Changing Race: Latinos, the Census and the History of Ethnicity 9780814771969

Citation preview

CHANGING RACE

CRITICAL AMERICA Richard Delgado and Jean Stefancic General Editors White by Law: The Legal Construction of Race Ian F. Haney López

Cultivating Intelligence: Power, Law, and the Politics of Teaching Louise Harmon and Deborah W. Post

Privilege Revealed: How Invisible Preference Undermines America Stephanie M. Wildman with Margalynne Armstrong, Adrienne D. Davis, and Trina Grillo

Does the Law Morally Bind the Poor? or What Good’s the Constitution When You Can’t Afford a Loaf of Bread? R. George Wright

Hybrid: Bisexuals, Multiracials, and Other Misfits under American Law Ruth Colker

Critical Race Feminism: A Reader Edited by Adrien Katherine Wing

Immigrants Out! The New Nativism and the Anti-Immigrant Impulse in the United States Edited by Juan F. Perea

Taxing America Edited by Karen B. Brown and Mary Louise Fellows

Notes of a Racial Caste Baby: Color Blindness and the End of Affirmative Action Bryan K. Fair

Please Don’t Wish Me a Merry Christmas: A Critical History of the Separation of Church and State Stephen M. Feldman

To Be an American: Cultural Pluralism and the Rhetoric of Assimilation Bill Ong Hing

Negrophobia and Reasonable Racism: The Hidden Costs of Being Black in America Jody David Armour

Black and Brown in America: The Case for Cooperation Bill Piatt

Black Rage Confronts the Law Paul Harris

Selling Words: Free Speech in a Commercial Culture R. George Wright

The Color of Crime: Racial Hoaxes, White Fear, Black Protectionism, Police Harassment, and Other Macroaggressions Katheryn K. Russell

The Smart Culture: Society, Intelligence, and Law Robert L. Hayman, Jr.

Was Blind, But Now I See: White Race Consciousness and the Law Barbara J. Flagg

The Gender Line: Men, Women, and the Law Nancy Levit

Heretics in the Temple: Americans Who Reject the Nation’s Legal Faith David Ray Papke

The Empire Strikes Back: Outsiders and the Struggle over Legal Education Arthur Austin

Interracial Justice: Conflict and Reconciliation in Post–Civil Rights America Eric K. Yamamoto

Black Men on Race, Gender, and Sexuality: A Critical Reader Edited by Devon W. Carbado

When Sorry Isn’t Enough: The Controversy over Apologies and Reparations for Human Injustice Edited by Roy L. Brooks

Disoriented: Asian Americans, Law, and the Nation State Robert S. Chang

Rape and the Culture of the Courtroom Andrew E. Taslitz

The Passions of Law Edited by Susan A. Bandes

Global Critical Race Feminism: An International Reader Edited by Adrien Katherine Wing

Law and Religion: Critical Essays Edited by Stephen M. Feldman

Changing Race: Latinos, the Census, and the History of Ethnicity in the United States Clara E. Rodríguez

C L A R A E . RO D R Í G U E Z

CHANGING RACE Latinos, the Census, and the History of Ethnicity in the United States

a New York University Press • New York and London

NEW YORK UNIVERSITY PRESS New York and London © 2000 by Clara E. Rodríguez All rights reserved Library of Congress Cataloging-in-Publication Data Rodríguez, Clara E., 1944– Changing race : Latinos, the census, and the history of ethnicity / Clara E. Rodríguez. p. cm. — (Critical America) Includes bibliographical references and index. ISBN 0-8147-7547-0 (pbk. : alk. paper) — ISBN 0-8147-7546-2 (cloth : alk. paper) 1. Hispanic Americans—Census. 2. Hispanic Americans—Race identity. 3. Hispanic Americans—Ethnic identity. 4. Categorization (Psychology). 5. Race—Social aspects—United States. 6. Ethnology—United States. 7. United States—Census. 8. United States—Race relations. I. Title. II. Series. E184.S75 R64 2000 305.8'00973—dc21 00-008629 New York University Press books are printed on acid-free paper, and their binding materials are chosen for strength and durability. Manufactured in the United States of America 10

9

8

7

6

5

4

3

2

1

Contents

ix xv

Introduction Acknowledgments

I

The Fluidity of Race

1

Latinos in the U.S. Race Structure

2

The Idea of Race

27

3

Stories of Self-Definition

47

II

3

Historical Constructions

4

Whites and Other Social Races

65

5

The Shifting Color Line

87

6

Race in the Americas

III

106

Race and the Census

7

The “Other Race” Option

129

8

Redefining Race in 2000

153

Appendix A: Data Limitations and the Undercount Appendix B: The Biological Concept of Race in the United States Appendix C: A Technical Oversight or Racial Flux? Appendix D: Free People of Color Notes References Index About the Author

177 182 187 193 199 229 265 283

vii

Introduction

Ethnicity is a hotly contested subject in the academy; even the term provokes intense scholarly debate. In addition, academic definitions and discussions of ethnicity are complex, with different disciplines emphasizing different aspects of the phenomenon. Anthropologists and sociologists focus on social and cultural factors and take for granted the psychodynamics of individuals. Conversely, psychologists place social phenomena in the background, stressing the importance of individual cognition and emotions (Leets, Clement, and Giles 1996). Social psychologists argue that all these dimensions must be linked through selfidentification and that culture and the individual must be considered together. Unfortunately, much of the scholarly writing on ethnicity is not theoretically rigorous. A recent analysis of ethnicity in the social science literature, reviewing 190 articles and 10,000 citations published between 1974 and 1992 (Leets, Clement, and Giles 1996), found that an overwhelming majority (82%) of the articles lacked any coherent theoretical foundation from which to view ethnicity.1 Moreover, the majority were not empirical, and many did not report how they had measured ethnicity. Most of the articles (43%) dealt with ethnicity only secondarily and usually measured ethnicity as a geopolitical category, for example, Hindus in India. Only 22 percent reflected multiple dimensions of ethnicity, acknowledged overlapping categories, or included objective and subjective components of ethnicity. Some scholars equated ethnicity with race. Generally, most investigators regarded ethnicity as an objective, self-evident social reality that needed little, or no further, elaboration.2 Ethnicity and race, however, have a fluidity and complexity that are not often acknowledged but nonetheless are evident. When we reflect on how or why we consider an individual or a group to be “ethnic,” we think, for example, of language or dialect; common cultural and/or geographic origin; religion; physical difference from us, such as height, ix

x

INTRODUCTION

skin color; food, music, and artistic preferences and creations; political interests in their country of origin and/or in the United States; institutions that represent and maintain the group; and an internal or external sense of distinctiveness. Indeed, all these variables surface when we consider the multifaceted population of Latinos in the United States. The experience of Latinos in the United States demonstrates that ethnicity involves both internal and external components, which are culturally, politically, and subjectively influenced and multileveled (Isajiw 1993:418ff; Leets, Clement, and Giles 1996; van den Berghe 1981:254– 261). Consider the situation of Jews in Europe during World War II and Latinos in today’s United States. During the war, Jews were regarded as a “race,” even though they are a religious-ethnic group. On a personal level, a person might have identified himself or herself as either a devout Jew or a secular person of Jewish ancestry. But on an instrumental level, this person might also have been a German. And the external identification of this person during the Nazi regime was both non-German and nondesirable. In the United States today, a person may be Puerto Rican or Mexican on a personal level, Latino on an instrumental level, and Hispanic to the government. Some people might classify this person as black, white, or Asian. Others think of Latinos as a brown race, and still others, as a multiracial ethnic group. This book discusses this distinction-plus-duality. For simplicity’s sake, and in order to appear compatible with what appears to be the prevailing language usage of most publishers, the terms “white” and “black” were not capitalized in this volume. This book emphasizes the multidimensional nature of individual racial identity (Hartman 1994, 1995). Hansen (1995) has provided examples of a number of these dimensions in the case of African Americans: self-definition (do I consider myself black?), perception (am I considered black by others?), and treatment (am I treated as if I were black?). He also points out that these three elements are not always congruent. Rand Reed (1994) added a few other related dimensions: What is the person biologically? Sociologically? When is race determined? At birth? Death? And by whom? By parents? By an unknown observer? These different levels coexist, with some more salient than others at different times. As befits a complex subject, Changing Race draws on empirical research and methodologies from many scholarly fields. For historical back-

INTRODUCTION

xi

ground, I drew on archival records of state and federal censuses and interpretive writings of the times. For a view into the shifting “official” definitions of race and ethnicity, I examined the standard reference works, such as Webster’s Collegiate Dictionary from 1898 to 1994. I also analyzed secondary sources for insight into the meaning of race and ethnicity from antiquity to the present. In addition, I investigated works on race in several Latin American countries. For a more contemporary analysis, I reviewed the relevant works in this area and also used several methodologies. For example, I used in-depth interviews for the case studies of identity in order to explore areas rarely covered in conventional social science research. According to Carlos Martin, a former colleague of mine at the Smithsonian Institution’s National Museum of American History, personal accounts often give an entirely different and more holistic view of “race” in general and of Latinos in particular. For patterns of Hispanic racial identification and the reasons for these patterns, I looked at national 1990 census data. I also used various forms of statistical analysis and original survey data from earlier works to provide quantifiable insights into the interplay of “identity” data and economic status. Finally, I relied on legal writings for the discussion of the “in-between” identities of several groups caught in the contradictions of racial identity in the United States. Conceptually, this book is positioned as follows: Academic disciplines offer at least four theoretical approaches to ethnicity, including the assimilationist and the pluralist models, the most common and traditional approaches in the United States. Assimilationists assume that ethnicity will be eliminated over time through assimilation, and pluralists believe that ethnic groups will change by adapting to or accommodating the host society. Sociobiological theorists assume that biological underpinnings and variables in social interactions are of prime importance. And Marxist theorists use a conflict orientation, in which every thesis produces an antithesis that results in a new synthesis that also produces a new antithesis, and so on. These (Marxist) conflict theorists can be subdivided into (1) dependency theorists, who focus on political economic relations, and (2) postmodern theorists (Leets, Clement, and Giles 1996). My theoretical stance in regard to identity, ethnicity, and race most approximates postmodernist theory. Postmodernist theorists argue that there is not a true and knowable self. Rather, one’s identity is relative and is constantly negotiated through relationships and situational contexts. Instead of a core of

xii INTRODUCTION

identity, or self, one has a plurality of selves, each of which surfaces in a particular situation. Thus, an individual is not committed to only one identity, and similarly, one’s ethnic identity is variable and subject to the active construction of the individual (Leets, Clement, and Giles 1996). In this volume, I stress the centrality of situational influences and the role of individuals in the active construction of multiple identities. That is, I see individual identity as relational and situational. But I also maintain that many people have a core of identity, or a self, that is made up of multiple identities—or, in the words of the postmodernists, a plurality of selves. I thus disagree with those scholars who believe that individuals are not committed to only one true identity. Rather, I maintain that individuals may have only one true and knowable self but that their ethnic identity also may be variable and subject to an individual’s particular construction of it and to the political and economic contexts in which this person functions.

PURPOSES In the first chapter, I explain that the concept of race can be construed in a variety of ways and that the experience of Latinos in the United States is a good example of the social constructedness of race. I also describe two situations seldom mentioned but often experienced. The first is that categories often come between people, and the second is that people often fall between categories. Examples of the first are particularly striking in the Native American community. As Forbes noted, placing Native Americans who have married African Americans into the “black” category has resulted in Native Americans’ being divided and losing some of their members (1990:44 ff). In some cases, even whole tribes were denied federal recognition as Indian tribes, for example, the Shinnecock Indians of Long Island and the Ramapo Mountain Indians in New Jersey, New York, and Connecticut. This categorization has also had a significant impact on the lives of these “red-black” peoples, not just in terms of identity, but also in terms of more immediate losses of land and treaty-protected rights and benefits. An example of how people fall between categories was most recently illustrated in the demands of multiracial individuals for a census category that would accommodate the many and increas-

INTRODUCTION

xiii

ing numbers of people who argue that they do not fit into any of the established census groupings. Although ostensibly removed from their real lives and everyday activities, categories and classifications do affect people in the United States. As this book shows, which and how people are counted has many ramifications. Moreover, although many believe that census data do not pertain to identity but, rather, are needed to address past discrimination, we also know that all race data lead to some sort of reification, which often affects the way in which peoples and individuals see themselves and others. That it is important to clarify these issues and the processes that lead to governmental classification is another reason that I wrote this book. A final purpose for writing this book was to shed light on an area fraught with conflict, emotion, and politics. Race in the United States is a complicated, political, and emotional subject. As a report by the U.S. General Accounting Office indicated, “[The] collection of these types of data is technically complex and publicly controversial” (1997:1). I hope that a better understanding of what has too often been used as a divisive and sometimes cruel issue can be addressed openly, honestly, and humanely. In this way, we can abolish racial hierarchies and become more respectful of one another’s unique and valued histories.

GOVERNMENT AND DICTIONARY DEFINITIONS OF RACE Much of this book’s focus is on official racial labels and categories, both past and present. I am aware, of course, that government records may reflect somewhat “arbitrary racial categories imposed by a white official or by white prejudice” and that this often contradicts other classifications, for example, familial traditions, that may have greater sociocultural and psychological meanings and that may also provide bonds and a sense of belonging that government categories do not (Forbes 1990:38, 41, 45). I also know that government categories are not always the best or most satisfactory measures of group affiliation. Nonetheless, government definitions or measures of race are the most geographically comprehensive, most readily available, and most numerically determinate tools we have. Census categories provide

xiv INTRODUCTION

insight into how a society’s ideologies and dominant ideas and beliefs are reflected in official government classifications. To a degree, they also represent public consensus on how populations are viewed and counted. I do recognize, however, that this is only one measure of race, which has its own inherent difficulties. Moreover, because they exercise a reflective as well as a regulatory role in society, census categories must be considered carefully and from new viewpoints. This is what I want to accomplish in Changing Race.

Acknowledgments

As I sat down to write these acknowledgments, I remembered when, not too long ago, the New York Women’s Agenda presented to me, and others, its STAR award and I was asked to say a few words. Ron Gault, a friend who was in the audience, later chided me, saying something like “God, you thanked everyone in the whole wide world . . . your mother, your father, all your sisters, your brothers, etc., etc.” And I told him, “Yeah, they all helped.” I also thought about my acknowledgments in an earlier book, in which I quoted Tato Laviera, a celebrated Puerto Rican poet, who once observed, “With every word I write I give thanks to 50 people.” It is the same feeling that I have now, that there are so many people to thank. Writing this book has been long and difficult, and many people helped me with their consistent and unquestioned support for “whatever it is you’re doing.” Here I count my family—my children Gelvi and José; my husband Gel; my mother Clarita; my sisters Minny and Myrna; my brother Jimmy; my cousin Lena; my nieces and nephew María, Michelle, and Tony; and my extended family, who are too numerous to list individually; and, finally, Rosa and Gloria, who also supported me. Other people helped me more directly, such as my talented editors at New York University Press: Stephen Magro, Niko Pfund, and Despina Gimbel, five anonymous reviewers, Fordham University for giving me some time off to complete the project, the Russell Sage Foundation and its staff for facilitating my work during the year I spent there as a visiting scholar, all the the authors cited in this work and those not cited but who contributed to its development, and the following individuals who contributed in unique and significant ways to its completion: Cristina Bryan, Katie Courtice, Gregory De Freitas, Richard Delgado, Vanessa Estrada, Norma Fuentes-Mayorga, Ian Haney-López, Charles Kamasaki, J. Kehaulani Kauanui, Ray Lohier, Terri Ann Lowenthal, Carlos Martin, Barbara Mundy, Nadine Naber, Jeff Passel, Olivia Carter-Pokras, Raedyn Rivera, Eric Rodríguez, Jean Stefancic, and Frank Torres. I thank you all. xv

PART I

THE FLUIDITY OF RACE

1 Latinos in the U.S. Race Structure

A C C O R D I N G TO D E F I N I T I O N S common in the United States, I am

a light-skinned Latina with European features and hair texture. I was born and raised in New York City; my first language was Spanish; and I am today bilingual. I cannot remember when I first realized that the color of one’s skin, the texture of one’s hair, or the cast of one’s features determined how one was treated in both my Spanish-language and English-language worlds. I do know that it was before I understood that accents, surnames, residence, class, and clothing also determined how one was treated. Looking back on my childhood, I recall many instances when the lighter skin color and European features of some persons were admired and terms such as pelo malo (bad hair) were commonly used to refer to “tightly curled” hair. It was much later that I came to see that this Eurocentric bias, which favors European characteristics above all others, was part of our history and cultures. In both Americas and the Caribbean, we have inherited and continue to favor this Eurocentrism, which grew out of our history of indigenous conquest and slavery (Shohat and Stam 1994). I also remember a richer, more complex sense of color than the simple dichotomy of black and white would suggest, a genuine aesthetic appreciation of people with some color and an equally genuine valuation of people as people, regardless of color. Also, people sometimes disagreed about an individual’s color and “racial” classification, especially if the person in question was in the middle range, not just with regard to color, but also with regard to class or political position.1 As I grew older, I came to see that many of these cues or clues to status—skin color, physical features, accents, surnames, residence, and other class characteristics—changed according to place or situation. For example, a natural “tan” in my South Bronx neighborhood was attractive, whereas downtown, in the business area, it was “otherizing.” I also 3

4

LATINOS IN THE U.S. RACE STRUCTURE

recall that the same color was perceived differently in different areas. Even in Latino contexts, I saw some people as lighter or darker, depending on certain factors such as their clothes, occupation, and families.2 I suspect that others saw me similarly, so that in some contexts, I was very light, in others darker, and in still others about the same as everyone else. Even though my color stayed the same, the perception and sometimes its valuation changed. I also realize now that some Latinos’ experiences were different from mine and that our experiences affect the way we view the world. I know that not all Latinos have multiple or fluctuating identities. For a few, social context is irrelevant. Regardless of the context, they see themselves, and/or are seen, in only one way. They are what the Census Bureau refers to as consistent; that is, they consistently answer in the same way when asked about their “race.” Often, but not always, they are at one or the other end of the color spectrum. My everyday experiences as a Latina, supplemented by years of scholarly work, have taught me that certain dimensions of race are fundamental to Latino life in the United States and raise questions about the nature of “race” in this country. This does not mean that all Latinos have the same experiences but that for most, these experiences are not surprising. For example, although some Latinos are consistently seen as having the same color or “race,” many Latinos are assigned a multiplicity of “racial” classifications, sometimes in one day! I am reminded of the student who told me after class one day, “When people first meet me, they think I’m Italian, then when they find out my last name is Mendez, they think I’m Spanish, then when I tell them my mother is Puerto Rican, they think I’m nonwhite or black.” Although he had not changed his identity, the perception of it changed with each additional bit of information. Latino students have also told me that non-Latinos sometimes assume they are African American. When they assert they are not “black” but Latino, they are either reproved for denying their “race” or told they are out of touch with reality. Other Latinos, who see whites as otherthan-me, are told by non-Latinos, “But you’re white.” Although not all Latinos have such dramatic experiences, almost all know (and are often related to) others who have. In addition to being reclassified by others (without their consent), some Latinos shift their own self-classification during their lifetime. I have known Latinos who became “black,” then “white,” then “human

LATINOS IN THE U.S. RACE STRUCTURE

5

beings,” and finally again “Latino”—all in a relatively short time. I have also known Latinos for whom the sequence was quite different and the time period longer. Some Latinos who altered their identities came to be viewed by others as legitimate members of their new identity group. I also saw the simultaneously tricultural, sometimes trilingual, abilities of many Latinos who manifested or projected a different self as they acclimated themselves to a Latino, African American, or white context (Rodríguez 1989:77). I have come to understand that this shifting, context-dependent experience is at the core of many Latinos’ life in the United States. Even in the nuclear family, parents, children, and siblings often have a wide range of physical types. For many Latinos, race is primarily cultural; multiple identities are a normal state of affairs; and “racial mixture” is subject to many different, sometimes fluctuating, definitions. Some regard racial mixture as an unfortunate or embarrassing term, but others consider the affirmation of mixture to be empowering. Lugones (1994) subscribes to this latter view and affirms “mixture,” mestizaje, as a way of resisting a world in which purity and separation are emphasized and one’s identities are controlled: “Mestizaje defies control through simultaneously asserting the impure, curdled multiple state and rejecting fragmentation into pure parts . . . the mestiza . . . has no pure parts to be ‘had,’ controlled” (p. 460). Also prevalent in the upper classes is the hegemonic view that rejects or denies “mixture” and claims a “pure” European ancestry. This view also is common among middle- and upper-class Latinos, regardless of their skin color or place or origin. In some areas, people rarely claim a European ancestry, such as in indigenous sectors of Latin America, in parts of Brazil, and in the coastal areas of Colombia, Venezuela, Honduras, and Panama (see, e.g., Arocha 1998; De La Fuente 1998). Recently, some Latinos have encouraged another view in which those historical components that were previously denied and denigrated, such as indigenous and African ancestry, were privileged (see, e.g., Moro: La Revista de nuestra vida [Bogota, Colombia, September 1998]; La Voz del pueblo Taino [The voice of the Taino people]), official newsletter of the United Confederation of Taino People, U.S. regional chapter, New York, January 1998). Many people, however—mostly non-Latinos—are not acquainted with these basic elements of Latino life. They do not think much about them, and when they do, they tend to see race as a “given,” an ascribed characteristic that does not change for anyone, at any time. One is either

6

LATINOS IN THE U.S. RACE STRUCTURE

white or not white. They also believe that “race” is based on genetic inheritance, a perspective that is just another construct of race. Whereas many Latinos regard their “race” as primarily cultural, others, when asked about their race, offer standard U.S. race terms, saying that they are white, black, or Indian. Still others see themselves as Latinos, Hispanics, or members of a particular national-origin group and as belonging to a particular race group.3 For example, they may identify themselves as Afro-Latinos or white Hispanics. In some cases, these identities vary according to context, but in others they do not. I have therefore come to see that the concept of “race” can be constructed in several ways and that the Latino experience in the United States provides many illustrations of this. My personal experiences have suggested to me that for many Latinos, “racial” classification is immediate, provisional, contextually dependent, and sometimes contested. But because these experiences apply to many non-Latinos as well, it is evident to me that the Latino construction of race and the racial reading of Latinos are not isolated phenomena. Rather, the government’s recent deliberations on racial and ethnic classification standards reflect the experiences and complexities of many groups and individuals who are similarly involved in issues pertaining to how they see themselves and one another (U.S. Department of Commerce 1995; U.S. Office of Management and Budget 1995, 1997a and b, 1999; these deliberations will be reviewed in chapter 8). Throughout my life, I have considered racism to be evil, and I oppose it with every fiber of my being. I study race to understand its influence on the lives of individuals and nations because I hope that honest, open, and well-meaning discussions of race and ethnicity and their social dynamics can help us appreciate diversity and value all people, not for their appearance, but for their character.

“OTHER RACE” IN THE 1980 AND 1990 CENSUSES It was because of my personal experiences that I first began to write about race (Rodríguez 1974) and that I was particularly sensitive to Latinos’ responses to the censuses’ question about race. The U.S. Census Bureau’s official position has been that race and ethnicity are two separate concepts. Thus, in 1980 and in 1990, the U.S. census asked people to indicate their “race”—white, black, Asian or Pacific Islander,

LATINOS IN THE U.S. RACE STRUCTURE

7

American Indian, or “other race”—and also whether or not they were Hispanic. (The two questions used in the 1980 and 1990 censuses are shown in figures 1.1 and 1.2). As table 1.1 shows, Latinos responded to the 1990 census’s question about race quite differently than did nonLatinos. Whereas less than 1 percent of the non-Hispanic population reported they were “other race,” more than 40 percent of Hispanics chose this category. Latinos responded similarly in the previous decennial census (Denton and Massey 1989; Martin, DeMaio, and Campanelli 1990; Rodríguez 1989, 1990, 1991a; Tienda and Ortiz 1986). Although the percentages of the different Hispanic groups choosing this category varied, all chose it more often than did non-Hispanics (see table 1.1, which shows a wide range in the proportion of Hispanic-origin groups choosing “other race” in the 1990 census). In addition, the many Hispanics who chose this category wrote—in the box explicitly asking for race—the name of their “home” Latino country or group, to “explain” their race—or “otherness.”4 The fact that these Latino referents were usually cultural or national-origin terms, such as Dominican, Honduran, or Boricua (i.e., Puerto Rican) underscores the fact that many Latinos viewed the question of race as a question of culture, national origin, and socialization rather than simply biological or genetic ancestry or color. Indeed, recent studies have found that many Latinos understand “race” to mean national origin, nationality, ethnicity, culture (Kissam, Herrera, and Nakamoto 1993), or a combination of these and skin color (Bates et al. 1994:109; Rodríguez 1991a, 1992, 1994a; Rodríguez and Cordero-Guzmán 1992). For many Latinos, the term race or raza is a reflection of these understandings and not of those often associated with “race” in the United States, for example, defined by hypodescent.5 Studies have found that Latinos also tend to see race along a continuum and not as a dichotomous variable in which individuals are either white or black (Bracken and de Bango 1992; Rodríguez and Hagan 1992; Romero 1992). This does not mean that there is only one Latino view of race. Rather, there are different views of race within different countries, classes, and even families. Latinos’ views of race are dependent on a complex array of factors, one of which is the racial formation process in their country of origin. Other variables also influence their views of race, for example, generational differences, phenotype, class, age, and education. But even though there is not just one paradigm of Latin American race, there are some basic differences between the way that

4. Is this person ———? Fill in one circle.

o o o o o o o o

White Black or Negro Japanese Chinese Filipino Korean Vietnamese Indian (Amer.) Print tribe ________.

o o o o o o o

Asian Indian Hawaiian Guamanian Samoan Eskimo Aleut Other—specify __________________

7. Is this person of Spanish/Hispanic origin or descent? Fill in one circle.

o o o o o

No, not Spanish/Hispanic. Yes, Mexican, Mexican-Amer., Chicano. Yes, Puerto Rican. Yes, Cuban. Yes, other Spanish/Hispanic.

FIG. 1.1. Two Questions about Race and Hispanic Origin on the 1980 Census

FIG. 1.2. Race and Hispanic-Origin Questions on the 1990 Census

8

LATINOS IN THE U.S. RACE STRUCTURE

9

Table 1.1 Racial Self-Classification by Selected Hispanic-Origin Groups, 1990

Mexican Puerto Rican Cuban Other Spanishb Dominican Ecuadoran Colombian Guatemalan Salvadoran Panamanian Total Hispanic Non-Hispanic Total population (millions)

White

Black

NAI

API a

Other

50.6 46.4 83.8 52.4 29.26 50.81 64.46 42.95 38.53 32.97 52.1 83.1 199.5

0.9 6.5 3.7 6.5 29.96 1.90 2.33 0.89 1.27 35.50 3.0 12.9 29.8

0.6 0.3 0.2 1.0 1.02 1.68 1.34 1.67 1.10 2.94 0.7 0.8 2.0

0.4 1.0 0.4 2.1

47.4 45.9 12.0 38.0 39.76 45.62 31.87 54.48 59.10 28.59 43.5 0.1 9.7

c c c c c c

0.9 3.1 7.2

Rows sum to 100% except for rounding. a API = Asian and Pacific Islander; NAI = Native American Indian. b Includes both those who gave a Latino referent and those who identified themselves only as Hispanic. c These two categories were combined because of small numbers. Source: 1990 PUMS (Public Use Micro Sample) 1% sample. (These numbers may not be identical to tables based on the 100% census survey or the 5% PUMS because of sampling variability.)

Latinos view race and the way that race is viewed overall in the United States. In the United States, rules of hypodescent and categories based on presumed genealogical-biological criteria have generally dominated conceptions of race. Racial categories have been few, discrete, and mutually exclusive, with skin color a prominent element. Categories for mixtures—for example, mulatto—have been transitory. In contrast, in Latin America, racial constructions have tended to be more fluid and based on many variables, like social class and phenotype. There also have been many, often overlapping, categories, and mixtures have been consistently acknowledged and have had their own terminology. These general differences are what Latinos bring with them to the United States, and they influence how they view their own and others’ “identity.” Although Latinos may use or approach “race” differently, this does not mean that “race” as understood by Latinos does not have overtones of racism or implications of power and privilege—in either Latin America or the United States. Indeed, the depreciation and denial of African

10

LATINOS IN THE U.S. RACE STRUCTURE

and Amerindian characteristics are widespread.6 Everywhere in Latin America can be found “a pyramidal class structure, cut variously by ethnic lines, but with a local, regional and nation-state elite characterized as ‘white.’ And white rules over color within the same class; those who are lighter have differential access to some dimensions of the market” (Torres and Whitten 1998:23). Even those countries that subscribe to a racial ideology of mestizaje7 often maintain racial and class hierarchies that favor upper-class interests and political agendas, privilege European components, ignore racialisms, and neutralize expressions of pluralism by indigenous or African-descended groups (Martínez-Echazábal 1998). That the awareness of these issues is increasing is evidenced by Torres-Saillant’s appeal to Dominican historians to embrace a narrative that “privileges the many rather than the few” (1998:140). As one Jamaican student traveling in the Spanish-speaking Caribbean noted, the attitude there toward race is similarly destructive but strikingly different from that in the United States. Unfortunately, time has not altered the fact that “color” and its associated connotations continue to convey and determine the treatment that many receive in the Americas and the Caribbean. When they migrate to the United States, some Latinos become more aware of the racism existing in their own country of origin, and other Latinos begin to question their conceptions of ethnic, racial, and national identities. Identities often thus become “a terrain of ideological contestation” (Duany 1998b:149; Foner 1998; Oboler 1995; Omi and Winant 1995; Torres-Saillant 1998). It was this ideological contestation that was manifested when Latinos checked the “other race” category and wrote in their national origins, ethnicity, and so forth on the decennial census forms. Thus, most of the 40 percent of Hispanics who marked the “other race category” and wrote in a Latino referent were asserting that they were “none of the above.” Others—non-Latinos— might fit them into one or more of the groups listed on the basis of color, phenotype, or biological or ancestral knowledge of “race” origin, but culturally or politically these Latinos did not see themselves as “white,” “black,” or “Asian or Pacific Islander”—or just one of these (Rodríguez 1992). According to their own, more culturally defined perspective of race, the “race” groups listed on the census were “social groups” but did not include their own social group. This is why many Latinos still mark “other” on census forms and fill in the space specifying their national origin. Still others disagree with the race structure mirrored in the cen-

LATINOS IN THE U.S. RACE STRUCTURE

11

sus’s race question and choose the “other race” category because they are more than “one of the above” race categories; that is, they are mestizo, mulatto, black Latino, or another mixture (Davis et al. 1998a; Rodríguez 1992; Rodríguez and Cordero-Guzman 1992; Rodríguez et al. 1991). Although the remaining 60 percent of Hispanics chose one of the census’s standard race categories, this does not necessarily mean that they all have assimilated or adopted the United States’ racial classification system. Rather, some Latinos believe that this is how they are seen and will always be seen in the United States and accept or understand that this is their race in this country. Others, however, choose one of the standard categories because that is what they are considered in their country of origin. As one Bolivian respondent explained in an interview conducted by the census, “I chose ‘white.’ I am considered white in my country” (Davis et al. 1998a:III-19).8 Still others are aware of the “official” pressure to mark one of the standard categories. As one Hispanic respondent in a census study indicated, “I do not consider myself white, but this is what the government says I am.” Another respondent said, “I don’t belong to any of these groups: probably I can be in ‘Some other race’ and say ‘Hispanic’; but I decided to use ‘White.’” Still another checked the white category but added, “I am a Hispanic white” (Davis et al. 1998a:III-20–21). These responses suggest that even though some Hispanics choose a standard race category, they believe that they also have other, or multiple, identities. Other Hispanics choose the standard race categories for the same reasons that members of other groups do. They determine that “biologically,” or in terms of “blood quantum,” they fit into a particular category (Davis et al. 1998b:48 ff). Finally, some Hispanics do not want to be (or admit to being) “other than white,” “other than black,” or “other than indio” (i.e., a member of an indigenous nation). That is, they identify culturally and/or politically with members of a particular category. Latinos’ responses to the census are discussed in more depth later. Suffice it to say at this point that in my many years of research in this area, I have noticed in my and others’ work that “race” is a recurring, sometimes amusing and benign, and sometimes conflictual issue.9 For Latinos, responses to questions of race are seldom as simple and straightforward as they tend to be for most non-Hispanic whites (Rodríguez et al. 1991). These “other race” responses presented a problem to the Census Bureau because they differed from previous responses and therefore

12

LATINOS IN THE U.S. RACE STRUCTURE

could not be easily fit into the existing race structure. What was to be done with the nearly 10 million Hispanics who answered the race question in this way? In what category were they to be placed? How were they to be reported or tabulated? In short, how was this group to be understood? When analyzing these results, references to this “data quality” problem were couched in terms of responses in “the other race” category. But the overwhelming majority (97.5%) who chose this category were “Hispanic,” and they accounted for 40 percent of the total number of Hispanics (U.S. General Accounting Office 1993:26). How, then, was this “other race” group (or Hispanic component) to be understood or accommodated in a country that for most of its history had employed an overarching dual racial structure with four presumed major color groups, that is, white, black, Asian or Pacific Islanders, and Native American Indian? This group, moreover, represented a growing number of people. In 1990, those who had checked the “other race” category represented the country’s second-fastest growing racial category (after Asian and Pacific Islanders) (Rodríguez 1991b:A14; U.S. General Accounting Office 1993). In addition, the population of Latinos was growing seven times faster than the population of the nation as a whole. Between 1980 and 1990, it had increased by half while the white (non-Hispanic) population increased by only 6 percent (U.S. Bureau of the Census 1991:table 1; U.S. Bureau of the Census 1993c:2).10 By 1999, the number of Hispanics in the United States (30 million) was greater than the total population of Canada. As we will see, the search for solutions to this and other problems has contributed to a radical reexamination of the concept of race by the U.S. government. This reexamination included numerous hearings, conferences, and massive studies of hundreds of thousands of households and resulted in the decision to reverse the Census Bureau’s twohundred-year policy. For the first time, in the 2000 census, respondents were allowed to choose more than one racial group when answering the question about race.

