging from 41 to 49 percent. Dominicans, as mentioned above, are the Hispanic subgroup least likely to identify as white, at 28 percent. If we take these same groups and explore the proportion selecting “some other race,” we see the mirror image. Cubans, for instance, selected “some other race” only 7 percent of the time. In contrast, 39 percent of Puerto Ricans and 47 percent of Mexican Americans preferred “some other race.” Central Americans selected “some other race” in percentages ranging from 43 to 57, and 59 percent of Dominicans preferred “some other race.”
It should not surprise us that almost as many Mexican Americans opt for the “some other race” option as for white. Both choices seem reasonably to be a product of Mexican Americans’ peculiar history as off-white—as sometimes legally white and almost always socially non-white. Given their history, it is predictable that many Mexican Americans would seek to formally identify themselves as racially white. Manifest Destinies has described Mexican Americans’ long history of utilizing whiteness (or claims to be white) to combat discrimination and racial subordination. On the one hand, claiming whiteness is a way of blunting the full impact of racial discrimination. Like others in American society, Mexican Americans also have internalized antiblack racism such that “white” could reflect “I’m not black.”50 Sociologist Joan W. Moore, one of the authors of the landmark 1970 study The Mexican American People: The Nation’s Second Largest Minority, recalls that, in the late 1960s, “many Mexican Americans rejected the term ‘minority’ and its implied association with black America.”51 Given their historic legacy as off-white, it should not surprise us that substantial numbers of Mexican Americans would today continue to identify as white.
That same history has influenced many Mexican Americans to reject a view of race as black and white and view themselves as non-white. Mexican Americans’ origins in this country as a colonized group, like Puerto Ricans, also contribute to their identification as non-white outsiders. Certainly the collective memory of events in the twentieth century that dramatically reinforced Mexican Americans’ second-class citizenship (such as the deportations of Mexican American citizens in the 1930s) may continue to haunt present generations. Mexican Americans participated disproportionately in World War II, and returning veterans formed the first “civil rights” organizations as we have come to understand that term.52 Inspired by the black civil rights movement, the Chicano student movement of the 1970s led many Mexican Americans to expressly reject a white identity in favor of a non-white identity. Part of the Chicano movement’s rhetoric was to embrace—rather than denigrate—Mexicans’ indigenous ancestry as part of the embrace of an oppositional, non-white identity.53
Given this history, Mexican Americans’ selection of “some other race” may reflect their recognition of a racial heritage that includes white, black, and Indian ancestry, as well as an affirmative rejection of white identity as a political statement. Certainly, Cuban Americans have little in common with this history, and that may account for their identification as white 90 percent of the time. Rather than a history as a conquered people, Cubans initially came to the United States as voluntary immigrants only fifty years ago. Most fled Cuba in the wake of the 1959 socialist revolution led by Fidel Castro in “the silk stocking exodus,” invoking their status as the most affluent members of Cuban society.54 Cubans entered American society during the midst of the civil rights movement, when the racial order was transitioning to one of formal equality and antidiscrimination policy that was mandated by law and they benefited accordingly. Sociologists Stephen Cornell and Douglas Hartmann have noted the dramatic differences in government policy aimed at Cuban and Mexican immigrants (including those Mexican immigrants who came to the United States at exactly the same time as the first Cuban immigrants).55 For example, in 1960, the U.S. government established the Cuban Refugee Program to help Cubans find jobs and housing, learn English, and generally adjust to a new society. No such effort was made on behalf of Mexican Americans as a group. Today, U.S.-born Cuban Americans have a higher median income than non-Hispanic whites.56
Empirical research lends support to my hypothesis about how today’s Mexican American racial identity—as divided between white and “some other race”—reflects the group’s historic legacy as off-white. In a study focusing on native-born Latinos (not only Mexican Americans), demographer Sonya Tafoya compared “white Hispanics” (that is, those who identified as both Hispanic and white on the census) and “some other race Hispanics” (those who identified as Hispanic and selected “some other race”). She found that white Hispanics had significantly higher incomes, were twice as likely to consider themselves Republicans, and were considerably less likely to believe that Hispanics experience significant discrimination in the United States.57 She posits that native-born Hispanics’ racial identification depends on their degree of inclusion in the American polity and economic success: to the extent that Latinos feel politically excluded and economically marginalized, they will opt for “some other race.”58
Tafoya’s study dealt generally with Latinos, and she did not provide national-origin breakdowns. However, she reported significant differences in racial identification between native-born Mexican Americans who live in Texas and those who live elsewhere. While about half of Mexican Americans generally identify as white, two-thirds of Mexican Americans in Texas do so.59 As Tafoya puts it, Texas “is the only state where a large Latino population was caught up both in Southern-style racial segregation and then the civil rights struggle to undo it.”60 In another study comparing Mexican Americans in Los Angeles and San Antonio, Texans were found to be more likely to identify as white but also more likely to speak Spanish, marry other Mexicans, and live in Mexican neighborhoods.61 These studies offer tentative support for my hypothesis about the continuing legacy of Mexican Americans’ off-white status. T
exas Mexicans identify as white not because they live white lives, but precisely because the opposite is true: they are embedded in Mexican American neighborhoods, speak Spanish, and probably experience continuing racism because of those facts. In this context, their claims to whiteness are defensive.
Sociologists Edward Telles and Vilma Ortiz have published a study of Mexican Americans that is methodologically innovative: they followed subjects longitudinally (they were interviewed initially in the late 1960s and then again in the late 1990s), and they considered the direct effects of generational status by collecting data on immigrants (first-generation Americans) and their specific descendants (children, grandchildren, and great-grandchildren, in some cases).62 Whereas surveys have found very low levels of ethnic identity among the grandchildren and great-grandchildren of European immigrants, Telles and Ortiz found persistent rates of ethnic identification as Mexican American/Latino even among the great-grandchildren and great-great-grandchildren of Mexican immigrants (third and fourth generations and higher in the survey).63 Based on their National Academies research, Tienda and Mitchell agree, concluding that Hispanics are increasingly viewing themselves in racial terms, “in strong contrast to the experience of earlier immigrant groups from Southern and Eastern Europe, whose social acceptance and cultural assimilation in the United States involved self-identification as white.”64
The presence of large numbers of immigrants complicates the picture for Latinos. On the one hand, a high proportion of relatively uneducated immigrants depresses Latino averages for educational attainment, income, and wealth acquisition.65 On the other hand, the so-called immigrant paradox may inflate Latino gains: the children and of Mexican immigrants outperform third- and fourth-generation Mexican Americans on a host of educational attainment indices.66 Telles and Ortíz found support for the immigrant paradox in their research tracking Mexican American families in Los Angeles and San Antonio over time. They concluded that Mexican children who were themselves immigrants and second-generation Mexican Americans (the children of immigrants born in the United States) had education performance levels comparable to those of first- and second-generation Italian Americans in the past, but for Mexican Americans in the third generation and beyond “educational assimilation is abruptly halted and slightly reversed.”67
Manifest Destinies, Second Edition Page 23