Latino Panethnicity - Reality Or Methodological Construction

LATINO PANETHNICITY: REALITY OR METHODOLOGICAL CONSTRUCTION?* EILEEN DIAZ MCCONNELL University of Illinois at Urbana-Champaign

SOCIOLOGICAL FOCUS Vol. 37 No.4: 297-312 November 2004

EDWARD A. DELGADO-ROMERO Indiana University

In Census 2000, approximately 15 percent of all Latinos identified panethnically; that is, they did not mark whether they were Mexican, Puerto Rican, Cuban, or another na.tional group but rather identified themselves in general terms such as "Hispanic" or "Latino." This was a 200 percent increase over the 1990 Census. Scholars have pointed to two explanations for Latino panethnic identification: first, that panethnicity is a meaningful identity stemming from shared backgrounds and structural commonalities and the need to unite politically and, second, that Latino panethnicity is a methodological artifact of the way that racial and ethnic data are currently collected. In this paper, we describe past research on Latino ethnic identity, document contemporary expressions of Latino panethnicity, and assess whether panethnicity is a genuine, meaningful identity for some Latinos and stems from methodological factors. Our assessment of Census 2000, the Census 2000 Supplementary Survey, the 2002 National Survey of Latinos, and the Alternative Questionnaire Experiment is that Latino panethnicity is both a substantive and a methodological phenomenon. Consequently, we encourage researchers to acknowledge the complexity and fluidity of Latino ethnicity, to recognize the influence of substantive and methodological factors on Latino panethnicity, and to incorporate multiple data sources in their research.

Census 2000 data have been used extensively in recent discussions about the U.S. Latino or Hispanic population, those persons described by the Office of Management and Budget (1997) as "Cuban, Mexican, Puerto Rican, South or Central American, or other Spanish culture regardless of race." Researchers, politicians, and members of the media have employed Census 2000 data to describe many recent trends, including the outnumbering of the African American population and the large growth of the Latino population in states such as North Carolina and Georgia. Another trend is of note: Approximately 15.3 percent of all Latinos identified panethnically in Census 2000, a 200 percent increase over the 1990 Census (Suro 2002). Panethnicity refers to the consolidation of groups who have had previously distinct ethnic or national identities "into a single racial or in the case of Latinos, * An earlier version of this paper was presented at the annual meeting of the American Sociological Association in August 2004 in San Francisco, CA. Please direct all inquiries to the first author at the Department of Sociology, University of Illinois at Urbana-Champaign, 326 Lincoln Hall, MC-454, 720 S. Wright, Urbana, IL 61801. E-mail: [email protected] paper was prepared with partial support by the U.S. Census Bureau to the first author. The authors gratefully acknowledge the technical assistance of Betsy Guzman, Roberto Ramirez, Kevin Deardorff, the suggestions of anonymous reviewers, Jorge del Pinal, Luis Falcon, and Guest Editors Rogelio Saenz and Edward Murguia.

297

298

SOCIOLOGICAL FOCUS

ethnic, category" (Omi 1999, p. 29). That is, in Census 2000, Latinos who identified panethnically did not mark whether they were Mexican, Puerto Rican, Cuban, or another national group. Instead, they checked the box for "other" and either did not write in a group or wrote in general terms such as "Hispanic" or "Latino." Researchers generally approach Latino panethnic identity in two ways. In the first approach, researchers consider Latino panethnicity to be a genuine identity stemming from shared commonalities and the struggle for power (e.g., Jones-Correa and Leal 1996; Landale and Oropesa 2002; Moore and Pachon 1985; Murguia 1991). Consequently, researchers have traced the development of panethnic identity among Latinos, examined whether it exists simultaneously with national origin identities, and if national groups with origins other than Spain, such as Brazilians, identify as Latino (Calderon 1992; Gimenez 1989; Hayes-Bautista and Chapa 1988; Itzigsohn and Dore-Cabral 2000; Lopez and Espiritu 1990; Marrow 2003; Padilla 1984; Portes and MacLeod 1996; Sellers et al. 1998). Thus, researchers treat Latino panethnicity as a meaningful construct. On the other hand, some researchers have argued that Latino panethnicity does not represent a valid type of identity (Gimenez 1992; Massey 1993). Instead, panethnicity represents a political and statistical construction, predominantly resulting from methodological issues in Census 2000 (Logan 2001; Martin 2002; Suro 2002). Their work, described later in the paper, suggests that Latino identification with panethnicity is driven by methodological factors. This paper addresses two important questions: Do U.S. Latinos, perhaps as a byproduct of acculturation and assimilation, forego distinct national, regional, cultural, and ethnic differences and embrace panethnic terms such as Latino or Hispanic? Alternatively, is Latino panethnicity a methodological artifact of the way racial and ethnic data are collected in the United States? To date, there has been no empirical research that simultaneously explores how substantive and methodological issues influence the panethnic identification of Latinos. We argue that considering both substantive and methodological arguments and employing census and non-census data sources is useful for understanding contemporary Latino ethnic identity. The first analyses compare panethnic responses in Census 2000 and an expanded sample of the American Community Survey (ACS), the Census 2000 Supplementary Survey (C2SS). Next, we incorporate relevant results from the 2002 National Survey of Latinos and the Alternative Questionnaire Experiment (AQE) to shed further light on the topic. Reviewing the patterns of Latino identification in a variety of surveys is a more comprehensive method for understanding the ethnic responses that Latinos actually use in different contexts. Finally, we discuss the implications of our results for future research about Latino identity. It is important to understand expressions of Latino ethnic identity, especially panethnicity. First, U.S. society has become increasingly diverse, both racially and ethnically. Race has long been understood in the United States as binary: whites and nonwhites. Naturally, the historical and numerical presence of Mrican Americans in the United States means that they have long been conceptualized as the predominant nonwhite group. However, the dramatic and visible growth of Latinos in the United States since the 1950s complicates the binary racial system in the nation. Further, census projections indicate that the Latino population will continue to increase; it is estimated that Latinos will comprise nearly 25 percent of the U.S. population by 2050

LATINO PANETHNICITY: REALITY OR METHODOLOGICAL CONSTRUCTION?