Demographic and Other Changes Also contributing to the question about the nature of race are broader demographic trends, such as immigration and the concentration (and consequently greater visibility) of racial and ethnic minorities

LATINOS IN THE U.S. RACE STRUCTURE

13

in populous states and metropolitan areas (Edmonston, Goldstein, and Tamayo Lott 1996). Added to this is the wide range of physical types of many immigrant groups, for example, Middle Easterners and Latinos, as well as the trend toward racial and ethnic intermarriage, particularly between those of high socioeconomic status (Edmonston, Lee, and Passel 1994; Kalmijn 1993; Rolark, Bennett, and Harrison 1994; Spickard 1989). These new trends contrast with past patterns, in which those in interracial unions were usually marginal, foreign born, or part of exploitative slave relationships (Berry 1963; Williamson 1984). Conversely, many of the children of these modern unions are attending university and will undoubtedly assume leadership positions in the future, in which their positions on multiracial identities will carry the weight of their class positions. The percentage of interracial marriages rose from 0.4 percent in 1960 to 2.2 percent in 1991 (Rolark, Bennett, and Harrison 1994), and the number of births to parents of two different races tripled, from 1.2 percent of all births in 1971 to 4.4 percent in 1995 (Atkinson, MacDorman, and Parker 1999).11 Indeed, the seriousness with which the proposal to include a multiracial category was received suggests that these forces have already influenced the way that race and ethnicity are viewed (see chap. 8). In addition to these demographic trends, the greater affirmation of a mixed-race identity and the increasing use and acceptance of selfidentification instead of observer identification have produced a more heterogeneous and more tenuous concept of race (Edmonston, Lee, and Passel 1994; Root 1992b, 1996) in the census and elsewhere. In this regard, it is interesting that in 1990, half (50.6%) the children of interracial unions were classified as “white” on the census form by their parent(s) (Bennett, McKenney, and Harrison 1995:table 5), whereas in the past, census takers would most likely have classified such children according to the race of the nonwhite parent.12 These trends are changing the “face” of the United States and will intensify in the twenty-first century, contributing to the growing trend to view race as many Latinos already do, as race-ethnicity.

Blurred Boundaries As increasing numbers of physically heterogeneous groups—such as Latinos—have become more concentrated and/or more visible,

14

LATINOS IN THE U.S. RACE STRUCTURE

questions of what constitutes “whiteness” and nonwhiteness have surfaced. Can individuals seen as white and those seen as nonwhite be members of the same race group? Where does whiteness—or blackness—begin? These questions have led to a reanalysis of whiteness and fundamental reconsiderations of race and ethnicity. (See, e.g., the following works, which examine how whites see themselves, how whiteness has been—or has not been—achieved by certain groups in American history and law, and how race and ethnicity are being rethought: Brodkin Sacks 1994; Delgado and Stefancic 1997; Ferrante and Brown 1998; Frankenberg 1993; Gallagher 1999; Haney López 1996; Ignatiev 1995; Waters 1990.) More and more native-born Americans see that many people’s racial/ethnic definitions of themselves are at variance with others’ definition of them. For example, white-appearing, third-generation Latinos, who sometimes no longer even speak Spanish, may insist they are “not white” or declare themselves to be “brown,” “black,” or “other.” Government officials, office managers, criminal justice administrators—that is, those who are responsible for counting race and ethnicity, are increasingly realizing that individuals—particularly the growing numbers of new and existing minorities—often define their “race” quite differently than they would be defined by others.13

THE PROPOSAL TO MAKE LATINOS A RACE In July 1993, the U.S. Office of Management and Budget announced that it would review the racial and ethnic categories used to collect government data (U.S. Office of Management and Budget 1997a). A number of proposals to amend the current categories were made. One proposal that received quite a bit of media attention was to add a “multiracial” category. Another proposal, even though it involved greater numbers of people, received considerably less attention: to make Hispanics a race.14 This proposal was subsequently referred to as “the combined question” because it would list “Hispanic” as a category along with the other race categories. That is, it would reclassify what the census had considered an “ethnic group”—in which Hispanics could be of any race—to a “race” group in which all Hispanics were of one race. What made this proposal curious was that Hispanics did not wholeheartedly initiate or support it, in contrast to other proposals con-

LATINOS IN THE U.S. RACE STRUCTURE

15

sidered at the time.15 Even more striking was the fact that evidently few Latinos noticed the lack of a Hispanic constituency. Although three Hispanic organizations were occasionally cited as supporting the proposal (del Pinal 1994; Wright 1994), a close look at their statements shows this was not exactly the case. Rather, their statements indicated reservations, questions, support for relabeling the race question “race/ethnicity,” and a need for more research (National Council of La Raza 1995; U.S. House Committee 1994k, 1994p). As the final chapter in this book makes clear, Hispanics were a significant but silent presence in the process, which was extraordinary given the striking population growth of Latinos in the United States. In March 1997, the Latino population was “officially” 29.7 million, or 11 percent of the total U.S. population (Reed and Ramirez 1998:table 1). This figure did not include, however, the 3.6 million Hispanics who lived in Puerto Rico (Hispanic Link, March 6, 1995, p. 1; Rodríguez 1994b) or those Hispanics who lived in the United States but were not counted. The U.S. Supreme Court recently decided in favor of total counts for the 2000 census, and not statistical sampling. The debate surrounding this highly politicized issue did not clearly explain the discrepancies that exist in each group with regard to the undercount.16 After Native Americans on reservations, who had an undercount rate of 12.2 percent, Hispanics had the highest undercount of all racial-ethnic groups, or 5.0 percent in the 1990 decennial census. African Americans followed with 4.4 percent, and non-Hispanic whites had an undercount rate of less than 1 percent (or 0.7%) (U.S. Bureau of the Census 1997:4). Moreover, about 4 million people, “most of them affluent whites living in suburbs that tend to vote Republican” were counted twice (Holmes 1999:24; and see app. A for a discussion of the undercount issue). But despite the undercount, the growth of the Hispanic population has been dramatic. Hispanic youths already outnumber black youths (Vobejda 1998:A2). Indeed, the U.S. Census projects that the Hispanic population will surpass the African American population by 2005, and it is expected to be about a quarter of the total U.S. population by 2050 (Day 1996:63,13; Larmer 1999). However, if immigration and birthrates continue to climb, some of these changes may occur much sooner than that. Notwithstanding the lack of support by this substantial and growing group, the proposal to make Hispanics into a separate race persisted

16

LATINOS IN THE U.S. RACE STRUCTURE

and became one of the primary propositions that the Office of Management and Budget examined in its extensive review between 1995 and 1997. The proposal was eventually dropped, however, when it became evident that making Hispanics into a separate race would result in fewer being counted—and in fewer whites being counted (U.S. Bureau of the Census 1996a, 1997).

MULTIRACIAL AMERICANS AND LATINOS The insistence on self-definition—particularly within one’s own linguistic and philosophical framework—is central to the challenges to racial construction in the United States today. The insistence on identity in one’s own terms is a major nexus between the issues raised by the multiracial movement and those raised by Latinos. Both groups seek, or have, definitions of self and their group that are often outside the biracial structure created in the United States. Furthermore, those who are “white” are dominant and thus determine who is “nonwhite” or “other.” Many Latinos, and many in the multiracial movement, are challenging these rigid categorizations, along with the implied racial hierarchy. Hispanics and those in the multiracial movement are often seen and defined as distinct groups, yet there are interesting overlaps. “Multiracial” Americans and those who go by the terms interracial, mixed race, and biracial are defined as “persons who identify with more than one race group” (Bennett, McKenney, and Harrison 1995:1). (Race group refers only to white, Asian or Pacific Islander, black, or Native American groups.) The census defines as “Hispanics” those who classify themselves as being of Hispanic or Spanish origin on the census, adding, “Hispanics may be of any race.” (The census defines origin as the ancestry, nationality group, lineage, or country of birth of a person or his or her parents or ancestors before their arrival in the United States [U.S. Bureau of the Census 1993b:B-12].) Yet many Hispanics claim a multiple “racial” ancestry. Indeed, in recent census tests, more Hispanics chose the “multiracial” category (6.7%) than did non-Hispanics (less than 1%), and about one-third of all those in the multiracial category were Hispanic (U.S. Bureau of the Census 1996a:13 and table 12). In addition, because many Latinos see race as a cultural construct, some consider themselves Latinos and “multira-

LATINOS IN THE U.S. RACE STRUCTURE

17

cial” because one parent is white, black, Asian, or Pacific Islander and the other is Hispanic or because each parent has a different Hispanic national origin.

HISTORICAL CONSTRUCTS An analysis of U.S. decennial census classifications shows the clear historical progression toward a more definitive bipolar structure. Although the taxonomy of race has changed, we can see in historical and legislative documents the evolution of two fundamental and socially constructed polarities that place “whites” at one end and “other social races” at the other. Although each of these polarities has been and continues to be fluid, this basic dichotomous structure has prevailed throughout most of the census’s two-hundred-year history. It is with this historically evolved bipolar structure that groups who have not been “quite white” or “quite black” have contended in the past, and it is in this structure that Latinos and other groups are entangled today. Although this bipolar structure has been overarching, providing the basic racial structure of the various “racial” groups, there is and probably always has been a great deal of heterogeneity within the two polarities. Moreover, the boundaries between these polarities have always been ambiguous and shifting. Finally, alterations of group and individual classifications have been both unofficial and legal and bureaucratic. For some people throughout U.S. history, the labels applied by the census and the identities created or used by the individuals and groups themselves have always differed. Furthermore, these externally created labels and identities have changed, so, for example, the Mohawks of the Hotinonshonni Confederacy refer to themselves—and recognize that they are also referred to—as “Iroquois,” “Native American,” or simply “Indian.”

IMMIGRANTS AND THE RACIALIZATION PROCESS In the past, new immigrants immediately underwent a racialization process, which conveyed an implicit hierarchy of color and power. The two elements of this racialization process were (1) the acceptance of and participation in discrimination against people of color (Bell 1992; Du

18

LATINOS IN THE U.S. RACE STRUCTURE

Bois 1962:700 ff; Morrison 1993) and (2) negotiations regarding the group’s placement in the U.S. racial-ethnic queue (Jacobson 1998; Rodríguez 1974; Smith 1997; Takaki 1994). Immigrants undergoing this racialization process discriminated implicitly or explicitly against others because of their color and status. Indeed, some immigrants realized that one way to become “white,” or more acceptable to whites, was to discriminate against others seen as “nonwhite” (Ignatiev 1995; Kim 1999; Loewen 1971). Kim (1999) reviewed the historical experience of Asian Americans being triangulated with blacks and whites through a simultaneous process of valorization and ostracism. This racial triangulation continued to reinforce white racial power and insulate it from minority encroachment or challenge. Some immigrants discriminated against blacks and/or other depreciated minorities by not living with “them,” not hiring “them” in enclave economies, or articulating prejudices against “them.” Institutionalized discrimination and normative behavior aided racialization so that, for example, it became difficult to rent or sell to members of certain groups because of exclusionary practices. Nearly all immigrant groups experienced this seldom-mentioned but indisputable dimension of the Americanization process. Critical to the racialization process was the belief that there was always some “other” group to which one was superior. Indeed, this process has been an effective means of protecting the status quo because it made it difficult to understand and pursue areas of common interest and resulted in divide-and-conquer outcomes.

Imputed and Self-Defined Race for Latinos Latinos—and many other groups—come to the United States with different views of race and with their own racial hierarchies. The relation of these people’s racialization to their hierarchies in the United States has not been widely studied. But it is clear that when they arrive, they too become part of a racialization process in which they are differentiated according to the official perception of their race, which may or may not be the same as their own perception. This racial reclassification immerses immigrants in a social education process in which they first learn—and then may ignore, resist, or accept—the state-defined categories and the popular conventions concerning race (particularly one’s own) (Rodríguez 1994a).

LATINOS IN THE U.S. RACE STRUCTURE

19

The racialization process also includes contradictory views of the way that Hispanics are generally regarded. At one extreme, Hispanics are a Spanish-speaking white ethnic group who are simply the most recent in the continuum of immigrant groups and are expected to follow the traditional path of assimilation. Another view holds that the term Hispanic—which has generally been unknown to new immigrants from Latin America—is subtly “colored” by negative and racial associations. For example, the stereotyped image (for both Hispanics and non-Hispanics) of a Hispanic is “tan.” Within this perspective, Hispanics are often referred to as “light skinned,” not as white. Yet many Hispanics would be seen as white, black, or Asian if it were not known that they were Hispanic. But seeing Hispanics/ Latinos as “light” clearly restricts their “whiteness” and thus makes them nonwhite by default, but not a member of other race groups. Thus, many Hispanics entering this country become generically “nonwhite” to themselves, or to others, regardless of their actual phenotype or ancestry. The United States’ racialization process affects all groups’ sense of who they are and how they are seen, in regard to color and race. There are few studies of this concerning Latinos, but some autobiographies suggest that the racialization process has had a significant impact (see, e.g., Rivera 1983; Rodriguez 1992; Santiago 1995; Thomas 1967). Whether this has been a dissonant impact and has affected Latinos’ mobility and the quality of life has not yet been determined. Some Latinos, influenced by movements such as the Black Power movement, Afrocentrism, Pan-Africanism and African diaspora philosophies, and the celebration of negritude, have come to see themselves, and sometimes their group, as black. Terms like Afro-Latino, black Cuban, and black Panamanian are now common, and some Latinos celebrate their African roots. Others focus on their Amerindian or indigenous component, while still others see themselves only as white or mixed or identify themselves only ethnically. A Dominican student of mine told me that each of her and her husband’s children claimed a different identity. So they had one black child, one white child, and one Dominican child. Each of the children had different friends and tastes. Many variables contribute to and interact with the racialization process to determine how individuals decide on their group affiliation. Generation, phenotype, previous and current class position, and the size and accessibility of one’s cultural

20

LATINOS IN THE U.S. RACE STRUCTURE

or national-origin group, as well as the relative size of other groups, all affect how individual Latinos identify themselves.

DISCRIMINATION Most Latinos believe that they are discriminated against as a group. In one of the largest and most comprehensive surveys of Latinos, 80 percent of Mexicans, 74 percent of Puerto Ricans, and 47 percent of Cubans reported “a lot” or “some” discrimination against their own group, a general perception that appeared to be unrelated to skin color (de la Garza et al. 1992:94–95). Falcon (1995), for example, found that Puerto Ricans’ phenotype was not related to their perception of group discrimination. Thus, although darker or more visible Latinos may experience more direct discrimination, looking white or light does not substantially alter their perception of discrimination. Indeed, it may sometimes have the opposite effect. That is, lighter Latinos may more often be in a position to observe discrimination. They may be assumed to be white and consequently be better able to see how others are treated or that they are treated differently from those who are darker. Moreover, all Latinos, regardless of color, may experience discrimination, for Hispanicity is based on more than skin color. Other clues, such as accent, residence, surname, or first name, can reveal that a person is Hispanic. Thus, despite an individual’s physical appearance as “white,” knowledge of this person’s Hispanicity often causes a readjustment of status. The perception shifts from “I thought you were one of us” to “You’re an other”—and even an accent is heard where it was not before. This type of redefinition or reclassification may be imposed more often on lighter Latinos and may make them just as conscious of discrimination as darker Latinos are. Therefore, even though “color” or phenotype is significant in an individual Latino’s experience, all physical types can and do experience discrimination.17 Considerable evidence shows that the discrimination Latinos perceive is very real, for example, disparities in judicial treatment (DíazCotto 1996:416–417; Haney López 1996:138–139, 252–253) and evidence of housing discrimination (Denton and Massey 1989; James, McComings, and Tynan 1984; Massey and Denton 1990; Yinger 1995). In New York City, black and Hispanic immigrants—particularly those from the

LATINOS IN THE U.S. RACE STRUCTURE

21

Dominican Republic—continue to live in the least desirable housing, pay among the highest percentages of income for rent, and have the lowest rates of home ownership compared with European, Russian, and Asian immigrants (Hevesi 1998; Schill, Friedman, and Rosenbaum 1998). Moreover, because of where they live, Hispanics and blacks in New York City—whether they are foreign born or native born—have less access to medical care, higher crime rates, and greater concentrations of poverty and housing-code violations (Rosenbaum et al. 1999). Individuals who are clearly identified as “Hispanic” by their names, résumés, accents, and, sometimes, stereotypical looks experience greater job discrimination than do equally qualified whites (Bendick 1992; Cross et al. 1990; Fix, Galsten, and Stryk 1993). Also, Hispanics experienced greater employment discrimination as a result of the 1986 Immigration Reform and Control Act (Bendick 1992; U.S. General Accounting Office 1990). With the passage of legislation sanctioning employers for hiring undocumented workers, many Hispanics who are citizens or legal residents were not hired for jobs for which they were qualified because employers thought they might have been in the United States illegally. Given these findings, it is not surprising that a review of judicial cases involving employment discrimination based on national origin found that most of the litigation pertained to Hispanics (del Valle 1993). Studies of employer preferences in hiring also suggest that discrimination against Hispanics is widespread in the labor market (Holzer 1997; Hossfeld 1994; Moss and Tilly 2000). In these studies, the employers interviewed had definite beliefs and preferences concerning the suitability of different groups for different jobs, including “negative attitudes” toward “workers of color” (Moss and Tilly 2000). According to Darity and Mason (1998:81), employers “set up a racial/ethnic gender ranking of potential hires” that favored white men and women workers over Hispanics and blacks. These studies underscore the disadvantages that race/color (and ethnic) markers can bring to employment and hiring practices (Darity and Mason 1998:81). The literature on the effect of labor market discrimination on earnings and occupational attainments has yielded a complex array of findings that reflect not just differing theoretical perspectives but also variations in sampling and methodology (Meléndez and Rodríguez 1992; Meléndez, Rodríguez, and Barry Figueroa 1991:293).18 More recently, the focus of labor market research has moved beyond measuring the

22

LATINOS IN THE U.S. RACE STRUCTURE

extent of in-market discrimination to the effect of premarket factors (e.g., human capital characteristics like educational attainment) and preemployment skills (e.g., punctuality). These researchers argue that Hispanics receive less compensation or are less often hired because they do not have the same preemployment skills as others and because premarket factors keep them out of the competition. But they gloss over the role of discrimination in premarket factors. For example, where one lives (or can live) influences early educational options and social, political, and personal networks. These, in turn, affect subsequent educational opportunities, which influence scores on tests, which influence educational options and outcomes. In addition, although the lack of preemployment skills is often mentioned as a reason for Hispanics’ lower incomes, there has been little systematic or scientific research on whether Hispanics as a whole have fewer preemployment skills. This explanation is reminiscent of earlier images of African Americans as lazy and shiftless when in fact more were working in the fields and other arduous occupations than others were.19 Similarly, Hispanics who often have poorly paid jobs without benefits, security, or full-time employment (Boisjoly and Duncan 1994) and are overly represented in “jobs others won’t do” are seen to lack preemployment skills. Yet in order to hold jobs, such as taking care of other people’s children, lawns, homes, meals, and apartment buildings and working in the food and textile industries, they must arrive on time and operate quickly and efficiently. From a more journalistic and contrastive perspective, Skerry (1990) contends that since Hispanics are not a race, they cannot be subject to racial discrimination in employment. Nonetheless, we have seen that although some Hispanics identify themselves as a cultural or ethnic group, others may see them as a “Spanish” race or as nonwhite. Whether ascribed race or self-reported race is more determinant of how Hispanics are treated in the United States has not yet been resolved or studied systematically.20 Some research, however, indicates that Hispanics who report they are black or are seen as black are more segregated and less successful in gaining access to predominantly Anglo residential areas than are their white Hispanic counterparts (Denton and Massey 1989; Massey 1988; Massey and Denton 1993:113 ff; Rosenbaum 1996). In addition, Latinos who classify themselves as white or are identified as white (or light) fare

LATINOS IN THE U.S. RACE STRUCTURE

23

better with regard to earnings, hourly wages, and other socioeconomic variables than do other Latinos (Arce, Murguía, and Frisbie 1987; Gómez n.d.; Katzman 1968; Relethford et al. 1983; Rodriguez 1990, 1991a; Telles and Murguía 1990). Moreover, “black Hispanics suffer close to ten times the proportionate income loss due to differential treatment of given characteristics than white Hispanics” (Darity and Mason 1998:72). The results of these studies suggest a need to continue collecting “race” data on Hispanics, for they indicate a possible economic rent, color credit, or tax paid, depending on perceived or imputed race. (The differences found within Latino groups, however, are less pronounced than those between white Latinos and non-Hispanic whites.) These findings parallel those found in the African American community, in which those with a lighter skin color had higher socioeconomic outcomes and those with a darker skin color were moderately associated with being working class and having a low income or little education (Hughes and Hertel 1990; Keith and Hering 1991; Krieger, Sidney, and Coakley 1998). Interestingly, as in the case of Falcon’s 1995 study of Puerto Ricans’ phenotype, color shade did not seem to be related to self-reported experiences of racial discrimination (Krieger, Sidney, and Coakley 1998).

AN UNEQUAL PLAYING FIELD Whether or not the result of discrimination, the demographic picture of Hispanics suggests that disparities exist in regard to standard socioeconomic indicators. For example, in 1996, more Hispanics were living in poverty than whites and even blacks. Hispanic men were more likely than white men to be employed, but they had higher unemployment rates. Despite the high numbers of Hispanics in the labor force, their income continued to be two-thirds that of whites, with family income slightly below the black average. Among married-couple families in which at least one person was working, Hispanics had the highest poverty rates and the lowest income levels, compared with both white and black families. Hispanics also paid a higher proportion of their income for housing than did either whites or blacks (National Council of La Raza 1997). Hispanics were less adequately covered by health insurance, having lower health insurance rates and pension benefits than did

24

LATINOS IN THE U.S. RACE STRUCTURE

either whites or blacks (del Pinal and Singer 1997:36–37; National Council of La Raza 1997; Santos and Seitz 2000). Finally, while high school completion rates have improved for whites and blacks, for the last thirty years, more Hispanics have continued to drop out of school. In 1994, this figure was 2.5 times the rate for blacks and 3.5 times the rate for whites. One in five Hispanics aged sixteen to twenty-four has left school (Secada 1998:5). Both U.S.-born and foreign-born Latinos continue to lag with regard to education (Chapa and Wacker 2000). And these results are not simply a transitory reflection of the increased number of unskilled Hispanic immigrants. Whereas other studies have concentrated on past and continuing structural, institutional, and discriminatory barriers that many Latino groups face (De Freitas 1991:4–5; 53–94; Morales 2000; Morales and Bonilla 1993; Rodríguez 1989:85–105; Torres 1995; Torres and Rodríguez 1991), at least two studies have concluded that the negative standing of Latinos relative to that of other groups cannot be attributed to immigration (Grenier and Cattan 2000; Valenzuela 1991). The economic boom at the end of the twentieth century has had a modest trickle-down effect. As compared with the past, Latinos today have a higher rate of home ownership, college completion, and earnings for college graduates, particularly for young Latinas (National Council of La Raza 1997; Reimers 2000). In addition, different pictures emerge when we examine diverse Hispanic groups by region, generation, and the like. For example, Cubans in Florida and Puerto Ricans in Texas typically live at a higher socioeconomic level than do Hispanics as a whole (García 1996; Pedraza-Bailey 1985; Portes and Bach 1985; Rivera-Batiz and Santiago 1997; Rodríguez 1991a:27, 46). Nevertheless, the broad indicators suggest that Hispanics’ general socioeconomic situation is not favorable. Moreover, the perception and evidence point toward discrimination. In other words, the playing field is not level, which further complicates issues of race.

The Reality of Race This book emphasizes the social constructedness of race and how Latino experiences in the United States illustrate race as a social construction.21 We should not, however, lose sight of the continuing significance of race. The research still shows that race and ethnicity influence

LATINOS IN THE U.S. RACE STRUCTURE

25

where and how people live, work, shop, and play and how they are treated in everyday social interactions and in institutions. Race is different, and it also feels different (Edley 1996). Racial/ethnic categories in the United States are still socially meaningful indicators of racial subordination, privilege, and denomination (Bhopal and Donaldson 1999: 784; Krieger, Williams, and Zierler 1999:782).22 My own life experiences have demonstrated the social constructedness of race, and this book shows that “race” is not fixed, is imperfectly measured, is at variance with scientific principles, is often conflated with the concept of “ethnicity,” and is under increasing scientific criticism and popular interrogation. Nonetheless, race is still real; it still exists.23 We may question its necessity, the right of anyone to establish such markers, and its validity as a scientific concept. We may see it as unjust and want to change it. But we must acknowledge its significance in our lives. It can be deconstructed, but it cannot be dismissed.

Race as a Changing Concept The concept of race is changing in the United States and Latin America and around the world. Increasingly, we find both exclusionist and inclusive definitions of racial and ethnic identities that go beyond nation-state boundaries, for example, in organizations such as the Aryan Nation and its international cousins, organizations for indigenous peoples worldwide, the various movements and organizations of African and African-descended peoples, and various diasporas. In our increasingly global world, all these definitions and movements help change race. Hanchard, speaking specifically about African-descended populations in the United States, argues that restrictions on their “citizenship and movement in the United States” have led “black political actors” to mobilize politically and transnationally (1999:1). Adding to the increased identification as African-descended populations are the affirming and reclaiming of ancestral identities that have always existed and were featured during the black power, American Indian, Chicano, Puerto Rican, and Asian American movements of the 1960s and 1970s. Similar restrictions and affirmation can be found among members of other populations when they travel throughout the world, and they also lead to greater and broader identification with ancestral groups. Opposing trends can be found as well, toward more restrictive

26

LATINOS IN THE U.S. RACE STRUCTURE

ethnic identifications and rivalries, such as in the ethnic cleansing in eastern Europe and Rwanda. Whether these trends will result in a more homogenous concept of race, built on U.S. race constructs, or in a greater variety of racial constructs remains to be seen. But even though the outcome may not yet be clear, it is clear that race is changing.

2 The Idea of Race

T H E R E S U LT S O F recent censuses, as well as the personal experiences of Latinos and non-Latinos, raise the question of what race is in the United States. Latinos’ wide range of physical types, their history, and their more “social” or cultural views of race have historically challenged U.S. racial constructions, and the government has had difficulty categorizing them. As the next chapters will show, Latinos are not alone in this regard, as other groups have had similar histories and present similar challenges. But it has been the increase in the two i’s, immigration and intermarriage, that has made questions of racial classification more salient and has led to the question of just what race is. In this chapter, we begin exploring these questions by examining “the many faces of race,” its multidimensional nature. We then turn to race as it has been commonly understood in the United States or, as U.S. courts have termed it, race “in the common understanding” (Haney López 1996:85, 91, 107). This understanding sees race as a “self-evident ‘fact’ requiring no protracted thought” (Hannaford 1996:3) and as existing in the same way in all places and times. We challenge this idea, however, when we examine (1) studies of “race” in the past, (2) how other governments count their populations, (3) the literature on “mixed” race, (4) changing U.S. census classifications, and (5) standard reference sources of racial definitions over time. Evident in these examinations are the fluidity and variability of race over time and place and its overlap with ethnicity, which is dependent on context.

THE MANY FACES OF RACE Race has many dimensions and so is often used and defined in different ways. For example, race can be as defined by official bodies, such as the census or state governments. This is state-defined race. Race also is the 27

28

THE IDEA OF RACE

perception or experience of laypersons. This is often referred to as popular race, folk race, or race “in the common understanding” (Jensen 1988; Wright 1994:50). Although state-defined race is often thought of as reflecting popular race, they often influence each other. For instance, law constructs race, and states can define, restrict, or privilege races through legislation (Haney López 1996:19). Race is also studied by scholars who examine racial ideologies or ideas in public pronouncements, policies, or literary works. This is referred to as ideological race (e.g., Graham 1990; Horsman 1981; Stanton 1960). In both academia and more popular circles, we find the “whatever you think it is” concept of race, which is often a shifting combination of all of the above and frequently translates into the “you know one when you see one” idea. Some people think of race as “identity” and “how you see yourself.” Others consider race to be determined more by “how others see you.” These two views sometimes conflict, hence, golf champ Tiger Woods’s dilemma in which his view of himself as being of mixed race conflicts with the view that many have of him as “black.”1 In reality, racial definitions are often both external (what others think) and internal (what the subject thinks). This external-internal axis is also described as “imputed versus self-defined race” or “objective versus subjective” definitions of race. Each of these different internal/external usages is strongly affected by cultural and class considerations. But these nuances or different definitions of race are generally not acknowledged in people’s everyday conversations.

RACE IN THE UNITED STATES In the United States, race as defined “in the common understanding” has usually been simple and straightforward. On the simplest, least reflective, and most practical plane of interaction in the United States, race is often thought of as one’s biological ancestry, manifested most clearly in skin color. Within this one-dimensional conception, color terms are frequently used to designate different “races.” Thus, there are ostensibly four color groups, roughly corresponding to geographic regions: black (Africa), white (Europe), red (North America), and yellow (Asia). In this color palette, what makes a person “white” is the absence of any “black” or nonwhite blood, and what makes a person “black” is the presence of “black” blood. White is white because it is not mixed

THE IDEA OF RACE

29

with any other color; it is “pure.” “Black” blood, however, is “potent” or “polluting,” in that it takes only a small amount of “black blood” (ancestry) to make someone “black.” This concept is referred to in academic circles as hypodescent, or “the one-drop rule.” (Davis 1992; Root 1992b; Williamson 1984; Wright 1994).2 Despite these different color terms, in all four cases, the white category is the norm or referent, and the other three groups are nonwhite. We can understand the significance and power inherent in this construction of groups if we imagine a similar classification schema that uses another defining category. In this case, we would have reds and nonreds or yellows and nonyellows. In broad and blunt terms, this is the way in which “race” has been simply understood in the United States; this has been its social construction. These four race-color groups have had and continue to have corresponding categories on census forms, one of which individuals must choose, and not more than one.3 The fact that the four categories have been presented as mutually exclusive conveys the impression that each of the groups is a “pure” race (Lee 1993). Although this has been the impression, the reality was and is quite different. Appendix B contains a closer review of contemporary critiques of “race” as constructed in the United States. In short, it has been challenged on the grounds that it is illogical and inaccurate; based on unscientific assumptions; more determined by contextual, political, economic, and social factors than generally acknowledged; and generally dismissed by scholars in the field (see, e.g., Begley 1995:67; Gregory and Sanjek 1994:6–7; Gutin 1994:73; Marks 1994, 1995; Rosin 1994; Sanjek 1994; Shreeve 1994:60; Washburn 1963; Wills 1994:81). More journalistic treatments have also found fault with this concept (see, e.g., the following extensive treatments: Barringer 1993; Discover November 1994; Lemonick and Dorfman 1999; Morrison 1993; Mother Jones October 1997; Newsweek February 13, 1995; Rosin 1994; The Sciences March/April 1997; Weissman 1990; Wood 1994; Wright 1994). The fact that this issue has appeared in the popular press indicates that it has gone well beyond modest academic contemplation. In addition to, or along with, this academic and journalistic examination has been an official, high-level, and massive reconsideration of racial-ethnic categories for the 2000 census, resulting in the decision to eliminate the Census Bureau’s “choose-only-one-category” standard that has contributed to the myth of pure races.

30

THE IDEA OF RACE

Given the U.S. racial classification system’s reliance on color, we should look at the critiques of the race concept’s emphasis on color and color differences among population groups. Population groups do vary by skin color. Thus, groups like the Scandinavians, who have lived for long periods in areas with little sun, have less melanin in their skin on average than do Africans who have lived for a long time close to the equator. Melanin is the chemical substance responsible for color in the skin. It protects humans from the sun’s ultraviolet rays, so in areas of the world where these rays are stronger, people’s skin contains greater amounts of melanin. Likewise, in areas where there are fewer ultraviolet rays, skin pigmentation is lighter.4 Consequently, skin color is an adaptive—evolved—characteristic, but it is independent of other genetic characteristics. That is, color can be inherited independently of hair texture or color. Geneticists examining the role of melanin have determined that it is most likely not related to differences affecting intelligence, personality, or ability (Wills 1994). Therefore, color—ostensibly the principal marker distinguishing groups or “races” in the United States—must act in concert with other variables to determine differences among groups, for by itself it does not seem to play a major role. Indeed, if the presence or absence of melanin were related to variables such as intelligence, we would expect that within population groups, tanning ability would reflect greater or lesser intelligence. But if you think for a moment about beach-tanning profiles, you will quickly dismiss this hypothesis!

RACE IN THE PAST Until recently, race in the United States was generally seen to be a universal given—uncomplicated, unchangeable, and unavoidable. Indeed, U.S. sociology textbooks often describe race as an “ascribed” characteristic, in contrast to “achieved” characteristics such as education and income. Some scholars contend, however, that what we understand as “race” today is what we used to understand as “ethnicity” (Bernal 1987; Dunn and Dobzhansky 1952:107–108; Sanjek 1994; Shreeve 1994:60; Snowden 1983; Thompson 1989). Other studies argue that race as we understand it in the United States today is a modern invention with no equivalent in pre-Columbian history (Bernal 1987; Hannaford 1996;

THE IDEA OF RACE

31

Snowden 1983; Thompson 1989). Hence, referring to “blacks” in ancient history is incorrect, because people then identified themselves and were identified primarily by religion, language, culture, and other variables, and not by color. Some writers maintain that ancient societies did not harbor the color prejudice of modern times (Hannaford 1996; Harris 1977; Jalloh and Maizlish 1996:9; Snowden 1983; Thompson 1989). In fact, Snowden believes that “this is the view of most scholars who have examined the evidence” (1983:63). In essence, the ancients made ethnocentric judgments about societies; subscribed to narcissistic canons of physical beauty; considered themselves civilized and others, barbarians; but did not regard black skin color as a sign of inferiority. The Greeks and Romans did not establish color as an obstacle to integration into society, and color was not the basis for judging a person. In regard to immigrants, Snowden argues that black émigrés were not excluded from the opportunities available to others of alien extraction, nor were they handicapped in fundamental social relations. Rather, they were “physically and culturally assimilated: in science, philosophy, and religion,” and “color was not the basis of a widely accepted theory concerning the inferiority of blacks” (1983:108). Egyptians, for example, saw their land as the only one that really mattered and considered outsiders to lack some elements of humanity. But “once a foreigner came to live in Egypt, learned the language and adopted Egyptian dress, he or she was accepted as one of ‘the people’” (Snowden 1983:89).5 Thompson, focusing specifically on the Roman Empire, supports Snowden’s view concerning the relative absence of color prejudice (1989:10 ff). Thompson asserts that “a black in possession of symbols of high status received appropriate deference from those of lower (genuine or apparent) status irrespective of colour and ethnic identity or origins.” This was the case “even if he happened to be passing through a district whose population lacked current familiarity with the sight of black faces.” The treatment he received depended above all “on the personal status and deference-position of each of the parties in the encounter, and there was considerable variety in the statuses and (positive and negative) deference positions of blacks.” There were, however, few blacks above the rank of plebeian (Thompson 1989:158–159).6 Thompson also agrees with Snowden that what “race” is today was not what it was in the past and so it would be wrong to apply today’s

32

THE IDEA OF RACE

assumptions to past relations (1989:10 ff).7 “Of course the notion of a collective mind precisely and exclusively linked at any given point in time with a particular skin colour (let alone the idea of an eternally fixed ‘white’ or ‘black’ or ‘yellow’ mentality) is an utter absurdity” (Thompson 1989:8). He points out that “differences in cultural habits, and in quality and scale of material goods” were as important in determining social distance as is “somatic distance” or “what is today popularly called ‘race.’” In our world, Thompson maintains, terms like European, whites, and blacks symbolize a particular cultural situation and power relationship. These symbolic assumptions were not relevant to the ancient Romans, however, who saw the majority of the world’s white inhabitants as “savages” and “benighted barbarians.” Therefore, to the Romans, “white people” was not a meaningful sociocultural category (Thompson 1989:10 ff).8 In their review of much of the classical literature, Harris (1977), Hannaford (1996), and Bernal (1987) agree that the idea of race as we know it today is not evident in these early works. Kinship, nationality, and cultural or religious identity had meaning then, but “skin colour in itself” had “no more meaning than height or weight” (Thompson 1989:8). Referring to later interpreters of these works—who, Hannaford (1996) and Bernal (1987) argue, did racialize early classical writings— Thompson says: “It is the mind of the observer that, drawing on past experience, renders pigmentation and other physical traits a repository of messages about personal beliefs, cultural habits, and social status, and makes these traits a focus of passionate sentiments transcending the merely aesthetic” (1989:8). Intermarriage was not prohibited and was common (Thompson 1989:40, 44, 95).9 In addition, the hypodescent rule did not exist. Rather, “in the Roman perceptual context the progeny (and even less so the more distant descendants) of an Aethiops did not necessarily fall into the category of Aethiops: some were perceived as ‘swarthy,’ some as ‘white,’ and some as Aethiops, the classification in all cases depending entirely on the individual’s physical appearance” (Thompson 1989:158). Snowden (1983) studied African blacks in northeast Africa and the Sudan and the Kushites in southern Egypt during ancient times. He examined meetings of blacks and whites, images, inscriptions, and literature and concluded that their views were largely the result of first impressions and a long history of contact and relations between the an-

THE IDEA OF RACE

33

cients and long-established African nations. Blacks had first been encountered as military men (often as part of conquering armies). Aethiops (Africans) were seen as civilization’s pioneers, astrologists, and writers and also were known to have had long-term territorial integrity, material resources, and trade. “There was clear-cut respect among Mediterranean peoples for Ethiopians and their way of life. And, above all, the ancients did not stereotype all blacks as primitives defective in religion and culture” (Snowden 1983:59).10 This contrasts with Americans’ first view of blacks—as slaves. Furthermore, in antiquity, slavery was independent of race or class, and most slaves were white. As Snowden (1983) put it, the ancients enslaved all conquered people, whereas the moderns, only the colored races.

