LATINO/"HISPANIC"-WHO NEEDS A NAME?* THE CASE AGAINST A STANDARDIZED TERMINOLOGY Martha E. Gimenez Originally published in _Thternational Journal of Health Services_, Vol 19, Number 3, Pages 557-571, 1989. Reprinted with permission. *I use quotation marks around "Hispanic" to indicate my critical stance toward the label. In the article, whenever the label is mentioned in the context of someone else's discourse, it will be written without quotation marks. ABSTRACT Public health specialists, policy makers, social scientists, and politicians, for different reasons, have welcomed the "Hispanic" label. The label presumably identifies an ethnic group that is also a minority group (i.e., a group historically subject to economic exploitation and racial discrimination). Consequently, its consistent use by federal and state agencies would produce large quantities of comparable data useful for research, and for policy making and implementation. Critics have argued that the label is racist, it mystifies the real reasons for the disproportionately high proportion of people of Mexican and Puerto Rican descent in disadvantaged social and economic conditions, and stands in the way of a fair implementation of affirmative action. Latino, a race-neutral term with historical roots, has been suggested as an alternative to be used in conjunction with national origin or regional forms of self-identification, In this article, I argue that any standardized terminology is unavoidably flawed and conducive to the development of racist or, at best, trivial stereotypical analysis of the data thus produced. The "Hispanic" label does not identify an ethnic group or a minority group, but a heterogeneous population whose characteristics and behavior cannot be understood without necessarily falling into stereotyping. The label should be abandoned; social scientists and policy makers should, instead, acknowledge the existence of six aggregates, qualitatively different in their socioeconomic stratification, needs, and form of integration in the U.S. economy: two minority groups (people of Mexican and Puerto Rican descent), and four immigrant populations (Cubans, Central American refugees, Central American immigrants, and South American immigrants). BEGIN TEXT INTRODUCTION In 1985 I found out that the affirmative action office of the university where I work was counting me as a "minority faculty," member of the so-called "Hispanic ethnic group." It was then that I became interested in the label and its implications for the people it identifies. I found its political construction and usage particularly worthy of examination because it abolishes, for all practical purposes, the qualitative historical differences between the experiences and life chances of U.S. minority groups of Mexican and Puerto Rican origin, and those of Latin American and Spanish peoples. The label imputes to Latin Americans a contrived "Hispanic ethnicity" while minoritizing them in the process (i.e., defining them as members of a minority group even though they have never been historically oppressed as such in the United States.) [Note 1: Working-class, unskilled Latin Americans, particularly undocumented workers, are economically exploited and disproportionately found in the worse jobs and in the poverty population, not because of their "Hispanicity" and "minority group status," but because of their lack of language skills and "human capital." On the other hand, middle-class, upper-middle-class, and petty bourgeois Latin Americans do very well, not in spite of their "ethnicity (i.e., they are not examples of "minority success" or "assimilation" to "majority" values), but because their social class entails ownership of economic and/or human capital.] Because the label is used in the context of affirmative action, it places professional and skilled immigrants in objective competition with members of the U.S. minority groups and forces them to pass, statistically, as members of an oppressed group. Although the political and ideological unintended functions of this label are numerous and complex, in this article I will examine primarily the adequacy of the theoretical and methodological grounds presented in its defense. I will argue that, far from being useful for social science research and effective policy making and implementation, the "Hispanic" label fulfills primarily ideological and political functions. It cannot replace preexisting theoretical (social class and minority group) and descriptive (national origin and socioeconomic status) categories of analysis; its presence in scientific and popular discourse adds nothing to knowledge while it strengthens racist stereotypes. The Terms of the Debate As I began to examine the theoretical and political significance of this label, a colleague told me about several relevant articles published in the American Journal of Public Health which proved to be very interesting and useful; to my knowledge, they form the only scholarly debate on this important issue (1-5). In these articles, both the defender and the critics of the "Hispanic" label agree on the need for a "standardized terminology"; i.e., an all-encompassing "Umbrella" term useful to identify all the populations it labels. The critics make a persuasive case for using the term "Latino," rather than "Hispanic," exposing the racist implications of the latter (1, p. 355; 5, p. 15), pointing out the problems it creates for implementing Civil Rights legislation (I; see also 6), and its roots in the history of U.S. economic and political domination over Latin America since the days of the Monroe doctrine (3). Acknowledging that Latino, like "Hispanic," is a generic term, to ensure comparability of samples and research findings they suggest that social scientists, in their work, should identify also national origin (e.g., Mexican, Puerto Rican, etc.), nativity (U.S. or foreign born), and/or generation. Since people living in the