'.

299

(U.S. Census Bureau 2004). Therefore, it is important to understand the ways that this important group is identified in research, how they identify themselves, and the implications of this identification for public policy and research. Second, there are broader societal implications of Latino panethnicity. If some Latinos use that term to describe themselves, is it due to recognized commonalities between Latino groups, increasing racial and ethnic intermarriage, or other forces? In addition, researchers, members of the media, and politicians commonly use terms such as "Hispanic" and "Latino" as if these terms are meaningful. Researchers also commonly combine diverse Latino groups into one group, a practice that raises the possibility that within-group differences of diverse populations are ignored (Murguia 1991). Additional systematic research helps determine for whom and in what contexts these categories can and should be used. Third, the Office of Management and Budget (OMB) determines the minimum standards for the collection of race and ethnic data by the federal government. Though it has been an interactive process, these standards mean that the federal government essentially decides how race and Hispanic origin are categorized and officially recognized in the United States. Consequently, decisions made by the OMB have far-reaching implications for demography, including how the U.S. population will be described and the number and type of racial and ethnic options that individuals can select. Finally, Census 2000, the C2SS, and other data sets are and will continue to be crucial sources of information for addressing important research questions about Latinos and other groups. Both public and private entities consider Census 2000 to be the predominant authority on demographic characteristics of populations in the United States in 2000. For example, Census 2000 has been used to identify the extent of Latino residential segregation and to confirm the increasing geographic dispersion of native and foreign-born Latinos in different areas of the United States (Guzman and McConnell 2002; Kandel and Cromartie 2003; Logan 2002; Singer 2004). Moreover, politicians, community organizations, researchers, and others use these data sources to make crucial political and economic decisions and to implement programs and policies targeted to specific racial and ethnic groups. Clearly, it is critical to understand how such data sources represent the ethnicity of Latinos living in the United States and how Latinos identify in light of the options available to them. RESEARCH ON ETHNIC IDENTITY AND PANETIINICITY

According to social identity theory, humans acquire particular social identities via membership in groups (Tajfel and Turner 1986). In turn, humans use these memberships to define their identities and status in society. One type of social membership, ethnicity, has three important dimensions: culture, identity, and minority status (Phinney 1996). The cultural dimension is based on shared historical or cultural characteristics, values, and customs (Min 1999; Parsons 1975). However, as with race, ethnicity often has different connotations and consequences for whites than for ethnic minorities. For whites, ethnicity could be viewed as a choice, a symbolic identity (Waters 1990). In contrast, non-whites often do not have this choice. Instead, ethnicity is an imposed identity equated with subordination,

300

SOCIOLOGICAL FOCUS

inferiority, minority status and marginalization (Trimble, Helms, and Root 2003). For many Latinos, even those who might choose to identify as white, their ethnic identity is commonly imposed by others. Some have argued that Latino panethnicity, the consolidation of a group formerly differentiated by distinct ethnic identities into one larger group, has been enforced by outsiders. Indeed, many characteristics categorized as Hispanic and/or Latino are generalizations that rely on overly broad categorizations of ethnic and cultural minorities, or "ethnic gloss" (Trimble 1991). The use of ethnic glosses can result in gross misrepresentations and serve to homogenize and understate the diversity within the population (Moore 1990; Oboler 1992). Some researchers have gone even further. Vidal de Haymes, Kilty, and Haymes (2002) have tied panethnicity to the racialization process in the United States, as the state benefits from recognizing and responding to one large bloc, rather than multiple groups. Others, however, believe that Latino panethnicity derives from Latinos themselves (e.g., Moore and Pachon 1985; Murguia 1991). For example, the Mexican philosopher Jose Vasconcelos (1957) argued that the synthesis and integration of diverse racial identities among Latinos and Latin Americans has led to a shared cultural identity, a cosmic race: La Raza Cosmica. More recently, scholars have generally pointed to Latino panethnicity as stemming from the following factors: (1) shared cultural and historical backgrounds; (2) racial mixing that resulted in mestizo (racially mixed) people; (3) structural commonalities such as race (Lopez and Espiritu 1990) resulting from the "continuing importance of race and the persistence of racial lumping in American society" (Espiritu 1992, p. 175); (4) an ambivalent relationship with U.S. foreign policies and political intervention in the history of Caribbean and Central and South American countries (Delgado-Romero and Rojas forthcoming; Garcia-Preto 1996); and (5) the increasing assimilation and incorporation of postimmigrant generations of Latinos in the United States (Jones-Correa and Leal 1996; Portes and MacLeod 1996). Researchers have examined whether panethnic identity may exist simultaneously with national origin identities for Latinos. In general, the research has shown that most Latinos prefer to identify with a specific country rather than with a general term (de la Garza et al., 1992; del Pinal and Singer 1997). However, researchers have also documented the presence of situational, shifting, and overlapping ethnic identities for Latinos, including panethnic and national origin identities (Calderon 1995; Cuello 1998; Rodriguez 2000). Indeed, ethnic identities may shift depending on the situation and/or context, as persons identify with numerous roles depending on the social interaction. Individuals hold multiple relevant and meaningful identities simultaneously (e.g., Greeley 1974; Okamura 1981; Yancey, Ericksen, and Juliani 1976). Thus, Latinos may identify with a variety of descriptors, including panethnic, national, and regional labels, but depending on the context may be more likely to provide one label or the other. In this case, panethnic identification may be a genuine reflection of how some Latinos want to identify at a particular point in time. Moreover, the social context in the United States for Latinos has changed in recent years. The growing numbers of Latinos and the popularity of Latino musicians, actors, and sports personalities, coupled with the increasing attention paid to the political and economic impacts of Latinos, may make it more socially acceptable to

LATINO PANETHNICITY: REALITY OR METHODOLOGICAL CONSTRUCTION?