UNRESOLVED ISSUES Embedded in these views of race in the past are a number of unresolved issues. One is relative size. Snowden (1983) examined—but took issue with—the postulate that the relatively small size of the black population at that time helped minimize the hostility toward them. He argues that the exact ratio of blacks to whites is not known because the ancients did not consider color sufficiently significant to mention it. Nevertheless, Snowden feels that blacks were more numerous than previously thought. The emphasis on counting “pure” Negroes, he argues, is misleading and tends to underestimate the number of blacks, just as such an approach in the United States now would tend to underestimate the number of blacks. He also notes that blacks are often depicted in pottery and artwork, suggesting that they were more numerous. But iconographical evidence of blacks has been neglected. Another issue is selectivity. Harris maintains that we have really two streams of information from the classical writers: one favorable to the people of Africa and one not favorable and that the unfavorable characterization has “had the greatest influence on the image and treatment of blacks in our own times” (1977:xx). Did the authors just cited ignore the unfavorable literature? Were the classic views of peoples farther south in Africa more negative (in comparison with the views of peoples closer to the Mediterranean)? Although the issue of whether darker-skinned groups were regarded less or more favorably than today has not been resolved, there

34

THE IDEA OF RACE

is general agreement that the ancients referred to as Aethiopes (Africans) had a highly developed civilization and that Africans and other Europeans had substantial and influential contacts before the development of Greek civilization (see, e.g., Harris 1977: 61–62, 67–147, 89; also see pp. 5–15 for a discussion of the derivation and meanings of the term Aethiopes). Another issue is that many of these sources refer only indirectly to blacks or Aethiops, even though they have been the focus of the research on blacks in antiquity. In no texts do the blacks speak for themselves. In addition, references to blacks are often made by those with, at best, second-hand information or by travelers. Therefore, one wonders how reliable these analyses are for interpreting color and race attitudes. Finally, there is the problem of accuracy concerning sources like works of fiction, scriptures, and pagan texts written from the perspective of the elite, who were remote from blacks (conversation with Prof. Lucius Outlaw at Haverford College). These issues are countered by the argument that the reason there are no texts in which blacks speak for themselves is that those peoples referred to as black today did not see themselves as such then, nor were they seen as such by others. Moreover, the application of this modernday, racialized lens to the past is problematic. An example is the Egyptians, who are sometimes referred to as African and therefore must be “black.” Yet Bernal (1987) asserts that because the Egyptians were geographically positioned at an important trading point, their population contained many different physical types. This assertion leads to two questions: One is whether darker-skinned Egyptians held or evinced a different identity but did not write about it. The second is whether Egyptians at that time saw themselves as a “black” or “nonwhite” group distinct from other groups. These issues continue to be debated today.11

THE SHIFT TO A RACIALIZED PARADIGM Regardless of how much the same or different race was in the past from what it is today, a number of scholars agree that the way in which race was conceptualized underwent a major shift, although not all agree on exactly how and when this shift began. Most believe that it was in place by the time that routes through Asia and the New World had begun to

THE IDEA OF RACE

35

be explored (see, e.g., Bernal 1987; Hannaford 1996; Johansen 1982:84; Sanjek 1994; Shreeve 1994; Snowden 1983; Thompson 1989).12 Bernal (1987) sees this as a shift from the ancient model to the Aryan model and argues that this shift was reinforced between 1785 and 1850 by the ascendant paradigms of progress, romanticism, and scientific racism. Also aiding this shift were the French Revolution and the consolidation of northern expansion into other continents (Bernal 1987:22 ff). According to Thompson, this shift encompassed not merely “the justification of the historically peculiar configuration of ‘white’ master/conqueror set against ‘coloured’ slave/subject” but also a rewriting of history that diminished and denigrated the role and contributions of Africans and Asians to civilization—to say nothing of those of Native Americans (1989:10–11). Also obscured in this process were the earlier relations between blacks and Europeans and the earlier conceptions of race. Thompson estimated that the shift occurred in the eighteenth century when Europeans (at home and in their colonies) began to attach greater significance to somatic distance than to religious and other cultural differences between themselves and other peoples.13 Thompson (1989) contends that in earlier times, non-Europeans were seen as “essentially” different but that they always believed that people of all ethnic categories could move from one socioeconomic category to another. Snowden contends that arguments for the “naturalness” and abolition of slavery were mustered only in the New World (1983:70 ff). Before this, a color association with slaves did not exist, since slaves were of many colors, although most were European and North African (Forbes 1988:101). According to Thompson, the European outlook changed when people became highly conscious of the distance between culture and technological and material power that came to separate the white from the nonwhite parts of the world. This was reinforced by the institution of all-black slavery and by European imperialism on other continents. This power distribution determined the world and caste into which people were born. Sanjek pointed out that by the 1700s, efforts were made to “fit” exploited peoples into “natural” schemes that would rationalize their oppressed position and included the devaluation of peoples of color (1994:1–17, esp. 5). A number of works have traced the emergence of these “scientific” racial classification efforts in western Europe and the United States during this period (see, e.g., Banton 1983; Barzun 1965; Bieder 1986; Freedman 1984; Gossett 1963; Gould 1981; Jordan 1968;

36

THE IDEA OF RACE

Sanjek 1994:5; Stanton 1960; Thomas 1989:29–31). Other works have critiqued and disproved the results of these earlier “scientific” studies (Gould 1978, 1981; see Sanjek 1994:5 for a review of this literature).

Mixture and “Pigmentocracies” Despite the racist paradigm that developed after the fifteenth century, people throughout the world began to interbreed, resulting in “varying social constructions” of racial identity in the United States, the Caribbean, Latin America, South Africa, and elsewhere (Sanjek 1994:4). These various constructions of racial identity were affected by the predominant ideological racial paradigm of the times. Racial rankings were understood and communicated through these paradigms and implemented through legal frameworks that specified racially determined limits to social interaction (Sanjek 1994). Later chapters analyze these social formations more closely. In short, although also affected by the same shift and therefore fundamentally racialist, the development of “race” in Latin America remained closer to earlier conceptions of race/ethnicity.

Governments Count “Peoples” around the World When we examine how other countries count their populations, we find that not all ask about “race” or “color”—indeed, only a minority does. Even ethnicity has not been found to be a universal population characteristic; nor is there general agreement on what constitutes ethnicity.14 Tamayo Lott (1997; 1992) notes that a dominant theme of a major international census conference was that ethnicity is constructed differently in each country—that for some race was a dimension of ethnicity, while for others ethnicity was a dimension of race. For nonwhites, race was more important than ethnicity because it represented their unequal power relations with Whites. Bates et al. (1995:433–35) reviewed a nonrandom sample of recent censuses in 45 countries. They found that tribe, nationality, linguistic group or dialect, district or country of birth, religion or sect, ethnic group, citizenship, indigenous or aboriginal origin, race, and skin color were among the criteria used (alone or in combination) to count populations in these countries. In half of the countries, questions about race

THE IDEA OF RACE

37

and ethnicity were not asked. They also found that direct inquiries specifically about “race or color” were “rare” and were concentrated in countries in or near the Caribbean region (Bates et al., 1995:434). Although this was not a representative sample, it does indicate that many governments do not ask about race/ethnicity, and that different cultural, political, ethnic, and religious criteria are used to measure and distinguish populations. The data collected by the UN (United Nations 1992) also reveal a great deal of fluidity and variability with regard to how different nation-states classify “race” groups in their countries. Different variables are used as bases for group classification, e.g., ethnic nationality, race, color, language, religion, customs of dress or eating, tribe, or various combinations of these factors (United Nations 1980:79; 1985). Indeed, it is difficult to make international comparisons because the data gathered are dependent on national circumstances (which are highly variable) (United Nations 1980; 1985). Many countries collect data on minority groups for reasons that are quite similar to those of the United States, i.e., because they are concerned with the equal participation of all groups and the equitable distribution of benefits to all groups (United Nations 1989:30–31). Finally, a survey of the censuses of 51 countries in the Americas over the last 40 years (Almey, Pryor, and White 1992) examined how race and ethnicity questions were asked and found that 31% (16) of the countries they surveyed did not include race/ethnicity questions on their national censuses, while 68.6% (35) did. The authors also found that many people did not want to, or were unable to, respond to ethnic background questions. The authors concluded that these results question the assumption often made that everyone has an ethnicity. Moreover, they found that there was “no consensus on what criteria determines ethnicity” (Almey, Pryor, and White 1992:3). Those countries that did distinguish drew from a variety of factors in different combinations to determine ethnicity, e.g., language, country of birth, residence, color, race, and religion. In addition, the study maintained that census questions on ethnicity elicited multiple responses and that these responses changed over time.15 In all three Americas, there was a striking fluidity and variability of race and ethnicity between countries and over time. Even in countries that share ostensibly similar “racial,” “cultural,”

38

THE IDEA OF RACE

or political identities, different approaches are sometimes taken. For example, English-speaking countries with common historical and political ties, e.g., Australia, Canada, and Britain, ask very different questions on their censuses. Britain, for example, has various categories of Blacks (Caribbean and African) and Asian Indians (Bangladeshi, Indian) (Sillitoe and White 1992; White and Pearce 1993). Australia does not ask about race, but has an open-ended question that asks about ancestry and a question that asks about aboriginal descent (Cornish 1992). These countries have also changed their questions and categories over time. The fluctuations inherent in the classification process can also be seen in the experience of other countries, which have altered their criteria to accommodate new populations and/or to be in accord with new political regimes, e.g., Russia and Malaysia. (Statistics Canada and U.S. Bureau of the Census 1993) Canada provides an interesting example of change and continuity. For example, it still collects data on their indigenous population and on the Métis, people who are the result of indigenous and European mixing. It has also included an ethnic origin question in all but its 1891 census. It recently reintroduced a race question after not having had one in its census for decades. On its 1996 census, it referred to races as “population groups” and further clarified that its term “the visible minority population” referred to those who were “non-Caucasian in race or non-white in colour.” Finally, they added that the term population group should not be confused with citizenship or nationality. Their race/population group question included the following detailed categories, with allowance to specify more than one: White, Chinese, South Asian (e.g., East Indian, Pakistani, Punjabi, Sri Lankan), Black (e.g., African, Haitian, Jamaican, Somali), Arab/West Asian (e.g., Armenian, Egyptian, Iranian, Lebanese, Moroccan), Filipino, South East Asian (e.g., Cambodian, Indonesian, Laotian, Vietnamese), Latin American, Japanese, Korean, Other—Specify_____. (1996 Canadian census form and instructions.) The results of these surveys suggest considerable change from country to country. Consequently, “persons migrating from one country to another are likely to encounter an official schema for classifying origin, race, or ethnicity, which is quite foreign to them.” (Bates et al 1995:435) Indeed, Duany’s 1997 study comparing Dominicans that migrated to Puerto Rico with those that migrated to New York City presents evidence for this.

THE IDEA OF RACE

39

THE SHIFTING LITERATURE ON MIXTURE IN THE UNITED STATES The literature on mixture in the United States also is changing. In particular, the extent to which the mixture of races has been acknowledged has changed, and the way in which it is discussed is different as well. Unfortunately, in the United States in the past, “mixture” tended to be either ignored or demeaned (Root 1992a, 1995). Early studies of communities of mixed groups tended to portray them as unfortunate or “pathetic folk of mixed ancestry who never know quite where they belong . . . neither fish nor fowl” (Berry 1963:vii). The term used to refer to communities in which two or more “races” had mixed was triracial isolates, a term that conveyed the marginality of such groups. These communities, which were referred to by names such as Melungeons, Brass Ankles, Croatan, or Red Bones, were described as existing in geographic isolation, having marginal status, and occurring in particular geographic situations, for example, in “forbidding swamps or inaccessible and barren mountain country.” There was general agreement in this literature that the members of these communities were reluctant to be identified as black and that they had relatively high growth rates (Berry 1963:32; Thornton 1987:210 ff). On the level of the individual, the most prevalent image was that of the tragic mulatto, the result of a slave woman raped by her master (Orbe and Strother 1996). Although later research has begun to explore instances of white women and black men who had children together (Hodes 1997), until recently the literature generally conceived of multirace persons as marginal, with “neither/nor status, cultural maladjustment, limited social assimilation, incomplete biological amalgamation, and pathological personalities” that were often the outcome of labyrinthine relationships between marginality and colonialism (Williams 1992:281). Mixed-race persons were thus viewed and treated as by-products of exploitative sexual unions between colonialists and members of indigenous or colonized groups (Williams 1992:281).16 Earlier studies of intermarriage in the United States also found a higher proportion of foreign whites and marginal whites in mixed marriages (Williamson 1984:112). But particularly during the country’s early formation, there must have been consensual unions in which women of color exercised some power or that involved white women and men of other races.

40

THE IDEA OF RACE

In recent years, a new paradigm has developed in which the earlier literature is challenged and new perspectives are emphasized (Root 1992b, 1995). For example, triracial isolates have been relabeled as pluralists, runaways, and refuseniks (Daniel 1992). Daniel maintains that “despite their patent Eurocentrism, these strategies . . . may be legitimately viewed as diverse tactics of resistance to oppression utilized by individuals of African descent. While some individuals may seek to confront oppression head-on, passers and pluralists seek to turn oppression on its head by subverting the racial divide” (1992:70). He further argues that other ways of subverting the racial divide have included “passing,” blue vein (i.e., when the skin color is light enough to reveal blue veins) societies, and the development of elite creole groups. In addition, the recent literature stresses “the complex realities” of multiracial people, the “multiple sensibilities” that come from their having both insider and outsider perspectives, and their highly developed capabilities to adapt “to various environments, different cultural settings, and paradoxical situations” (Williams 1992:283). The new literature has also examined areas formerly neglected, signaling a new perspective on mixture. For example, scholars have studied settings in which intermarriage is common, for example, Hawaii (Grant and Ogawa 1993; Labov and Jacobs 1986); the significance of color differences in the African American community (Russell, Wilson, and Hall 1992); the merging of diverse racial-ethnic groups under panethnic categories (Lopez and Espiritu 1993); children of mixed parentage (Cauce et al. 1992; Field 1996; Jacobs 1992; Johnson, R. 1992; Root 1992c; Taylor-Gibbs and Hines 1992; Tizard and Phoenix 1993); philosophical dimensions of mixture (Zack 1995); critiques (Jacobs 1992; Johnson, D. 1992; Miller 1992; Nakashim 1992; Root 1992a; Valverde 1992); explorations of the applicability of conventional psychological theories to mixed-race families (Stephan 1992); and the fluidity and shifting contexts of racial constructions (Gregory and Sanjek 1994). Contributing to the changing paradigm reflected in this literature is the increase in the numbers and types of individuals in interracial families. In the past, it was lower-class whites who intermarried, but today, it is upper-class whites (Spickard 1989) and higher-status blacks (Kalmijn 1993) who are intermarrying.17 In addition, in the past, mixedrace children were born to black women and white men, whereas today, most are born to white women and black men (Williamson 1984:112). Recently Kalmijn (1993), analyzing 1970–1980 marriage license data in

THE IDEA OF RACE

41

thirty-three states, found that the incidence of intermarriage has risen since the 1960s, especially between high-status black males and lowerstatus white females. Kalmijn’s research (1993) reinforces the link earlier found between status and race in intermarriage (Davis 1941; Heer 1966; Merton 1941; van den Berghe 1960). In other words, those minorities with a higher socioeconomic status marry members of the majority with a lower socioeconomic status, thereby exchanging one person’s “racial caste prestige” for the other’s socioeconomic prestige (Kalmijn 1993:122–123). Spickard (1989:349) noted that leftist intellectuals, college professors, artists, entertainers, and people with elite international careers are those most likely to intermarry. Rolark, Bennett, and Harrison’s analysis (1994) of multiracial responses on the 1990 census found that the number of interracial marriages rose from 0.4 percent in 1960 to 2.2 percent in 1991. Two-thirds of these marriages were between black males and white females.18

CHANGING CLASSIFICATIONS IN THE UNITED STATES In the United States, racial classifications have also been more variable and fluid than generally acknowledged. Definitions of race vary by state and sometimes are regionally based, often changing over time and in response to political and legal events (Bell 1973; Davis 1992; Domínguez 1986; Haney López 1996; Schafer 1993; Williamson 1984). Davis offered an early example of these changes: in Virginia, the definition of who was “white” became more and more restrictive in order to limit intermarriage. That is, in the 1800s, a “white” person was anyone who was less than one-quarter Negro; these people could marry other whites. By 1924, however, legislators prohibited anyone with “a single drop of Negro blood” from marrying a white person (Cose 1995:70). Although there is little awareness in everyday speech of the lack of uniformity or cohesion that has existed or exists in state-defined race in the United States, racial criteria are not as clear-cut or unchanging as many believe. The U.S. government and, more specifically, the Census Bureau, which is responsible for counting people by race, do not have a single criterion or principle to determine different races. Rather, they currently use several, for example, national origin, tribal affiliation and membership, and physical characteristics (McKenney and Bennett 1994:16). Moreover, these criteria are not applied in the same way to all

42

THE IDEA OF RACE

groups (Hahn 1992). For instance, whereas tribal affiliation is critical to identifying Native Americans, it is not used at all to identify whites or blacks. In addition, the format and terminology that the U.S. census has used in the question on race have changed over time (Lee 1993). In some censuses, groups have been assigned racial categories. For example, mulattoes were a racial category in the 1850, 1860, 1870, 1890, 1910, and 1920 censuses. “Mexicans” were a “race” in 1930 but not before or after then. Finally, before 1980, “race” was generally determined and/or reported by the census interviewer. Since 1980, however, people have chosen their own race from a list of categories. Moreover, the way that people see themselves and the way that the census takers or others record their race are not always the same. Although the public has adhered to a rather rigid belief in race as a biological fact, some census officials have come to believe that “race” is more social than biological. Accordingly, as early as the 1950 census, “race” was explained as follows: The concept of race, as it has been used by the Bureau of the Census, is derived from that which is commonly accepted by the general public. It does not, therefore, reflect clear-cut definitions of biological stock, and several categories obviously refer to nationalities. . . . Although it lacks scientific precision, it is doubtful whether efforts toward a more scientifically acceptable definition would be appreciably productive, given the conditions under which census enumerations are carried out. (U.S. Bureau of the Census 1953:35, italics added)

Here, the census is acknowledging that “race” is a social construction and not a scientific criterion, as the courts had earlier concluded. In addition, the census now admitted the significance of geographic “context” or “social setting” in making racial determinations and conceded that it was difficult to classify individuals by race without sufficient numbers of their group in the area where they were being interviewed.19

THE CONFLATION OF RACE AND ETHNICITY More recently, the U.S. census has become aware of the overlap of race and ethnicity, as reflected in a recent census report noting that this issue

THE IDEA OF RACE

43

should be examined further for the 2000 census (McKenney and Bennett 1994:23–24). This is a significant departure from the census’s past position, treating “race and ethnicity as two separate concepts” (McKenney and Bennett 1994:16; U.S. House Committee 1994f). The idea that “race” and “ethnicity” overlap is not new, but demographic changes (e.g., increasing numbers of Latinos, Asians, and mixed-race individuals) and greater “ethnic” self-identification have brought this idea to the forefront. In particular, it is the experience of Latinos in the United States that most clearly illustrates the interrelatedness of “race” and ethnicity. Even dictionary definitions of race reveal the overlap between race and ethnicity. Moreover, according to Webster’s Collegiate Dictionary, this overlap has existed for some time. In all the editions between 1898 and 1994, most definitions of “ethnic” refer to races, and some definitions of “race” sound like ethnic definitions (Rodríguez and López-Hernández 1995). These definitions also have shifted over time, from a singular and narrow biological definition of race to the inclusion of more cultural and social definitions (see table 2.1). For example, in the first edition (1898), race is biologically defined as meaning “the descendants of a common ancestor; lineage; breed; stock” (Rodríguez and López-Hernández, 1995; Webster’s Collegiate Dictionary 1898:660). In 1936, “the race of doctors” is used as one example of “race” (Webster’s Collegiate Dictionary 1936). In 1973, “the English” is cited as an instance of “race” used as “a class or kind of people unified by community of interests, habits or characteristics.” This example is repeated in the 1983 edition and appears most recently in the 1994 edition (Webster’s Collegiate Dictionary 1973:950; 1983:969; 1994:961). In sum, some definitions of “race” begin to sound like definitions of “ethnicity” or culturally distinguishable classes. At present, primary definitions of race still refer to “a breeding stock of animal,” and secondary definitions refer to “a class or kind of individuals with common characteristics, interests or habits . . . or a kind of people unified by community of interests, habits, or characteristics.” In essence, the biologically based definition of “race” as a “breeding stock” has been retained, but additional definitions have been added to include other, more cultural or social definitions. The inclusion of these broader, somewhat overlapping definitions of race and ethnicity in a commonly used dictionary is somewhat surprising, particularly because ethnicity in the United States (and in the U.S. census) has generally been considered separate from race.

44

THE IDEA OF RACE

RACE AS RACE OR AS ETHNICITY Today, most scholars agree that “race” is determined by context, just as ethnicity is.20 Indeed, some scholars believe that race and ethnic group are the same. Spickard (1992:23), for example, argues that race and ethnic group are the same, because both are defined on the basis of social, not biological, criteria. Both race and ethnic group claim descent from a common set of ancestors, have a common sense of identity, share the same culture from clothing to music to food to language to child-rearing practices, build similar institutions like churches and fraternal organizations, and pursue common political and economic interests. Other scholars, though, insist that race or legal minority status is quite different from ethnicity (Cox 1948:317–320; 392–401; Mullings

Table 2.1 Dictionary Definitions of “Race” and “Ethnic,” 1898–1994 Year of new edition

Race

Ethnic

1898, 1st ed.

The descendants of a common ancestor; lineage; breed; stock.

1. Belonging to races or nations; based on distinctions of race; ethnological.a

1910, 2d ed.

No change.

No change.

1916, 3d ed.

The descendants of the same ancestor; a family, tribe, people, or nation taken as of the same stock; a lineage; breed; also, a class of individuals with common characteristics, interests, or the like.

1. Heathen; pagan. 2. Pertaining or peculiar to race; pertaining to groups of mankind discriminated by common customs and character.b

1931, 4th ed.

No change.

No change.

1936

The descendants of a common ancestor; a family, tribe, people, or nation, believed to belong to the same stock; a lineage; a breed; also a class or kind of individuals with common characteristics, habits, or the like.

1. Neither Jewish nor Christian; pagan. 2. Of, pertaining to, or designating races or groups of races discriminated on the basis of common traits, customs, etc.

1949, 5th ed.

No change.

No change.

1963, 6th ed.

1. A breeding stock of animal. 2a. A family, tribe, people, or nation belonging to the same stock. b. A class or kind of individuals with common characteristics, interests, or habits.

1. Neither Jewish nor Christian; heathen. 2. Of or relating to races or large groups of people classed according to common traits and customs.

THE IDEA OF RACE

45

Table 2.1 (continued) Year of new edition

Race

Ethnic

1973, 7th ed.

1. A breeding stock of animal. 2a. A family, tribe, people, or nation belonging to the same stock. b. A class or kind of individuals with common characteristics, interests, or habits. 2b. A class or kind of people unified by community of interests, habits, or characteristics. The English.

1. Neither Jewish nor Christian; heathen. 2. Of or relating to races or large groups of people classed according to common traits and customs. Minorities. 3. A member of an ethnic group; esp. a member of a minority group who retains the customs, language, or social views of his group.

1983, 8th ed.

No change.

1. Ethnic: a member of an ethnic group; esp. a member of a minority group who retains the customs, language, or social views of his group. 2. Ethnic: 1. Heathen. 2a. Of or relating to large groups of people classed according to common racial, national, tribal, religious, linguistic, or cultural origin or background. b. Being a member of an ethnic group. c. Of, relating to, or characteristic of ethnics.

1993, 9th ed.

No change.

1. Heathen. 2a. Of or relating to large groups of people classed according to common racial, national, tribal, religious, linguistic, or cultural origin or background. b. Being a member of an ethnic group. c. Of, relating to, or characteristic of ethnics.

a b

The definition given is for ethnical. The word ethnic does not appear until the 1936 edition. The word ethnical is no longer listed.

Source: Webster’s Collegiate Dictionary (Springfield, MA: Merriam-Webster, 1898, 1910, 1916, 1931, 1936, 1959, 1963, 1973, 1990, 1994).

1978; Ogbu 1978; Sanjek 1994:8 ff; Steinberg 1981). They do not accept that ethnicity can be substituted for race because the concept of ethnicity does not convey or imply the context of discrimination associated with race in the United States. Also, ethnicity is a term that historically has been used to refer mainly to people of European origin and not those of African origin. Conversely, those who maintain that race is a social fabrication of little scientific or practical value contend that racial categories only reinforce our beliefs in this falsehood and that ethnicity should be used only, for example, to classify people (Patterson 1997).

46

THE IDEA OF RACE

This conflation of and/or confusion between race and ethnicity has been at the root of some of the positions taken throughout the United States’ history. For example, in the debate between Takaki (1994) and Schlesinger (1992), the issue was presented as follows: In the history of the United States, were diverse ethnic (meaning race and ethnic) groups assimilated into one melting pot? Or is it a segmented, racialized history comprising an official, articulated white history and a neglected, as yet largely unwritten, history (or histories) of not-white groups—with two melting pots? At present, the issue of whether race and ethnicity are independent or overlap (and to what degree) has still not been resolved. Lee (1993) noted in her review of census categories over time that historically, race and ethnicity have been confused. Indeed, race and ethnicity are often discussed as if they were separate, somewhat independent concepts, which seems still to be the census’s official position. Hispanics, for example, are regarded as members of an ethnic group that can be of any race. But the recent proposal to make Hispanics a race suggests pressure to view Hispanics as a race and thereby to fold them into the United States’ racial structure. The common juxtaposition of Hispanics with groups such as whites, blacks, and Asian and Pacific Islanders reinforces this intention.

3 Stories of Self-Definition

T H E R E A S O N T H AT both Latinos and non-Latinos choose particular

categories on the census has only recently received much research attention (see, e.g., Elias-Olivares and Farr 1991; Kissam, Herrera, and Nakamoto 1993; Martin, DeMaio, and Campanelli 1990; McKay and de la Puente 1995; Rodríguez and Cordero-Guzmán 1992; Tucker et al. 1996:22–28). Even if we assume that Latinos who pick the traditional U.S. race categories do so for the same reasons that non-Latinos do, this still does not explain those who do not identify with “any of the above.” The following case studies look at why some people choose the “other race” category and how they decide on their particular “identities.” Although the case studies are not meant to be representative, they are not unusual. These personal accounts are from a sample of sixty Latinos living mainly in the Northeast (see Rodríguez et al. 1991 for a more detailed discussion of this research project). The respondents were selected by snowball sampling, and the researchers chose a spread of class levels and national origins. This method ensured finding a diverse group of Latinos willing to be interviewed at length on an issue that is, for many, sensitive. As might be expected, given the northeast slant of the sample, 61 percent consisted of Latinos of Caribbean origin, mainly from the Dominican Republic (33%) and Puerto Rico (28%). Five percent were “mixed,” that is, had parents from different Spanishspeaking countries, and 1 percent had one non-Latino parent. The remaining 34 percent came from Central and South America. Of the thirty-six women and twenty-four men, 8 percent described their backgrounds as upper or middle class, 45 percent as working class, and 15 percent as lower class. The remaining 32 percent did not indicate their class background. One-quarter of the sample was raised in Spanishspeaking countries, but only 15 percent of the interviews were conducted in Spanish. All but two of the fifteen trained interviewers were bilingual and generally were members of the groups they interviewed.1 47

48

STORIES OF SELF-DEFINITION

Most of the pre- and posttest interviews took place between 1989 and 1990, but a few were conducted later. The respondents were not paid and were interviewed in a variety of home, school, and office settings, where privacy, adequate time, and relaxed comfort had been arranged before the interview. The research team and I created an extensive, detailed, bilingual questionnaire and tested it for a year and a half. It contained 107 openended and structured questions and covered a wide variety of areas. The interviewers were instructed to determine the respondent’s phenotype—as white, black, or “other”—before beginning the interview. The respondents were asked first to fill out a duplicate of the 1980 census question on race, and then they were asked to explain why they had answered the question as they did. The next questions were (in order of appearance): How would you describe yourself racially? What do you consider yourself to be: white, black, or other? What color are you: white, black, or other? How do you think North Americans see you: white, black, or other? How would you describe yourself over the telephone to a person who has never met you but who has arranged to meet you in a crowded place? How would you classify, in terms of color and features, the other people in your family on this (five-point) scale? Do you think your identity has changed over time? The case studies presented here are of persons (not using their real names) who identified themselves as “other race.” During the interview, the first four described here appeared to be at ease with their “racial” identities. All four checked the “other race” category on the census question and supplied a Latino referent. All four viewed their “race” in different terms: three according to their Latino heritage and one according to his Latino and black heritage. Based on their phenotypes, the first was categorized as “white,” the second and third as “black,” and the fourth as “in-between.” In fact, this fourth person explained that depending on the eye of the beholder, he can be white, mixed, black, or “Hispanic.”

JOSÉ PETERSON OR JP: THE HYPHENATED AMERICAN This respondent was named José Peterson because this composite name has both Anglo and Hispanic elements. We chose the nickname JP, with its business-tycoon connotation, because at the time of the interview, he

STORIES OF SELF-DEFINITION

49

perceived himself to be (and was understood to be) assimilating into white corporate America. He is single, twenty-five, and a college graduate working as an administrator in the arts. JP’s parents migrated from Puerto Rico to New York when they were in their early twenties and settled in a section of the city with a high incidence of violent crime. JP lived here for the first eight years of his life until his parents moved to a more stable working-class area where he continues to live. His family speaks both Spanish and English at home, but JP says his Spanish is not good. JP indicated that his background was working class because his father did manual labor. He is the fourth child in a family of five siblings and the first to go to college. On the census race question, JP checked “other” and specified “Puerto Rican American.” He explained that he attributed his Puerto Rican heritage to his parents but that he identified as American because he was born in the United States. He added that he was bicultural because “various aspects of both the American and Puerto Rican cultures” influenced him. But when asked questions that he interpreted as referring to his physical appearance, JP answered consistently and unequivocally that he was “white.” For example, when asked how he would racially identify himself—as white, black, or other—he answered white because of his European (Spanish) background. He also referred to being “white” when asked about his color, how North Americans viewed him, and how he would describe himself over the phone. He classified everyone in his family as white except for two grandmothers who, he explained, had “Indian blood.” Finally, JP was classified by the Puerto Rican interviewer as “white.” When he understood the questions to be asking about his physical appearance, JP consistently answered that he was white. But he did not feel it necessary to explain why he did not then select the white race category. When he understood the questions to be asking about his cultural identity, however, he said he was a “hyphenated American.” In effect, when JP answered the census question on race, he assumed that the categories represented other major social-cultural-racial-political groups in the United States, and he supplied his own (hyphenated) group. Although he clearly saw himself as physically white, he identified himself as “other (Puerto Rican–American) race.” Even though JP identified as “other race,” his adaptation to the U.S. racial system followed the familiar immigrant assimilation model. He

50

STORIES OF SELF-DEFINITION

answered the questions in much the same way that second-generation European Americans usually answered them. For example, Greek Americans or Italian Americans would see themselves as culturally the product of both the old country and the United States, as JP did. But earlier immigrants probably would not have mentioned that their grandmother had Indian blood.

CELIA: LATINA, BLACK, AND PROUD . . . AND NOT AFRICAN AMERICAN Celia was named after the salsa singer Celia Cruz because of the many similarities she appeared to share with her. Celia identified herself as “other race,” and on the census question, she wrote in “black Hispanic Panamanian.” Celia came to the United States from Panama when she was eighteen years old. She has four children, all over the age of seventeen, and has been married to her husband for more than twenty-five years. She rose through the ranks to become an account coordinator at the same place where she has worked for more than twenty years. Celia lived through the racial insensitivity of the 1950s, the racial awakening of the 1960s, and the renewed racial hatreds of the 1980s. Yet she says that throughout the constant racial turbulence, she always knew “who she was.” She emphasized that she is both Hispanic and black and has strong roots in both identities. Celia often mentioned her love of Hispanic culture and her pride in being born and raised as a Panamanian. Her feelings about the “black” portion of her self-identification are equally strong, and on the family chart, she labeled all the members of her family as black. Her tone was unwavering on this point; her identity has never changed. Celia realized that most North Americans saw her to be like “any other black” but noted that she felt uneasy with American-born blacks. She sensed that they strongly disliked blacks from other countries. To illustrate, Celia described an experience she had had at a playground with two of her children when they were young. She said that when she began talking and a black American woman there heard her accent, the woman verbally abused her, and so Celia left the playground. Celia is a black Hispanic Panamanian and proud of it. But she is not a black American, and she does not see herself as black according to U.S. defi-

STORIES OF SELF-DEFINITION

51

nitions of blackness. So she did not check off the “black” category on the census question.