Southwest identify themselves in a variety of ways (e.g., Hispano, Mexicano, Manito, Chicano, Raza, etc.), regional variations ought to be taken into account as well. The defender (4) considers "Hispanic" a better term on scientific, political, and pragmatic grounds. Pragmatically, because the statistics compiled by the federal government and government agencies use the Hispanic label, social scientists and policy makers should avoid using a different term. This would create confusion, establishing social scientists in the role of relabeling millions of people who self- identified as such in the 1980 Census. Also, inhabitants of the Southwest who identify themselves with a regional label find Latino unacceptable, while readily agreeing to the Hispanic identifier. Unlike Latino, therefore, the Hispanic label ensures greater population coverage. Scientifically, it is important to have comparable data. [Note 2: Federal, state, and private agencies, on the other hand, often do not use the same codes for race, thus making it difficult to compare data from different sources. See, in this respect, refrerences 7 and 8.] The use of the label in data gathering ensures coverage and consistency; the fact that it is used in the collection of vital and health statistics creates the possibility for trend analysis (4, p. 70). Consistent use of the label is also important from the standpoint of policy making and implementation because the goal of public health specialists is "to make progress in standardizing . . . ethnic and social classification systems so that we may move forward in our understanding of the health needs of all our populations" (4, p. 69). Politically, the label identifies a minority group subject to severe discrimination. Defending the minority status of "Hispanics," Trevino argues that "despite the fact that Hispanics had lived in the U.S. for more than 400 years, [they] were still less educated than Blacks, about as poor, had no more luck in getting good jobs, received less health care" (4, p. 71). To replace Hispanic with Latino would undermine, in Trevino's view, affirmative action protection for Hispanics because, academically defined, Latino designates "the peoples, nationalities or countries such as the French, Italian, Spanish, etc. whose languages and culture are descended from the Latin"; this would make eligible under affirmative action people whom the term Hispanic currently excludes (4, p. 70). Regardless of their differences about the relative merits of Latino versus "Hispanic" as umbrella terms, critics and defender agree about the need to have, in addition to a standardized terminology, as much information as possible about the population under study, to identify needs, factors affecting health, access to health services, and so forth. In spite of the arguments advanced in its support, however, the label does not help either social scientists or policy makers because it only creates an artificial population; i.e., a statistical construct formed by aggregates of people who differ greatly in terms of national origin, language, race, time of arrival in the United States, culture, minority status (see, for example, 3; 5; 9, pp. 9-10; 10), social class, and socioeconomic status. The empirical referent of "Hispanic" fully justifies these assessments: "[T]his statistical construct has hardly any relation to the real world" (10); "[it] vastly oversimplifies the situation. The heterogeneity of the Hispanic population reduces the term to a merely heuristic device" (9, p. 9). It is here that the main theoretical and methodological problems are located. Succinctly stated: what can this, or any other "umbrella" term, identify? Is it a minority group? Is it an "ethnic" group? What is the meaning of the data gathered about this population? THE PROBLEM OF INTERPRETATION: WHAT DO DATA AND RESEARCH FINDINGS ABOUT "HISPANICS" ACTUALLY MEAN? It is fascinating to observe how those writing about "Hispanics" (discussing public policy issues, or reporting research findings and vital statistics) do so while fully cognizant (with exceptions) of intrapopulation variations of such magnitude as to render statements about "Hispanics" in general either meaningless or suspect. These are some representative statements: The birth rates for 1983 and 1984 for the "Hispanic" population "were about 50 percent higher than those for the non-Hispanic population. . . . [T]he fertility rate for all Hispanic women was 42 percent above the rate for "non-Hispanic" women" (11, p. 2). "[T]he median age of the Hispanic population is 6-7 years below that of the non-Hispanic population ... almost 8 years below that of the White population, and almost 2 years below that of the Black population" (9, p. 21). "Hispanic population increases at 5 times rate of rest of U.S." (12). These are descriptive statements, useful (it may seem) for comparing "Hispanics" with whites, blacks, and Asians. What is the meaning and purpose of such comparisons? Is it possible to account for those patterns, or to make such comparisons, without creating stereotypes? The demographic characteristics (age and sex tructure) and composition (income, education, and occupational distribution) of the "Hispanic" population can be described. Any attempt to account for those characteristics, or for differences in fertility (or anything else) must rest, however, either upon well-established empirical generalizations (e.g., the inverse relationship between income, education, and fertility), which do not necessitate the racial/ethnic classification of the population, or upon stereotypical generalizations about "Hispanic" culture. The mass media and politicians exploit data about the youth, higher fertility, and growth rate of the "Hispanic" population in ways that, ultimately, intensify racist fears among those worried about low white fertility, increase the likelihood of conflict with blacks (who see their communities competing