301

identify panethnically than in earlier decades. Irrespective of the exact reason, many would argue that the panethnic responses of some Latinos reflect deliberate and meaningful choices. DATA AND METHODS

This paper examined the substantive and methodological arguments for Latino panethnicity using four sources of data. First, we compared Census 2000 and C2SS data about the panethnic Hispanic population. Comparison of these two data sources was especially useful for exploring the impact of methodological issues on panethnic identification because both data sources used identical wording regarding Hispanic origin but differed methodologically. The Census 2000 data used here came from an internal Census Bureau file that has been restricted to the household, non-institutionalized civilian population so that the population universe matched C2SS data. (See McConnell and Guzman 2003 for more details on this file.) The figures for each data set have been rounded to the nearest thousand for ease of interpretation. Standard errors and 95 percent confidence intervals were calculated for C2SS data because this data source included sampling error. The comparison of panethnic responses in Census 2000 and C2SS with other data showed how Latinos responded to a variety of questions about ethnic identity. Therefore, we included two other sources of data that provided information about Latino panethnicity: the 2002 National Survey of Latinos (Pew Hispanic Center/Kaiser Family Foundation 2002) and the Alternative Questionnaire Experiment (Martin 2002).

CENSUS 2000

The U.S. Gonstitution calls for a counting of the U.S. population every 10 years for the purposes of political apportionment. The collection of data about Latinos in the decennial census has changed since census takers first categorized individuals of Mexican origin as a racial group in 1930 (Chapa 2000; Lee 1993; U.S. Bureau of the Census 2000). In 1977, the Office of Management and Budget (OMB) developed five basic racial categories and two ethnic categories, "Hispanic origin" and "not of Hispanic origin," to be used in 1980 and 1990. Twenty years later, the OMB revised the standards to collect information that better "reflects the increasing diversity of our Nation's population stemming from growth in interracial marriages and immigration" (OMB 2000, p. 6). The 1997 revisions to the federal standards indicated that respondents could select one or more of five revised racial categories: American Indian or Alaskan Native, Asian, black or African American, Native Hawaiian or Other Pacific Islander, or white. As in the 1977 standards, individuals selected whether they were Hispanic. The text of the Hispanic question in Census 2000 was, "Is this person Spanish/Hispanic/Latino? Mark the "No" box if not Spanish /Hispanic /Latino." Respondents could choose between a response indicating that they were not Hispanic or that they were Mexican, Mexican American, Chicano, Puerto Rican, Cuban, or other Spanish/Hispanic/Latino. This question was asked before the question about race.

.........

LATINO PANETHNICITY: REALITY OR METHODOLOGICAL CONSTRUCTION?

303

Table 1 shows that Census 2000 counted nearly 10 million "other Hispanic or Latinos," that is, those who responded they were not Mexican, Puerto Rican, or Cuban. Approximately 6.1 million Latinos, nearly 16 percent of all Latinos, did not identify with a Latin American country, but rather checked the box for "Yes, other Spanish/Hispanic/Latino" and left the provided boxes blank or else wrote in general terms such as "Hispanic" or "Latino." Another 2.1 percent of all Latinos were categorized as "all other Hispanic," as they provided responses that were not elsewhere classified, such as "Tejano." Table 1 also documents that Latinos who wrote in a panethnic term were far more likely to write in "Hispanic" than any other term. Indeed, approximately 7.0 percent of all U.S. Latinos wrote in this term. The other terms, "Spanish" and "Latino," were far less common. Finally, 5.0 percent of all Latinos, approximately 1.7 million, identified as "other" Hispanic by checking the box but did not specify further. CENSUS 2000 SUPPLEMENTARY SURVEY

Pinpointing the differences between panethnic responses in Census 2000 and C2SS and the direction of such differences was useful because the same Hispanic origin question was asked in each. Other aspects of the research designs, such as methods to reduce non-response, differed between the data sources. The purpose of C2SS was to test whether it was possible to collect census long-form type data about the U.S. population more frequently than every 10 years. Data were collected from approximately 700,000 households between January and December of 2000. Table 1 shows that approximately 8.2 million Latinos, or 23.7 percent of all Latinos, reported that they were Spanish/Hispanic/Latino but not Mexican, Puerto Rican or Cuban. More importantly, 9.7 percent of all Latinos provided responses that were not associated with a Latin American country in C2SS. Of these, approximately 1.1 million identified as "Spaniard," "Spanish," or "Spanish American," with "Spanish" as the most common response. Another 2.3 million individuals responded in ways that the U.S. Census Bureau categorized as "all other Hispanic." Unfortunately, the C2SS data are not disaggregated further, so it is not possible to determine what proportion wrote in terms such as "Hispanic" or "Latino." THE 2002 NATIONAL SURVEY OF LATINOS

The 2002 National Survey of Latinos (NSL), conducted by the Pew Hispanic Center/Kaiser Family Foundation (PHC/KFF 2002), is especially useful for studying the ethnic identification of Latinos. The 2002 NSL randomly selected a sample of approximately 3,000 Latino respondents and was weighted to be representative of Latino adults in the United States. The survey, administered by telephone between April and June of 2002, asked about preferences in identifying with national origin and panethnic terms. The question used to identify respondents as Latino was: "Are you, yourself, of Hispanic or Latin origin or descent, such as Mexican, Puerto Rican, Cuban, Dominican, Central or South American, Caribbean or some other Latin background?" The 2002 NSL collected more information than either Census 2000 or C2SS about the ethnic identity of Latinos, including preferences in identification and if

304

SOCIOLOGICAL FOCUS

Latinos ever used other terms to describe themselves. Indeed, the unique contribution that this data source provides is that it probed beyond the first answer, allowing for multiple responses to the ethnic identity question. The left side of Table 2 shows that approximately 22.0 percent of Latinos identified with origins other than Mexican, Puerto Rican or Cuban, and the majority of those identified with a specific Latin American country. Six percent of Latinos identified in other ways: 2.8 percent with Spain and 3.2 percent with other types.