FAT JOE (LATINO) Fat Joe (Latino) was given this name because of his two passions, his love for his Puerto Rican culture and his love of rap music—he had been both a disk jockey and a rap artist. Born and raised in the suburbs of Washington, D.C., Fat Joe was twenty-seven, recently married to an African American model, temporarily working as a cab driver in Atlanta, and about to complete a master’s degree in the social sciences when he was interviewed in New York City. Both his parents had immigrated to Washington in the 1960s from Puerto Rico, where both had been teachers. His father had been recruited to work in a government agency. Fat Joe described his family and upbringing as middle class and recalled his growing up as being filled with “good memories” and as having friends that he had kept despite moving from one suburban neighborhood to another. Few Puerto Ricans lived in these neighborhoods; rather, they were made up mostly of other Latinos, blacks, and whites. The main language he spoke at home was Spanish, but he and his wife speak English. Over the years, he has frequently visited Puerto Rico, but he has never lived there. At the time of this interview, Fat Joe was intending to obtain a more advanced graduate degree. He chose the “other race” category on the census question and wrote in “black Puerto Rican.” When asked why he had answered in this way, he first said that was how he always answered this question (about racial classification). But he then added that if it had been an official census question, he would have answered “black.” Or if the list included “Puerto Rican,” he would have marked that. He would not have checked “Hispanic” because that category was too vague and general for him. His responses illustrate the significance of context to Latinos responding to questions of race. Fat Joe’s first response also indicates his strong identity as both black and Latino, which was repeated in subsequent questions. When asked how he would describe himself racially, he answered, “As a Puerto Rican of African descent.” When asked how he thought North Americans saw him, he said that everybody saw him as a black

52

STORIES OF SELF-DEFINITION

American until they talked to him and found out about his background. Presented with three categories, white, black, and other, he chose “black” but added that he was of a “brown” color. To him, “black” meant that it racially described his body as being of African descent. He said he identified as black because his ancestors were brought over from West Africa to the Caribbean. When asked how he would describe himself to someone over the phone, he replied, “Black, 5’10”, 220 pounds, bald head and beard.” Only when asked the general nonracialized question of how he identified himself did he say, “By my name.” Fat Joe described all the members of his family, including himself, as “black” except for a maternal grandmother, whom he said was “white.” He also indicated that they all would see themselves as “black.” When asked what he felt his roots were, he answered, “Puerto Rican, I guess.” He considered himself to have been “raised as a Puerto Rican” but also stated that African Americans, white Americans, and their culture had “rubbed off me as well.” He said he could “relate” in a lot of ways, except when it came to food. The people in the southern United States could not understand why he could not relate to collard greens—because he had been raised with arroz and abichuelas (rice and beans) and pasteles. Asked to recall experiences outside his family in which people reacted to his race, Fat Joe said there were hundreds. For example, as a senior in high school, he was at the beach when a bunch of white guys sped by screaming “nigger.” Contrary to the experience of many Latinos, Fat Joe (Latino) is consistently assumed to be black and seldom anything else. Thus, only in specific contexts is he thought to be Latino. As he said, only if he is in a Latino store and reading something in Spanish will someone speak to him in Spanish. People are always surprised—and doubtful—when he tells them he is Puerto Rican, for he is generally assumed to be any other kind of Caribbean but Puerto Rican. Although this used to bother him, he says that now that he has studied the history of these areas, he understands why people react this way, that there are “phenotypically more African-looking people in the Dominican Republic and Cuba than in Puerto Rico.” Fat Joe indicated that he has probably assumed a black identity on occasion, noting that “if an African American refers to me as a ‘brother,’ I acknowledge the background.” However, he has known others who

STORIES OF SELF-DEFINITION

53

were darker than he who would have taken offense at the reference. His assumption of this identity did not have an emotional impact on him— it was “just what I am.” Fat Joe’s identity has changed over time. He has no specific memory of when he first became aware of his color or of his being Puerto Rican, but he assumes that he must have become aware of color when he first started to play with other children, when he was three or four, and of his Puerto Rican ethnicity when he brought friends home. He described the two neighborhoods where he had been raised as having few Puerto Ricans, but he also noted that his family’s circle of friends included many Puerto Ricans. He admitted that he had changed the way he viewed his color, facial features, and hair texture, explaining that when he was younger, he was sensitive about his more African phenotype, even though his hair was less kinky than that of his friends. Colorism and “joking” about “too kinky” hair seemed to be part of growing up among his friends. In fact, he joked that now he would love to have kinky hair—any kind of hair. Fat Joe reached an important turning point in the formation of his identity when he started applying to college. His parents sat him down and told him that he should take advantage of the fact he was phenotypically black and Puerto Rican. So he added his mother’s last name, a Spanish surname. He attended a large public university and believes that he was counted as Hispanic and not as black, partly because he was the only one of his friends who received handouts on Hispanics. But in high school and college, he did not spend much time on Puerto Rican culture; rather, he “delved into the black side of me.” He studied black literature and participated in African American activities. Midway through college, he decided to major in African American studies and began to pay more attention to his Latino side. Then, when he had to pick a topic for his senior thesis, he chose a connection to his Puerto Rican culture. This focus in African American studies was distinctive, and he thought it would be helpful. Fat Joe (Latino) seems to have had a variety of experiences, and when he was interviewed, he seemed comfortable with his identity as a black Latino and interested in celebrating, and knowing more about, both these heritages. His feelings were echoed by the Costa Rican–born Delina D. Pryce, who pointed out, “Being Latina and black are not mutually exclusive, but mutually complementary. Being black and Latina

54

STORIES OF SELF-DEFINITION

has influenced and shaped my views, my thoughts, my experiences— who I am. Never would I deny either because they’re both me. And I like me. All of me.”

MR. ARCO IRIS’S RAINBOW IDENTITIES Mr. Arco Iris is yet another representative of the “other race” category. His name means “rainbow” in Spanish, and he is always addressed as “Mr.” because at sixty-two, he has a respected and established position as a professional in the criminal justice system. Born and raised in East Harlem and the South Bronx (predominantly black and Hispanic neighborhoods), he is the son of parents who migrated from Puerto Rico to New York before World War II. He described the household in which he was raised as Spanish speaking and lower middle class. He considers his roots to be in Harlem, for he has lived in Puerto Rico only a short time. Although Mr. Arco Iris is fluent in Spanish, he is more comfortable speaking English. His wife is West Indian, and they have three children. In response to the census race question, Mr. Arco Iris checked “other” and wrote “Puerto Rican” in the space next to it. But when answering “How would you describe yourself racially” and “What do you consider yourself to be,” he stated, in both instances, “I am a mixture of black, white, and possibly Indian.” He noted that his racial identity had changed over time. “As a child, I perceived myself as a Puerto Rican and distinctly apart from black and white. But as I grew, I understood Puerto Rican as a mixture, and I could identify with both blacks and whites.” The way he viewed his ancestry also has changed: “I would have considered myself more white up to the age of nine. As I got older, I developed a broader definition of race and acknowledged greater mixture.” Mr. Arco Iris described his color as “brown” and explained that North Americans tend to see him as a “brown-skinned Puerto Rican or a light-skinned black.” His interviewer described him as “not white/not black.” On a five-point color scale, Mr. Arco Iris labeled his mother as a one (light) and his father as a five (dark), and he identified himself as a four. This was darker than the interviewer’s view of Mr. Arco Iris, as a three (intermediate in color). When asked why he characterized himself as darker than North Americans might see him, Mr. Arco Iris stated that “four is more biologically accurate” and further ex-

STORIES OF SELF-DEFINITION

55

plained that he identified himself as dark out of respect for and loyalty to his brown-skinned father. Mr. Arco Iris and others like him identify strongly with color and express a preference for racial diversity and mixture. They take pride in themselves as combinations of African, European, and Native American.2 The identification of Mr. Arco Iris’s race also varies according to the eye of the beholder. He noted that since childhood he has been regarded as white, black, Greek, Arab, and Asian. These many instances of mistaken identity have prompted him, perhaps more than many others, to think about his identity. His racial identification mirrors this selfreflection, representing a unique innovation and resolution within an essentially biracial system. These four persons, with different backgrounds, are phenotypically white, black, and in-between. They emphasize not just physical but also cultural variables. Furthermore, they are at ease with how they have resolved their identities, even though some data gatherers thought their responses showed that they misunderstood or were confused by the question. At this point in their lives, they seemed confident and indicated that it was all right to be “other,” although they did not mention this specifically. The following two stories reflect the more conflictual and stressful dimensions of Latino identity resolution.

JOSÉ ALI: THE PRESSURE TO BE BLACK The name José Ali is a combination of the common Latino name José with Ali, borrowed from Muhammad Ali, the former heavyweight champion. José Ali is a Dominican, twenty-four years old, single, and a full-time student at a public university. He has a part-time job in an advertising firm where the majority of his coworkers are white. His parents are immigrants from the Dominican Republic. José Ali was raised in New York and has visited the Dominican Republic only once, when he was five. He lived in a predominantly Hispanic neighborhood until he was eight years old, and Spanish was the only language spoken at home until he was twelve. He later moved to another area of New York with a large African American population. He describes his family as working class: his father worked in the metal goods industry, and his mother was an office worker. He does not have a Hispanic surname.

56

STORIES OF SELF-DEFINITION

José Ali answered “other, Hispanic” to the census race question and explained: “By inheritance I am Hispanic. However, I identify more with blacks because to white America, if you are my color, you are a nigger. I can’t change my color, and I do not wish to.” He consistently alluded to his identification as black when answering other racial items in the interview, for example, “Hispanic, yet identifies as black” and “I describe myself as black.” When asked what the word black meant to him, he replied, “As other people see me.” Finally, when asked, “Why do you see yourself as black?” his answer was, “Because when I was jumped by whites, I was not called a ‘spic,’ but I was called a ‘nigger.’” During the interview, José Ali noted that he assumes everybody at his job believes he is black and he does not “want to burst their bubble.” He said that he goes along with their assumption as long as he is treated well but admitted that he accepts this identity because it would take him too much time to explain why he is culturally not an African American. He pointed out that “when you are seen as a certain race, you are also seen culturally the same.” But when people assume that he is an African American, they are “disregarding my own feelings. They don’t ask, they simply assume.” (The Dominican interviewer described him as “a stereotypically dark-skinned Latino or a light-skinned AfroAmerican.”) Asked if his identity had changed over time, José Ali answered yes. “I realized that although I feel Hispanic, I was not seen as Hispanic or Latino, but as black. Now, I agree with whoever thinks I’m black. There is no point in trying to prove that I’m not black . . . after being practically attacked by whites because of the way I look. I decided to accept the fact that no matter who I feel to be, I am categorized as black.” Thus, even though José Ali says he is “other race, Hispanic,” his responses reveal the pressures that some Latinos feel to identify as an American black. This conflict was first described in the literature on the Puerto Rican migration (Colon 1982; Iglesias 1980).3 (It was perhaps best portrayed in Piri Thomas’s 1967 Down These Mean Streets, and it has been discussed most recently by Brady 1988, Santiago 1995, and Comas-Díaz 1996. For recent discussions of race and gender in Puerto Rico, see Barbosa 1991, Ramos Rosado 1987, and Valcarcel 1994.) This imposition of the black-white racial order on Latinos separates them into “whites” and “blacks” and in the process attempts to create new African Americans and so-called hyphenated (European) Americans. Latinos understand this phenomenon as their being iden-

STORIES OF SELF-DEFINITION

57

tified racially but not culturally. Other Latinos in the sample felt similarly confused or pressured to be “white.” Consequently, today Latinos are pressured to be categorized according to their color rather than their national heritage and culture.4

VICTORIA: A CELEBRATION OF COLOR The next case illustrates the tensions inherent in assimilation through education. In contrast to José Ali, Victoria first strongly assimilated into “whiteness.” Then after a period of conflict and struggle, she acknowledged her resentment of this assimilation and began to celebrate her color. She was named Victoria because she seemed to have been victorious in overcoming the obstacles that caused her pain and confusion. Victoria is a single, thirty-year-old Chicana graduate student who was born and raised in a small town on the U.S.-Mexican border. Almost all the town’s residents are Mexican and work in the fields, although her parents do not. Her father has a working-class occupation, and her mother is a homemaker. Victoria has several sisters, and her family is Protestant. She has been to Mexico only once, when she was twentythree, and she described this trip as consciousness raising. During her interview, Victoria consistently placed herself in an intermediate position, choosing “other” on the census question and specifying “Hispanic.” She gave her color as “brown” and said that North Americans saw her as “other,” not “white” or “black.” Even when asked how she would describe herself over the phone to someone she had never met (but would meet), she said she always was careful to note her tan coloring. Even though the interviewer thought that Victoria could be regarded as white with a summer tan, she consistently identified herself as “not white”; indeed, a nonwhite color seemed to be an important part of her racial identity. (But nonwhite was apparently not the same to her as black.) Victoria saw herself as Hispanic because she was not white and not black and because historically she (and her group) had had a different relationship to those two groups. When Victoria finished elementary school, she went to a junior high school where she was placed on the accelerated track. Here most of her classmates were Anglos. She describes this period as when she went from being a “smart Chicana” to being a “smart white.” Most of her friends were white, and her sisters would make fun of and mimic her

58

STORIES OF SELF-DEFINITION

“whiteness.” Victoria remembered that a group of mejicanos (Mexicans) once showed their disapproval of her hanging around with whites by calling her a Tía Taca (the equivalent of Uncle Tom). At the white parties, Victoria said she knew that she was not a beauty because of her skin color, so she compensated by developing a good “personality.” Her awareness of being different (and less attractive or acceptable) because of her skin color was so acute that when she became part of a group traditionally made up of “pretty girls,” she assumed she was included because she was “nice.” She also remembered that she always wanted to be a cheerleader, but somehow this was not what Mexican girls did. When Victoria went to the local community college, she continued to excel academically and was very active in student government. She also recalled the following experience that subsequently made her feel very ashamed: One day the dean patted her on the shoulder and told her, “I’m so glad you’re not like the other Mexicans,” considering this a compliment. Asked how she felt about the remark, Victoria said that it made her uncomfortable but remembered that she had looked up and smiled. Until she went to another college, in California, Victoria did not realize the significance of the dean’s remark. When she did, she first reacted with fury at having denied her heritage and having accepted the implication that her accomplishments were an exception to the rule. She later also resented what she perceived as the limitations of Mexican culture. As she explained, she traveled a long road in a short time, from being identified as white to being proud of being Mexican to being angry at Mexican patriarchy. In essence, Victoria saw her education as a vehicle that helped her escape certain sexual and racial boundaries, but she also felt that while doing so, she had had experiences that damaged her self-image, such as when she was treated as a credit to her race.5 Clearly, family dynamics and other antecedent factors influence how people decide on a racial identity. Victoria is an interesting example of these dynamics. She described her family as having considerable physical variation and herself as the darkest one. Accordingly, she rated herself a four on the scale while giving everyone else in her family an average of 2.5. Although the interviews were not calculated to elicit deep psychological motivations, it seems that Victoria’s selfdescribed position as “the darkest in her family” may have influenced her drive to be high achieving and her desire to be “white.” She said during her interview that she did not remember openly saying

STORIES OF SELF-DEFINITION

59

that she wanted to be “white,” that at the time she felt she was just following her intellectual interests. On reflection, however, she said that she wanted to be white culturally and that her desire to be “white” was subconscious. Excelling in school and being accepted by whites may have added to the low value she felt as a dark woman. Whether her educational accomplishment was recognized by others in her life (Latinos and nonLatinos) is unclear, because their responses were filtered through Victoria’s eyes. But she appears to have been praised by white officials and rejected by her more “Mexican” family and community. It also is likely that some whites resisted or resented her efforts and that some Mexicanos were proud of her achievements. What is perhaps most important here is that Victoria’s drive to achieve may have stemmed from her perception that she was not highly valued because she was dark. In effect, her academic achievement was compensation. Whether it was also a cry for greater acceptance (by whites and perhaps other Latinos) we do not know. Another less openly acknowledged and perhaps more positive side effect of being the darkest in the family surfaced in Victoria’s account. She described in detail the treatment that Blanca, her “light-skinned, green-eyed sister,” received as la favorita (the favored one). Victoria referred to the “privileges of color” that Blanca enjoyed and how this contrasted with her own treatment and that of her other sisters. For example, in the division of household chores, she and her sisters always were given the harder, less desirable work, for example, carrying out the trash. She also noted that even today, at family gatherings, the dark sisters brought the food that took hours to prepare, and la favorita brought the paper goods. Victoria also noted the impact of this differential treatment on the development of her sisters. She believed that because Blanca was so “privileged,” that is, because she was protected and treated as fragile and delicada (delicate), she never developed the independence and strength that her other, darker, sisters did. In effect, Victoria’s darker sisters also compensated for their skin color by developing other areas of their personalities and lives. According to Victoria, they knew they would always have to struggle in life, which helped them deal with adversity. Blanca, however, never ventured beyond her traditional suburban existence, did little that was innovative or challenging, and did not develop the strong character of the others. According to Victoria, she

60

STORIES OF SELF-DEFINITION

was psychologically and constitutionally weaker than the others, who were active in community and activist movements. Victoria did not have strong negative feelings about her differential treatment. Instead, they seemed to be a dull dislike of the attitudes implicit in this behavior, that is, the depreciation of color and the fascination with or glorification of European physical types. She did not appear angry at Blanca. When asked about this, she explained that when she was growing up, family love overrode these dynamics and that other factors also influenced her treatment, for example, age, position in the family, and sexuality. Consequently, despite the depreciation of dark color that Victoria perceived in her family, it did not divide the family, nor were the darker members excluded. Rather, color appears to have been the basis of an implicit hierarchy that challenged those who were darker to compensate in various ways. Victoria’s maturation involved viewing differently both her cultural identity and her feelings about her color. She came to appreciate the beauty of her darker color, and she credited Chicano men with helping her appreciate and celebrate her dark color.

MARIO: SELF-IDENTIFIED AND IMPUTED RACE Mario represents another, quite different, example of the influence of family dynamics. Mario is a common name used by both Italians and Puerto Ricans. He identified himself on the census as “other race” and wrote in “Puerto Rican.” As an infant, he was adopted and raised by a Puerto Rican couple. Biologically, however, Mario was the child of an African American mother and an Italian American father. Even though he is aware and proud of his black and European ancestry, he characterizes himself as Puerto Rican because of his immersion in the Puerto Rican culture, but others not familiar with his background might identify him as a black American. For Mario, as for many others, his racial and ethnic identity is only partially in the eye of the beholder. Rather, to him his identity is how he has lived rather than his biological ancestry. In his mind, his subjective view of his racial-ethnic identity supersedes any race classification that others may ascribe to him because of his appearance or his biological ancestry. Although in our sample, subjective and external views often played off against each other, in Mario’s case, his subjective view was

STORIES OF SELF-DEFINITION

61

determinant. In the interview, he identified himself simply and solely as Puerto Rican, whereas some of the respondents in the preceding examples were torn between self-identity and that imputed by others. These case studies are examples of the various identities of those who say they are “other race” and specify a Latino referent. Some of them easily accept being “other,”6 but others feel the pressure to be white, black, or brown and to assume the multiple identities they sometimes develop. Some persons in the sample described the resolutions to these pressures as “intense.” As these case studies showed, racial identities change over time— for example, from white to brown or from tan to white as the respondents progressed from childhood to adulthood7—and according to context—for example, in their home, their job, or their school. Forces such as socioeconomic class, phenotype, family, the United States’ racial structure, and experiences in school, jobs, and social settings also are important determinants of racial identity. Consequently, the way in which race is constructed in the family, school, or society influences the way in which Latinos identify themselves and may create multiple identities, for example, white at home and brown on the job, or vice versa. These case studies also challenge the way in which race and racial identity are generally defined in the United States, where phenotype and genotype are primary and racial identity is unchanging and unambiguous. Although all these respondents answered “other race” on the census question on race, many Latinos choose one of the standard U.S. race categories. But even in these situations, identity is complex. For example, in the larger sample, one of the respondents who identified herself as “white” on the census question described herself racially as “an American with Cuban blood” and stated that North Americans saw her as “Hispanic.” She also defined the term white as “the comparative color of my skin to other groups. It is not my background, my race, my attitude toward others or my income.” She distinguished between her white skin and her culture, explaining that this is “what I appear to be” (white) to other groups. This view might be quite different from that of non-Latinos who identified themselves as white. The larger sample contained even more dramatic examples showing that Latinos’ racial identity is complex and fluid. For example, a professional woman from Puerto Rico stated that she was “white” in

62

STORIES OF SELF-DEFINITION

her home and neighborhood but, because of her lower socioeconomic origin, was “less white” in the traditional upper-class Puerto Rican family into which she had married. In the United States, where she was labeled “Hispanic” or “Puerto Rican” because of her accent, name, and cultural style, she was “nonwhite.” The case studies also demonstrate the resistance of many Latinos to bipolar racial classifications, despite the pressures from both inside and outside their culture. Finally, the case studies demonstrate that Latinos’ racial identity is not just genetically determined but that it depends on many variables, including phenotype, social class, language, phenotypic variation within the family, and neighborhood socialization. In addition, for Latinos, “race” is individually as well as socially constructed. A story told by one of the interviewers in the research project seems a fitting end to this chapter, for it illustrates people’s creativity when defining themselves in particular political and economic contexts. The interviewer was a young Latina, of Caribbean origin, raised in a barrio in New York City during a time when the phrase “black is beautiful” was popular. She recalled that as a child, she and her three sisters, all of whom were different colors, would walk down the street, arm in arm, chanting “black is beautiful,” “white is wonderful,” and “trigueña is terrific.” She could not remember other people’s reactions or even whether they (the sisters) spent much, if any, time discussing or analyzing this. But she remembered that it happened.

PART II

HISTORICAL CONSTRUCTIONS

4 Whites and Other Social Races

A S A F O R M E R census official once pointed out, decennial censuses often reflect a country’s historical needs, and the information collected is deemed necessary for the national interest and for the needs of small geographic areas (Estrada 1993:497). These “historical needs” are seen differently by different political groups, however, and when governments try to create statistical representations of its populations, the process is predicated on political and ideological choices. Thus the resulting categories generally reflect a political consensus on who is to be counted, how, and how often (Lee 1993). These categories describe the population(s) from the perspective of those who have the power to select them, and in turn, they influence the way that populations see themselves. Over time, U.S. decennial census classifications have moved toward a more sharply defined bipolar structure. Basically, two socially constructed polarities have evolved that contain “whites” at one end and “other social races” at the other. Although each polarity has been and continues to be fluid, this dichotomy has prevailed throughout most of the census’s two-hundred-year history. At various times, different divisions are featured, for example, white, black, and mulatto or white, black, Chinese, and Indian. But even in these instances, the basic dichotomous structure of “whites and other social races” has been retained. It is this bipolar structure that groups—those not quite white or black—have contested in the past, and it is this structure that Latinos today resist (Halter 1993; Leonard 1992; Loewen 1971; Smith 1993). This chapter traces the decennial censuses’ changing classification of race. Besides some surprising changes over time, we will see the evolution of this bipolar structure. Among the surprises are that the U.S. Constitution did not refer to color or race when it set forth the criteria on which the census was to be based. Rather, the initial distinctions pertained to free or slave status and taxed Indians. 65

66

WHITES AND OTHER SOCIAL RACES

By 1790, however, when the first census was taken, the color term white was introduced, and so it was color—and not race—that became the primary term of classification. Color remained an essential category of the census for more than a century and a half and preceded race as a category by nearly one hundred years. The concepts of color and race were officially joined in the twentieth century and are the foundation of the bipolar structure that evolved.

THE EARLY CENSUSES Because the United States of America was conceived as a democratic and representative government, its people had to be counted. In addition, all the states had to agree on who was to be counted and how. The 1787 Constitution of the United States established the outline of such a count in its criteria for apportionment, an immediate outcome of which was the structure of the census with regard to race. According to article I, section 2, clause 3 of the U.S. Constitution (the apportionment rule), the population was to be counted every ten years, and this became the mandate for the decennial census. The same paragraph specifies: Representatives and direct Taxes shall be apportioned among the several states which may be included within this Union, according to their respective Numbers, which shall be determined by adding to the whole Number of free Persons, including those bound to Service for a Term of Years, and excluding Indians not taxed, three fifths of all other Persons. (cited in Anderson 1988:9)

What is interesting about this excerpt is its vagueness. With the exception of the oblique reference to Indians, “race” is not explicitly mentioned. “Free Persons” does not specify “whites,” and persons of African descent are not directly mentioned. Nonetheless, it is understood that “three fifths of all other Persons” refers to slaves, who were of African descent. That is, for apportionment purposes, persons in this category were to be counted as three-fifths of a white person. In addition, indentured servants, most of whom were from Europe, were to be counted as free persons. The

WHITES AND OTHER SOCIAL RACES

67

Constitution clearly states that untaxed Indians—most likely the majority of Indians then—were not to be counted. But the implication was that taxed Indians would be counted. These were generally Indians who lived in European settlements and were no longer affiliated with a tribe. They may also have included Indian women who had married white men. As Anderson noted, the apportionment rule incorporated into the census and the political fabric of the new nation a tradition of differentiating “these three great elements of the population”—the free, slave, and Indian populations (1988:12). The method used to determine apportionment was tantamount to deciding who was to be acknowledged, and how. These decisions reflected how various groups of people were viewed at that time, which groups were considered to be part of the constituent population, and which were not. In the first census of 1790, being indentured or being a Native American did not prevent one from being counted. As long as they paid taxes, Indians could be represented.1 Slaves, however, were not counted or represented. Thus the two main non-European components of the U.S. population were recognized in different ways. Being counted, however, was not by itself assurance of equal citizenship rights, for free white women were counted but could not vote, and free white men who did not own property could not vote. But not being counted meant that a person had no official place in society and being calculated as a fraction of a free person meant that one was regarded as a different or lesser kind of person.

The Initial Reference Point The 1790 census was taken one year after President George Washington was inaugurated and included the population of the original thirteen colonies plus the territories of Maine, Kentucky, Vermont, and Tennessee (U.S. Bureau of the Census 1989:1). The questions asked the name of the head of the family and the number of persons living in each household who were free white males or free white females, both over and under sixteen years of age; all other free persons; and slaves. The gender and age of the slaves or “other free people” were apparently not important (U.S. Bureau of the Census 1978:1).2 The interest in “free white males over the age of sixteen” reflected the need “to assess the

68

FIG. 4.1. Census Schedule Forms, 1800–1820. U.S. Statutes at Large, 6th–12 Cong., 1799–1813, vols. 2 and 3, ed. Richard Peters, Esq. (1846 and 1856; reprint, Buffalo: Dennis & Co., 1963).

1800

Schedule of the whole number of persons within the division allotted to A. B.

Schedule of the whole number of persons within the division allotted to A. B.

1810

Eleventh Congress. Sess. II. Ch. 17. 1810.

Sixth Congress. Sess. I. Ch. 12. 1800.

1820

Schedule of the whole number of persons within the division allotted to A. B.

Sixteenth Congress. Sess. I. Ch. 24. 1820.

WHITES AND OTHER SOCIAL RACES

69

country’s industrial and military potential” (U.S. Bureau of the Census 1989:1). Between the drafting of the Constitution in 1787 and the taking of the first census in 1790, the term white became an explicit part of the first category to be measured. The “slaves” category remained unchanged, and the third category was labeled “all other free persons.” Theoretically, those in political charge could have chosen another definition for the first category and, consequently, themselves. That is, they could have chosen “free English-speaking males over sixteen” or “free males of Christian descent” or “of European descent.” But they chose color. Having named the central category “white” gave a centrality and power to color that has continued throughout the history of the census.

A Definitive Color Line In the census’s first four decades, local authorities took the census, and so the information was not uniform. Hence, the categories used on the national census frequently differed from those used on the state census. Finally in 1830, uniform census forms were introduced, although congressional records between 1800 and 1820 already included schedules recommended for taking the census (see figure 4.1). By 1840, the census categories had established a number of patterns, and significant changes had been made as well. As table 4.1 shows, between 1790 and 1840, the categories of “free whites” and “slaves” stayed the same. But in 1800 and 1810, the 1790 category “all other free persons” was changed to “all other free persons, except Indians not taxed.” In 1820 it was subsumed under “free colored persons,” and in 1830 it disappeared altogether.3 The “free colored persons” category was retained in the 1840 census. In 1830, when uniform census forms were introduced, the color line was also more clearly established. The original color-free category “all other free persons” that appeared in the first three censuses had disappeared. The major divisions were now more explicitly “colored”: whites, who were free, and coloreds, who were free or slave. (See appendix C for a discussion of why the first three censuses did not contain a color term and why the original “all other free persons” category was replaced by the “free colored” category.)

Table 4.1 Labeling Citizens and Others in Early Censuses, 1790–1840 1790

1800

1810

1820

1830

1840

Free white males and females

Free white males and females

Free white males and females

Free white males and females

Free white males and females

Free white males and females

Slaves

Slaves

Slaves

Slaves

Slaves

Slaves

All other free persons

All other free persons, except Indians, not taxed

All other free persons, except Indians, not taxed

Free colored persons (all other free persons, except Indians, not taxed)

Free colored persons

Free colored persons

Foreigners not naturalized

Foreigners not naturalized

Foreigners not naturalized

Sources: Return of the Whole Number of Persons within the Several Districts of the United States (1802); “Aggregate Amount of Each Description of Persons within the United States of America, and the Territories Thereof, Agreeably to Actual Enumeration Made According to Law, in the Year 1810” (1810); Census for 1820 (1821); U.S. Dept. of State (1832a and b, 1835, 1842); U.S. House of Representatives (1895); U.S. Statutes at Large (1846, 1856); U.S. Bureau of the Census (1967, 1978, 1989); Heads of Families at the First Census of the United States Taken in the Year 1790: New York (1992).

THE EVOLVING BIPOLAR STRUCTURE After 1820 and the shift to color categories, the elements of culture, language, religion, and mixture were compacted into a choice between white and colored. By 1830, a bipolar structure—of “whites” and “nonwhites”—was clearly taking shape. As table 4.1 shows, the categories used between 1790 and 1840 were based on the three criteria of freedom, birthplace, and color (on one side were “free white males and females,” “all other free people,” “free colored,” and “foreigners, not naturalized”; and on the other side were “slaves”). By 1830, however, the data on “free people of color” and “slaves” were combined in some instances. For example, a table in the 1830 census reporting “the number of deaf and dumb” combined “slaves and colored persons” in one count, and “aliens and foreigners not naturalized” were included in the white count (U.S. Dept. of State 1832a:42–43). Moreover, “aliens and foreigners not naturalized” were included in the “total white” count (U.S. Dept. of State 1832b:48–51). The bipolar structure of white and colored became more explicit. 70

WHITES AND OTHER SOCIAL RACES

71

Although the censuses between 1830 and 1860 reported the numbers of whites, slaves, and free colored in separate columns, some of the tables combined slaves and free colored.4 The hypodescent rule also became more explicit,5 and “other races” were put into the not-white or colored column. By the third decennial census (1820), whiteness was more precisely defined, with the addition of the “foreigners not naturalized” category6 (see table 4.2). This category distinguished the foreign (most likely white and free) from the native-born white and free. Its introduction suggests a distinction between the “whites” in the power structure, who were citizens by birth, and “probationary whites,” who were not (Ignatiev 1995; Jacobson 1998). In the 1850 census, the category of “free whites” was changed to simply “whites,” which suggests that by this time it was evident that all the people in this category were free. As table 4.2 shows, from 1820 to 1880, census forms continued to ask for “color,” but by the twentieth century, they shifted away from the term color and substituted race. As table 4.3 indicates, “mixed” persons were counted, as there appeared to be a growing concern with measuring mixture more accurately, particularly after the Civil War. A category of “other races”—for example, Chinese, Indians, and Japanese—was added. Finally, more information, such as exact age, was collected for all persons, regardless of race or color. Beginning with the 1850 census, census takers were instructed to gather information on the color, age, sex, and other characteristics of each slave and free colored person. This was a major shift, because previously these groups had simply been listed as household members and information about them was not collected. Now these two “not-white” groups, the slaves and the free colored, were to be described as fully as the white group.

Mulattoes As table 4.3 shows, in both the free and slave populations, mulattoes were counted for the first time in 1850, with similar procedures used to count both the slaves and the free colored.7 According to the published data, mulattoes never constituted a large proportion of the total recorded “Negro” population—less than one-fifth in all but one year (Miller 1991:table 2; Williamson 1984:102). But given the difficulties of measuring those who attempted to “pass” and of “accurately” measuring “mulatto-ness,” these figures are not reliable.

72

■

1790

■

■

1800 a 1810

■

1820

■

1830

■

1840

■

1850

■

1860

■

1870

■

1880

■

1890b 1900

■

1910

■

1920

■

1930

■

1940

■

1950

■

1960

■

1970

■

1980

■

1990

Sources: Return of the Whole Number of Persons within the Several Districts of the United States (1802); “Aggregate Amount of Each Description of Persons within the United States of America, and the Territories Thereof, Agreeably to Actual Enumeration Made According to Law, in the Year 1810” (1810); Census for 1820 (1821); U.S. Dept. of State (1832b, 1842); U.S. House of Representatives (1895); U.S. Bureau of the Census (1932, 1943, 1953, 1963, 1973, 1978, 1989).

The 1800, 1810, and 1820 censuses contain the category “all other persons except Indians not taxed.” But starting with the 1820 census, that category was placed under a new, broader category, “free colored persons.” b In 1890, the category stated “whether white, black, mulatto, quadroon, octoroon, Chinese, Japanese, or Indian.” Figures for these groups were reported separately and there was also a “total colored” column that provided the total for all these groups. See U.S. House of Representatives 1895. c In the 1960 and 1980 censuses, an interrogative category was used: Is this person . . . ?

a

Is this person . . . ?c

Race

Color or race

Color

Free colored

All other free persons

Census Categories

Table 4.2 The Shift from “Color” to “Race” in Decennial Censuses, 1790–1990

WHITES AND OTHER SOCIAL RACES

73

Table 4.3 Labeling Mixture and Other Races, 1850–1880 1850 a

1860

1870

1880

Whites Free blacks Free mulattos Slave blacks Slave mulattos

Whites Free blacks Free mulattos Slave blacks Slave mulattos

Whites Blacks Mulattos

Whites Blacks Mulattos

Indiansb Chinese

Indians Chinese Japanese c

a

By 1850, gender (referred to then as “sex”) was being recorded for most groups. Counts for Indians and Chinese were reported in the 1860 census. But it was in 1870 that categories for these groups appeared on the census form. c A category for the Japanese was not listed separately on the census form in 1880, but some of the tables did report separate figures for the Japanese (see, e.g., U.S. House of Representatives 1883:table 1a, p. 3). The preface to the 1880 census also describes Whites and Coloreds and indicates that Asiatics includes Chinese, Japanese, East Indians, etc. (U.S. House of Representatives 1883:xxvi). b

Sources: U.S. House of Representatives 1883:xxvi, 1895; U.S. Statutes at Large 1856; U.S. Bureau of the Census 1967, 1978, 1989.

The concern with correctly measuring color surfaced after the Civil War when the slave category became an anachronism, and it is evident in the instructions given to enumerators during this period. In the 1850 and 1860 censuses, the enumerators had been instructed to write “B” for “black” and “M” for “mulatto” and to leave the space blank for “white.” But in 1870, the census takers were instructed: “It must not be assumed that where nothing is written in this column ‘white’ is to be understood.” This may have corrected what must have been a problem in the previous censuses, that leaving the space blank might have enabled some people of mixed ancestry to “pass” into the “white” category. Thus, when in doubt about the “color” of difficult-to-classify individuals, the enumerators might have been inclined to leave the designation blank, resulting in their being counted as “white.” This type of “passing” may have been more tolerated under slavery, when the number of free people of color was relatively small and the condition of slavery served as a primary marker of status, color, and race. But “passing” was not tolerated after Emancipation, when status could not be determined as easily and light-skinned former slaves might try to pass into the white category.