for scarce resources with an ever-growing "minority" group), and strengthen stereotypes about "Hispanics"' cultural traits and the perception that their presence will contribute to increase social problems and tax payers' burdens: e.g., growth of the "Hispanic" population will make it more difficult to eradicate poverty, will increase welfare expenditures, will increase the demands for health care and other social services; given their high fertility, they will be the largest minority group by the year 2060 (13). Poverty, however, is not something inherent in people's genes or culture; it depends on class location and individual resources. Furthermore, if immigration from Mexico and Central America continues to be composed mainly of poorly educated, low-skilled workers and displaced peasants, the percentage of "Hispanics" below or close to the poverty level will remain high and might even increase. This kind of inference, however, is precluded by generalizations stressing the "Hispanicity" of the population that, regardless of their authors' motives, have inherently misleading latent or unanticipated ideological effects. In the popular consciousness, as well as among social scientists, "Hispanicity" seems to be equivalent, at best, to "traditional culture" and, at worst, to the culture of poverty. A statement taken from research on young mothers illustrates this point (14, p. I 1): [T]he findings for Hispanic mothers, who report generally higher fertility expectations and lower educational expectations than do other mothers, suggest that these women represent a relatively unacculturated subgroup, with more traditional attitudes toward motherhood and higher education for women. The overgeneralization about "traditional" culture cited above is quite typical of the mass media and the "modernization" school of U.S. social scientists, still caught in the simplistic understanding of historical change as a process of modernization; i.e.; change from traditional (e.g., "Hispanic") to modern (e.g., U.S.) culture. In fact, the assumption of the "traditional" nature of "Hispanic" culture or "ethnicity" is built in research that compares "Hispanic" (or Puerto Rican, Mexican, Cuba, or Central and South American) fertility (or any other demographic or social pattern) with white and black patterns. Typical of this kind of reasoning is the following (15, p. 376): Hispanic origin is now an important control variable in the analysis of religious [fertility] differentials. Hispanics are a growing proportion of the U.S. population, they have high fertility, and about three fourths of them are Catholic. Whether or not Hispanics were included had a dramatic effect on the size of the religious differential . . . .As the size of the Hispanic population grows, and more are foreign born and from high fertility societies, we can expect this effect to become more pronounced. Within the parameters of this discourse, high fertility reflects religion and the culture of "high fertility societies." A more sophisticated theoretical analysis of fertility, as a rational household survival strategy within given structural conditions of existence (see, for example, 16), would focus researchers' attention on the social class, socioeconomic strata, and actual opportunity structure confronting some "Hispanic" women in this country. What appears as an effect of religion and/or the culture of high fertility or "traditional" societies might be the outcome of conditions of existence similar, in their demographic effects. to those conducive to high fertility in the lower strata of the working class, and in rural and urban subsistence sectors everywhere (17). Everyone writing about "Hispanics," especially social scientists and policy makers, ought to pay attention to Cafferty's and McCready's warning about the dangers entailed in assuming "certain behavioral characteristics based on group identity. [T]he serious thinker expresses legitimate concern when he worries that any examination of Hispanics in the United States may result in negative stereotyping" (18, p. 5). If general statements about "Hispanics" are problematic, are statements about national origin aggregates (e.g., Mexican, Puerto Rican, Cuban, Central and South American) any better? Most social scientists supplement general statements with observations and data about the different aggregates included under the umbrella term. For example, the report about births of "Hispanic" parentage also indicates that, although "Hispanic" women begin child-bearing at young ages, the percentage of live births to teenagers was higher among Puerto Rican (21.3 percent) and Mexican (18.0 percent) mothers than among Cuban (8.2 percent) and Central and South American (8.1 percent) mothers. Mothers' educational attainment also varied according to national origin, with the highest level found among Cuban, Central and South American mothers, and the lowest among Mexican mothers (I 1, pp. 2-3). Those populations also differ in their age structure; Puerto Ricans are the youngest group (median age, 20.7) and Cubans the oldest (median age, '33.5) (9, p. 22). To the extent that researchers use national origin not simply descriptively, but as a proxy for culture as the main independent variable, there is a real danger of developing stereotypes about each of those populations. For example, why are teenage births so few among Cuban and Central and South American mothers? Would a purely demographic explanation (e.g., age structure) suffice? Why is educational attainment greater in these groups? Census data show substantial differences in levels of income and educational attainment among the national origin groups in which data about "Hispanics" are usually classified. Data about the foreign born are limited, but what are available show striking differences between them (Table 1) and indicate that a great deal of important information is lost in the construction of "Hispanics" and national origin groups. Each national origin aggregate is different from the others in terms of historical origins, minority status, problems with language, class structure, socioeconomic stratification, and last but not least, reception by the "host" country. (Compare the reception given to the Cuban bourgeoisie and middle-class who fled to Miami after Castro took power with the treatment given to the people living in Mexican and Puerto Rican territories after the annexation, or to Mexican and Puerto Rican manual workers migrating to the United States throughout the 20th century.) This means that it is as misleading to make general statements about "Hispanics" as it is to make them about, for example, Puerto Ricans, Argentines, Cubans, and so forth. To avoid the possibility of constructing stereotypes in the process of interpreting data, social scientists need to go beyond cultural explanations to examine class and socioeconomic status differences within each aggregate. Were they to do so, they would most likely find greater behavioral similarities between "Hispanics" and non-"Hispanics" of the same social class than between "Hispanics" of different social classes and/or national origin. On the other hand, similarities between "Hispanics" and non-"Hispanics" of similar class and socioeconomic status should not be reduced to the result of "acculturation"; they are, after all, something to be expected given the relationship between class location and people's life chances. Excessive reliance on culture as the major explanatory variable limits researchers' ability to make sense of information already available. For example, critical of the widely held belief that culture is the main barrier between "Hispanics" and successful "Americanization," Cafferty states: "the economic success of Cuban immigrants in Miami and of some Dominican immigrants in New York suggest that Hispanic culture, as such, is no obstacle to achievement" (19, p. 41). This statement fails seriously to challenge cultural explanations for "Hispanics"' relatively low socioeconomic standing; what had to be indicated also is the fact that some immigrant populations are more successful not in spite of their culture, but because of the resources (economic and/or human capital) they bring with them. Generalizations about national origin groups, e.g., "Mexican immigrants earn less and achieve lower occupational levels than others" (1 9, p. 41 ), if isolated from additional data, cannot but unwittingly create stereotypes about those groups. To sum up: the problem facing social scientists and public health specialists in trying to make sense of the data collected by federal, state, and other agencies is a problem not only of comparability but of meaning. The avowed aim of using a standardized terminology is improvement in the identification of an ethnic group that is also, presumably, a minority group (an issue to be examined in the next section). However, the heterogeneity of the population included under the umbrella term undermines the validity of defining it, for social research and policy purposes, as an "ethnic group" (i.e., a group with common cultural characteristics). Table 1 Selected characteristics of the population by selected country of birth: 1980* ------------------------------------------------------------------------------ Median Nationality Total High household and selected persons, school College Prof. Service income, country of birth thousands grad grad specialty occup. 1970 dollars ------------------------------------------------------------------------------ Native 212,466 67.7% 16.3% 12.3% 12.7% $17,010 Foreign born Mexico 2,199 21.3% 3.0% 2.5% 16.6% $12,747 Colombia 144 62.8% 14.6% 8.1% 15.8% $15,883 Dominican Rep 169 30.1% 4.3% 3.1% 18.5% $10,130 El Salvador 94 41.4% 6.5% 2.6% 31.7% $12,261 Ecuador 86 56.0% 9.3% 5.3% 14.7% $15,402 Guatemala 63 42.7% 6.9% 3.9% 27.9% $13,385 Cuba 608 54.9% 16.1% 9.2% 12.2% $16,326 Argentina 69 70.9% 24.2% 16.3% 13.1% $18,892 ------------------------------------------------------------------------------ *Source: U.S. Bureau of the Census. Statistical Abstracts of the United States: 1987, Ed. 107, p. 30. Washington, D.C., 1986, To speak about "Hispanic" fertility, child-rearing habits, health subculture, migration patterns, etc., is to engage in empty talk, at best, or in stereotyping. The heterogeneity of national origin groups, in turn, undermines generalizations about the entire group. It seems that social scientists doing research about "Hispanics" are beginning to recognize the problems inherent in relying on ethnicity as the main independent variable, and the advantages in studying social class variations in behavior, attitudes, etc. As one social scientist acknowledges, " [T] he problem with ethnicity is that it has been overused as the sole explanation for all types of behavior among Hispanics" (20, p. 180). To state, for example, that "Hispanic men do not readily accept the notion that they are ill and, therefore, will not visit the physician in the same proportions" (20, p. 161) (as whites, presumably) is as stereotypical as saying, for example, that Mexican men distrust modem medicine. It is important, on the other hand, to learn that "Mexican-American men with low education [have] high levels of distrust of modern medicine and doctors; and that age, sex, education, and income [are] powerful factors in explaining utilization of health services" (20, p. 161). What is needed to make sense of the data collected under national origin categories is a breakdown of the information on the basis of not only nativity and length of stay in the United States, but social class and socioeconomic status as well. In the United States there is a great deal of overlap between race, ethnicity, and class, so that a large proportion of "Hispanics" are located in the lower strata of the working class. This situation leads to the masking of the