The 2002 NSL also solicited information about which terms that Latinos first or only provide when asked about ethnic identity. The right side of Table 2 shows that approximately 54.0 percent of Latinos first or only identify with a country of origin; however, 24 percent would be more likely to use "Latino" or "Hispanic" primarily to identify themselves. The survey probed further, asking respondents whether they ever described themselves using terms such as "Latino or Hispanic" or "American." The results show that the majority of Latinos at some point in time or in some contexts identify with a country of origin. However, as Table 2 shows, even Latinos who identify with a country of origin are also likely to describe themselves using terms such as "Latino" or "Hispanic." Thus, the majority of Latinos do identify in both specific and in general ways. The survey also documented that the majority of respondents did not have a preference between the two primary panethnic terms, Hispanic or Latino. Of those who did have a preference, individuals were nearly three times as likely to prefer Hispanic over Latino. THE ALTERNATIVE QUESTIONNAIRE EXPERIMENT

The U.S. Census Bureau conducted the Alternative Questionnaire Experiment (AQE), an experiment that compared responses to mailed, short-form questionnaires that used the 1990-style and the Census 2000style items. This experiment was illuminating because the Census 2000 Hispanic question was different from the 1990 question in several important ways: (1) The Hispanic origin question was placed before the race question; (2) the term "Latino" was added to the question; (3) the examples of "other" Hispanic groups such as Argentinean were dropped; (4) the

uH--

LATINO PANETHNICITY: REALITY OR METHODOLOGICAL CONSTRUCTION?

305

format shifted from matrix to individual person spaces; and (5) the instructions changed from "Is this person of Spanish/Hispanic origin?" to "Is this person Spanish/Hispanid Latino?" (Martin 2002). Approximately 10,500 randomly selected U.S. households received the 1990-style short forms, and approximately 25,000 randomly selected households received the Census 2000-style short-form questionnaires. (See Martin 2002 for details about the AQE.) Using information adapted from Martin (2002), we found that Latinos sampled in the AQE overwhelmingly preferred to identify with a specific group: Approximately 70.0 percent and 73.0 percent of those responding to the 1990-style census short form and the 2000-style census short form, respectively, provided a specific ethnicity. Few identified panethnically: Only about 5.2 percent of all Latinos who responded to the 1990-style question wrote in terms such as "Hispanic," "Latino" or "Spaniard." However, about 12.2 percent of Latinos responding to the 2000-style question wrote in such terms. Further, approximately 5.0 percent (1990-style form) and 7.3 percent (2000-style forms) of Latinos identified as Hispanic but either did not write in a term or wrote in terms that were not categorized. Overall, this experiment indicated that the Census 2000 Hispanic question solicited far more panethnic responses than the 1990 question. Nevertheless, a substantial proportion of the population using the 1990-style question, 5.0 percent, identified in panethnic terms. This is not an insignificant proportion.

REALITY OR METHODOLOGICAL CONSTRUCTION?

Researchers have pointed to two potential explanations for Latino panethnic identification. As described earlier, some Latino respondents might not write in a group on Census 2000 forms as a deliberate choice. On the other hand, Latino panethnic responses could be primarily influenced by methodological factors. PANETHNICITY AS A GENUINE IDENTITY

The contemporary data sources surveyed here provide important clues about the frequency and validity of panethnic identifications for Latinos. Every data source documents that some proportion of Latinos identify panethnically. For example, Census 2000 indicates that 8.6 percent of all Latinos write in terms such as "Hispanic" or "Latino" while, according to C2SS, 6.6 percent of all Latinos do so. Moreover, the 2002 NSL estimates that 24.0 percent in the sample first or only use panethnic terms such as "Latino" or "Hispanic," while 81.0 percent have ever used panethnic terms to describe themselves. The AQE shows that 5.2 percent of Latinos using the 1990 style question wrote in terms such as Hispanic, Latino, or Spanish. Clearly, the 2002 NSL provides the most convincing evidence that contemporary panethnic labels are used by Latinos in addition to national origin labels. In fact, the NSL results are even stronger when viewed simultaneously with Jones-Correa and Leal's (1996) reexamination of the Latino National Political Survey (LPNS) conducted in 1989 and 1990 by de la Garza and colleagues (1992). Jones-Correa and Leal (1996) found that 41.7 percent of Latinos in 1989 and 1990 had used a panethnic identifier at some time. In 1989-1990, 2.9 percent of LPNS respondents provided a panethnic identifier as their first and only identifier (Jones-Correa 1996), but in 2002, fully 21.0

306

SOCIOLOGICAL FOCUS

percent of Latinos used terms such as "Latino" or "Hispanic" first or only to describe themselves. Such data suggest that Latinos are using panethnic labels more now than even a decade ago. Further evidence that Latino panethnicity is a meaningful construct is the variety of panethnic terms that Latinos use. Census 2000 and the 2002 NSL show that "Hispanic" is a much more commonly used panethnic term than "Latino," a finding confirming past studies showing that "Hispanic" tends to be the most common panethnic label used by Latinos (Granados 2000; U.S. Department of Labor 1995). Explanations for its popularity include the longer history of this term and the federal government's standard use of the term since the early 1970s (Flores-Hughes 1996). "Latino" is also a meaningful term, though less common. The label typically refers to those of Latin American descent living in particular areas of the United States and could be considered to be more inclusive, politicized, and politically progressive than the term "Hispanic", (Comas-Diaz 2001; Cuello, 1998; Hayes-Bautista and Chapa 1987). Patterns of preference for a particular ethnic designation vary by region and state in the United States, also indicating that panethnicity is a meaningful identity for Latinos. For example, according to Census 2000, Texas Latinos account for approximately 19.0 percent of the total U.S. Latino population, but are 34 percent of all U.S. Latinos identifying as "Hispanic." Texans overwhelmingly prefer the term "Hispanic" (804,000) to "Latino" (35,000). The pattern also holds in New Mexico and California. However, those in other states, such as New York, prefer the "Latino", label. Indeed, Latinos in New York comprise 8.1 percent of all U.S. Latinos but were 31.7 percent of those writing in the term on the census form. The 2002 NSL data showed a less consistent pattern: 53 percent of Latinos did not have a preference for one label over the other. Clearly, researchers must further explore Latino preferences for such terms and connections of such labels to individual and group identity and experience. PANETHNICITY AS A METHODOLOGICAL CONSTRUCTION