74

WHITES AND OTHER SOCIAL RACES

The instructions for the 1870 census also advised enumerators to be “particularly careful in reporting the class Mulatto. The word is here generic, and includes quadroons, octoroons, and all persons having any perceptible trace of African blood.” In addition, “Important scientific results depend upon the correct determination of this class” (U.S. Bureau of the Census 1989:26). The concern with mixture (understood mainly as the proportion of “black blood”) reached a peak in the 1890 census, which counted quadroons (one-quarter “black blood”), octoroons (one-eighth “or any trace of black blood”), mulattoes (threeeighths to five-eighths black), and blacks (three-quarters or more) (Wright 1956:187).8 This more complicated racial scheme was unworkable for the census, however, and it was omitted from the next one (Miller 1991:1; U.S. Census Office 1901:cxi).9

THE GROWTH OF A RACIST IDEOLOGY The statement that “important scientific results” depended on the correct classification of “mulattoes” and “blacks” suggests that the census may have been influenced by the then popular theories of scientific racism, which held that group differences could be “scientifically” attributed to “race.” It is widely believed today that in the nineteenth century, a racist ideology (based on color differences) developed that served the purpose of rationalizing expansion, slavery, and class differences (Banton 1983; Barzun 1965; Bernal 1987; Bieder 1986; Freedman 1984; Gossett 1963; Gould 1981; Johansen 1982:84; Jordan 1968; Sanjek 1994:5; Snowden 1983; Stanton 1960; Thomas 1989: 29–31; Thompson 1989). Horsman, who examined writings, politicians’ speeches, and newspaper coverage of the period, found that after 1815, the “supposed lessons of the American experience hastened the collapse of Enlightenment theory and helped produce scientific theories of black and Indian inferiority. Along with this debasement of other races was to come an enhancement of the white race as superior ” and more explicitly stated census concerns about the mixing of the races (1981:115). By 1850, the census publications already manifested a strong identification with northern Europe and a desire to preserve, legitimize, or develop a northern European “racial” identity for the United States. As the 1850 census stated: “The great mass of the white population of this country is of Teutonic origin, with a considerable admixture of Celtic”

WHITES AND OTHER SOCIAL RACES

75

(U.S. Secretary of the Interior 1853b:10). It was reasoned that with a predominantly northern European population, the United States would be able to compete with its northern European counterparts, particularly since it was located on much the same latitude and had a climate similar to that of Europe (U.S. Secretary of the Interior 1853b:10–11). The assumption was that climate had determined and would continue to determine the evolution and progress of the different human races. The United States, imagined as a country whose population was of primarily Anglo-Saxon and Teutonic origin, was seen as having a particular destiny. According to the 1850 census, As has been truly observed, “a race of men launched upon the tide of existence, have, by virtue of all the conditions, a determined course to run, which will make its own way, and fulfil its own destiny, in accordance with a system of laws as unalterable and supreme as those which control the physical universe.” (U.S. Secretary of the Interior 1853b:10)

Another assumption was that the same laws of life would prevail on both sides of the Atlantic and “produce like results upon both continents” (U.S. Secretary of the Interior 1853b:10). To this end, the life expectancies of American whites were computed and compared with those in Europe and were found to be the same for the “different branches of the Teutonic family of nations, in temperate climates.” The statistics were compared for England and Massachusetts, which are on the same latitude, and for those of Maryland and France, which are also on the same latitude. That this type of discourse should appear in the census volumes was unusual, as they tended to be rather bureaucratic and devoid of editorial positions. The departure probably reflected the intensity of these issues before the Civil War. This view of a future predestined by geographic location, the migration of northern Europeans, and climatic features was undoubtedly the basis of the concern with the growth of the “colored” population, which had evolved in more southern latitudes, in different climates, and from seemingly less advanced people. This concern also reflected a perceived threat to the numerical and political dominance of whites and to the clear demarcation of the “races.” Whites may also have feared that the colored (both slave and free) population might retaliate

76

WHITES AND OTHER SOCIAL RACES

against what one census publication referred to as the “governing race” (U.S. Secretary of the Interior 1852:20). These concerns about the races’ mixing and the growth of the “colored” population had surfaced earlier in census documents. For example, in the preparation for the 1830 census, Congress specifically asked the census for projections of this population’s growth and its impact on the white population. The 1830 census accordingly prepared tables comparing the 1790 and 1830 populations, in which it combined both the free colored and slave populations (U.S. Dept. of State 1835). Also in 1850, the census produced tables showing the ratio of increase of the white, free colored, and slaves since 1790 (U.S. Secretary of the Interior 1853a:ix, lxxxvii). Concern with the growth of the colored (both slave and free) population may also have been rooted in the fear that they might retaliate against the “governing race” (U.S. Secretary of the Interior 1852:20). Indeed, the demographic picture of the populace at the start of the census taking shows that the “governing race” was not so much in the majority (in all areas) as subsequent history texts suggested. In 1790, those seen to be the military and commercial guardians of the society were not the overwhelming majority of the population. Free white males over the age of sixteen constituted only 20.7 percent of the total population, and slaves and “all other free persons” of all ages accounted for 17.8 percent and 1.5 percent, respectively10 (Heads of Families, 1908/1992:8). The distribution of blacks by state at this time also shows that some had very high proportions of “Negroes” (U.S. House of Representatives 1895:xcvi). For example, in 1790, blacks (both slave and free) constituted 44 percent of South Carolina, 41 percent of Virginia, 36 percent of Georgia, 35 percent of Maryland, 27 percent of North Carolina, 22 percent of Delaware, and 19 percent of Ohio Territory (Reference Library of Black America 1990:483). These demographic findings may have fueled the concern of many about the growth of the “colored” populations. This concern continued throughout the nineteenth century. The last census before the Civil War, in 1860, contained a table comparing the growth rates of the free colored, slave, and white populations by state and territory between 1840 and 1850 (Kennedy 1862:table 1, p. 130) and a table showing the percentage increase of the free colored and slave populations between 1790 and 1850 (Kennedy 1862:17). Then, in the

WHITES AND OTHER SOCIAL RACES

77

1870, 1880, and 1890 censuses, maps were included that showed the density of the colored population and the proportion of colored in the total population (U.S. House of Representatives 1883, 1895; U.S. Secretary of the Interior 1872b).11 Occasionally, the census expressed concern with the growth of the colored population. For example, when relating the history of African Americans in Maryland, a special census volume stated: “The tendency of the colored race to encroach upon the numerical superiority of the white continued for twenty years longer, until, in 1810, they were found to have attained the ratio of 38.22 in a hundred of the entire population, and the whites had declined correspondingly to 61.78” (U.S. Secretary of the Interior 1852:20). It added that during the last twenty years, the number of colored had been more than double that of whites but that a way had been found a way to check this growth: There was in 1810, reason for apprehension that, in another half century, the blacks would become the preponderating race. There is reason to believe that this alarming tendency was checked by the introduction of new pursuits of industry, giving employment to a portion of the native population, which would otherwise have sought it beyond the limits of the State and inviting into it emigrants from abroad. (U.S. Secretary of the Interior 1852:20)

These efforts, plus the encouragement of migration from Great Britain and Germany, “rescued the whites from the peril, which seemed to be impending, of a loss of their numerical predominance” (U.S. Secretary of the Interior 1852:20).

Free People of Color Free people of color were a challenge to the distinction between the slave and free populations. On the one hand, they were free and therefore perhaps entitled to the same rights as nonslaves. On the other hand, they were of African descent and thus “not equal” to free whites. Because of the striking rise in their numbers in the two decades before 1820, it became important to count them more precisely.12 (See appendix D for a more detailed discussion of the concerns with the growth of this group.)

78

WHITES AND OTHER SOCIAL RACES

Other Races Before 1870, the census form offered a choice between two categories, whites, defined in terms of the absence of any “black blood,” and colored, defined by its presence. Then in 1870, categories for Chinese (and later Japanese) were added to the census form in response to the increasing numbers of Asian immigrants toward the end of the nineteenth century.13 The addition of Native Americans reflected the growing recognition of their dependence on the U.S. government after they were relocated onto reservations (Lurie 1974). In 1870, data were gathered according to color (i.e., “whites,” “Chinese,” “Indian,” and “colored”—blacks and mulattoes) but were reported separately by group. Thus, under the heading “Color,” the enumerators were to write in “white (W),” “black (B),” “mulatto (M),” “Chinese (C),” or “Indians (I)” (U.S. Secretary of the Interior 1872a:18, 20–21; 1872b:606–609). According to Carlberg (1992), these additions may have been “color” groupings, but they did not represent the population referred to as “colored” at the time. In other words, they were not “colored” (as understood then), but they also were clearly not “white.” This method of separately reporting information on the other races was continued in the 1880 census,14 but some tables and the introductory section of the 1890 census contain a footnote that the “colored population” included “persons of negro descent, Chinese, Japanese, and civilized Indians” (U.S. House of Representatives 1895: 400–401, clxxx, 681).15 Thus, it appeared that the earlier “white” and “colored” dichotomy had begun evolving into a “white” and “other than white” dichotomy, with many more categories in the “other than white” group.

Whiteness and Birthplace The large influx of immigrants from southern and eastern Europe also led to a concern about how they were affecting the population at large. Hence, the 1880 census listed the proportion of “defective, dependent, and delinquent classes”—including the mentally disabled, retarded, blind, and deaf—among the native born, foreign born, whites, and colored (U.S. House of Representatives 1883:926 and table ix). In addition, the census gave the distribution of native-born colored in the population according to state or territory of birth (U.S. House of Representatives 1883:477 and table xii).

WHITES AND OTHER SOCIAL RACES

79

This concern with immigrants and their impact on the total population was reflected again in the 1890 census, in the more elaborate maps, charts, and sections on the foreign-born population. Maps and tables also showed the distribution in the United States of “Natives of the Germanic Nations” and of “Greco-Latins.” Pie charts described changes in the U.S. population over time in the birthplaces of nativeborn parents, foreign parents, foreign born, and colored. These were called the four “elements at each census” and were accompanied by tables and discussions of the marital status of each (U.S. House of Representatives 1895:clxxix, 394, 681 ff). These reports suggest the continuing concerns with preserving a national identity as a basically northern European people. The United States’ bipolar structure was still in place at the end of the nineteenth century, although it had become more complex. Whites were clearly the central category by which others were defined—as either white or not white. Now, however, there were “other races” and also more information on everyone. The official definition of “mulatto” was someone with any perceptible trace of African blood, which was an important step in the development of the hypodescent rule. At that time, the rule distinguished mulattoes from blacks, but eventually it would define all blacks (Grieve 1996:56). Concern with the impact of immigration on the total population continued. Questions about immigrants and their racial origins were the subject of the government’s massive Dillingham Report, which focused on immigration at the turn of the century (U.S. Immigration Commission 1911). The report used the phrase “races and peoples” throughout and entitled its ninth volume Dictionary of Races or Peoples. This reflected the ambiguity of whether Europe’s linguistic groups were racial groups or peoples. Nonetheless, these various white peoples were eventually accepted as Caucasian or American white (Jacobson 1998). As the century drew to a close, questions of who was white and who could be a citizen also began to be litigated in the courts, thereby defining whiteness even more narrowly (Haney López 1996). The persistence of the 1790 federal law requiring that naturalized citizens be white, in combination with other state laws that required one to be a citizen in order to own property, vote, hold office, and the like, continued to restrict the rights of many nonwhite immigrants and to bring them to court in an attempt to be designated either “white” and/or a citizen.16

80

WHITES AND OTHER SOCIAL RACES

By the end of the nineteenth century, it was clear from these court cases that a basic racial structure of whites and not-whites had evolved.

THE SECOND CENTURY Color or Race, 1900–1940 The 1900 census dropped the 1890 attempt to count the black population by blood quantum of one-eighth and so forth and admitted that these figures had been “of little value” (U.S. Census Office 1901:cxi). But it still counted mulattoes and blacks as two subcategories of Negro,17 and a footnote to the “Negro” column indicated that this category included “all persons of Negro descent.” In addition, a special census of Native Americans asked how much “white blood” they had (U.S. Bureau of the Census 1989:46). Thus, “blood” (and its effect on color) still seemed to be an important, if not the principal, basis for establishing a person’s color, which in turn determined his or her “race.” The division between “white” and “other-than-white” became much more clear-cut in the 1900 census. Now the data on blacks, Chinese, Japanese, Indians taxed, and Indians not taxed were listed under the broader “Colored” column (U.S. Census Office 1901:483). Likewise, in the introduction to the 1900 census, whites and colored were carefully distinguished: “From these tables it appears that the population of the entire area of enumeration in 1900 is composed of 66,990,788 white persons and 9,312,589 colored persons, the latter figure comprising . . . persons of negro descent, . . . Chinese, . . . Japanese, and . . . Indians” (U.S. Census Office 1901:cxi, numbers omitted). Curiously, the 1900 census also added the term race to color and introduced the phrase color or race, which was used on all the census forms for the next forty years (U.S. Bureau of the Census 1978, 1989). Used together, these terms reinforced the singularly physical interpretation of racial construction in the United States. Nevertheless, both the white and other-than-white race groups were in fact social and political constructions. Although the 1900 census text was clear with regard to the division between races, it was slippery when classifying “mixed” or in-between groups. It noted, for instance, that the Croatans in North Carolina had been counted as white in 1890 and as Indian in 1900 (U.S. Census Office 1901:cxxiv).

WHITES AND OTHER SOCIAL RACES

81

The 1910 census addressed the issue of in-betweenness more directly. It gave new instructions to the enumerators that formed the basis for the following decennial censuses: “For all persons not falling within one of these [race or color] classes they should write ‘Ot’ (for other). . . . For census purposes, the term ‘black’ (B) includes all persons who are evidently full-blooded negroes, while the term ‘mulatto’ (Mu) includes all other persons having some proportion or perceptible trace of negro blood” (U.S. Bureau of the Census 1989:ii, 50). Perhaps influenced by the then politically ascendant eugenics movement, which also was influencing immigration legislation (Jacobson 1998:133; Marks 1995:87 ff), the 1920 census reported in its introduction the “color or race” of the people in the United States’ outlying possessions—Guam, American Samoa, the Panama Canal Zone, the Virgin Islands, the Philippine Islands, and “Porto Rico” (Puerto Rico) and also a special census of “Porto Rico” (U.S. Dept. of Commerce, Bureau of the Census 1922:11). In addition, the census more explicitly defined blood “purity” and categories; for example, “The term ‘white’ as used in the census reports refers to persons understood to be pureblooded whites.” Also, the “colored” applied to blacks, Indians, Chinese, Japanese, and “all other,” who were “Filipinos, Hindus, Koreans, Hawaiians, Malays, Siamese, and Maoris” (U.S. Dept. of Commerce, Bureau of the Census 1922:10; 1921:16). The 1920 census still had a few tables counting mulattoes,18 but volume 2 acknowledged the “considerable uncertainty” concerning “the classification of Negroes as black and mulatto,” since the “accuracy of the distinction” depended largely on “the judgement and care employed by the enumerators” (U.S. Dept. of Commerce, Bureau of the Census 1921:16–17). Furthermore, figures for the same county varied greatly depending on whether the census enumerators were black (as in the 1910 census) or white (as in the 1920 census). Black enumerators found a higher proportion of mulattoes. This awareness that racial perception was influenced by variables such as the interviewer’s race or the community’s acceptance probably helped move the 1920 census to abandon the distinction between “mulatto” and “black,” thereby moving the hypodescent rule to another level. Anyone with any black ancestry was now simply “black” or “Negro.” Paradoxically, the awareness that racial classification was socially constructed, that is, influenced by personal and social factors, led to a more rigid adherence to genetic ancestry, which further reinforced the

82

WHITES AND OTHER SOCIAL RACES

hypodescent rule. The hypodescent rule also separated “race” from “ethnicity,” for regardless of ethnicity, one’s race was the main determinant of one’s status. Thus, race was the primary means of identification, and ethnicity was subordinated, obscured, or combined with race. Williamson (1984), Davis (1992), and Domínguez (1986) discussed this shift in the hypodescent rule and the involvement of both the government’s definitions and people’s own self-affirming and self-determining actions. Williamson, for example, argued that in the shift from a three- to a two-tier racial structure, a “new people” was born. A fusion of Europeans and Africans, they were proud and articulated their identity most eloquently in the Harlem Renaissance of the 1920s (p. 111). When the Negro culture was embraced, “negritude” was also redefined, and “the beauty of all colors and features” was recognized (p. 58). Accordingly, race was redefined as based on descent and cultural definitions rather than appearance. African Americans found strength in their blackness and in that strength lay the power to stand apart from the world (p. 187). As Williamson noted, “The drive for a biracial society had reached its culmination . . . not by white dictation . . . but . . . by the eager embracement of ‘blackness’ by American Negroes” (p. 3).19 The 1920 census also confirmed the hypodescent rule by specifying how mixed-race people were to be classified: “A person of mixed blood is classified according to the nonwhite racial strain or, if the nonwhite blood itself is mixed, according to his racial status as adjudged by the community in which he resides.” The examples provided made clear that “regardless of the amount of white blood,” a person with a mixture of “Negro” or “Indian” blood was to be classified “either as an Indian or as a Negro, according to his racial status in the community in which he lives.” Finally, the white population was divided into four groups depending on birthplace (U.S. Dept. of Commerce, Bureau of the Census 1922:10).20 As table 4.4 indicates, although other categories were added throughout the twentieth century, the censuses taken between 1900 and 1940 varied little from the basic structure established in 1920. This structure contained three divisions (whites, Negroes, and other races) within a basically bipolar population of whites and colored. “Other races” included all those who were not white or Negro, for example, Japanese, Chinese, and Indians—all those who were nonwhite or colored. There was, however, one interesting deviation. In 1930, “persons of Mexican birth or parentage who were not definitely reported as white

WHITES AND OTHER SOCIAL RACES

83

Table 4.4 Census Race and Color Categories, 1890–1990 1890

White

■

Black

1900

1910

1920

1930

1940

1950

1960

1970

1980

1990

■

■

■

■

■

■

■

■

■

■

■ ■

■

■

■

■

■

■

■

■

■

Negro Black/Negro Mulatto

■

Quadroon

■

Indian

■

■

■

■

■

■

■

■ ■

American Indian

■ ■

■

Aleut

■

■

■

Eskimo

■

■

■

Indian (Amer.)

Asian or Pacific Islander

■

Chinese

■

■

■

■

■

■

■

■

■

■

■

Japanese

■

■

■

■

■

■

■

■

■

■

■

Filipino

■

■

■

■

■

■

■

Hindu

■

■

Korean

■

■

■

■

■

■

■

■

Vietnamese

■

■

Asian Indian

■

■

Guamanian

■

■

Samoan

■

■

■

■

Hawaiian

■

Part Hawaiian

■

■

Mexican Other

■

■

■

■

■

■

■

Source: Adapted from Sharon M. Lee, “Racial Classifications in the U.S. Census: 1890–1990,” Ethnic and Racial Studies 16 (1) (January 1993): 78.

or Indian were designated Mexican” and tabulated with “other races,” such as Native American, Japanese, or Chinese (U.S. Bureau of the Census 1932:1).21 The 1940 census, however, reversed this policy regarding Mexican classification, stating that “persons of Mexican birth or ancestry who were not definitely Indian or of other nonwhite race were

84

WHITES AND OTHER SOCIAL RACES

returned as white” (U.S. Bureau of the Census 1943:3). Thus, within a decade, Mexicans were shifted from their own “Mexican” category to being included in the “white” category—unless they appeared to census interviewers to be “definitely Indian or of other Nonwhite races” (U.S. Bureau of the Census 1943:3; see table 4.4).

Fluctuating Labels, 1950–1990 Until 1940, “color or race” was consistently used as a label to describe groups, but in the second half of the century, this practice changed. As table 4.2 shows, in 1950, the census form used only “race.” In both 1960 and 1980, it simply asked, “Is this person . . . ?” and provided a list of categories.22 In 1970, it was “color or race,” and in 1990, it was again “race.” As this book goes to press, the question in the 2000 census will be, “What is this person’s race?” and in a major departure from the census’s two-hundred-year history, more than one response will be allowed. After World War II, the census first tried to explain the concept of race, and the 1950 census admitted that the concept lacked scientific precision and was based on public opinion (U.S. Bureau of the Census 1953:35). The census also recognized the importance of context in determining race: “Experience has shown that reasonably adequate identification of the smaller ‘racial’ groups is made in areas where they are relatively numerous but that representatives of such groups may be misclassified in areas where they are rare” (U.S. Bureau of the Census 1953:35). Similar admonitions were repeated in the 1960 census (U.S. Bureau of the Census 1963:xx).23 These questions mirrored the scientific and international community’s broader questioning of the concept of race, in the wake of the atrocities committed during World War II in the name of racial purity (see UNESCO 1952). The question of “who was black” in the United States also was examined more closely and was found to have different answers in different states. In the years leading up to the 1954 Brown v. Board of Education decision outlawing segregation in the public schools, it was evident that legal definitions of a “black” person varied as well. As Haney López (1996:118–119) noted, some states used a broad “one-drop” rule; for example, in Alabama and Arkansas, anyone with one drop of Negro blood was black. Texas used the “all persons of mixed blood descended from negro ancestry” standard. Tennessee followed the same rule but

WHITES AND OTHER SOCIAL RACES

85

included mestizos; it defined “blacks in terms of mulattos, mestizos, and their descendants, having any blood of the African race in their veins.” A number of states—Florida, Indiana, Missouri, Nebraska, North Carolina, South Carolina, and North Dakota—followed a more precise and simple one-eighth rule, and Oregon had a one-quarter rule. Utah law used a similar blood-quantum approach that distinguished among mulattoes, quadroons, and octoroons. Other states relied on what could be established. For example, Georgia referred to “ascertainable” nonwhite blood. Kentucky relied on a combination of any “appreciable admixture of Black ancestry and a one-sixteenth rule.” Louisiana adopted an “appreciable mixture of negro blood” standard, and Mississippi combined an “appreciable amount of Negro blood” and a one-eighth rule. Maryland used a “person of negro descent to the third generation” test. Interestingly, Oklahoma, the home of many resettled Indian nations, referred to “all persons of African descent,” adding that the “term ‘white race’ shall include all other persons,” which suggests that Native Americans and others were now “white.” Virginia appeared to differentiate black Indians from blacks when it defined blacks as those in whom there was “ascertainable any Negro blood: with not more than one-sixteenth native American ancestry.” By 1970, the census appears to have begun departing from what it admitted was a very unscientific, contextually dependent, and opinionbased approach and shifted to a self-classification of race. Although the census forms as a whole were still administered by census takers, the 1970 census noted that information on race was “obtained primarily through self-enumeration” and that respondents self-classified themselves “according to the race with which they identify themselves” (U.S. Bureau of the Census 1973:5).24 By 1980, census forms were mailed, and the recipients chose their race from the categories supplied. Self-classification continued in the 1990 and 2000 censuses.

THE LONG ROAD TO TODAY From our current vantage point, it may seem surprising that throughout the census’s two-hundred-year history, color and not race has usually been the term of reference. “Race” appeared on the census form only at the start of the twentieth century when it was included with

86

WHITES AND OTHER SOCIAL RACES

“color” on the 1900 census. It was not until 1950 that “race” appeared by itself. In contrast, “color” was in the census legislation from its inception. Thus “color” was an integral characteristic of the census, persisting for more than 150 years in census forms, introductions, and instructions to census takers25 (U.S. Bureau of the Census 1978, 1989). Why was the term color retained for so long on the census? It might have been inertia or a reflection of the then commonly used term colored people. Is the concept of color still commonly accepted today, even though the term has been discontinued officially? Finally, is the history or legacy of this concept connected to the fairly recent introduction of the term people of color? This term is used, particularly in academia, to define or unite what are, in effect, “other social races.” The category that we think of today as “race” has undergone several transformations. Nonetheless, many people believe that racial classifications are static and biologically based. These views were encouraged by the government’s policy requiring individuals to choose only one category to identify themselves, which reinforced the impression and myth of “pure” races (Lee 1993). Since these categories were based on supposed color differences, census classifications also reinforced a presumable biological basis for what were really social distinctions and definitions. According to Lee (1993), the concepts of race and ethnicity have been confused as well, viewing what are in effect “social groupings” as biological races. This view began to change in 1950 with the census’s tacit admission that “race” is not a scientific concept but that it is often socially determined. This view has continued to change, and the basic bipolar, hierarchical racial construction is being challenged as the result of a series of events, such as increased and more diverse immigration, greater intermarriage, more global and intense economic competition, new scientific and technical discoveries, changes in the socioeconomic positions of “other social races,” and new views of race.

5 The Shifting Color Line

D E S P I T E T H E O V E R A R C H I N G bipolar structure that emerges from

our review of census documents, there is and has probably always been a great deal of heterogeneity within the two polarities. Moreover, the lines between the two have not always been definite but have fluctuated. For example, some individuals and groups in the “other social races” have occasionally been classified as “white,” and mixed-race persons have always blurred the boundaries of these socially constructed polarities. Some people and groups have tried to alter their classification, and the census itself has changed the labels it uses to describe various groups. Among the groups that have legally contested their racial classification or had it changed are Filipinos, Afghans, Chinese, Japanese, Koreans, Syrians, Burmese, Mexicans, Hawaiians, Native Americans, and certain mixtures (Haney López 1996). Sometimes the rulings regarding their racial status have been both curious and conflicting. For example, the 1854 case of People v. Hall ruled that Chinese immigrants in California were “generically ‘Indians,’” and the 1893 case of Saito v. U.S. ruled that Japanese immigrants were “Mongolian” (Almaguer 1994:10). Armenians were first classified as “Asiatic” until a federal court ruled in 1909 that they were white (Haney López 1996:130–131; Takaki 1994:15). This chapter focuses on the changes in the census classifications of Native Americans, African Americans, Asian Indians, and Hispanics. These groups’ experiences are a good illustration of the shifts in racial placement and labeling by the census over time, the groups’ challenges of their racial classification, and the influence of political factors on racial classification. In particular, the experiences of Native Americans, African Americans, and Asian Indians illustrate the historical relationship between challenges to racial classification and the awarding of citizenship. The Hispanic experience—although less contentious in this regard—nonetheless highlights the extent to which “mixture” has been 87

88

THE SHIFTING COLOR LINE

perceived as problematic for full U.S. citizenship. All groups exemplify the historical difficulty that the census has had dealing with mixture and with groups who have not fit neatly into discrete categories of color. Finally, these groups’ experiences underscore the extent to which classifications have been influenced by, and have influenced, political considerations.

NATIVE AMERICANS The generic term used to refer to those peoples present when Europeans first arrived in North America has been modified only slightly over the last one hundred years (see table 4.4). This vastly diverse set of multilingual, multicultural peoples were first misnamed “Indians” by Christopher Columbus, who thought that he had reached India. It is a label that persists even today, although Native Americans is preferred. This persistence is perhaps reflective of the tendencies in this country’s racial structure to ignore differences among those classified as “not white.” Although the census did distinguish between “domesticated” or taxed Indians, referred to tribes in an “advanced state of civilization” who owned slaves (Kennedy 1862:11), and described blood quantum, the name used to describe the group as a whole has tended to stay the same. From 1860, when the census first counted untaxed Indians, to 1940, Native Americans were simply “Indians”; for the next twenty years, they were “American Indians”; and then between 1970 and 1990, they were listed as “Indians (Amer.).” In the 2000 census, the category is American Indian or Alaska Native. The censuses have always collected tribal identification but only occasionally have reported it. The U.S. Constitution states that taxed Indians are to be counted as equal to “whites” for apportionment purposes. Thus, Native Americans may have first been counted as white—if they paid taxes. Then when all Indians were first counted separately in the 1860 census, they were classified as a not-white, not-Negro group within the “other races” category, along with the Chinese.1 Beginning in 1970, they have been listed, along with Eskimos and Aleuts, in their own “Native American Indian” race category. The Constitution does mention taxed Indians. The fact that the federal government did not report taxed Indians separately led to the as-

THE SHIFTING COLOR LINE

89

sumption that taxed Indians had been included in the white counts. A later census showed that this was the practice: “A few domesticated or taxed Indians” had been earlier “included in the tables of the whites” (U.S. Secretary of the Interior 1853b:ix). The 1790 census form for New Hampshire, however, showed that taxed Indians were included in its “free colored” column, not in the “white” column (U.S. Bureau of the Census 1967, 1989:276). It is not clear how these taxed Indians from New Hampshire were reported in the national figures. But this New Hampshire census suggests that how taxed Indians were counted in these earlier censuses varied by locality. Very likely, how taxed Indians were counted was determined by factors such as phenotype, the extent to which they had assimilated and/or intermarried, and how much wealth and property they had acquired. Eventually, the category of “taxed Indian” ceased to have any “practical relevance” and became “an anachronism” (Pevar 1983:155). In 1935, the U.S. Supreme Court ruled that all Indians were subject to federal taxation (Superintendent v. Commissioner), and in 1940, for apportionment purposes, all Indians were included in the total number of persons (Clemence 1981). Although some Native Americans still are not taxed (those on reservations do not pay federal taxes), the census has long ceased to distinguish between those taxed and untaxed. Of greater importance perhaps, from our present-day perspective, is that untaxed Indians were not counted. The 1850 census contained the first estimate of untaxed Indians (De Bow 1854a:41, 1854b), and the 1860 census also included figures on Indians (Kennedy 1862:134–135). But not until 1870 was there a serious attempt to count such Indians, in order to measure the country’s “true population” (U.S. Secretary of the Interior 1872a:22).2 By 1890, the census reported that there were more untaxed Indians (189,447) than taxed (84,160) (U.S. House of Representatives 1895:cxxiv). The 1900 census was the first to classify systematically all Indians residing in the United States, taxed and untaxed. In addition to the early differentiation between taxed and untaxed Indians, Native Americans were also separated according to blood quantum. In the 1860 census, for example, “half-breeds” were listed separately from Indians. Again, how they were counted on the local level varied. In Wisconsin and in New Mexico Territory, they were tabulated both separately and in the white column, whereas in California, half-breeds and Chinese were listed separately under the white column (Kennedy 1862:134–135).

90

THE SHIFTING COLOR LINE

By 1870, the census admitted that “Indians” had intermixed to the extent that there were few persons of “pure Indian race”3 (U.S. Secretary of the Interior 1872a:19). Consequently, the census wondered how half-breeds should be classified racially. It began by defining the term as popularly understood, that is, as including “persons with any perceptible trace of Indian blood, whether mixed with white or with negro stock” (U.S. Secretary of the Interior 1872a:19). It then asked: “Shall they be regarded as following the condition of the father or of the mother? Or, again, shall they be classified with respect to the superior or to the inferior blood?” (U.S. Secretary of the Interior 1872a:19). Although the census clearly regarded Indians as a different race and half-breeds as having both “superior” (white) and “inferior” (Indian) blood, it stated that the criteria applied to “the former slave population” should not be applied to Indians (U.S. Secretary of the Interior 1872a:19).4 Curiously, the census finally chose a socially dependent criterion that classified half-breeds as white if they lived with whites and had the “habits of life” and “methods of industry” of whites. But if they lived in Indian communities, they were to be classified as Indian.5 This approach was referred to as the “most logical and least cumbersome treatment of the subject,” especially if the census was “to trace and record all the varieties of this race” and considering the “small and fastdecreasing numbers” (U.S. Secretary of the Interior 1872a:19). Thus, although Native Americans were referred to as a race, the hypodescent rule was not strictly applied to them because behavior and community recognition were considered in determining the race of half-breeds.6 In chapter 4, we discussed the censuses’ difficulty—especially toward the end of the nineteenth century—ascertaining the extent to which persons of African descent were “mixed.” The censuses also wanted to gauge the extent of Indians’ “white” or “black” blood. Accordingly, in 1890, the census questioned Native Americans living on reservations about this (Thornton 1987:217), and as noted earlier, the 1900 special census of American Indians asked them how much “white blood” they had (U.S. Bureau of the Census 1989:46). The census used the amount of “Indian blood” as a basis not just for counting Indians but also for awarding treaty rights and defining identity. Wilson (1992:108–125) and Jaimes (1994:41–61) maintain that gauging this socalled blood quantum has had a deleterious effect on Native Americans and is at variance with the Indians’ own definitions of themselves. As Wilson noted, this blood quantum criterion imposes “non-Indian racial

THE SHIFTING COLOR LINE

91

(and racist) assumptions onto Native American thinking.”7 Before the Europeans arrived, people intermarried across tribes but did not use the concepts of “half-“ and “quarter-breeds” or blood quantum (Wilson 1992:109, 116). Nonetheless, this blood quantum approach has divided—and continues to divide—the Native American community, as individuals debate what it is to be “Indian” and who is more “Indian,” based on that person’s perceived blood quantum (Jaimes 1994).8

AFRICAN AMERICANS The census history of African-descent persons is similar in some ways to that of Native Americans. Both were subdivided into two groups— one into free and slave and the other into taxed and untaxed. Initially, the “free colored” and the “taxed Indians” were small groups that were between whites and their respective unfree and untaxed groups in terms of rights of citizenship. In addition, blood quantum was used in both cases to subdivide and classify the groups. Finally, in both cases, one generic term was applied to all persons regardless of their highly diverse cultural and linguistic backgrounds. But the two broad groups also had some important differences. Most persons of African descent were first counted in a category that referred to their state of enforced lifetime bondage, that is, as “slaves” and thus as three-fifths of a person. Even within this category they were counted not as individuals but as part of a household. This classification reflected their legal status as “property.” For example, the data might show that the Henderson household contained one white male over sixteen years of age and one white female over sixteen, four children of different sexes, and five slaves. The gender and age of slaves were not reported separately until 1820, and other information was not available until 1840, when each slave was given a number (names were not listed). In 1820, a separate “free colored persons” category was introduced and retained until slavery was abolished. Both slave and free African Americans were subdivided according to their mixed heritage. Between 1850 and 1920, they could be either “black” or “mulatto,” and in 1890, smaller fractions of “black blood” were requested. In contrast to Native Americans, the generic term used by the census to refer to persons of African descent did change substantially over

92

THE SHIFTING COLOR LINE

time. In the early censuses, the category used was “black.” Later, the category “Negro” was used, and it included both “blacks” and “mulattoes.” Then between 1930 and 1960, the category Negro was used by itself. Beginning in 1970, the category “black or Negro” was used, and in the 2000 census, the category is “black, African Am., or Negro.” The inclusion of “African American” is significant, for it is the first time the group has been given a label that suggests geographic origin rather than color or race. Only one other contemporary race term does not refer specifically to geographic origin, the “white” census category.