effects of social class and socioeconomic status under the cover of "ethnicity" (a code word primarily used to refer to the culture of those considered "nonwhites"), to the point that researchers-normally trained in viewing class and culture only as analytically different things-feel that it is very difficult to separate the effects of each (see, for example, 20, pp. 179-183). Dialectically, however, culture is not a thing one learns or unlearns (thus becoming "acculturated"); it is the lived experience of people shaped by their location in the class and socioeconomic stratification systems. Therefore, research should be aimed not at the assessment of the amount of variance to be explained by class or by culture, but to establish the complex connections between culture, behavior, and their objective basis in people's class location. But this approach, which would result in better social research and more effective policies, cannot be pursued because researchers are trained, to some extent, to consider social class superficially (i.e., reducing it to socioeconomic status), mechanically "controlling" for education, income, and/or occupation whenever possible. Furthermore, even if concerned with social class and socioeconomic status differences, social scientists are constrained by having to use data specifically designed to construct "ethnicities" while providing scant information about class and socioeconomic status indicators. As long as social scientists and policy makers spend time and resources in the construction of standardized terminologies for the identification of politically constructed "ethnicities" (e.g., whites, blacks, "Hispanics," Asians) through the racialization and ethnicization of national origin, improvements in data collection are not likely to contribute to credible, nonstereotypical research findings and effective policy making and implementation. In fact, standardized terminologies pose problems for dealing with data about all the "racial" and "ethnic" categories currently in vogue. The problems inherent in lumping together populations heterogenous in terms of class, socioeconomic status, culture and national origin obtain also for the white, black, and Asian populations. THE COERCIVE NATURE OF SELF-IDENTIFICATION: DOES IT CONFIRM THE EXISTENCE OF A SINGLE MINORITY GROUP? Forty percent of those whom demographers (21) analyzing Census responses classify as "'Hispanics' using secondary identifiers" gave a negative answer to the Spanish/Hispanic origin Census question [Note 3: The Census self-identification question is as follows: Is this person of Spanish/Hispanic origin? No (not Spanish/Hispanic). Yes, Mexican, Mexican American, Chicano. Yes, Puerto Rican. Yes, Cuban. Yes, other Spanish/Hispanic.] instead, they wrote their country of origin in the space left for "other" in the question designed to establish race. [Note 4: The Census question is as follows: Is this person White Asian Indian Black or Negro Hawaiian Japanese Guamanian Chinese Samoan Filipino Eskimo Korean Aleut Vietnamese Other ________ Specify ___________ Indian (Amer.) Print Tribe: _____________________] Those who exhibited the greatest "consistency," self-identifying as "Spanish/Hispanic" in addition to secondary "Hispanic" identifiers (place of birth, ancestry, Spanish "race," surname, and language) (21, p. 8), were primarily of Mexican and Puerto Rican origins and had considerably lower socioeconomic status (in terms of income, occupation, and education) than the "inconsistent" 40 percent mentioned above. "Inconsistent" respondents were primarily Central and South Americans; a small percentage were Puerto Ricans and Cubans, and a higher percentage, Mexicans (21, pp. 9-17). These facts, from my standpoint as a sociologist who is also Latin American (specifically, Argentine), indicate that a large percentage of Latin Americans not only know precisely who they are but prefer their historical identities (e.g., Mexican, Colombian, Argentine, etc.) to a label devised by some government officials, conservative politicians, and academics. (I am indebted to Rodolfo Alvarez (Department of Sociology, University of California, Los Angeles) for calling my attention to the relationship between the invention of "Hispanics" and the"Nixon Southern Strategy.") Those respondents are telling U.S. social scientists and politicians who presume to know better - e.g., "[A]fter more than thirty years of experimentation with enumeration strategies, we should by now know what a Hispanic is" (4, p. 71) - that they are, in fact, wrong. That, however, is not the way U.S. social scientists interpret their response. In their eyes, it is a case of response error (9, p. 10) or worse. According to Trevino, this means that they (the "inconsistent" respondents) "perceive Hispanic ethnicity to constitute a race" and "use of interviewer-observed race would result in most of this 40 percent being reclassified as White. . . . (M]ost of these Hispanics are White Hispanics who do not believe or understand they are White" (4, p. 70). In fact, as a casual reading shows, the question intended to elicit "race" is poorly constructed; it cues people for racial categories and national origin (2 1, p. 5). Trevino's interpretation is exceedingly problematic in its implications, and as patronizing as that offered by Tienda and Ortiz (21, pp. I 1-1 5): [T]hese individuals were likely to be Hispanics with ambivalent ethnic identities who misreported their origin either because they objected to the lack of response choices on the full-enumeration item (e.g., no Venezuelan, Argentine, etc, choices), or who deliberately denied their Hispanic origins . . . .