The U.S. Census Bureau conducted numerous evaluations of the race and ethnicity data in Census 2000 (del Pinal 2004). At least four studies to date suggest that methodological factors account for the large numbers of Latinos who provided panethnic responses in Census 2000 (Cresce and Ramirez 2003; Logan 2001; Martin 2002; Suro 2002). These researchers noted that panethnic responses in 2000 were most likely elicited because of the format and wording of the question and other methodological issues. For example, AQE data strongly suggests that that the responses to the Hispanic origin question in Census 2000 are an artifact of the question. Martin (2002) suggested that such results were "due to the combined effects of the modified question wording, the vague 'print group' instruction, and the elimination of examples in the Census 2000 questionnaire" (p. 592). Simulations and extrapolations from several data sets also suggest that Latino panethnic responses are methodologically driven. For example, Suro (2002) applied the national origin distributions from C2SS data to the total Hispanic population size counted by Census 2000 to derive alternative numbers of Hispanic groups. He found that applying C2SS estimates to Census 2000 counts would reduce the proportion of

LATINO PANETHNICITY: REALITY OR METHODOLOGICAL CONSTRUCTION?

307

persons identifying in a panethnic way in Census 2000 by about 45.4 percent. He concluded that a substantial proportion of those who identified panethnically in Census 2000 were actually Mexican, with others coming from countries such as Guatemala, Honduras, and EI Salvador. This finding was consistent with the results from Logan's (2001) application of pooled samples from the 1998 and 2000 Current Population Survey to the Census 2000 results at the census tract level. Logan's simulation showed that large proportions of "other Hispanics" in Census 2000 were Mexicans, Dominicans, Salvadorans, Colombians, and Guatemalans. In another simulation, Cresce and Ramirez (2003) used place of birth and ancestry questions from Census 2000 sample data to simulate the national origin of those who reported a panethnic term in the Hispanic origin question in Census 2000. They found that the most common simulated origin of those who identified in panethnic terms was Mexico. Their research suggested that factors influencing the panethnic responses in Census 2000 included both the format of the question and other contextual factors such as the "census effect," a phenomenon described by Tortora, Miskura, and Dillman (1993) as the "combination of media attention, advertising, and cultural sense of participation that seems to build during each census year" (p. 123). Thus, studies that explore the impact of methodological factors on responses to the Hispanic origin question have tested different wording and examples, applied estimates derived from large-scale surveys, or conducted simulations using other available Census 2000 data (Martin 2002; Suro 2002). Yet, unlike the current study, none of this research compared Hispanic group data in Census 2000 with other data sources with identical population universes. A reexamination of Table 1 is warranted. The standard errors and z scores (not shown) calculated for C~SS show that there are statistically significant differences between Census 2000 counts and C2SS estimates of the "other than Mexican, Puerto Rican, or Cuban" and "other than Latin American" groups, among others. Approximately 9.7 percent of Latinos in C2SS identified panethnically compared to 17.7 percent in Census 2000. That these differences are statistically significant is important because Census 2000 and C2SS questionnaires used the exact ordering and wording of the Hispanic question. Neither question in the mailed questionnaire included examples beyond Mexican, Puerto Rican, and Cuban, a point that undoubtedly led to more respondents providing panethnic labels with this question than if the -additional examples provided in the 1990 Census had been used instead Martin (2002). Clearly, question wording can not be the only explanation for Latino panethnic identification, because if it fully explained panethnicity, then Census 2000 and C2SS would have relatively similar panethnic patterns. Additional methodological factors may playa role. (See McConnell and Guzman 2003 for a detailed discussion of methodological differences between Census 2000 and C2SS.) The most important differences between the two data sources are the predominant modes of data collection and strategies to reduce non-response. Although Census 2000 and C2SS were primarily mail-back surveys, both surveys used phone and personal contact to increase response rates. The majority of respondents to Census 2000 filled out forms mailed to their homes without assistance from a Census field representative. However, C2SS procedures called for in-person interviewing of one-third of those who did not respond to the initial form; thus, a much larger proportion of the total C2SS sample had contact with field

308

SOCIOLOGICAL FOCUS

representatives, who were instructed to prompt for specific responses to each question. The C2SS in-person interview question did include examples of Hispanic groups (del Pinal 2004). Consequently, C2SS participants were prompted for specific responses while the majority of Census 2000 respondents filled out the forms without interacting with field representatives. In addition, Census 2000 had vast resources to reduce coverage error, including a carefully constructed and frequently updated Decennial Master Address File and a combination of mail, phone, and personal contacts to encourage households to return the census forms (U.S. Department of Commerce 2000). Census 2000 also worked with community organizations and local governments and used a private marketing campaign targeted specifically to Latinos and other racial and ethnic groups (U.S. Department of Commerce 2000). C2SS procedures called for fewer strategies to encourage participation but did include the mailing out of pre-notices, reminder notices, second mailings of the survey, and attempts to administer the survey by phone and in person to those who had not responded. OUR ASSESSMENT