BIRTHRIGHT, CITIZENSHIP, AND COLOR Citizenship is related to the question of classification, for in the United States, classification as white meant that a person could be a citizen by birthright or as a result of naturalization. Although the states had the right to restrict citizenship, they could not grant it to nonwhites. Citizenship is perhaps a society’s most basic and significant definition of rights and equality. Although the U.S. Constitution does not explicitly define citizenship, it does give Congress the power to naturalize aliens. One of the first laws that Congress passed was the Naturalization Law of 1790, which required that naturalized citizens be white. Thus, almost from the nation’s inception, the general outlines of citizenship were in place: a white person who was born in the United States was automatically considered a citizen and, if foreign born, could become a citizen. For a nonwhite person, however, citizenship was not a birthright, and a nonwhite, foreign-born person was prohibited by law from becoming a citizen. Consequently, neither African Americans nor Native Americans born in the United States could automatically become citizens. Although the path to full citizenship was different for Native Americans and African Americans, for both groups, citizenship was initially given to those in between, that is, to free people of color and to taxed Indians. A number of scholars have argued that these in-between groups did not enjoy a full citizenship status equal to that of whites; rather, it was a second-class citizenship status that was given to (and sometimes withdrawn from) them (see, e.g., Aptheker 1968; Fishel and Quarles 1970; Franklin 1967; Kettner 1978).

THE SHIFTING COLOR LINE

93

Native Americans The relationship of the Native American nations to the new United States government changed over time. In the colonial period, government officials dealt with independent and unconquered tribes on the fringes of the white settlements as “sovereign political communities.” After 1776, “the central government assumed primary authority over Indian affairs—or at least over tribes outside the boundaries of existing states” (Kettner 1978:288, 291).9 Since Native Americans were part of these sovereign political nations, they were initially seen to be “aliens” and not citizens. Furthermore, the naturalization laws that allowed European “aliens” to become citizens excluded Indians. Some “Indians,” however, did become citizens in accordance with their “individual circumstances.” This meant that some de-tribalized Indians were absorbed into the white population as citizens. Others negotiated separate agreements and relationships with the British monarchy, the different colonial governments, or, later, the U.S. government through their tribal governments. The extent of this “absorption” is not well documented. Kettner, for example, cites one source that found increasingly “separate and unequal treatment of Plymouth’s Indians” and not absorption. But many were undoubtedly absorbed as they intermarried and as white settlements gradually took over their lands and their status as tribes or sovereign political entities was challenged (Kettner 1978:289–299). During the nineteenth century, this early status of “sovereign nations” gradually eroded (Johansen 1982; Lurie 1974). In 1831, the U.S. Supreme Court rejected the Cherokee Nation’s argument that it constituted a “foreign state” in the sense in which this was understood in the U.S. Constitution. In Cherokee Nation v. Georgia, the Court argued that Indian nations were “domestic dependent nations” occupying a “state of pupilage.” According to Kettner (1978), this domestic, dependent nation status ultimately served the purposes of those who wished to maintain control over the Indians without fully incorporating them into the community of citizens. Being “domestic” allowed for the extension of white laws over the Indian nations, for as “nations,” they were not given citizenship or protection.10 The federal courts and executive branch concurred in excluding tribes and tribal members from citizenship. Even after the passage of

94

THE SHIFTING COLOR LINE

the Fourteenth Amendment, which granted citizenship to all ex-slaves, the federal courts continued to rule against birthright citizenship for Native Americans. Instead, they were deemed to be perpetual inhabitants with few rights—not citizens. Finally, in the Dred Scott case of 1857 (also important to determining the citizenship status of African Americans in the mid-nineteenth century), the Supreme Court argued that Indians were “aliens incapable of qualifying for naturalization because of the naturalization law’s color restrictions” (Kettner 1978:294–296).11 Nevertheless, despite the federal courts’ decisions and the naturalization laws restricting citizenship to free white immigrants during the nineteenth century, a number of treaties and statutes considered awarding citizenship to Native Americans under certain conditions. For example, the Cherokee treaties of 1817 and 1819 provided for land grants to heads of families “who may wish to become citizens of the United States” (Kettner 1978:292). Unfortunately, citizenship granted in this way eroded tribal landownership systems and often led to “the destruction of the tribal organization and government” (Kettner 1978:293). After the removal of the Cherokees to Oklahoma in the 1830s, North Carolina agreed to consider in the same way as other citizens those Cherokees who remained. Other examples are the treaties with the Delawares in 1778, which envisioned the admission of a separate Indian state as part of the Articles of Confederation, and the Cherokee treaties of 1785 and 1835, which raised the possibility of congressional representation. However, neither of these last two provisions ever took effect (Kettner 1978:291, 294). It is not known how many Indians became citizens through treaties and by breaking relations with their tribes, but the commissioner of Indian affairs reported in 1891 that before 1887, only 3,072 Indians had been admitted to citizenship through such treaties and congressional acts (Kettner 1978:293). With the Dawes Act of 1887, however, these numbers increased dramatically. This act admitted to citizenship those Indians who severed their relationship with their tribe and accepted grants of land in severalty. (It also resulted in the destruction of much communal tribal ownership.) Additional legislation raised the numbers further; for example, an 1888 law allowed both Indian women who married citizens and Indians who enlisted to fight in World War I to become citizens. By the time the act to make all Native Americans citizens was passed in 1924, two-thirds of them had already been admitted to citizenship through these acts and treaties (Kettner 1978:300).

THE SHIFTING COLOR LINE

95

African Americans Whereas the question of citizenship for Native Americans was a moot issue by those who saw Indians as belonging to (or having allegiance to and citizenship in) another “nation”—albeit a domestic, dependent one—individual African Americans did not have allegiance to a comparable foreign organization. Moreover, before Emancipation, most African Americans were slaves, who were neither aliens nor citizens but property. Kettner suggested this was a legal convenience, for if slaves could be seen as property, “judges could avoid fitting them into established categories of membership or non-membership” (1978:301). Immediately after the American Revolution, there were moves toward manumission, and during the first decades of the nineteenth century, slavery declined in the North, and the federal government formally outlawed it. But in the South, the rapid rise of cotton production and the continued fear of an ever-expanding black population led to a reversal of these early antislavery tendencies. Consequently, by the 1830s, local laws in the South became primary, and federal laws, secondary. Slavery and noncitizenship thus remained sanctioned by law in the Southern states and by the federal government’s policy of compromise and withdrawal (Kettner 1978:302, 311). The phrasing of the issue of slavery and citizenship before the Civil War shows how entrenched slavery had become in some areas. The question at that time was not whether free Negroes were citizens but whether their status was that of a former slave or a free person, that is, whether they were property or persons. Those who held that they were property argued that the manumission of slaves was an individual master’s decision; therefore, the state could not bestow citizenship. Since it was the master’s right to relinquish ownership of his property, his property did not have the right to be a citizen, regardless of whether he or she was a slave or an ex-slave. Thus, from this perspective, even free Negroes were property, but without owners to command them. In time, Southern states such as Tennessee and North Carolina retreated from these explicitly dehumanizing stands and emphasized more active discrimination against free blacks and mulattoes as indication of their separate status as a “degraded race” or a “third class.” In circular fashion, they cited the discrimination as justification for their continued separate (and consequently discriminatory) treatment of blacks with regard to citizenship rights. Similar arguments were used in

96

THE SHIFTING COLOR LINE

the U.S. Supreme Court in the Dred Scott case. In contrast, the Northern courts favored citizenship for free Negroes, but they also supported discriminatory legislation for them (Kettner 1978:315, 320). The issue of native-born free Negroes raised other questions. One concerned gradations of rank within citizenship status—did free Negroes have second-class status? Another question was whether the states could adopt a definition of citizenship that differed from that of the federal government. The latter question was particularly relevant to the acquisition of new territories. If a slave moved with his master to a territory or state where slavery was not legal, was he still a slave there? When he returned? These issues came to a head in the Dred Scott case. Dred Scott, a slave in Missouri, was taken by his master to a nonslave territory for a number of years. When Scott returned to Missouri, he sued for his freedom in the state courts. When he lost his case, he appealed to the federal courts, which would hear cases only when the litigants were “citizens of different states.” (During this period, the state still determined citizenship.) Thus, the first question was whether Dred Scott was a citizen of Missouri. The U.S. Supreme Court (with the majority of its justices from the South) decided that he was not a citizen because he was a Negro and that residence in a free state or territory did not result in a slave’s emancipation. The language used by the Court in this case was particularly inflammatory. One justice referred to Negroes as “natural-born subjects” and “not citizens.” In rendering the Court’s decision, Chief Justice Roger B. Taney stated: In the opinion of the court, the legislation and histories of the times, and the language used in the Declaration of Independence, show, that neither the class of persons who had been imported as slaves, nor their descendants, whether they had become free or not, were then acknowledged as a part of the people, nor intended to be included in the general words used in that memorable instrument. (Dred Scott v. John F. A. Sandford 60 U.S. 393, 10 [1856])

Justice Taney added that blacks had been “regarded as beings of an inferior order, and altogether unfit to associate with the white race, either in social or political relations.” Moreover, they were seen to be “so far inferior, that they had no rights which the white man was bound to respect.” The rationale for not granting citizenship rights was the prevailing condition of such people. Justice Taney noted, “Indeed when we

THE SHIFTING COLOR LINE

97

look to the condition of this race in the several States at the time, it is impossible to believe that these rights and privileges were intended to be extended to them” (Dred Scott v. John F. A. Sandford, 60 U.S. 393, 10, 13 [1856]). He concluded that “the negro might justly and lawfully be reduced to slavery for his benefit” (cited in Blaustein and Zangrando 1968:162). Some scholars have cited the Dred Scott decision as “the most farreaching judicial statement of the nineteenth century” with regard to race relations and as “the case that set the stage for the Civil War” (Blaustein and Zangrando 1968:146). The case clarified in 1857 the national status of both slaves and free Negroes. Justice Taney referred to the Constitution to justify his decision, which, he said, differentiated between “the citizen race, who formed and held the Government, and the African race, which they held in subjection and slavery, and governed at their own pleasure” (Dred Scott v. John F. A. Sandford, 60 U.S. 393, 17 [1856]). In essence, those of the African race, whether slave or free, were never intended to be citizens. Rather, the citizen race was the white race. Only after the Civil War and the passage of the Fourteenth Amendment was the principle of birthright citizenship finally affirmed for African Americans. The 1790 legislation was amended in 1870 to permit the naturalization of “persons of African nativity” and “persons of African descent” (Kettner 1978:331, 345). Native Americans had to wait until 1924 for legislation to make them citizens, and both groups are still struggling for equal rights. For Native Americans, the price of citizenship was the surrender of their tribal lands, tribal relationships, and tribal culture. For African Americans, citizenship was granted initially by local regulations or agreements, which were replaced much later by federal policies. In both cases, the federal policies differed from those for Europeans, and more restrictive local policies and needs often drove more restrictive federal policies. In both cases, the citizenship status of those in between—that is, taxed Indians and free people of color—often varied and was ambiguous. In some states, they could be citizens, but in others, they could not. In addition, the type of rights they had varied by state and changed over time in some states (Kettner 1978:301). The status and rights of taxed Indians and free blacks were also undoubtedly related to wealth, property ownership, intermarriage, phenotype, and acculturation. Finally, in both cases, these in-between groups disappeared.

98

THE SHIFTING COLOR LINE

Given that their own rights to equality were often challenged, it is interesting that both Native Americans and free people of color owned slaves, in the Americas and in Africa. Some scholars argue that it was a different type of slavery. The traditional servitude that existed in African feudal society before the onset of the European slave trade did include people with virtually no freedom, but with few exceptions, “servants were regarded as human beings and not chattel. They could marry, own property, maintain their family unity, freely worship their god, and sometimes they became military commanders and even rulers.” This kind of servitude is “not to be confused with American slavery in which the slave was regarded as chattel, and in some cases defined as property” (Harris 1972:73). Among the indigenous peoples of North America, slaves were often captives of war and could be Indian, white, or black. Only a few free people of color in the United States owned slaves, but a number of Indian nations did keep numerous slaves. Indeed, the 1860 census devoted a section to Indian slavery, and tables in its appendix listed the number of slaves held by “Indian tribes west of Arkansas, comprising the Choctaw, Cherokee, Creek, and Chickasaw nations” (Kennedy 1862:10–11). These were southeastern tribes that had been removed from slave-owning states and been resettled, mainly in Oklahoma. (The census acknowledged that these groups were but a “small portion of the Indian tribes within the territory of the United States” [U.S. Secretary of the Interior 1853b:11].) The census calculated that slaves formed about 12.5 percent of the total Indian population in these nations12 (U.S. Secretary of the Interior 1853b:11). Some scholars contend that slavery among Indians and free people of color differed from that among European-descended peoples. The Cherokee Nation, for example, had early been a “haven for escaped black men,” who often served as English-language “interpreters for full-blooded masters.” Some also taught the Cherokees how to cultivate the soil (Strickland 1975:79, 82). The Cherokees eventually promulgated laws similar to those of the states in which they resided. The laws tended to favor the Cherokees’ planter class, but they were “at such variance with the needs and expectations of the majority of the tribe that the laws were widely ignored” (Strickland 1975:83). Initially, the Cherokees’ regulations regarding slavery resembled “more closely [those for] tenant farmers or hired servants, with little restriction on private life but a clear separation between the red and black races.” As

THE SHIFTING COLOR LINE

99

“plantation agriculture began to emerge, the role of the slave began to conform more closely to that of blacks in the southern cotton kingdoms” (Strickland 1975:79–80). Legal restrictions on slaves were borrowed from those of Alabama and Georgia. However, the “evidence clearly demonstrates that most of these restrictions were ignored” and that agricultural crops were shared, slaves were allowed to keep guns and were educated, and it was not uncommon for them “to possess horses, cattle and swine” (Strickland 1975:81–82). There is mention of “only one minor slave uprising” (Strickland 1975:84). Slavery varied among the Indian nations, for example, some intermarried to a greater degree and had fewer slaves, and the Seminoles—who had also been “transplanted from slaveholding states”—had no slaves and intermarried with ex-slaves (Katz 1986; Kennedy 1862:11). The majority of free people of color had a personal interest in their slaves. For example, they might have been married to a slave; the slaves might have been the children of a free father; or they might have been “close friends who by law would have to leave the state if freed” (Fishel and Quarles 1970:128). In some instances, large numbers of slaves were owned just for economic benefit (Fishel and Quarles 1970; Franklin 1967:224 ff). Many free people of color also protested slavery and helped slaves escape through their benevolent societies, schools, churches, and the abolitionist movement (Aptheker 1968; Du Bois 1972:235–272; Fishel and Quarles 1970:128–132; Foner 1964; Rawick 1972:109–113; Rose 1965).

Asian Indians How individuals or groups are classified by their government is relatively unimportant if the rights of all members of the society are truly equal, regardless of race, color, ethnicity, class, or gender. It is only because these rights, practices, and privileges have not been equal in the United States that such classifications have become important.13 As we have seen, census categories have reflected, sustained, and, in some cases, established certain power relations because of the rights associated with being classified as white in the United States. The extent to which this has been the case is illustrated by the example of immigrants from Asia who, for 162 years could not become citizens (and therefore could not own land) in some states. The federal government’s 1790 naturalization law specified that only persons classified as free “white”

100

THE SHIFTING COLOR LINE

immigrants could become naturalized citizens (Kettner 1978:331, 345; Leonard 1992; Takaki 1994).14 The fluctuating racial classification of Asian Indians is an interesting example of both the extent to which racial definitions and classifications can change and the role of political factors in influencing racial classifications. In some censuses, Asian Indians were counted as “white” and in others as “other race.” In the 1910 census, for example, Asian Indians were counted and classified as “other race,” but a footnote explained that “pure blood hindus” were ethnically white and had been so declared in several naturalization cases, but, it continued, in the popular conception they were not seen as “white.” Consequently, they were included in the 1910 census with the “other races” category, along with the Chinese, Japanese, Filipinos, and others (Jensen 1988:252). A 1923 Supreme Court case involving Asian Indians used the same reasoning. Bhagat Singh Thind v. U.S. 261US204 concerned the question of whether Asian Indians were “white” and therefore eligible for citizenship. The Court concluded that although scientific and linguistic evidence indicated that Asian Indians were Caucasian, the common understanding of people in the United States was that “white” meant European and Caucasian, not just Caucasian. Accordingly, at least sixty-five Asian Indians were denaturalized between 1923 and 1927 (Haney López 1996:91). In effect, the 1923 Supreme Court’s decision legitimized the government’s refusal to accept scientific definitions of race and to opt instead for a definition of race that was more socially acceptable at the time. The “race” of Asian Indians could be defined in two ways: one was seen to be scientific, and the other was based on what it was believed “the common man” thought. It was the second one that counted. In making this decision, the Court reversed the position it had taken just a year earlier. In Ozawa v. United States (1922), a Japanese immigrant contended that since the color of his skin was white—indeed, whiter than that of many white persons—he should be classified as “white.” The Court then unanimously ruled that “the words ‘white person’ are synonymous with the words ‘a person of the Caucasian race’” (cited in Haney López 1996:85). Thind thus used this ruling to argue that since he was a Caucasian, he was therefore white and eligible for citizenship. However, the Supreme Court’s decision on Thind made

THE SHIFTING COLOR LINE

101

clear that “white” was what people believed it to be—or, as the Court argued, “what the common man thought” (Haney López 1996:107). In essence, race was socially determined. Subsequent census classifications of Asian Indians reflected the Court’s decision. The 1930 and 1940 censuses included a separate “Hindu” category (see table 4.3). Curiously, Asian Indians who were Muslim or Christian were placed with Hindus in the “Hindu” category. Their racial classification at the local level varied, depending on their skin coloring, the county, and the observer classifying them. For example, during this period, clerks issuing marriage licenses to Punjabis in California sometimes wrote “brown,” sometimes “black,” and sometimes “white” for the Punjabi grooms (Leonard 1992:68).15 Toward the middle of the twentieth century, as India and Pakistan moved toward independence, politics influenced racial classification. In 1945/46, after extensive lobbying by Asian Indians, legislation was passed enabling them to become citizens.16 It was after this time that their classification as “white” commenced.17 Indians were subsequently counted as white in the census until 1980, when a separate “Asian Indian” category was created. In 1990, Asian Indians became a subcategory under the generic “Asian and Pacific Islanders” (API) category. In the 2000 census, they are listed along with other groups from Asia or the Pacific Islands but without the pan-ethnic API label.18 Thus, Asian Indians have progressed from being an undefined racial category to being “other race”—that is, Caucasian but not white “in the common understanding” (or not European white)—to being “legally white,” to being listed as their own “Hindu” race category without a generic label or group, to being part of the “Asian and Pacific Islander” race group, to again being listed as a race along with other Asian and Pacific Islander groups but not under this generic label.

Hispanics The classification of Hispanics has also fluctuated in the U.S. census, not just because of “racial” classification changes, but also because the cultural criteria, such as language, surname, and “origin,” to determine Hispanicity have changed.19 As noted in chapter 4, the 1930 census created a “Mexican” category for the race question (see table 4.3). Thus, in 1930, first- and second-generation Mexicans were of the “Mexican race” unless they were determined by the (usually white) census

102

THE SHIFTING COLOR LINE

interviewer to be definitely white, Negro, Indian, Chinese, or Japanese. Those in the Mexican category were considered part of “other races” along with groups such as the Japanese, Chinese, and Native Americans. In 1940, the census dropped the Mexican category and stated that all Mexicans were to be reported as “white” unless they were determined by the census interviewer to be “definitely Indian or of other Nonwhite races” (U.S. Bureau of the Census 1943:3).20 So Mexicans moved from being Mexican unless determined otherwise in the 1930 census to being white unless determined otherwise in the 1940 census. This criterion, established in 1940, was also applied to other Hispanics who immigrated to the United States in greater numbers after World War II—for example, Puerto Ricans in the late 1940s and 1950s, Cubans during the 1960s, and Dominicans and Central and South Americans in the late 1960s and 1970s. Accordingly, in the 1960 census, the instructions for determining race or color by observation directed that “Puerto Ricans, Mexicans, or other persons of Latin descent would be classified as ‘white’ unless they were definitely Negro, Indian, or some other race” (U.S. Bureau of the Census, 1989:78). In the 1970 census, enumerators asked respondents to choose a category for race. If the respondents wrote in, for instance, “Mexican” or “Puerto Rican,” the enumerators moved them according to their appearance into the racial categories listed (Lee 1993). Consequently, before 1980, most Hispanics were classified as white.21 But in 1980 and 1990, when mail-back questionnaires were instituted, Hispanics were permitted to classify themselves, and they reported a variety of racial and ethnic groups. With regard to the changing cultural criteria used to define Hispanics, in 1940 the census used a linguistic definition to determine who was Hispanic, and “persons of Spanish mother tongue” were reported.22 In the 1950 and 1960 censuses, the language criterion was replaced by “persons of Spanish surname.” In the 1970 census, in response to pressure from the Hispanic community for a Hispanic self-identifier (Choldin 1986), a subgroup of individuals were asked “about their ‘origin,’” and respondents could choose among several Hispanic origins listed on the questionnaire (U.S. Bureau of the Census 1993c). (As the next chapter explains, political factors also played a role in the decision to include a Hispanic identifier in the 100 percent count of the 1980 census.) Thus, between 1940 and 1970, Hispanics were counted according to three different cultural criteria, linguistic (1940), surname (1950 and 1960), and origin (1970).

THE SHIFTING COLOR LINE

103

In sum, over time, both various cultural criteria and various racial classifications were used to classify Hispanics. In 1930, Mexicans were a “race” within the “other races” category unless the census interviewer determined they were white, black, or Native American. Between 1940 and 1970, Mexicans and other Latinos were “white” unless they clearly appeared to be Indian or Negro,23 and between 1980 and 2000, they could be “of any race” they chose. In the 2000 census, the format used to count Hispanics is essentially what it was in 1990, except that the question about whether or not a person is Hispanic comes before the race question. In contrast to the other groups discussed in this chapter, citizenship issues for Hispanics have been more a matter of defining citizenship than of securing it. Perhaps somewhat incongruously, citizenship was granted to many Spanish-speaking persons as a result of the treaties signed after the United States invaded Florida, the Southwest, and Puerto Rico. Many questions, however, have been raised about whether this citizenship by conquest was an equivalent or a second-class citizenship, whether legal repression occurred after conquest, and whether this citizenship included cultural citizenship, that is, the right to speak Spanish and maintain one’s culture (Acuña 1988; Cabranes 1979; Flores and Benmayor 1997:1–23; Hernández 1997). The legal case of Rodríguez, a “pure-blooded Mexican” who applied to become a naturalized citizen illustrates the ambivalence and tenuousness attached to this citizenship by conquest, particularly in regard to “color.” Although a Texas court granted Rodríguez’s request in 1897 to be granted citizenship because of the treaties’ existence, it also remarked that “if the strict scientific classification of the anthropologist should be adopted, he would probably not be classed as white” (cited in Haney López 1996:61). This decision and whether “a person of [Mexican] descent may be naturalized in the United States” were later questioned in the courts (Haney López 1996:242, n. 37). Thus, although the Texas court did allow a “pure-blooded Mexican” to naturalize, in rendering its judgment, it reinforced the more general rule that color was still a bar to citizenship for nonwhites such as the Chinese and Japanese. Although the census never tried to measure specific mixtures among Hispanics, the changing instructions to enumerators regarding how they were to classify Mexicans and the current census position that Hispanics can be of any race suggest that Hispanics are at least a mixed

104

THE SHIFTING COLOR LINE

lot. But are they a mixed lot in the same way that the United States as a whole is a mixed lot, or are they seen as mixing a lot? Horsman (1981:chaps. 11, 12, 13) argues the latter, maintaining that Americans saw Mexicans as less fit because they had intermarried so much with Indians and thus were not capable of governing the southwestern territories. Although such a perspective can be seen to justify the expansion of the United States into the Southwest, some academics believe that this perspective also influenced how all Latin Americans were viewed. For example, Hayes-Bautista and Chapa contended that with the promulgation of the Monroe Doctrine and the rise of Manifest Destiny, Latin Americans were racialized into a homogenous group (of Latinos) that transcended the boundaries of Latin American nations. In essence, with the annexation of other people and the incorporation of foreign territories, racial identification replaced national identification, and the “conquered race” was relegated to a lower social class level than that of the “conquering race.” With Latin Americans continually cast as persons belonging to a less advanced and different race, the confusion of race for nationality continued. Hayes-Bautista and Chapa believe that the general North American public assumes that the “race” of Latin Americans is a reality and that it is the antithesis of the civilized United States population (1987:62–63). Whether Latin Americans were racialized into a Latino group during the nineteenth century or later has not been resolved. What is clear is that political factors have been important to the definitions of both citizenship and racial classification in the United States. It is also clear that Mexicans and other Latinos have confounded, and continue to confound, the bipolar structure that evolved in the United States. In part, the reason is that they do not fit easily into the bipolar structure—nor in some cases do they wish to be—because of their varying phenotypes, mixture, and perspectives on race. Hispanics, perhaps more than other groups, best illustrate the permeability and shifting lines of the bipolar structure.

RACE IN REAL LIFE, THE ACADEMY, AND THE CENSUS Historically, there have been many shifts in racial classification (Anderson 1988; Forbes 1988; Lee 1993), even though the general impression is that the concept of “race” has been unequivocal and unchanging in the

THE SHIFTING COLOR LINE

105

United States. As a recent extensive review of this subject pointed out, there is a “widespread popular perspective that race is biologically determined and permanent and that ethnicity is culturally determined and equally permanent” (Edmonston, Goldstein, and Tamayo Lott 1996:18). Yet a primary perspective in the social sciences now views race and ethnicity as social constructions. Indeed, Almaguer holds that it has become “axiomatic in sociological research to view racial categories as sociohistorical constructs whose meanings vary widely over time and space” (1994:9).24 This contrast between the popular and the academic perspectives of race is apparent at a time when the significance of racial classification has shifted. In the past, nonwhite petitioners to the courts often argued—as did Plessy in Plessy v. Ferguson—that they should be classified as “white” so that they could be given the rights of whites. Thus, race and ethnic definitions were often ways of excluding individuals from equal membership in the society. More recently, defining groups has been a way of including them and ensuring that particular groups are not discriminated against. But whether race and ethnic classifications are used to include or to exclude groups, the basic bipolar structure— that of whites and other social races—has prevailed. Nonetheless, those in between have always been more dialectically engaged—individually and as groups—in contesting, resisting, rejecting, ignoring, transforming, or being transformed by census categories than is generally believed and than census documents might indicate. Indeed, these census-based historical analyses may project a smoother sense of history than what the lived experience has perhaps been. The reason is that in addition to being labeled by the census (and by others in more casual situations), individuals also identify themselves racially and ethnically for reasons of pride and to express group affiliation, and these self-classifications may differ in meaning as well as in actual terminology from those used by the census for both these in-between groups and other groups.

6 Race in the Americas

I N L AT I N A M E R I C A and the United States, Europeans, Africans,

Asians, and indigenous peoples mixed and produced “new people.”1 Few people, however, would deny that the population stew in the United States is quite different from that in Latin America. Indeed, a number of scholars have noted the difference between the processes of racial formation in the north and the south (Degler 1959; Denton and Massey 1989; Ginorio 1979; Harris et al. 1993; Petrullo 1947:16; PittRivers 1975; Rodríguez and Cordero-Guzmán 1992; Wade 1985; Wagley 1965). Snowden argues that the concept of “race” in Latin America is similar to that of the ancient peoples in the areas surrounding the Mediterranean (1983:97), and others contend that in Latin America, “race” is more like a “social race” or an ethnicity (Pitt-Rivers 1975; Wagley 1965). However, the concept of race in Latin America has been quite different from the ancient view, in that it also implies a “pigmentocracy,” a racist paradigm in which honor, status, and prestige are signaled by skin color and phenotype. The whiter one’s skin and the more European looking one is, the greater is one’s claim to honor and privilege. Conversely, the darker one’s skin is, the more closely associated one is with African and Amerindian peoples—that is, the conquered and the laborers.2 The recent literature has highlighted this difference and consequently stressed the similarities in this regard between racial formation in the United States and Latin America.

POLARITIES ALONG THE NORTH-SOUTH AXIS Latin America is a very large and extremely heterogeneous area. For a variety of reasons, studies of race by North Americans have usually focused on just a few regions—for example, Brazil, the Spanish 106

RACE IN THE AMERIC AS

107

Caribbean, and Mexico.3 Consequently, the following generalizations must be read with a number of provisos in mind. First, not all Latin American countries have been adequately researched. Second, much of what is reviewed here was written in English and thus does not cover the literature written in Spanish or Portuguese that is not well known in the United States. Third, even the Spanish and Portuguese studies may not accurately incorporate or even consider the views of the less educated or nonelite. Finally, at present, travel, communications, and the exchange of peoples and goods between Latin America and the United States are at all-time highs and are expected to increase in this era of global transformation. Because race is a social construct, it has been and will continue to be influenced by these changes. Table 6.1 indicates the broad differences found in the literature. The first of these four major differences between the north and south is the tendency in Latin America to see “race” as a social-racial construction and in the United States to see it as a genealogical concept. In the Spanish Caribbean and Latin America, ancestral “blood” is only one variable determining one’s race. Moreover, race is not necessarily passed down from generation to generation, as is implicit in a system based on hypodescent or genetic inheritance. In the Spanish Caribbean, the parents of a white child may be black or an intermediate shade. Accordingly, in the Caribbean and Latin America, phenotype is often viewed as an “individual marker,” whereas in the United States it is a group marker determining one’s reference group (Wright 1994). Of course, in almost all Latin American countries, certain phenotypes are associated with particular linguistic or social/cultural groups, with cultural types, and/or with stereotypes, for example, the tall Otavalo Andean Indians of Ecuador and the African-descended people of the Chocó in Colombia (see Arocha 1998 for an interesting analysis of the relationship of the latter to issues of inclusion in Colombia). A second, related, dimension is that race is not always based on just color. Other physical and social characteristics, such as facial features, hair texture, social class, dress, personality, education, linguistic identity, cultural modes of behavior, relation of the referent to the speaker, and context are important to “racial classification” (Rodríguez and Cordero-Guzmán 1992; Sanjek 1971:1128). Hence, a person who would be considered white in the Spanish Caribbean might be considered black or nonwhite in the United States because of his or her color. Race in the Caribbean and Latin America is highly dependent on context

108

RACE IN THE AMERIC AS

Table 6.1 General Differences in Racial Constructions United States

Latin America

Type of social construction

Genealogical-biologicalhypodescent

Social-racial

Categories

Few, discrete, mutually exclusive

Multitude, overlapping

Role of color

Basic variable

One of many variables

Fluidity over time

Some fluidity

Substantial fluidity

Nomenclature

Unstable for mixtures or nonwhites

More stable

and situation (Harris et al. 1993; Johnson et al. 1997; Rodriguez and Cordero-Guzmán 1992). Third, in many parts of Latin America, race is more openly reported as able to change over time and space.4 That is, in some countries, a person may be born “brown” but become “white” with upward mobility, whereas in the United States, race is more static and is often considered to be an ascribed characteristic. Reflecting this more fluid conception of race in the Spanish Caribbean and Latin America are a variety of racial terms, often overlapping and without clear demarcation. In Brazil, for example, an open-ended question about race in a survey can yield more than 140 categories of answers (Sanjek 1971). The various terms used to refer to racial types or categories indicate the different conceptions (and constructions) of race. Hence, the racial taxonomies differ, although in both the north and the south, white has generally been seen as preferable to or better than black because it was the color of those who conquered and colonized and were the “governing race,” as they were labeled in the 1850 U.S. census (U.S. Secretary of the Interior 1852:20). In each country of the Spanish-speaking Caribbean, a greater number of terms are consistently and commonly used—for example, moreno, indio, jabao, and trigueño—for what in the United States might be called “black” or “intermediates,” a term not often used here. Some of these terms (e.g., trigueño or moreno) also are ambiguous or have many meanings, referring at times to those regarded as white or black in the United States.5 In addition, a variety of terms are used to refer to those who are blancos (whites) or to their particular color, for example, blancusina/o

RACE IN THE AMERIC AS

109

(very white), cano/a (white, as in gray or white hair), rubia/o (blond), guera/o (blond), colorá/ado/a (reddish), and jincha/o (pale). In some groups, the very term blanco or blanquito (white) is also increasingly used to refer to removed, powerful, or upper-class persons, regardless of their color, thereby underscoring the relationship between perceived color and power. Finally, descriptive terms are used to refer to skin color that is not white-white, for example, piel canela (cinnamon skin), trigueño claro (light trigueño), and trigueño oscuro (dark trigueño). Alvar found eighty-two racial terms used throughout Latin America—and since the same term often has more than one meaning, he listed 240 definitions for them (1987:89–215). Many of these terms have been used for a long time and have different meanings in different countries. The following example from his work illustrates how complex these terms are. The term puchuelo was first cited by Father Morell in a work published in 1776, but he noted that it was coined much earlier. In Peru and Venezuela, puchuelo is defined as the result of a cross between a European and an ochavona. This cross is said to produce a person of raza totalmente blanca (of the totally white race). Puchuelo is also defined as the cross between a white person and a person who is cuarterona de mestizo (one-quarter mestizo). If the term is modified by de negro, “of blacks,” in Mexico the term means the child of a white and an ochavona negra (or octoroon) (Alvar 1987:185). Some people may legitimately object to works like Alvar’s that analyze and itemize minute differences in conceptions of race as trivializing the inherent brutality of slavery and racism. However, such works also demonstrate that both researchers and the lay public were aware of the magnitude of phenotypic diversity and the complexity and fluidity involved in creating such differences. In contrast, in the United States, despite early and regional or local variations—for example, in New Orleans—race is generally determined by perceived or imputed biological inheritance. The “rule of hypodescent,” according to which one drop of “black” blood makes a person “black,” has been applied most recently and most rigidly to African Americans.6 Because race and color are often used synonymously, African Americans are considered to be “black” regardless of their appearance or other factors. Similarly, Asians are considered “yellow” and Native Americans “red.” Although other social variables are often part of racial determination in the United States, for example, accent or speech or dress style, the basis is ancestry and color. Moreover, in the

110

RACE IN THE AMERIC AS

United States, persons born black remain black no matter what they achieve socially. Until recently, intermarriage between whites and nonwhites (other than blacks) has resulted in their children’s race being defined as partialized, as in “half-breeds” or “part Asians,” for a number of generations. Assimilation often has meant hyphenated American status for some groups, minoritization for other groups, and total Americanization for those who most resemble Europeans. Although African Americans have also developed a variety of terms to refer to color tones (Russell, Wilson, and Hall 1992) and some terms such as mulatto and half-breed have been used by governmental bodies in the past, the emphasis has been on constructing race categories as if they were “pure” (Lee 1993). Even in the African American community, individuals never become fully white, as they do in some Latino communities. Indeed, when a person of African American descent becomes white, it is because the individual is “passing,” that is, leaving the black community. Although some people in the African American community are seen as “white” by African Americans, whites, and others, to use this as a self-designation or category would be seen as denying their group, ancestry, or “true” identity. In contrast, in Latino communities, it is not uncommon to refer to individuals as Latino and white (in color), without any denial of ethnicity implied. At the same time, the terms white and Latino can be juxtaposed as two distinct cultural-racial groups, so to be one is not necessarily to be the other. In contrast, in the United States, individuals are rarely considered both black/African American and white (in color); they tend to be seen as mutually exclusive. It is only with the recent increase in intermarriage that the children of such unions have begun to use terms such as biracial and multiracial for themselves and that these terms have become common ways of describing individuals of “mixed” heritage. Those persons without immediate “mixed” ancestry have not generally been so described, although increasingly many are claiming all their ancestries. In essence, as noted in chapter 4, the racial taxonomy of the United States has reflected a small number of intermediate racial categories that have fluctuated (in both official and everyday use) over time, whereas in the Spanish Caribbean and other parts of Latin America, many intermediate and stable categories have persisted over time. As noted earlier, the extent to which these different constructions of race influence one another because of immigration to the United States,

RACE IN THE AMERIC AS

111

transnational migration movements, and increased communications between both hemispheres is not yet clear.