[I]nconsistent "Hispanic" [sic] respondents . . . appear to exhibit ambiguity about their "Hispanicity" [sic]. Using the status inconsistency perspective, they suggest that it is desire to assimilate and upward mobility that presumably lead people to hide their ethnic identity: "ambiguity in their social identity derives from their desire to be recognized by the majority group (non-Hispanics) based on their socioeconomic credentials" (21, p. 15). This interpretation overlooks the politically constructed nature of the label (22, 23), which makes it unrealistic to expect a universally favorable reception. That many respondents chose to write their national origin where they did, while declining to accept a "Spanish/Hispanic origin," reveals, in all likelihood, neither error, ignorance, or an effort to hide an embarrassing "ethnic" identity, but rejection of the coercive nature of the self-identification question. The question forces respondents to agree to having "Spanish/Hispanic" origin, something which for a substantial number of people makes no sense, both in terms of their actual ancestry and/or in terms of their historical sense of who they are and/or (in the case of Latin Americans) their nationalist allegiance to their country of origin. The status inconsistency perspective is a subtle exercise in "majority" power. It "scientifically" neutralizes the assertion of an alternative identity (or the scholarly critique of the label, as the case may be), ignoring its historical structural determinants and reducing it to the effect of psychological states: i.e., suppression of, or ambivalence about, "real" "ethnic identity," an identity that exists mainly in the eyes of the "majority" beholder. On the other hand, the relatively low socioeconomic status of those who "agreed" to being labeled "Hispanics" and the fact that 74.2 percent of them were of Mexican American (59.8 percent) and Puerto Rican (14.4 percent) origin (21, p. 13)-two historically evolved U.S. minority groups with origins in colonial conquest-suggest that the ideological construction of reality to which people are exposed through the mass media and state power (via the Census itself) might be more effective among those politically less powerful because of their minority status within the U.S. ethnic stratification system. The preceding discussion highlights some of the problems entailed by the use of a self-identification question to enumerate an ethnic group that, in this case, is presumably a minority group. The Civil Rights approach to the amelioration of problems created not only by discrimination but by the normal functioning of the capitalist economy, generates self-interest in agreeing to being identified as a member of a minority group; under such conditions, "consistent responses" are far from yielding scientifically useful information (21, p. 20: 23). Given the characteristics of the population included under the umbrella term, the "Hispanic" label does not identify a minority group; it only adds together a variety of peoples, 25 percent of whom (Cubans, other Latin Americans, and a proportion of "Other Spanish") "have not lived in the U.S. for more than 400 years" and cannot claim to have been historically subject to racial discrimination and economic oppression in the United States. Like the Cuban bourgeoisie, or the small number of Argentine immigrants whose median household income in 1980 was higher than that of the native-born population (Table I ), some are, in fact, quite privileged. This situation has not remained unnoticed (9, p. 10; 1, p. 355): [N]ot all Hispanics agree that they themselves are part of a minority group, and some who claim minority status for themselves would reject it for certain others (for example, they might reject it for well-educated professionals who immigrate from South American countries). [Note 5: An interesting illustration of the effects of including well-educated South American professionals in the "Hispanic minority group" is the recent award of a minority fellowship, by Boulder's local newspaper, to a high school senior "minority" student, the talented and multilingual (speaking English, Spanish, French, and German) Argentine-born son of two Argentine university processors (24). This example clearly shows how the statistical definition of minorities makes a travesty of the concept and subverts the goals of policies devised to do away with discrimination.] Continued use of the term "Hispanic" [sic] or "Spanish Origin" [sic] denies the very basis upon which discrimination has been based, and confuses the basis for civil rights and affirmative action efforts. This situation is politically counterproductive; it sets the basis for political opportunism, it strengthens the perception of people in racial terms, and because it minoritizes foreign technical workers, scientists, and professionals, it creates a misleading appearance of minority advancement (6, pp. 46-52). Trevino dismisses this issue too lightly (4, p. 70); not only can foreigners legally seek minority status protection under affirmative action (though they are protected against discrimination in employment by the Civil Rights Act of 1866), but also minority status is routinely imposed upon them by employers, whether employees argue to being thus labeled and counted or not. As Lowry points out, "[T]ypically, an ethnic identity is assigned to each employee by his employer, based on whatever clues can be found in physiognomy, speech patterns, name and place of birth. Employees rarely know how they have been classified" (23, pp. 61-62). Third-party identification is another important source of data unreliability whose effects in the construction of "ethnicity" are, for all practical purposes, impossible to assess or eradicate. In vital statistics, both in birth and death certificates, third-party identification creates populations different from those identified by Census data. To make matters worse in terms of comparability and quality of data, there are three different methods to assign race/ethnicity to the newborn: the National Center for Health Statistics instructions, "Hispanic" parentage, and the race/ethnicity of the mother and father. Depending on the