Based on our review of the data sources included in this paper, we believe that Latino panethnicity is a genuine identity for some Latinos. We base this judgment on a number of factors. First, every data source, irrespective of mode, possible prompting, and wording, shows that some proportion of Latinos identify panethnically. The 2002 NSL provides the strongest support for this judgment: The overwhelming majority of Latinos have used panethnic labels to describe themselves. Latinos actually use panethnic terms. Second, Latinos tend to prefer one term over another, suggesting that the choice of label is an important one. Third, Census 2000 data show that there are variations by state in the preference of either "Hispanic" or "Latino." Variations in labels and by states suggest that these labels are deliberate identifications by some Latinos. Fourth, past research reviewed here indicates that Latinos often do hold panethnic identities simultaneously with national origin identities. Nevertheless, we believe that much of the Latino panethnic reporting in Census 2000 was due to methodological factors associated with how the Hispanic origin question was asked. The most persuasive evidence is of the AQE results showing a dramatic drop in panethnic responses when using the 1990 Census-style question compared to the Census 2000-style question. We agree with Martin (2002), who argued that the Census 2000 wording of the question and the lack of examples seemed to encourage respondents to give a panethnic response. And yet, C2SS data showed that far fewer Latinos gave responses other than Mexican, Puerto Rican, Cuban, or other Latin American origins (3.3 million) compared to Census 2000 (6.1 million), even though both used the same Hispanic question in the mailed survey. This finding suggests that additional methodological issues, such as mode of data collection, may also account for some of the panethnic responses. Perhaps most importantly, this paper confirms the difficulty of empirically "pinning down" Latino ethnic identities. Clearly, using one question to study racial or ethnic identity is problematic. It is not realistic to think that we can capture the essence or complexity of people's ethnic identity from one closed-ended question.

LATINO PANETHNICITY: REALITY OR METHODOLOGICAL CONSTRUCTION?

309

Using participants' self-reported race and/or ethnicity as the sole means to infer an understanding about the psychological construct of identity in research is not a scientifically sound strategy (Helms 1994). Instead, researchers should aspire to find the salience, centrality, or meaning of race or ethnicity to research participants (e.g., Sellers et al 1998; Sellers et al 1997). Clearly, studies such as 2002 NLS are important for learning more about the highly contextualized and fluid nature of Latino identities. Further, our review of different data sources suggests that the best way to address questions of Latino ethnicity and panethnicity is to triangulate using several different sources of data. Employing just one data source, especially Census 2000, is inappropriate for exploring the ethnic identification of Latino groups. The purpose of the decennial census is to count the total U.S. population; the counting of Hispanic groups is not part of the Constitutional mandate. Thus, the Census Bureau is more concerned with counting the total number of Hispanics rather than enumerating specific national origin groups, regardless of how people identify. For both this reason and what past studies and the current project show, researchers should recognize the limitations of using Census 2000 to study specific Latino groups. According to Etzioni (2001), "America has, by and large, dropped the notion that it will tell you what your race is" (p. 13). We disagree. As Rodriguez (2000) showed, the history of Latino data collection by the U.S. Census Bureau indicates that how Latino (and other) data are collected can be influenced by political pressure. The OMB standards govern the minimum standards for how race and ethnicity are treated and distributed in federal data sources. The OMB directs how and what data will be collected and how it will be expressed, such as the terminology employed, the number and type of racial groups, which groups are distinct enough to warrant separate categories, and the official designation of Latinos as an ethnic group independent of the race groups. Thus, we believe that unquestioned acceptance and usage of OMB standards to study such a complex topic can result in limited and methodologically driven views of Latino ethnicity. CONCLUSION

This paper focused on a very interesting pattern in Census 2000: The 15.1 percent of all Latinos identified in general terms such as "Hispanic" or "Latino." Approximately 5.3 million Latinos identified panethnically in 2000, more than double than in 1990. Scholars have pointed to two explanations for Latino panethnic identification. First, panethnicity is a meaningful identity stemming from shared backgrounds, structural commonalities, incorporation into U.S. society, and the, need to unite politically. Panethnicity may exist simultaneously with a more specific identity based on a national origin. The second explanation of Latino panethnicity is that it is a methodological artifact of the way Latino data are collected. Our assessment, based on analyses of Census 2000 and C2SS, and a review of the 2002 National Survey of Latinos and the Alternative Questionnaire Experiment, is that Latino panethnicity can be both a substantive and a methodological phenomenon. Panethnicity is an option that Latinos do use. Each data source shows that a proportion of Latinos identify in that way. Therefore, we disagree with others (Logan '2001; Suro 2002) who seem to imply that responses in Census 2000 are a

310

SOCIOLOGICAL FOCUS

methodological construction. Yet the comparison of Census 2000, C2SS and the AQE suggests that some Latino panethnic reporting is due, in part, to methodological factors. Clearly, the results show that one question is not sufficient to explore the diversity and complexity of ethnicity. Therefore, we caution researchers interested in Latino issues, ethnicity, and identity to think carefully about the implications of methodological issues on their choice of data and methods. We are concerned about overstatements of Latino panethnicity, such as the categorization of this phenomenon as the "melting down of national identities" (Glenn 2004). Instead, we encourage researchers to incorporate multiple data sources in their studies of the complexity and fluidity of Latino panethnicity and to acknowledge that this topic can be both substantively and methodologically driven. More research is needed to develop a more complete understanding of the ways in which an important component of the Latino experience in the United States, Latino panethnicity, is conceptualized, collected, manipulated, and expressed.

Eileen Diaz McConnell is an assistant professor in the Department of Sociology at the University of Illinois at UrbanaChampaign. Her current research focuses on the social demography of native and foreign-born Latinos, the role of public institutions in the social construction of race and ethnicity in the United States, and the consequences of racial and ethnic stratification for the mobility of Latinos. Edward A. Delgado-Romero is an assistant professor in the Counseling Psychology Program at Indiana University, Bloomington. He is also an affiliate faculty member of the Latino Studies program at IUB. His interests are in Latino/a identity, multicultural psychology, and the extent to which counselors address race and racism in the process of therapy.