DIFFERENCES AND SIMILARITIES Although the differences between north and south pertain to racial constructions, they may also obscure the similarities. First, both Americas have histories of indigenous conquest, slavery, and immigration. Second, in both Americas, race has been constructed to reflect and support class and power relations. Each country in Latin America has developed its own racial constructions, but in all cases, they have tended to benefit those in power. The ideological and practical racial distinctions of the colonial structure of Latin America as a whole has favored the conquerors and colonizers. As Spickard noted, From the point of view of the dominant group, racial distinctions are a necessary tool of dominance. They serve to separate the subordinate people as Other. Putting simple, neat racial labels on dominated peoples—and creating negative myths about the moral qualities of those peoples—makes it easier for the dominators to ignore the individual humanity of their victims. (1992:19)

A racial hierarchy was and is still evident today in Spanish-speaking America, which has been reinforced in the Spanish-language media, particularly in the ever-popular television novelas or soap operas that air in both Latin America and the United States. In most novelas, the protagonists and major characters are usually played by northern European–looking actors, and the marginal and lower-status service roles, such as maids and chauffeurs, are given to darker-skinned, non-European actors (Subervi-Vélez et al. 1997:234–235).

Reasons for the Differences The reasons offered for these differences are too numerous to be explained fully here, so I will only summarize a few of them. One is that Spain’s contact with North Africa made the Spanish more tolerant of different color groups than the northern Europeans were. That is, Mediterranean peoples tended to see darker-skinned people as white or

112

RACE IN THE AMERIC AS

more like them than northern Europeans did. According to Forbes, Spaniards were used to a great variety of colors but did not associate them with a concept of separate “races.” Thus, the Spanish Mediterranean world used a variety of color terms and had an awareness of many gradations in human physical types and subscribed to the general view that “human types changed gradually and blended into one another.” According to Forbes, this view was typical not just of Spain but also of most observers before 1900 (Forbes 1988:268).7 Both Sanjek (1994) and Forbes (1988), however, contend that the Spanish way of viewing race shifted over time toward a more racialized view. Sanjek argues that despite the initially different views of Spain and northern Europe regarding color and race, by the late seventeenth century, all European countries looked down on both Africans and Native Americans and were reluctant to sanction intermarriage or to admit persons of mixed background to the full entitlements enjoyed by those of solely European ancestry (1994:1–17). “As the centuries of dispossession and enslavement of these peoples wore on, the ordinariness and economic utility of such treatment were accepted more and more” (p. 5). Spain adopted the Roman slave law codes, which were developed when a person of any race could be a slave. In this context, slavery was “an unfortunate accident that could befall any luckless one.” This conception of slavery as accidental and not racial “automatically endowed [African/black slaves] with the immunities contained in the ancient prescription” (Degler 1959:28). Consequently, the Spanish conceived of slaves and Indians as vassals or royal subjects and thus as having certain rights. This differed from the North American conception of slaves as property. This does not mean that the Spanish treatment of slaves was more benevolent but, rather, that it was sanctioned and conceived of differently. Indeed, Hoetink found in his study of the Caribbean that “there is no clear connection between the type of slavery practices, that is, whether ‘cruel’ or ‘mild,’ and the positions attained by free blacks or colored in the society” (1985:8). A third difference noted is the influence of the Spanish Catholic Church, which had a central role in the conquest of Latin America. In Latin America, it also promoted the conversion, baptism, and attendance of slaves at integrated religious services, whereas in the United States, the churches for blacks and whites were separate.8 For some, the role of the Catholic Church is seen to be analogous to that of the Span-

RACE IN THE AMERIC AS

113

ish legal code. That is, in theory it promoted a positive cultural attitude toward persons of color but in practice failed to carry it out (Denton and Massey 1989; Rodríguez and Cordero-Guzmán 1992). The economies of many Latin American countries were more mixed and less dependent on slavery. As a result, slavery was less important as an institution, and there were fewer slaves, both absolutely and proportionately. (Brazil and the Caribbean were the major exceptions.) This, together with the immigration of many Europeans and the substantial numbers in some countries of indigenous peoples, may have led to a conception of race that was fluid instead of dichotomous (Duany 1985; Hoetink 1985). Duany (1985) illustrated the significance of economic development in racial formation by comparing the history of race relations in nineteenth-century Cuba and Puerto Rico. He contends that race relations in Cuba, which had a plantation economy, distinguished more rigidly between the white planters and the nonwhite plantation workers or slaves. In Puerto Rico, however, the absence of an extensive plantation economy created a large intermediate group of free colored persons, which facilitated social-racial mobility (Hoetink 1985:14; Williams 1984).9 The gender ratio was also quite different in early Spanish America. As Rodríguez and Cordero-Guzmán found, “the greater migration of European women and families to North America as compared with Latin America—where men predominated and European women were scarce—may also have influenced the relations between races and the consequent conceptions of race that evolved” (1992:527). Indigenous and African women may more often have been mates of European men. The children of such unions were, in some cases, recognized and educated, and they contributed to the formation of the criollo class (Burkett 1978). In addition, the development of a large, free, African-descended class may also have produced greater differentiation. As appendix D explains, the number and proportion of “free people of color” in the United States was never very large, peaking at less than 15 percent of all African Americans.

RACE IN EARLY SPANISH AMERICA Undoubtedly all these explanations contributed to the distinctive constructions of race in Latin America, as compared with those in the

114

RACE IN THE AMERIC AS

United States. When and how did these differences begin? Forbes suggests that the Latin American tendency to view individuals in a “progression of colors” reflected a more Mediterranean worldview in which numerous shades were associated with Europeans, Africans, and so forth (1988:268). This approach, he argues, was brought by the Spaniards to the Americas and predated the extensive mixing that took place there. Even though his assumption has not been extensively researched, if we examine the records of early Spanish America, we find a society with many, fluid, and overlapping race categories, determined by various physical, social, and economic variables. Early Spanish colonial records use color terms to describe Europeans, a practice not followed in the British colonies. Evidently, race was determined by a variety of factors, such as reputation, legal process, choice, acculturation, and calidad (quality) and many different terms were used to describe people physically.10 In addition, and again somewhat in contrast to the British colonies, mixture (or mestisaje) was recognized in the writings and paintings of the time. Finally, the early Spanish American literature refers to lower-class Spaniards (or whites) as a caste, suggesting that castes were not based just on color. An example of how substantial the color variations were even among those classified as “Spaniards” is a 1677 roster of colonists bound for New Mexico. It lists as Spaniards those individuals described as having “fair skin,” others as having “dark complexions,” and still others listed as being “mestizos” or “dark.” Spaniards also were classified by national origin, and so there were European Spaniards, Mexican Spaniards, and Spanish Indians (Gutiérrez 1991:197). Color was apparently an adjective that could be applied to persons of different nationalorigin groups. These references contrast with the later practices in the British colonies, whose European-descended population seldom referred to degrees or modifications of color. Greene and Harrington’s 1966 compilation of population estimates in the British colonies before 1790 reflects this convention and indicates the common use of terms such as people, souls, inhabitants, or whites—with the last two terms sometimes modified, as in European or white inhabitants—to count populations.11 Color terms were not used to describe Europeans, nor were those of “mixed race” generally reported as such. With some minor exceptions, the basic divisions were—as in the first decennial census—whites, slaves (or blacks), and Indians, by tribe.12

RACE IN THE AMERIC AS

115

Settlers in early Spanish America also emphasized racial classification according to reputation or social acceptance (Gutiérrez 1991). For example, in his analysis of sixteenth- to eighteenth-century northern New Spain (now New Mexico), Gutiérrez discovered that a Juan Sandoval was listed “by appearance of white racial status.” Another man was described as “mestizo, according to reputation,” and still another was promoted to lieutenant because “he is known as a white man” (1991:198). Categories such as español, mestizo, and mulato were sometimes used interchangeably with descriptions of physical color—like blanco (white), pardo (roughly brown or gray), and prieto (black)—even though color had no real legal definition. Consequently, a person could be described as español mestizo (or a mestizo Spaniard). Comments that a person “appeared to be,” “was reputed to be,” or “was known to be” of a certain race also indicated that classification depended somewhat on social perception and acceptance. This in turn suggests that racial mixing and “passing” may have been prevalent on this remote fringe of northern New Spain (Gutiérrez 1991:198). Except at the extreme ends of the color scale, however, there was no direct correspondence “between race and actual physical color” (Gutiérrez 1991:197). Other scholars writing about early Spanish America have noted this malleability of “race” (Carroll 1991). MacLeod, for example, writing about Central America in the seventeenth century, notes that many Indians kept their “race” but became non-Indian through dress and the adoption of language and cultural customs—becoming culturally mestizos or Ladinos (1973:308, 383). Katzew reports that in Mexico during the eighteenth century, a number of newly wealthy families who were descendants of Indians and slaves purchased certificates of legal “whiteness” (called gracias al sacar, which is translated literally today as “thanks to be taken out” but which may have had a different meaning at the time) (1996:12). At the same time, others manipulated their racial identities for other purposes, as when mestizos identified themselves culturally with Indians and adopted Indian hairstyles, language, and the like in order to avoid paying tribute. Likewise, blacks adopted Indian and Spanish customs. Furthermore, according to Gutiérrez, in this early period in Spanish America, a person’s status was based not solely on race but also on calidad (1991:202 ff). Calidad and color were often closely related. Spaniards prized their honor, and many were persons of calidad because they lived

116

RACE IN THE AMERIC AS

among (and were above) genízaros (detribalized Indians), who had been dishonored by their enslavement, and among the Pueblo Indians, who had been conquered. Gutiérrez pointed out that much of what it meant to be “honorable” was a projection of what it meant to be a free, landholding citizen of legitimate white ancestry. Conversely, those without honor were slaves, outcasts, or Indians. Nonetheless, the concept of honor was not necessarily rooted in racial-physical difference but was, rather, “a complex measure of social status based on one’s religion, ethnicity, race, occupation, ancestry and authority over land” (Gutiérrez 1991:206). Consequently, the resulting social order tended to favor those most akin to the European conquerors yet still allowed non-Europeans to improve their position. This order differed somewhat from the system that evolved in the United States, which based social and racial status strongly, if not solely, on biological descent or appearance. Thus, calidad was not a concept that had an exact equivalent in the United States’ racial formation process. Moreover, even though concepts similar to calidad undoubtedly could be found in the United States, for example, “god-fearing” and “honest,” these were not generally used in racial classifications. As a minimum, it required “whiteness” to be a citizen, so people were first members of a race and then were god-fearing, honest, or whatever. Another difference is that in North America, mixtures were described only biologically, whereas in Latin and Central America, cultural descriptors were never completely abandoned (Forbes 1988). Latin Americans distinguished first between those of legitimate birth raised by the Spanish and those raised by Native Americans. This distinction recognized that cultural factors or socialization influenced the identity of the “hybrids.” Early North American colonists followed this same path, but by the 1800s, and especially after the Civil War, “greater and greater emphasis was placed upon wholly biological or ‘racial’ categorization and differentiation in North America.” (Forbes 1988:269; see also Davis 1992; Logan Alexander 1991; Williamson 1984) A number of scholars have noted the early use of various and changing terms to physically describe “mixes” of people as well as the conquering Spaniards (Alvar 1987; Forbes 1988; O’Crouley 1972; Rodríguez, R. 1991:24). O’Crouley, for example, in his description of New Spain in 1774, lists and defines several common terms used to describe mixtures.13 Gutiérrez’s 1991 analysis of marriage and baptismal records in sixteenth- to eighteenth-century northern Mexico uncovered various

RACE IN THE AMERIC AS

117

terms used to describe brides, grooms, babies, and parishioners. This contrasts with the United States, where official documents or common parlance did not use many terms to refer to mixtures of people. Many of these terms are still commonly used in Latin America, for example, mestizo. Others, however, are no longer employed, for example, castizo or genizaro.14 MacLeod also noted the diverse categories in early Central America, for example, Spanish and Ladinos, mestizos, blacks, mulattoes, Indios, English (1973:228). Finally, the casta paintings commissioned during the 1700s by wealthy Spaniards and criollos illustrate the numerous terms used to describe the different “mixes” (Katzew 1996).15 Although in the United States, terms referring to “mixtures” were few, in some parts of the Spanish-speaking Americas, numerous terms were used to refer to different kinds of mixes, for example, children of indigenous and African parents (Forbes 1988:130). Gutiérrez (1991), examining colonial records in Mexico dating between 1690 and 1846, also found a variety of terms used for different mixes. For example, coyote and lobo (wolf) were widely used to refer to the “half-breed” children of Indian slave women born in captivity.16 Color quebrado was a broad term that did not specify the nature or extent of racial mixture but, rather, meant “broken color” or “half-breed.” The precise degree of racial mixture was not indicated. Hence, in contrast to the United States, the blood quantum was not ranked, although like the United States, the sense that mixture diminished “purity” was present. Gutiérrez (1991) noted the various classifications of Indians and Spaniards in northern New Spain. If an Indian spoke Spanish, he or she was known as an indio ladino. Indios were Pueblo Indians who lived in their own towns and were economically and politically independent. A genízaro was a detribalized Indian who lived in a Spanish town. This term is no longer used, but at the time it appeared as a column heading in census counts. According to Gutiérrez, the status of genízaros was similar to that of domestics or slaves (1991:150). With regard to early views of the influence of “non-Spanish blood” on future generations, the picture is quite complex. On the one hand, Spanish colonial records indicate that (black) race was not necessarily transmitted from one generation to another (Forbes 1988:121). Although the records might classify a mother as negra (black), they might also classify her daughter as lora (brown)17 or might not indicate color at all, which usually meant white. Gutiérrez’s 1991 analysis of early records also indicates that race was not necessarily inherited. In the

118

RACE IN THE AMERIC AS

early period, marriage statistics seldom gave race, and mixing appeared to be common. This contrasts with how the children of African slaves or free people of color were classified in the United States. Writers on the early Spanish American period also have not found evidence of a strict view of hypodescent, as was prominent in the north. Sometimes, however, racist views were clearly articulated, and the biological and cultural supremacy of the Spanish and Europeans was often explicitly stated or assumed. Also, some commentators of the time distinguished between the influence of “black blood” and that of “Indian blood” on “Spanish blood.” For example, in Pedro Alonso O’Crouley’s description of eighteenth-century New Spain, he, as a Spanish merchant from Cadiz, accepts without question the superiority of Spaniards and refers without hesitation to the more indelible stigma of mixture with Negroes as opposed to Indians (1972:20 ff). Another Spanish merchant writing at about the same time affirms this view and stresses even more the supremacy of the white pole to the black (Katzew 1996:10–11). The question, of course, is whether such texts reflected the prevailing customs, the views of the elite class, or the observations and prejudices of these particular upper-class Spanish observers. The casta paintings offer a similarly complicated view. What is unusual about these paintings is that they depict the complexity of intermixing. Indeed, this appears to be the paintings’ purpose. Thus, we see a variety of mixtures, from children who appear to be white but whose parents are described as not white, to those who appear to be black but whose parents appear to be white. The paintings’ depiction and explication of mixture are not found in the same degree in the north, where mixing also occurred. It also is curious that in these paintings, the Spanish who intermarry or interbreed are depicted as being of both genders. Similarly, the Indians or blacks are not always the female slaves or Indian princesses commonly found in U.S. literature or folklore. Finally, in the early eighteenth century, each of these mixed persons is portrayed as wealthy, whereas the later casta paintings show them in less affluent circumstances. Analysts today see the projections of wealth as reflecting the insecurities of the criollos and Spanish elite in the American colonies who were attempting to convince Europeans and themselves of the stability and prosperity in the New World. Although we do not know whether these paintings were more ideal than real or whether these terms were commonly used at that time, they nonetheless present

RACE IN THE AMERIC AS

119

a striking visual contrast to how mixture was projected (or not) in the north.18 The casta paintings also show that the results of intermixture differed depending on whether a black or an Indian was mixing with a Spaniard. As O’Crouley pointed out, a white and an Indian could have a “white” child, and thus the Indian stigma would disappear “because it is held as systematic that a Spaniard and an Indian produce a mestizo; a mestizo and a Spaniard, a castizo; and a castizo and a Spaniard, a Spaniard.” Thus, if intermarriage continued with the Spanish, the Spaniard would return. If, however, “Spanish stock is mixed with Indian several times over, there is also a return to Indian” (O’Crouley 1972:20). Some of the paintings depict this process. Intermarriages with blacks also are described and depicted in the casta paintings as producing white-appearing children, for example, albinos and moriscos, but not as producing a “Spaniard.” The paintings also depict a return to “black,” that is, torno atras (a return backward). Thus, as the two Spanish merchants of the time maintained, white blood was not “redeemable” (i.e., recoverable) with blacks. Blacks and Indians could return to their original types, but Spaniards could return to Spaniards only if the mixture had been with Indians. It is likely, however, that as the mixtures continued to mix with other mixtures, the utility or relevance of these classifications or theories diminished.19 After the third generation (when everyone has eight grandparents), it was difficult to categorize the racial mixture definitively. In fact, the difficulty of classifying these mixtures was already evident in at least two of the terms used in the casta paintings, tente en el aire (hold yourself in midair) and no te entiendo (I don’t understand you). All of this impeded “the creation of a fixed system of classification and representation” (Katzew 1996:10). Consequently, even though Spanish commentators may have employed a version of hypodescent at the time, very likely only those concerned about maintaining the “purity” of their European ancestry and their “blood” claim to upper-class status or power would have worried about such distinctions (Gutiérrez 1991:292). Indeed, the casta paintings’ racial classifications may have been attempts to clarify and stabilize what was an increasingly fluid society whose social and racial boundaries were uncertain (Katzew 1996). Suggesting the uncertainty of such boundaries are references in the literature of that time to “castes,” which included whites and Spaniards.

120

RACE IN THE AMERIC AS

MacLeod noted, for example, the concerns in seventeenth-century Central America about the growing number of castes, which included free Negroes, mulattoes, mestizos, and déclassé white vagabonds (1973:141–42, 192, 211–213, 235, italics added).20 Siguenza y Góngoro, writing in the seventeenth century about the deplorable drinking habits of the Indians and other groups of the Mexican population, described these groups as “composed of Indians, of Blacks both locally born and of different nations in Africa, chinos, mulattos, moriscos, mestizos, zambaigos, lobos, and even Spaniards . . . who are the worst among such a vile mob (cited in Katzew 1996:12, italics in original). Finally, MacLachlan and Rodríguez wrote that although the colonials of New Spain (Mexico today) assumed that there was an ethnic hierarchy (and historians accepted this assumption), the notarial records indicate otherwise (1980:223). The records state that by the seventeenth century, wealth and status were not confined just to whites and that the poor included all racial groups. In fact, some European immigrants remained quite poor, and mestizos of means occasionally had Spanishborn servants. In essence, from early on, Latin America more freely acknowledged the influence of culture, class, and other social factors in determining race. As a consequence, many Latin American countries developed overlapping racial categories on a continuum from light to dark, from European to indigenous, or from white to black, rather than discrete, mutually exclusive racial categories. Scholars of early Spanish America explain that the racial system in place then had many of the same features found today. It was not bipolar (i.e., it had more than two categories), and it was—as the preceding examples illustrate—apparently fluid, dependent on social perception, and quite complex. We also see in place by 1744 the use of intermediate terms such as pardo, a polite description of individuals known to be mulattoes, and the use of the term moreno for those who were negros (O’Crouley 1972). As noted earlier, these terms are still common in many Latin American countries today. At the same time, pigmentation was emphasized, and implicit and sometimes explicit racism dominated the determination of one’s social status. Biological descent was only one variable entering the racial calculus, but as in the United States, it may have been more stringently applied to those with African ancestors than to those with Indian ancestors. During the early Spanish colonial period, other characteristics such as class, physical type, social networks, or hairstyle and dress also

RACE IN THE AMERIC AS

121

were important indicators of social status and ethnic identity (Gutiérrez 1991:205 ff). Yet despite what may have been great fluidity in early colonial Spanish America, the race order relied on the existence of oppressed “others” in order to define “the included.”

Racial Configurations A comparison of differences and similarities between north and south may obscure the differences among the different countries in Latin America itself. As we saw in chapter 2, the various countries’ racial and ethnic categories differ, as do the concepts and their definitions (Rout 1976:185–312). Some countries do not ask about race and ethnicity, and sometimes these change over time (Almey, Pryor, and White 1992; Bates et al. 1995:433–435; Lee 1993; Martin, DeMaio, and Campanelli 1990; Miller 1991; Statistics Canada and U.S. Bureau of the Census 1993).21 That is, race has been conceived differently in each country in Latin America because each has had a different history (Scott 1995:56). For the countries in Central and South America, this variability depends on their history of settlement as well as political considerations and policies concerning the collection (or noncollection) of race and ethnic data. Almey, Pryor, and White (1992) examined how, during a fortyyear period in the twentieth century, the censuses of fifty-one countries classified their populations with regard to race and ethnicity. They found that the census forms of those countries whose settlers had a predominantly European cultural background (e.g., Argentina, Chile, Costa Rica, and Uruguay) had no questions on race/ethnicity (Almey, Pryor, and White 1992:7). But the censuses of Central American and Andean countries usually did include questions on ethnicity and race. Moreover, the censuses of countries in the Americas that had slave and plantation economies generally did ask about ethnicity and race but used different terms for racial categories. For example, Cuba and Brazil used color terms to distinguish groups such as black, pardo (similar to brown), white, and yellow.22 The British West Indies, where Chinese and East Indian indentured labor was an important part of the country’s history, listed separate categories for these groups. Similarly, those countries where Syrians, Lebanese, and other Arabs immigrated in substantial numbers had separate categories for these groups. Some countries included a category for “mixed,” and over time, others

122

RACE IN THE AMERIC AS

replaced “black” with “African.” In the 1980 censuses for Belize, Barbados, and the Dominican Republic, the Portuguese were put into a category different from that for whites (Almey, Pryor, and White 1992). Countries such as Bolivia, Guatemala, and Panama collected data on their large indigenous and nonindigenous populations. But according to Almey, Pryor, and White (1992), the data on indigenous populations were not consistent. In general, such information was collected only sporadically, and the categories changed over time. Moreover, those nations with small and rapidly disappearing indigenous populations, for example, Brazil and Chile, did not attempt to use the national census to identify and count them (Almey, Pryor, and White 1992:8). Thornton noted that Belize counted Mayans and Caribs, those who are mixed Native American and black, and that some countries counted separately those who speak another language (1987:222). Settlement history is not the only variable determining how questions of race and ethnicity are asked (Almey, Pryor, and White 1992) or how racial ideologies are expressed. Government policies, conditions of the nation-state, balance of power, and external views of race also influence how countries come to see or measure race. These factors—because they vary by country—also lead to a multiplicity of racial ideologies and policies (Graham 1990). For example, scholars have concluded that between the late nineteenth and early twentieth centuries, Cuba and Argentina developed racial ideologies that openly or explicitly emphasized and glorified whiteness and whitening (Andrews 1980; Helg 1990; Rout 1976:193 ff); Brazil celebrated its “Racial Paradise or Racial Myth”;23 Puerto Rico talked about a sense of cryptomelanism (Serreno 1945; for a more extensive review of the literature on race in Puerto Rico, see Rodríguez 1996); Mexico celebrated “mestizaje and indigenismo” (Gutiérrez 1991; Knight 1990); and Venezuela created a complacent café con leche society (Wright 1990).24 These racial ideologies also influenced cultural self-definitions, preserved power relationships, shaped policies, and controlled the oppressed.25 Although these characteristics apply to particular countries, regional exceptions within countries and overlaps between countries can be found as well. Moreover, a particular ideology may also be found in another country; for example, El Salvador may also have a café con leche society. In addition, racial formation is constantly evolving, so the characteristics of one period may change.26 Moreover, these characteristics are drawn from analyses of writings on race, which often reflect the

RACE IN THE AMERIC AS

123

class biases of upper-class intellectuals and political elites in these countries. Popular views of race may therefore be quite different from these descriptions, although one could argue that a country’s racial ideology eventually affects everyone. The process is circular. “Race” is created by cultural practices; it is articulated in a particular way by writers; once constituted, it speaks to and about culture; and it influences the way people see themselves (Nobles 1995:128). But these countries also have much in common. All have a legacy of slavery and the oppression of non-European peoples, although some countries were more dependent on slaves than others were. They all also responded to the development of racialist theories in the nineteenth century and to the shift in the balance of power during this period, with its attendant competition for political and economic dominance. Some countries responded similarly, for example, Cuba and Argentina (Helg 1990), others differently, for example, Mexico (Knight 1990). But the racialist theories and the popular thinking of the time touched them all. This thinking, in turn, influenced each country’s policies, especially with regard to immigration and national conceptions of race and identity. In effect, all the countries emphasized the desirability of whiteness and European immigration policies (Graham 1990). In the same way that all Latin American countries were affected in the past by Eurocentrism and racism, they continue to be affected in the present by new movements, for example, Afrocentrism, Latinismo, and worldwide indigenous rights movements.

Racial Legacies Despite its different historical constructions, “race” in the various Latin American countries has been more fluid and has led to the creation of more categories than the binary division adopted in the United States. Moreover, race in these countries has not been solely determined by genetic inheritance but has been much affected by other variables such as class, phenotype, language, and degree of assimilation. When viewed through the U.S. racial lens, this view of race is more akin to ethnicity, culture, or national origin, but from the Latin American perspective, it is simply raza, “race.” Recently, this view was manifested in the responses of many Hispanics to the U.S. census’s questions about race. As noted earlier, at least 40 percent of all Hispanics in the United States responded that they

124

RACE IN THE AMERIC AS

were “other race,” and many of them wrote in a Latino referent, for example, their national origin or a cultural or ethnic label. Moreover, some of the findings of studies conducted by the census echo many of the differences between the U.S. and Latin American views of race just described.27 For example, and as will be discussed in the next chapter, several ethnographers found that Hispanics, especially recent immigrants, generally do not view race as a dichotomous variable but, rather, as a continuum (Bracken and de Bango 1992; Rodríguez and Hagan 1991; Romero 1992). Respondents also conceptualize race “as a constellation of national origin, skin color and culture” (Bates et al. 1994:109). Finally, it was not only particular Hispanic groups that had difficulty with the race and Hispanic-origin questions, but all Hispanics, regardless of national origin (de la Puente 1993:37–38). A study by Kissam, Herrera, and Nakamoto (1993) found that many Hispanics understand “race” to be national origin, nationality, ethnicity, or culture and that for many Hispanics, “race” and “ethnic group” are closely related.28 According to this study, while the meanings of each of the three terms race, Hispanic origin, and ethnic group varied extensively from respondent to respondent, in contrast, the concept of national origin was, for most, well understood. The study’s focus groups confirmed the researchers’ in-depth interview findings (p. xi) and concluded that for the Hispanics they interviewed, race was closely tied to national origin and cultural identity and only weakly to phenotype or genotype (p. 32). The study also provides some intriguing discoveries about education and race. It found that education influenced the respondents’ answers, with those with more education responding in the way the census anticipated, especially to the race question. The study stated that in the cognitive interview, race “provided one of the most frustrating barriers for low-literate Hispanic respondents” (Kissam, Herrera, and Nakamoto 1993:22). Many of the less well educated respondents scanned the first three racial-group terms, blanco (white), negro (black), and “indio” (Indian). They then eliminated each and in some cases wrote in a Hispanic term. Because of the overlap for many Hispanics among race, ethnicity, and national origin, even well-educated Hispanics “expressed annoyance when they realized that their preferred racial group term was part of the amorphous group of ‘otro grupo racial’ (other race)” (p. 23). The authors decided that for the respondents, choosing the “other race” category conveyed a “disturbing and sometimes in-

RACE IN THE AMERIC AS

125

sulting connotation to Hispanic immigrants about their role in ethnic interactions in the United States” (p. 23). According to the authors, the respondents felt that they had no “label” of their own and only a generic “other race” at the end of the list. The researchers concluded that the message to the Hispanic respondents was that they were less important than other races and that from the perspective of most of the study participants, the “census’s implicit conceptual framework . . . [was] considered inadequate” (p. x).29 These findings prompt several questions. How does the way that a group is seen in the United States compare with the way it sees itself? How long does it take for groups to understand their placement or classification on government forms? How might they resist, such as when people who have always seen themselves as “Cuban” or “Peruvian” are told that they are “Hispanic” and that this is the same as “Argentinean”? This is a repetition of the earlier immigrant experience in which Sicilians became “Italians” and Cherokees became “Indians.” But today, these findings involve for Latinos, at least, an apparent change in definitions of race. Many Latinos may come to the United States believing that race, ethnicity, and hispanidad (Hispanicism) all are related (because this is how “race” is socially constructed in Latin America), but they soon learn that for U.S. census purposes, these are supposedly distinct concepts; that contrary to what the ancients and other cultures believed, a race group is not the same as an ethnic group. Moreover, race is primarily biological or color based. The following anecdote from one of the census studies illustrates another dimension of the racialization process. It suggests that racial perceptions change over time in the United States. A focus group predominantly made up of immigrant Hispanic women was confused about the racial question. A more acculturated Hispanic woman in the group told the others, “What they want you to put down is ‘white’” (conversation with de la Puente, January 6, 1993). This conflict between the U.S. census’s articulation of “race” and the respondents’ views will be discussed in more detail in the next chapter.

PART III

RACE AND THE CENSUS

7 The “Other Race” Option

HISPANICS AND THE U.S. CENSUS As chapter 1 pointed out, the 1980 census represented a historical break. For the first time in its two-hundred-year history, the census asked the respondents to indicate their race and also if they were of Hispanic or Spanish origin (see figure 1.1 in chapter 1 which shows these questions).1 Both questions produced surprising results, which provide a focus for a larger discussion about who reported they were “other race,” what contributed to their responses, and what issues their responses raised. More broadly, these results and the explanations of them provide dramatic evidence of the fluidity and social construction of race, as well as of the persistence and pull of the United States’ bipolar racial structure. From the perspective of the mainstream press, one of the most significant results of the 1980 and 1990 censuses was that the number of people who checked categories that were “other than white” was much higher than in the 1970 census. By 1990, one of every four Americans identified himself or herself as either Hispanic or not white, that is, of non-European descent or race. Time magazine used this subject as a cover story, in which it coined the colorful (and subsequently much used) metaphor “the browning of America” (Henry 1990). Implicit—but largely unnoticed—in this phrase was the assumption that all Hispanics, Asians, Pacific Islanders, African Americans, or any others who checked any category other than “white” were “brown.” The reverberations from this interpretation were many. For example, some corporations changed their marketing plans to be more inclusive, and the definition of “American” was scrutinized even more carefully than it had been during the 1960s. Although these revelations about the United States’ changing “racial-ethnic” composition received 129

130

THE “OTHER RACE” OPTION

considerable media attention, practically no notice was taken of Hispanics’ responses to the race question. In 1980, however, Hispanics’ responses to the question differed radically from those of the non-Hispanic population, in that a substantial number of Hispanics (7.5 million, or 40% of all Hispanics) chose the “other race” option (Denton and Massey 1989; Martin, DeMaio, and Campanelli 1990; Rodríguez 1989; Tienda and Ortiz 1986). Many of them wrote in an explanation in the box asking for race. That is, after checking the “other race” box, they specified that they were Dominican, Honduran, Boricua, or some other cultural or national-origin term. (Another 57.7 percent of Hispanics indicated that they were white, and 4.6 percent indicated that they were black, Asian, or Native American.)2 In contrast, less than 3 percent of the non-Hispanic population in all states said they were “other race” (Rodríguez 1989). Despite changes in the 1990 census’s race question, this pattern was repeated. A comparison of the race questions used in 1980 and 1990 shows that some of the changes were calculated to reduce the likelihood that respondents would confuse race with national origin.3 For example, the label race was included in the race question and was also added to the category of “other.” Indeed, in contrast to the 1980 question, in which the term race did not appear at all, in 1990 it appeared five times in the question (see figure 1.2). Nevertheless, the proportion of Hispanics who replied that they were “other race” did not decline; rather, it increased by almost 3 percent in the 1990 census. In fact, the “other race” category was the second largest racial category (after “Asian and Pacific Islanders”), increasing by 45.1 percent between 1980 and 1990 (Rodríguez 1991b:A14; U.S. General Accounting Office 1993). Moreover, the overwhelming majority (97.5%) of those in the “other race” group were Latino (U.S. General Accounting Office 1993:26).

“OTHER RACE” RESPONSES The literature initially offered two explanations for why so many Hispanics chose the “other race” option. One was that they had misunderstood or had had difficulty with the question. Although most articles did not specifically refer to Hispanics’ “misunderstanding,” many did refer to the “difficulty” that the race item posed or stated that Hispanic

THE “OTHER RACE” OPTION

131

respondents had “difficulty” responding to questions about race (McKenney and Bennett 1994:21; McKenney and Cresce 1993:173–222; McKenney et al. 1993; U.S. House Committee 1994f:9). One article referred to how Hispanics had “inappropriately identified their race as ‘other’ (for example, some white Hispanics reported their race as other . . .)” (Buehler et al. 1989:458). This article went on to say that the Census Bureau “corrected” the classification of race for many persons and created “race-corrected data” that were later used in their calculations (Buehler et al. 1989:458). The second explanation was that the “other race” response represented “mixed-race” individuals, that is, mulattoes and mestizos. In essence, Hispanics who said they were “other race” and identified their national origin were seen to be either “mixed up” or “mixed.”4 Neither interpretation questioned the validity of the race question. Further research showed that neither of these explanations was entirely satisfactory. Indeed, later analyses showed that it was not just Hispanics who had “difficulty” with the race question. Census reinterview studies (i.e., studies that later interviewed those persons who filled out a census form) concluded that in addition to both foreign-born and native-born Hispanics, many other foreign-born persons had difficulty reporting in the race items (McKenney and Bennett 1994:22). In addition, other studies indicated that these responses resulted not from a misunderstanding of the question but from a different understanding of race. As noted earlier, Kissam, Herrera, and Nakamoto (1993) observed that many Hispanics chose the “other race” option because they viewed race as culture, national origin, ethnicity, or nationality or a combination of these and skin color (Bates et al. 1994:109; Rodríguez 1992, 1994a; Rodríguez and Cordero-Guzmán 1992; Rodríguez et al. 1991). These different understandings of race also were evident in many colloquial expressions, such as Mexican Americans’ references to themselves as raza and the common use of the term raza to refer to culture and national origin, for example, la raza dominicana (the Dominican race), la raza colombiana (the Colombian race), and la raza italiana (the Italian race). There were other indications that Hispanics had different understandings of race. For example, in the 1990 census race question, twothirds of those who did not specify their race did write in their Hispanic ethnicity (U.S. Office of Management and Budget 1995:44689). That is, they were not a race, they were Hispanic. Hispanics’ more cultural or ethnic view of race was also revealed in the census’s methodological

132

THE “OTHER RACE” OPTION

and ethnographic studies investigating the reporting and undercount issues (see, e.g., Bracken and de Bango 1992; Romero 1992; Rodríguez and Hagan 1991; and Elias-Olivares and Farr 1991). As the Interagency Committee for the Review of the Racial and Ethnic Standards of the Office of Management and Budget later noted, “Hispanics tend to see race as a continuum and use cultural frames of reference when discussing race” (U.S. Office of Management and Budget 1997a:36909). The second assumption made about Hispanics’ divergent pattern of racial responses was that those who reported they were “other” were mestizo or mulatto.5 This, too, has not been fully supported in subsequent research, although it does account for some responses. Given a facsimile of the 1980 census question on race, Latino respondents were asked first to answer the question and then to explain their choice of category.6 Of those who chose “other race,” only 11.5 percent referred to biological “race” or mixture (Rodríguez 1992). Another 15.4 percent gave both physical and cultural reasons. The majority (63%) stated that they had chosen the “other race” option because “this was their culture,” and/or they referred to their family, birthplace, socialization, or political perspective. Representative answers were “Because that’s what my parents and family are”; “I have always known this is my culture”; “That’s where my roots are”; and “Although I was born in the United States, my parents are Dominican.” Thus, it does not appear that the majority of those Hispanics who say they are “other” do so because they see themselves as “mixed” (Rodríguez 1992). The assumption that “other race” represents mixed race may be true for some Latinos, but it cannot be assumed that it represents the view of all.