method, the "ethnicity" of the infant will vary; comparisons between "Hispanics" and non-"Hispanics" will yield different results (25, 26). Insistence on considering "Hispanic" anyone who has at least one "Hispanic" parent or ancestor betrays a remarkable obsession with racial purity and racial classification that should not remain unnoticed or escape criticism at a time when racism is, presumably, under attack. It also indicates allegiance to a reified concept of culture, as if it were genetically inherited. In the light of the problems examined here and in the preceding section, continued use of this label can only have political motivations; e.g., the cultural or racial legitimation of economic success or failure, or the belief-among some minority leaders- that greater numbers mean, necessarily, greater power. As the example of South Africa indicates, numbers and political strength are not necessarily equivalent. The differences among "Hispanics" are greater than their imputed commonalities; it is unlikely that they may become united as a single political force, although they may form local alliances around single-issue objectives (for an assessment of the political potential of the label, see 27). CONCLUSION The "Hispanic" label is eminently political: it identifies neither an ethnic group nor a minority group. It is the temporary outcome of political struggles between the major parties to win elections, particularly in the Southwest, and will serve its role as long as political alignments, the terms of acceptable political discourse, and the definition of legitimate channels of access to social and health services, education, and the road to upward mobility for minority groups remain unchanged. Central to the dominant political discourse is the notion that the "majority" has access to health, social services, education, and employment opportunities through the impersonal mechanism of market allocation. Minorities, on the other hand, are disproportionately poor, are less educated, earn lower incomes, and are relatively excluded from the better paying jobs not because they are also disproportionately working class (where they occupy the lower strata), unemployed, or underemployed, but because of their culture and nonmarket processes such as, for example, racial discrimination and segregation. Their problems, it follows, require nonmarket solutions (e.g., affirmative action and development of policies designed to maximize their access to needed social and health services, education, etc.) contingent upon the identification of the population whose needs have to be served. Individuals and groups, therefore, have to accept whatever legal identity and social status they are given to qualify for benefits and have access to legal protection; at tills time, that means accepting the "Hispanic" label. Public health officials, policy makers, and social scientists, on the other hand, are concerned with the quality and comparability of the data. However, as I have argued in this article, the label is far from being appropriate for social research, and for policy making and implementation; on the contrary, it has created an irresolvable tension between political and research needs that, in the long run, will result in ineffective policies and the accumulation of data of doubtful significance. Cafferty's and McCready's assessment of the label is correct (28, p. 254): [P]olicies are created for Hispanics which help some and harm others because there are . . . no "generic" Hispanics. . . . "Hispanics" [sic] is much too generic a term for policy makers and . . . much greater information and insight must be generated in order to enable our contemporary system to make intelligent and productive responses to the needs of these citizens and newcomers to the country. The real issue, in the last instance, is not whether Latino of "Hispanic" is a better umbrella term, but whether it is wise to have an umbrella term at all. In my view, the answer is self-evident. Regardless of politicians' concerns for numbers, social scientists and policy makers must seriously confront the problems attached to this and any other umbrella term: the stereotyping and minoritization of foreigners; the transformation of minority groups into mere statistical categories, thus subverting the historical reasons for their situation and their claims upon the resources of the state; the creation of a synthetic or artificial "ethnicity"; the production of data difficult to interpret in nonracist or stereotypical fashion, and so on. There is a simple alternative to the umbrella term: to acknowledge the existence of qualitative differences in history, culture, class, and social stratification, and racial/ethnic composition of populations that ought to be publicly named by their real historical names, and understood (through social research) and treated (through social and health policies) in their own right. These populations are the following: +Two minority groups: people of Mexican and Puerto Rican descent who, because of the historical conditions surrounding their entry into the United States, the integration of the U.S. economy with that of Mexico and Puerto Rico, and the presence of migration flows and counterflows, constitute two special populations with features and problems of their own. Because of the heritage of economic exclusion and racial discrimination, it is in their context that affirmative action makes sense. +Four additional aggregates: Cuban immigrants, Central America refugees, Central American immigrants, and South American immigrants. Privileged classes within these immigrant populations do very well; those who do not, do poorly because of lack of human capital. The determinants of the social stratification and the needs of these populations are different from those subject to generations of racial/ethnic discrimination. Policies designed to serve their needs should, therefore, differ from