REFERENCES Calderon, Jose. 1992. "Hispanic" and "Latino": The Viability of Categories for Panethnic Unity." Latin American Perspectives 75:37-44. Calderon, Fernando. 1995. "Latin American Identity and Mixed Temporalities; or How to Be Postmodern and Indian at the Same Time." Pp. 55-64 in The Postmodernism Debate in Latin America, edited by J. Beverly, J. Oviedo, and M. Aronna. Durham, NC: Duke University Press. Chapa, Jorge. 2000. "Hispanic/Latino Ethnicity and Identifiers." Pp. 243-246 in Encyclopedia of the U.S. Census, edited by Margo J. Anderson. Washington, DC: Congressional Quarterly Press. Comas-Diaz, Lillian. 2001. "Hispanics, Latinos, or Americanos: The Evolution of Identity." Cultural Diversity and Ethnic Minority Psychology 7:115-120. Cresce, Arthur R. and Roberto R. Ramirez. 2003. "Analysis of General Hispanic Responses in Census 2000." U.S. Census Bureau, Population Division. Working Paper No. 72. Cuello, Jose. 1998. "Latinos and Hispanics: A Primer on Terminology." . de la Garza, Rodolfo, Louis Desipio, F. Chris Garcia, John Garcia, and Angelo Falcon. 1992. Latino Voices: Mexican, Puerto Rican and Cuban Perspectives on American Politics. Boulder, CO: Westview Press. del Pinal, Jorge and Audrey Singer. 1997. Generations of Diversity: Latinos in The United States. Population Bulletin, vol. 52. Population Reference Bureau. del Pinal, Jorge. 2004. "Race and Ethnicity in Census 2000." Census 2000 Testing, Experimentation, and Evaluation Program. Topic Report Series, No.9. Delgado-Romero, Edward A. and Angela Rojas. In press. "'Other' Latinos: Counseling Cuban, Central and South American Clients." In Cross Cultural Psychotherapy: Toward a Critical Understanding of Diverse Client Populations, edited by Charles Negy. Boston, MA: Allyn and Bacon. Espiritu, Yen Lee. 1992. Asian American Panethnicity: Bridging Institutions and Identities. Philadelphia: Temple University Press.

LATINO PANETHNICITY: REALITY OR METHODOLOCHCAL CONSTRUCTION?

Etzioni, Amitai. 2001. "Inventing Hispanics." The Brookings Review 20:10-13. . Flores-Hughes, Grace. 1996. "Why the Term 'Hispanic'?" Hispanic 64. Garcia-Preto, Nydia. 1996. "Latino Families: An Overview." Pp. 141-154 in Ethnicity and Family Therapy, 22nd edition, edited by M. McGoldrick, J. Giordano, and J. K Pearce. New York: The Guilford Press. Gimenez, Martha E. 1989. "Latino/Hispanic - Who Needs a Name? The Case Against a Standarndarized Terminology." International Journal of Health Services 19:557-571. . 1992. "U.S. Ethnic Politics: Implications for Latin Americans." Latin American Perspectives 19:7-16. Glenn, David. 2004. "Scholars Cook Up a New Melting Pot." Chronicle of Higher Education 50:AI0. Gracia, Jorge J. E. 2000. Hispanic/Latino Identity: A Philosophical Perspective. Malden, MA: Blackwell Publishers. Granados, Christine. 2000. " 'Hispanic' Vs. 'Latino.' " Hispanic Magazine. <www.hispanicmagazine.com/2000/dec/Features/latino.html>. Greeley, Andrew. 1974. Ethnicity in the United States. New York: Wiley. Guzman, Betsy and Eileen Diaz McConnell. 2002. "The Hispanic Population: 1990-2000 Growth and Change." Population Research and Policy Review 21:109-128. Hayes-Bautista, David and Jorge Chapa. 1988. "Latino Terminology: Conceptual Bases for Standardized Terminology." American Journal of Public Health 77:61-68. Helms, Janet E. 1994. "How Multiculturalism Obscures Racial Factors in the Therapy Process: Comment on Ridley et al. (1994), Sodowsky et al. (1994), Ottavi et al. (1994), and Thompson et al. (1994)." Journal of Counseling Psychology 41:162-165. Itzigsohn, Jose and Carlos Dore-Cabral. 2000. "Competing Identities? Race, Ethnicity, and Panethnicity Among Dominicans in the United States." Sociological Forum 15: 225-247. Jones-Correa, Michael and David L. Leal. 1996. "Becoming "Hispanic": Secondary Panethnic Identification among Latin American-Origin Populations in the United States." Hispanic Journal of Behavioral Sciences 18:214-215. Kandel, William and John Cromartie. 2003. "Hispanics Find a Home in Rural America." AmberWaves: The Economics of Food, Farming, Natural Resources and Rural Amerwa. U.S. Department of Agriculture, Economic Research Service. . Landale, Nancy S. and R. S. Oropesa. 2002. "White, Black or Puerto Rican? Racial Self Identification Among Mainland and Island Puerto Ricans.” Social Forces 82:231-254. Lee, Sharon. 1993. "Racial Classifications in the U.S. Census: 1890-1990." Ethnic and Racial Studies 16:75-94. Logan, John. 2001. "The New Latinos: Who They Are, Where They Are.” Lewis Mumford Center for Comparative Urban and Regional Research University at Albany. Logan, John R. 2002. "Hispanic Populations and Their Residential Patterns in the Metropolis." Lewis Mumford Center for Comparative Urban and Regional Research University at Albany. Lopez, David and Yen Le Espiritu 1990. "Panethnicity in the United States: A Theoretical Framework." Ethnic and Racial Studies 13:198-224. Marrow, Helen. 2003. "To Be or Not to Be (Hispanic or Latino): Brazilian Racial and Ethnic Identity in the United States." Ethnicities 3:427-464. Martin, Elizabeth. 2002. "The Effects of Questionnaire Design on Reporting of Detailed Hispanic Origin in Census 2000 Mail Questionnaires." Public Opinion Quarterly 66:582-593. Massey, Douglas S. 1993. "Latinos, Poverty, and the Underclass: A New Agenda for Research." Hispanic Journal of Behavioral Sciences 15:449-475. McConnell, Eileen Diaz and Betsy Guzman. 2003. "Evaluating Hispanic Data." . Min, Pyong Gap. 1999. "Ethnicity: Concepts, Theories, and Trends." Pp. 16-46 in Struggle for Ethnic Identity: Narratives by Asian American Professionals, edited by Pyong Gap Min and Rose Kim. Walnut Creek, CA: AltaMira Press. Moore, Joan. 1990. "Hispanic/Latino: Imposed Label or Real Identity?" Latino Studies Journal 1:33-47. Moore, Joan and Harry Pachon. 1985. Hispanics in the United States. Englewood Cliffs, CT: Prentice Hall. Murguia, Edward. 1991. "On Latino/Hispanic Ethnic Identity." Latino Studies Journal 2:8-18. Oboler, Suzanne. 1992. "The Politics of Labeling: Latino/a Cultural Identities of Self and Other." Latin American Perspectives 19:18-36. Office of Management and Budget. 1997. "Revisions to the Standards for the Classification of Federal Data on Race and Ethnicity." October 30,1997. . . 2000. Provisional Guidance on the Implementation of The 1997 Standards For Federal Data On Race And Ethnicity. December 15, 2000. . Okamura, Jonathon. 1981. "Situational Ethnicity." Ethnic and Racial Studies 4:452-465. Omi, Michael. 1997. "Racial Identity and the State: The Dilemmas of Classification." Law and Inequality: A Journal of Theory and Practice 25:7-23. Padilla, Felix. 1984. "On the Nature of Latino Ethnicity." Social Science Quarterly 65:651-664.

311

312

SOCIOLOGICAL FOCUS

Parsons, Talcott. 1975. "Some Theoretical Considerations on the Nature and Trends of Ethnicity." Pp. 53-83 in Ethnicity, Theory, and Experience, edited by Nathan Glazer and Patrick Moynihan, Cambridge, MA: Harvard University Press. Pew Hispanic Center/Kaiser Family Foundation. 2002. 2002 National Survey of Latinos. . Phinney, Jean S. 1996. Understanding Ethnic Diversity: The Role of Ethnic Identity." American Behavioral Scientist 40:143-152. Portes, Alejandro and Dag MacLeod. 1996. "What Shall I Call Myself? Hispanic Identity Formation in the Second Generation." Ethnic and Racial Studies 19:523-547. Rodriguez, Clara. 2000. Changing Race: Latinos, the Census, and the History of Ethnicity in the United States. New York: New York University Press. Sellers, R. M., M. A. Smith, J. N. Shelton, S. J. Rowley, and T. M. Chavous. 1998. "Multidimensional Model of Racial Identity: A Reconceptualization of African American Racial Identity." Personality and Social Psychology Review 2:18-,'39. Sellers, R. M., S. J. Rowley, T. M. Chavous, J. N. Shelton, and M. A. Smith. 1997. "Multidimensional Inventory of Black Identity: A Preliminary Investigation of Reliability and Construct Validity." Journal of Personality and Social Psychology 73:805-815. Singer, Audrey. 2004. "The Rise of New Immigrant Gateways." Brookings Institution Center on Urban and Metropolitan Policy. Metropolitan Policy Program. . Suro, Roberto. 2002. "Counting the "Other Hispanics": How Many Colombians, Dominicans, Ecuadorians, Guatemalans and Salvadorans are there in the United States?" Pew Hispanic Center Report. Suro, Roberto and Audrey Singer. 2002. "Latino Growth in Metropolitan America: Changing Patterns, New Locations."Center on Urban and Metropolitan Policy and The Pew Hispanic Center. Washington, DC. lTaifel, Henri and John C. Turner. 1986. "The Social Identity Theory of Intergroup Behavior." Pp. 2-24 in Psychology of Intergroup Relations, edited by S. Worchell and W. G. Austin. Chicago: Nelson-Hall Publishers. Tortora, Robert D., Susan M. Miskura, and Don A. Dillman. 1993. "Onward Towards a 2000 Census Design: Research Results." Pp. 120-128 in Proceedings of the Section on Survey Research Methods, American Statistical Association. . Trimble, Joseph E. 1991. "Ethnic Specification, Validation Prospects and Future of Drug Abuse Research." International Journal of the Addictions 25:149-169. Trimble, Joseph E., Janet E. Helms, and Maria Root. 2003. "Social and Psychological Perspectives on Ethnic and Racial Identity." Pp. 239-275 in Handbook of Racial and Ethnic Minority Psychology, edited by G. Bernal, J. E. Trimble, A. K. Burlew, and F. T. L. Leong. Thousand Oaks, CA: Sage. U.S. Census Bureau. 2001. Census 2000 Summary File 1. (SF1). .2004. "Table 1a. Projected Population of the United States, by Race and Hispanic Origin: 2000 to 2050." U.S. Interim Projections by Age, Sex, Race, and Hispanic Origin. . U.S. Department of Commerce. 2000. Census 2000 Operational Plan. December 2000. . U.S. Department of Labor. 1995. "A CPS Supplement for Testing Methods of Collecting Racial and Ethnic Information: May 1995." Bureau of Labor Statistics. Press Release USDL 95-428. . Vasconcelos, Jose. 1957. "La Raza Cosmica: Mision de la Raza Ibero-Americana." Pp. 903-942 in Obras Completas, vol. 2, Libreros Mexicanos. Vidal de Haymes, Maria, Keith M. Kilty, and Stephen N. Haymes. 2002. "Another Kind of Rainbow Politics." Journal of Human Behavior in the Social Environment 5: 175-188. Waters, Mary C. 1990. Ethnic Options: Choosing Identities in America. Berkeley: University of California Press. Yancey, William L., Eugene P. Erickson, and Richard N. Juliani. 1976. "Emergent Ethnicity: A Review and Reformulation." American Sociological Review 41:391-403.