those designed to ameliorate the historical effects of discrimination (29). These six populations are themselves stratified on the basis of class and socioeconomic status; this ought to warn social researchers and policy makers against making generalizations, assuming a common culture or "ethnicity" as the major explanatory variable. To identify these populations in terms of national origin is easy (except in the case of undocumented workers and refugees) and, politically, less laden with racist innuendos than the effort to minimize their differences by placing them all under a common label. The need to identify Spanish-speaking populations to provide better health care and other social services does not justify the use of a label that, because it racializes national origin and triggers the perception of recipients in terms of stereotyped "Hispanic" traits, may generate a "blaming the victim" understanding of their problems and the provision of low quality services. In the last instance, access to good health care is not a function of race, ethnicity, or language skills; it is a function of social class and location in the socioeconomic stratification system, a social science truism that bears repeating over and over in a social, academic, and policy-making context that downplays the existence of class differences and their impact upon people's life chances. Advocates of a standardized terminology should assess its short-term benefits for gathering data of doubtful quality in the light of its long-term political, ideological, and scientific costs, and give it up. Acknowledgments - I would like to thank Estevan T. Flores, Benjamin F. Hadis, and Richard Rogers for their helpful suggestions. 1 Hayes-Bautista, D. E. Identifying "Hispanic" populations: The influence of research methodology on public policy. Am. J. Public Health 70: 353-356, 1980. 2. Hayes-Bautista, D. E. On comparing studies of different raza populations. Am. J. Public Health 73: 274-276, 1983. 3. Hayes-Bautista, D. E., and Chapa, J. Latino terminology: Conceptual basis for standardized terminology: Conceptual basis for standardized terminology. Am. J. Public Health 77: 61-68, 1987. 4. Trevino, F. M. Standardized terminology for standardized populations. Am. J. Public Health 77: 69-72, 1987. 5 .Yankauer, A. Hispanic/Latino-What's in a name? Am. J. Public Health 77: 15-17, 1987. 6. Gimenez, M. E. Minorities and the world-system: Theoretical and political implications of the internationalization of minorities. In Racism, Sexism, and the World-System, edited by J. Smith et al., pp. 39-56. Greenwood Press, Westport, Conn., 1988. 7. Trevino, F. M. Uniform minimum data sets: In search of demographic comparability. Am. J. Public Health 78: 126-127, 1988. 8 .Westermeyer, J. Problems with surveillance methods for alcoholism: Differences in coding systems among federal, state, and private agencies. Am. J. Public Health 78: 130-133, 1988. 9. Sullivan, T. A. A demographic portrait. In Hispanics in the United States, edited by P. S. J. Cafferty and W. C. McCready, pp. 7-32. Transaction Books, New Brunswick, N.J., 1988. 10. Petersen, W. Who's what? 1790-1980. Wilson Q. 9: 110, 1985. 11. National Center for Health Statistics. Births of Hispanic parentage, 1983 and 1984. Monthly Vital Statistics Report. NCHS, Washington, D.C., 1987. 12. Hispanic population increases at 5 times rate of rest of U.S. Daily Camera (Boulder), September 11, 1987, pp. 7A-9A. 13. Beck, J. Hispanic immigration tests U.S. Daily Camera (Boulder), September 11, 1987, p. 4A. 14. Mott, F. L. The pace of repeated childbearing among young American mothers. Fam. Plann. Perspect. 18: 5-12, 1986. 15. Mosher, W. D., et al. Religion and fertility in the United States: The importance of marriage patterns and Hispanic origin. Demography 23: 367-379, 1986. 16. Mamdani, M. The Myth of population control. Monthly Review Press, New York, 1972. 17. Mamdani, M. The ideology of population control. In And The Poor Get Children, edited by K. L. Michaelson, pp. 39-49. Monthly Review Press, New York, 1984. 18. Cafferty, P. S. J., and McCready, W. C. Introduction. In Hispanics in the United States, edited by P. S. J. Cafferty and W. C. McCready, pp. 1-6. Transaction Books, New Brunswick, N.J., 1988. 19. Cafferty, P. S. J. The "new" immigration. In Hispanics in the United States, edited by P. S. J. Cafferty and W. C. McCready, pp. 33-48. Transaction Books, New Brunswick, N.J., 1988. 20. Giachello, A. 1. Hispanics and health care. In Hispanics in the United States, edited by P. S. J. Cafferty and W. C. McCready, pp. 159-194. Transaction Books, New Brunswick, N.J., 1988. 21. Tienda, M., and V. Ortiz. "Hispanicity" and the 1980 Census. Soc. Sci. Q. 67: 3-20, 1986. 22. Choldin, H. M. Statistics and politics: The "Hispanic" issue in the 1980 Census. Demography 23: 403-418, 1986. 23. Lowry, 1. S. The science and politics of ethnic enumeration. In Ethnicity and Public Policy, edited by W. A. Van Horne, pp. 42-62. University of Wisconsin System, Milwaukee, Wis., 1984. 24. Daily Camera (Boulder), April 1988. 25. Sullivan, T. A., et al. Alternative estimates of Mexican American mortality in Texas. Soc. Sci. Q. 65: 609-617, 1980. 26. Rogers, R. G. Ethnic Differences in Infant Mortality: Fact or Artifact? Unpublished manuscript, 1988. 27. Mufioz, C. Chicano pohtics: The, current conjuncture. In The Year Left 2, An American Socialist Yearbook, pp. 35-52. Verso, New York, 1987. 28. Cafferty, P. S. J., and McCready, W. C. Conclusion. In Hispanics in the United States, edited by P. S. J. Cafferty and W. C. McCready, pp. 253-257. Transaction Books, New Brunswick, N.J., 1988. 29. Bodas, G. J. Jobs and employment for Hispanics. In Hispanics in the United States, edited by P. S. J. Cafferty and W. C. McCready, pp. 146-157. Transaction Books, New Brunswick, N.J., 1988. Direct reprint requests to: Dr. Martha E. Gimenez Department of Sociology University of Colorado Boulder, CO 80309