THE GOVERNMENT’S RESPONSE Soon after the 1980 census’s results were published, few argued that the “other race” response represented a different understanding of race. Indeed, there was relatively little reaction after 1980 to the “other race” response, especially compared with the reactions ten years later after the 1990 census, which led to public hearings and extensive research on this issue.7 This delayed reaction was reflected in a high-level summary by the U.S. Office of Management and Budget (1995), which stated that “a high percentage of Hispanics selected ‘other race’ in the 1990 decennial race question” (U.S. Office of Management and Budget 1995:44690, ital-

THE “OTHER RACE” OPTION

133

ics added). It did not say, however, that this response was first given in the 1980 census. It was only later that the U.S. General Accounting Office noted that “in both the 1980 and 1990 Censuses, the Bureau found that Hispanics had difficulty classifying themselves by race” (1997:8).8 Would the situation have been handled or interpreted differently if the issue had not been race? In other words, if 40 percent of a group (or more than 9 million people in 1990) had responded to any other question on the census (e.g., marital status, income) in a way that differed significantly from expectation or from the rest of answers, would the assumption have been that the group had difficulty with or misunderstood the question? More likely, such a result would have meant that the question may not have been relevant to the group or that the response was not a misunderstanding but perhaps a different understanding of the question. The following example illustrates this point. Consider gender identity. If 40 percent of any group of people stated that they were neither male nor female, it seems likely that (1) such an answer would have been noticed before the same result was obtained again ten years later and (2) a search would have been undertaken to determine why the people answered in this way before concluding they did not understand the question or that all of them were hermaphrodites. But after the 1980 census results were known, alternative explanations were not seriously explored at the time. Only when the results were repeated in the 1990 census—despite attempts to discourage the “other race” response—were other explanations sought. Despite the substantial changes in thinking about these issues, the notion that Hispanics were confused lingered. For example, in hearings held in 1993 to reassess racial and ethnic standards, many of the expert witnesses and federal representatives accepted the conclusion that Hispanics were confused by or had difficulty classifying themselves in terms of the race categories (see, e.g., U.S. House Committee 1994d:234, 1994j:75; 1994m:55, 1994r:95. These hearings will be the focus of chapter 8.). Moreover, as recently as 1997, according to the U.S. General Accounting Office, the Census Bureau’s evaluations in 1980 and 1990 found that “Hispanics had difficulty classifying themselves by race” and that this difficulty led to inconsistent reporting (1997:8). The U.S. Office of Management and Budget’s interim report, while acknowledging that “the [race] question may not be operating as intended,” still stated that “some research supports the public comments that some respondents are confused about how to respond to separate race and

134

THE “OTHER RACE” OPTION

Hispanic origin items” (1995:44679). It then cited the high proportion of Hispanics who marked “other” in the race question9 and reminded readers that in the census reinterview studies, many Hispanics changed their classification when questioned by census personnel, again suggesting confusion. (The report did not note the possible role of the census interviewer in influencing this shift to another means of classifying race.) In 1999, the results of cognitive interviews conducted by the census led the Office of Management and Budget to decide that “there was confusion regarding the separation of Hispanic or Latino origin from race” (U.S. Office of Management and Budget 1999:10).

REASONS FOR CHOOSING THE “OTHER RACE” CATEGORY Many Hispanics chose “other race” because they saw that their national origin, ethnicity, and so forth were different from the other choices on the census, that they were not the same as or just “white,” “black,” “American Indian,” or “Asian and Pacific Islander.” But context—often referred to as “external or methods effects”—also influenced how Hispanics responded to the race question.10 For example, whether the interviewer was Anglo or Hispanic or whether the respondent answered the questionnaire in private, over the phone, or in person might have affected the answer. The structure of the question, that is, whether it was open-ended or closed, and the options offered also influenced the responses. In regard to questions about race, some groups are unwavering. For example, regardless of who asks the question or how the question is asked, whites always say they are white, and African Americans always say they are black (U.S. Office of Management and Budget 1995:44675).11 But Hispanics are different, and their answers to questions of race often vary considerably according to context.12 Many studies refer to this variability to as inconsistency because respondents do not consistently give the same answer (U.S. General Accounting Office 1993:26). The following section reviews the research on how Hispanics’ responses to questions of race are affected by (1) who asks and who answers the question, (2) the format of the question, and (3) the context in which the question is asked.

THE “OTHER RACE” OPTION

135

Who Asks and Who Answers the Question The question of who determines “race”—that is, whether it is determined by the person being questioned or by someone else—is important to determining Hispanics’ racial classification. Three examples are illustrative. The introduction of self-identification in the census caused the proportion of “white” Hispanics to drop from 93.3 percent in the 1970 census to 57.7 percent in 1980. At the same time, the proportion of “other race” Hispanics rose from only 1 percent in the 1970 census (when census personnel determined racial classification) to 40 percent in 1980. The proportion of “black” Hispanics remained the same in both years (Rodríguez 1991c:65, 81). Second, in the 1990 census, 43 percent of Hispanics were “other race,” 52 percent were white, 3 percent were black, 1 was percent API, and less than 1 percent was American Indian (del Pinal 1994:4). But in March of the next year, when the Current Population Survey (CPS) was taken, Hispanics were 96 percent white and 1.5 percent other race (del Pinal 1994:2; U.S. Bureau of the Census 1992:3) (see table 7.1). This discrepancy in racial classification occurred because in the CPS data, a census interviewer determined racial classification and Hispanic origin, and the “other race” category was not even on the form. (The “other race” category was used only when respondents refused to be placed into those specified on the form.) Consequently, in this case, the use of census interviewers to classify Hispanics resulted in significantly more “white” Hispanics and significantly fewer Hispanics in the “other race” category. (The proportions classified as black, Native American, or Asian and Pacific Islander also decreased.) Third, in the census’s Content Reinterview Study, personal interviews were conducted with those who had earlier submitted their information on the decennial census form. In a study of those who reported that they were “other race” in the 1980 census, only 10 percent were similarly classified when reinterviewed by census personnel (McKenney, Fernández, and Masamura 1985). Martin, DeMaio, and Campanelli cited as a possible cause for this inconsistency “interviewer behavior in the reinterview study” (1990:554). Although race was to be self-reported by respondents using a flashcard listing the race categories, they stated that the interviewers might have changed the “other race” responses to “white” for respondents who “looked white.” Or

136

THE “OTHER RACE” OPTION

Table 7.1 Hispanics by Race in Current Population Survey and 1990 Census CPS a (%)

1990 Census (%)

White Black American Indian Asian and Pacific Islander Other race Total

95.7 2.2 0.6 0.3 1.5 100.3

51.7 3.4 0.7 1.4 42.7 99.9

Total population

21,437

22,354

a

March 1991 Current Population Survey.

Sources: Jorge del Pinal, “Social Science Principles: Forming Race-Ethnic Categories for Policy Analysis,” paper presented at the Workshop on Race and Ethnicity Classification, 1994, p. 4. 1990 census data from 5% PUMS (Public Use Micro Sample) sample.

maybe Hispanic respondents answered differently in a personal interview (probably conducted by a white interviewer) than they did on the census questionnaire. These examples show the distinction between “self-determined” and “imputed” race. In the case of Hispanics, how people see themselves and how they are identified by others (imputed race) may be quite different. Consequently, what Latinos say they “are” in standard U.S. “racial” terms is not necessarily what they are perceived to be by others. For example, even though a Latino may write on a census form that he or she is “white,” the same person may be viewed by landlords, employers, or institutional service personnel as “black.” Conversely, a Hispanic who has always thought of himself or herself as dark might be considered by others as “white.” This is called perceptual dissonance (Rodríguez 1974, 1992), and these differences between how Latinos classify themselves and how they are identified by interviewers using the same categories have been found even when the interviewers have been Latinos as well (Falcon 1995; Ginorio 1979; Ginorio and Berry 1972; Martínez 1988; Rodríguez 1974; Rodríguez and Cordero-Guzmán 1992; Tumin and Feldman 1961). Other studies examining ancestry, self-classification as Hispanic, and interviewer identification have found that self-identified Hispanics are usually labeled as white by interviewers (see Drury, Moy, and Poe 1980; Hahn, Truman, and Barker 1996). In funeral homes, Poe and colleagues (1993) found that Hispanics were misclassified as non-Hispanic

THE “OTHER RACE” OPTION

137

on 19 percent of death certificates. Nationally, more than half of all Hispanic infants who died within a year (between 1983 and 1985) were classified as either “white” or “black” on their death certificates, as were 20 percent of Mexican, 48 percent of Puerto Rican, and 67 percent of Cuban infants (U.S. Office of Management and Budget 1997a:36910). This may seem morbidly amusing, but it has serious practical implications for calculating infant mortality rates, for example. Other health surveys have found similar discrepancies between self-classification as “Hispanic” and interviewer identification as “white” or “black” (see Lindan et al. 1990; Massey 1980).

The Format of the Question The format of the question also has an impact on how Hispanics respond. For example, the mere presence of an Anglo interviewer was found to influence responses. On the March 1980 Current Population Survey, Hispanics identified themselves overwhelmingly as “white” to a census interviewer who presented them with four non-Hispanic choices (Chevan 1990). One month later, however, when Hispanics were filling out the census form in the privacy of their own homes and were offered an “other race” alternative, 40 percent chose the “other” option (and wrote in a Latino referent). Thus, in the course of one month, the proportion of white Hispanics fell from 97 percent to 55.6 percent. How the question is structured or phrased also affects how Hispanics respond. As Chevan (1990) noted, when Latinos are faced with rigid categories that do not include either a Hispanic or an “other” category, most Hispanics choose “white.” This was discovered to be the case in the 1990 census reinterview studies. Many of those Latinos who had said on the census form that they were “other race” shifted themselves into the “white” category when interviewed later by census personnel (McKenney et al. 1993). Rodríguez and Cordero-Guzmán (1992) found that if an open-ended question were used to ask Puerto Ricans their “race,”13 a wide variety of responses were elicited, but few said they were white (11.1%) or black (1.6%). The majority (57.5%) responded with ethnic descriptors, for example, “Puerto Rican,” “Spanish,” or “Latino.” Part of the question’s context is the presence of other cultural groups in the census’s race question, for example, Chinese and Japanese. Some

138

THE “OTHER RACE” OPTION

have argued that this may have prompted Hispanics to respond “other” and write in a Latino referent (Lowry 1982; Tienda and Ortiz 1986). This is quite plausible, for the appearance of these groups on the census form might activate a sense of race more akin to that developed in Latin America, that is, race as cultural or social (U.S. Bureau of the Census 1992:3; Wagley 1965). The presence of such cultural groups on the census form became part of the context to which Latinos responded, and this induced them to respond culturally as well. This is consistent with the view that for Hispanics, racial self-classification is very dependent on context. In this case, the question included other cultural groups like theirs, and so they responded to the question by also identifying themselves culturally.14 The format—where the question is placed on the census form and how the question is asked—also influences the responses. The format and sequencing of the questions and the presence of other cultural groups in the question were the main reasons offered that Hispanics checked the “other race” category. In other words, because the format of the question did not include the word race, the respondents might have mistakenly assumed that the question was asking about ethnicity. In addition, because the question about race preceded the question about Hispanic origin, Hispanics might have assumed that the race question was asking about their “Hispanicity” or Hispanic origin. Consequently, in 1990 the word race was inserted into the race question numerous times, and in 1997, it was decided to place the Hispanic question before the race question in the 2000 census because government research showed that “Hispanics appear less confused by the race question and do not select the ‘Other’ race category as often” when this is done (U.S. Office of Management and Budget 1997a:36940). Even though formatting issues did influence the “other race” responses, they did not account for all the other race responses. As we have seen, when the term race was reinserted into the question in 1990, the proportion of Hispanics saying that they were “other” rose. Moreover, when Martin, DeMaio, and Campanelli (1990) reversed the sequence of the race and Hispanic-origin questions in experimental tests (the Hispanic question was asked first), the percentage of Hispanics born in the United States who reported “other race” dropped. Foreignborn Hispanics, however, continued to report that they were “other.” More recent experiments in which the Hispanic and race items were reversed resulted in fewer persons reporting they were “other,”

THE “OTHER RACE” OPTION

139

but the category of “other race” was not eliminated (Bates et al. 1994; McKenney et al. 1993; U.S. Office of Management and Budget 1995:44679). The most recent and largest government surveys to date obtained substantially the same findings (U.S. Bureau of the Census 1996a, 1997; U.S. Office of Management and Budget 1997a:36912). As Bates and colleagues (1995:452, 455) concluded, “Other remained the preferred race for a large minority.” These studies and their results will be discussed again in chapter 8.

The Context in Which the Question Is Asked The context in which the question is asked also influences responses. From the respondents’ perspective, different contexts may have particular consequences, as illustrated by the Puerto Rican woman who commented at a seminar on race: “The only time I respond that I am ‘white’ on a questionnaire is when I’m applying for a mortgage or a loan.” Other situational factors affect how Latinos and other groups respond to questions about both race and ethnicity. Johnson and colleagues, for example, found that 40 percent of their mixed-race and Hispanic respondents changed the way they reported their racial/ethnic background depending on the context, social situation, options on application forms, or “perceived advantages in applying for scholarships, loans, school admissions, housing and employment” (1997:15). Changes in self-awareness and identification also were responsible for changes in reported identity. Hispanics with two Hispanic parents were “much less likely (12.5%) to indicate ever having identified themselves differently” (p. 15). Esbach and Gomez (1998) found similar contextual shifts among Hispanic youth, with their identifying more consistently as Hispanic in urban areas and less consistently in areas with few Hispanics and also among English monolinguals.

WHO CHOOSES THE “OTHER RACE” RESPONSE? Based on the presumption that the process of Americanization will produce Americans with similar racial views, we would expect that those Hispanics reporting that they are “other race” would be more likely to have been in the United States for the least amount of time and to be foreign born, have limited English skills and education, and low levels of

140

THE “OTHER RACE” OPTION

acculturation. This is not always the case, however. The relationships between these variables and how Latinos racially classify themselves are complicated. Bates and colleagues’ multivariate analysis did not find “conclusive evidence” that educational level, knowledge of English, foreign or U.S. birthplace, or level of acculturation (i.e., whether or not respondents were immigrants) were consistently associated with choosing an “other race” response (1995:452–454). Moreover, my analysis of the 1990 census data shows a similarly complex picture. The following section examines the largest Hispanic-origin groups (those with substantial proportions of U.S.-born individuals over the age of 18) with regard to each of the preceding variables.15

Those Least Educated Do those who choose the “other race” category tend to be less educated? Yes, but many of those with more education also check “other race.” As figure 7.1 indicates, the proportion of those who say they are “other race” does decline for all the major Hispanic-origin groups as educational attainment rises, but each group has a different slope and profile.16 In addition, self-classification as “other” remains substantial in all groups, regardless of age or educational attainment.17 It is also considerably larger than the 2 percent or less of non-Latinos who choose the “other race” category. In part, the higher numbers of Mexicans and Puerto Ricans indicating they are “other race” is associated with the fact that they are younger as a group. With greater age and education, Mexicans and Puerto Ricans report they are “white” more often and “other race” less often. For example, in the 1990 census data, more than 73 percent of college-educated Mexicans (aged 45 to 90) reported they were “white,” compared with only 41 percent of young Mexicans (aged 18 to 26) who had an eighth-grade education or less. The respective figures for Puerto Ricans were 70 percent and 44 percent. Correspondingly, fewer of this same older, educated group of Mexicans reported they were “other race” (22%), while more (58%) of the younger, less educated group chose this option. The most notable finding in this analysis was that the tendency to self-classify as “other race” persisted for substantial numbers of all Hispanic groups, despite increasing age and education. Even in the oldest and most educated group (aged 45 to 90), for which classification as

THE “OTHER RACE” OPTION

141

FIG. 7.1. Self-Classification as “Other Race” and Educational Attainment. 1990

Public Use Micro Sample (PUMS) 5% sample.

“other” showed the most precipitous decline as educational attainment rose, 22 percent of Mexicans and 25 percent of Puerto Ricans with a graduate school education still reported that they were “other.” In addition, although the cell sizes were small and additional controls were necessary, there was an intriguing rise in “other race” reporting and fall in “white” classification for the youngest and most educated group, that is, those under age thirty-five with some graduate education. These results are consistent with my earlier work (Rodríguez 1989, 1990, 1991a) on Puerto Ricans in New York in the 1980 census. They suggest that Hispanics’ tendency to choose the “other race” option is not necessarily the consequence of low educational attainment.18 The association between higher education and more frequent selfclassification as “white” raises questions about causation. The data indicate that higher education does not always mean a greater likelihood of classification as white, but the trend is in this direction. Are those with more education also those seen as white and/or those who see themselves as white? Or is it that the very process of higher education— in both Latin America and the United States—induces a change in racial

142

THE “OTHER RACE” OPTION

classification, so that individuals come to see and to identify themselves as white?

Immigrants or U.S. Born? Given that the foreign born tend to self-classify themselves differently than do the U.S. born (Martin, DeMaio, and Campanelli 1990; McKenny and Bennett 1994:22), we would expect more foreign-born Hispanics to report they are “other race” and more native-born Hispanics to classify themselves in traditional U.S. racial terms, that is, as “white,” “black,” or whatever. This is the case, but what is striking is that there are still relatively high proportions of U.S.-born Hispanics who choose the “other race” category, especially Mexicans and Puerto Ricans. In the three Hispanic-origin groups, with large proportions of U.S.born individuals over the age of eighteen—that is, Mexicans, Puerto Ricans, and the “other Spanish/Hispanic” (OSH) group—more U.S.-born Latinos classified themselves as “white,” “black,” or “API/NAI” than did their foreign-born counterparts.19 Solid proportions of the U.S.-born persons in these groups, however, still classified themselves as “other race,” for example, 41.62 percent of U.S.-born Mexicans and 42.66 percent of Puerto Ricans born in the states. Figure 7.2 shows this distribution for the largest group, Mexicans.20 De la Garza and colleagues obtained similar results from their sample of 2,817 Latinos in the United States. They found, for example, that 53 percent of foreign-born Mexicans and 44 percent of native-born Mexicans chose the “other” category and supplied a Latino referent (1992:22–23). Finally, my analysis (Rodríguez 1989, 1990) of earlier 1980 census data on Puerto Ricans in New York City found that 48 percent of both those born in Puerto Rico and those born in the states chose the “other race” category.21 In sum, it appears that although a majority of U.S.-born Latinos choose traditional U.S. race categories, many also choose other race.

Those Who Speak Spanish at Home Knowledge of the Spanish language is important to racial classification, for it conveys terms, concepts, and perspectives that do not exist

THE “OTHER RACE” OPTION

143

FIG. 7.2. Racial Self-Classification of Mexicans, 18 and Over, U.S. Born and For-

eign Born, 1990. API includes Asian and Pacific Islanders; NAI includes Native Americans. 1990 Public Use Micro Sample (PUMS) 5% sample.

in English. For example, Spanish-language terms for “intermediate” racial types like trigueño or moreno are incorporated and influence one’s worldview. This may, in turn, make Latinos more likely to report that they are “other race,” for they view “racial” identities through different lenses. The 1990 PUMS (Public Use Micro Sample) data do seem to support the hypothesis that those who speak only English at home are more likely to classify themselves in standard U.S. race terms, that is, as “white,” “black,” or API/NAI.22 This was true for almost all the Hispanic-origin groups examined. For example, the percentage of Mexicans who “speak only English at home” and classify themselves as “other” (37.3%) was considerably smaller than those who speak Spanish at home (50.5%). For Puerto Ricans, the pattern was the same, with the respective figures being 27.8 percent and 50.3 percent.23 Consequently, within these admittedly smaller “English-only” Latino groups, more persons classified themselves in traditional U.S. race terms, as “white,” “black,” or “API/ NAI.”24 These results are consistent with earlier work that found a higher percentage of those who spoke only English at home reporting

144

THE “OTHER RACE” OPTION

that they were white or black (Rodríguez 1989, 1990). The relationship of language to racial classification clearly is complex. As Bates and colleagues noted, although a knowledge of English was not by itself consistently associated with choosing the “other race” response, speaking only English at home did seem to influence the choice of more traditional U.S. race terms. Therefore, speaking Spanish at home may also be associated with classifying oneself as other race, but more research is needed in this area (Bates et al. 1995:452–454). The extent to which the Spanish language is retained and its racial terminology and constructions are applied is, in turn, affected by a Latino’s exposure to primary and secondary language environments. For example, socialization (in a Spanish-language environment) may be more important than birthplace to determining the racial self-classification of Hispanics. McKenney and colleagues’ finding (1993) that speaking another language at home is associated with “inconsistency”—that is, with changing one’s racial classification, regardless of one’s proficiency in English—also suggests that exposure to Spanish-language environments is important to retaining intermediate and alternative views of race. The amount of travel to one’s country of origin may also affect language retention, assimilation, and racial self-classification. The circular migration and transnational migrations and communities associated with many Hispanic groups, for example, Puerto Ricans, Dominicans, Colombians, and Mexicans, are important in this regard. Other variables may be significant. Rodríguez and CorderoGuzmán (1992), for example, found that age, education, and the respondents’ perception of North Americans’ racial perception of them were strongly related to Puerto Ricans’ racial self-classification.25 More research is needed to understand the role of these and other variables in determining racial self-classification and identity. For example, does living in areas with a high proportion of Hispanics influence how Hispanics racially classify themselves? What is the effect on Latinos’ selfclassification of living near large (or small) proportions of other groups?26 How does national origin, phenotype, or family color constellation—for example, being the darkest sibling in a family—affect racial self-classification? In particular, we need to investigate why many Latinos born in the United States continue to classify themselves as “other” and to write in a Latino referent.

THE “OTHER RACE” OPTION

145

Those Who Have Been Here the Longest The research on Latinos who have lived in the United States for a long time is very limited, and the few data we do have—although often qualitatively rich and provocative—are preliminary and only raise more questions. The assumption is if race is socially constructed, then “persons migrating from one country to another are likely to encounter an official schema for classifying origin, race or ethnicity which is quite foreign to them” (Bates et al. 1995:435). The question is whether the way that immigrants report their race changes as their time in the United States increases. The following studies used different samples, time periods, and methodologies. But because of these differences and the complexity of the question, the findings must be interpreted cautiously. Controlling for age, I found (Rodríguez 1989, 1990) that older, mainland-born Puerto Ricans in New York City classified themselves as black and white more often than did younger, mainland-born Puerto Ricans. I obtained the same result for Puerto Ricans born in Puerto Rico but now living in New York, that the older they were, the more often they classified themselves as white or black. However, Rodríguez and Cordero-Guzmán (1992), controlling for age, education, and how the Puerto Rican respondents thought they were “racially classified by North Americans,” found that “length of time in the United States” was only moderately related to self-classification as white rather than other race. Taking an ethnographic approach, Duany examined Dominican migration to the United States and Puerto Rico and found support for the restructuring of “cultural conceptions of racial identity,” particularly regarding migration to the United States (1998b:148). He also found that transnationalism, that is, “back-and-forth” travel patterns, eroded hegemonic discourses in race and ethnicity in both the sending and receiving countries (Duany 1998b) and contributed to redefinitions of identity (Duany 2000). Bates and colleagues (1995) found that more recent immigrants to the United States reported that they were “other race” more often than earlier immigrants did, but these results were not statistically significant. Latinos who had lived longer in the United States were also less affected by the reversal of the race and Hispanic-origin questions and

146

THE “OTHER RACE” OPTION

tended to choose the white and black categories (Bates et al. 1995).27 In essence, reversing the sequence of the race and Hispanic-origin questions appears to have affected the racial self-classification of Latinos born in the United States, but not of those born abroad. This is consistent with the hypothesis that greater exposure to the United States increases the tendency to choose the white or black category. If we compare the racial classification pattern of Hispanic immigrants who arrived in the United States before 1980 with those who arrived after 1980, the pattern of racial classification varies considerably by group. With regard to Mexicans, Puerto Ricans,28 Colombians, and Salvadorans, there is little difference in the racial configuration before and after 1980 (see table 7.2). At the same time, there were substantial changes among Cubans,29 Dominicans,30 and Panamanians.31 Other groups also show some, but relatively trivial, changes.32 It is tempting to conclude that these observations of racial classification patterns reflect an alteration (or the lack thereof) in the racial classification of immigrants the longer they live in the United States. It also is possible, however, that the changes represent a shift in the nature of the migrations before and after 1980 or that changes in the respective countries are contributing to the differences in racial self-classification patterns, for example, changes in the way in which “race” is addressed by these countries’ political leaders. Taken as a whole, the results of these studies and of more journalistic writings (see, e.g., Escobar 1999), suggest that the amount of time spent in the United States may not be sufficient by itself to determine racial self-classification. Just as important may be which years were spent in the United States or in the country of origin. The childhood years? Young adult years? Teenage years? Also important is the respondents’ subjective assessment of their time spent in the United States or their country of origin. This, in turn, may be influenced by other variables, such as relative socioeconomic status, discrimination, aspirations, and phenotype. Finally, whether immigrants who have arrived more recently have a different racial-classification pattern than do those who arrived earlier varies by country of origin. That is, the nature and timing of immigration streams affect the racial classification patterns of Hispanic groups. The clearest example of this among Latino groups are the Cuban migrations before Castro and soon after Castro and the Mariel boat lift (García 1996; Pedraza-Bailey 1985; Portes and Bach 1985).

THE “OTHER RACE” OPTION

147

Table 7.2 Percentage Differences, by Racial Classification and Hispanic-Origin Group, of Immigrants Arriving before and after 1980

Mexican Puerto Rican Ecuadoran Colombian Guatemalan Salvadoran Other Latin American Other Spanish/Hispanic Cuban Dominican Panamanian Non-Hispanic Total

White (%)

Black (%)

-1.98 -1.06 -3.37 +1.58 -3.49 -0.32 +4.02 -9.64 -7.81 -0.28 +0.54 -36.69 -23.77

+0.31 -0.19 +0.14 +0.82 -0.74 +0.2 +2.2 +2.44 +2.6 +5.6 -22.62 +5.06 +2.34

API/NAI (%) Other (%)

+0.09 +0.27 +0.49 -0.24 -0.35 +0.35 +0.39 -1.63 +0.15 +0.19 +2.2 +31.29 +14.59

+1.76 -1.13 +2.73 +1.01 +2.39 -0.23 -5.85 +8.83 +5.06 -5.52 -19.88 +0.33 +6.85

Total

1,733,947 251,486 58,421 121,534 120,658 264,228 322,604 150,093 174,328 153,155 264,228 3,902,646 7,272,937

Figures indicate the percentage difference between pre-1980 and post-1980 immigrants. For example, the percentage of Mexicans reporting that they were white was lower for the post-1980 immigrants (43.37%) than for the pre-1980 immigrants (45.35%). Source: 1990 (Public Use Micro Sample) 1% sample.

ALTERNATIVE VIEWS OR DENIAL We also need to understand better whether the phrasing of the question about racial classification may frame the response. For example, as noted earlier, an open question on race produced a majority of responses that referred to what is considered in the United States as “ethnicity” but generated a small fraction (12.7%) of responses that conformed to traditional U.S. race categories, that is, white and black (Rodríguez and Cordero-Guzmán 1992). A closed question produced greater numbers of respondents classifying themselves as white or black, but the same proportion indicating they were “other.” In this closed question, however, fewer of those who said they were “other” (20%) used ethnic referents, and the majority (54.5%) used physical referents; for example, they said they were intermediate or trigueño (a wheat-colored individual). If we think of the questions as representing contexts to which individuals respond, did the open-ended question allow the Latinos in this study to express their own view of race, and did the closed question restrict their response? Even though the open-ended question in the 1992 Rodríguez and Cordero-Guzmán study asked specifically about race, it

148

THE “OTHER RACE” OPTION

yielded several responses that ignored physical attributes, suggesting that the respondents saw their “race” through their cultural frame of reference. The closed question and the presence of categories for white and black, however, introduced a more North American racial context in which the respondents answered in more physical-racial terms about themselves. But we do not know to what extent these responses may have been influenced by the presence of a Latino interviewer. From another perspective, we might argue that Latinos’ choice of an ethnic descriptor, as opposed to a racial descriptor, reflects the disinclination of many Hispanics to identify as black and that many more Latinos would be classified or identified as black by others than this study or the census figures indicate. According to Rodríguez and Cordero-Guzmán (1992), when the question was closed and the respondents were asked whether they considered themselves to be white, black, or other, the number who said they were white increased substantially, from 11.1 percent to 38.8 percent, and those who said they were black increased slightly, from 1.6 percent to 5.1 percent. The small proportion who said they were black raises the question of whether they had an aversion to classifying themselves as black or whether they really believed that they were simply not identified as black. We also need to understand better why so few in the study responded “white” in the first open-ended question. When Rodríguez and Cordero-Guzmán (1992) asked the respondents how they thought North Americans viewed them, the proportion answering “black” doubled, from 5.1 to 11.9 percent, and the proportion assuming they would be seen as “white” fell from 38.8 to 30.5 percent. This suggests that more in the group thought they would be seen as “darker” when viewed through North American eyes. But the proportion who thought they would be seen as “other” stayed about the same. What changed was how they defined “other.” The number of “intermediate” or trigueño terms that had been chosen by 30.6 percent of the group dropped, which is to be expected, since such terms or concepts are not common in English. The largest percentage of the group that chose the “other” category (28.5%) thought that they would be seen as “other, Spanish” or “other, Puerto Rican,” and another 11.5 percent did not specify the kind of “other.” This shift in responses does indicate coexisting dual racial contexts and the respondents’ awareness of them. If there are dual contexts in which the ways of viewing race (or understanding the question about race) differ and if many Hispanics see

THE “OTHER RACE” OPTION

149

race as culture, national origin, and so forth, do they distinguish between the concept of race and that of culture? In the United States, these two concepts are distinct. In one small study, in which the respondents were asked to define race and culture (Rodríguez 1992), many saw race as inseparable and indistinguishable from culture, and others defined race in geographic terms (as “where they came from”). Still others recognized an intellectual distinction between race and culture but ignored it in their everyday lives or when describing their own identity or themselves. Some respondents saw race as “independent of culture” and responded to questions of race and ethnicity in the way in which the census expected.33 McKay and de la Puente’s analysis (1995) of cognitive interviews contained conceptual questions about race and ethnicity. But they decided that their respondents found the questions too difficult. Their results thus differed from those of my 1992 study, which might have been because of different methodologies and samples. McKay and de la Puente’s 1995 study used seventy-four respondents from various racial and ethnic backgrounds who had been recruited by community organizations, and my 1992 study used fifty-eight Latinos predominantly from the Northeast. My questions were “How do you define race?” and “How do you define culture?” and McKay and de la Puente’s were “Please tell me what you think is the most important characteristic that defines race,” and “Do you think there is any difference between race, ethnicity, and ancestry?” These results are intriguing but require more research with larger numbers to find out how other Latino groups in other parts of the country would respond to standardized questions. What seems apparent at this point is that many Latinos chose (and will choose) the “other race” category. Did they do so because they felt culturally or socially “none of the above” or because they refused to become officially “brown” in the eyes of North Americans? And what determines racial reporting in the other categories?

ASIANS, HAITIANS, JAMAICANS, AND RACE Hispanics were not the only group that did not answer as expected or that were seen to have problems with the race and ethnicity questions on the census. For example, in the census’s ethnographic studies, Straus

150

THE “OTHER RACE” OPTION

noted that of twenty respondents classified as Native Americans in the Chicago census, only thirteen classified themselves as such in subsequent interviews (1991:10). Wingerd found that the Haitians she studied checked “other race” or left it blank (1995:17), and Bunte and Joseph discovered that the Cambodians they studied were confused because they were not specifically listed in the Asian and Pacific Islander category (1992:10–11). All these results led to the conclusion that “the way people view their own ethnic or racial identity and the way they perceive the identity of others is a complex psychological and sociological phenomenon that needs to be better understood before modifications are made to the race, Hispanic origin and ancestry questions on the census form” (de la Puente 1993:38). The results of another not-yet published study by the census were similar (de la Puente 1993). In this study, foreign-born blacks and Asian and Pacific Islanders in Washington, D.C., San Francisco, and Miami were interviewed in English and were part of focus groups. This study also focused on cognitive understandings of race and ethnicity, in order to determine the extent to which an English-speaking, foreign-born black’s, Asian’s, or Pacific Islander’s self-concept of race and ethnicity would result in misreporting, nonresponses, inconsistent responses, or other problems. It was important, therefore, to understand the respondents’ thought processes and the terms they used when answering questions about race and ethnicity. For example, how did Haitians or Jamaicans view such terms as “African American,” “Afro-American,” or “black”? Respondents were also asked about the Hispanic question, to determine how much their culture and beliefs about race and ethnicity influenced their definitions of race, ethnicity, and ancestry. The preliminary results indicate that these non-Hispanic groups also had different views of race, similar to those discussed in chapter 2 as having existed in ancient times and also to those found in research with Hispanics (Rodríguez 1991c, 1992, 1994a; Rodríguez and CorderoGuzmán 1992). In focused, cognitive interviews with twenty Filipinos, Chinese, Koreans, and Vietnamese, de la Puente (1993) found that the respondents: • Tended to interpret “race” as “national origin.” • If born in the United States, checked “other” and wrote in, for example, Chinese American for race, to acknowledge their “dual” origin.

THE “OTHER RACE” OPTION

151

• Had some awareness of the U.S. view of race but retained their own views for self-definition. • Had trouble understanding terms like ethnic origin and ancestry because of their limited proficiency in English. When the respondents finally understood that the census wanted historical information in regard to ancestry, they supplied their national origin, for example, “Chinese.” Once we realize that the way race is currently viewed in the United States is not necessarily the way it is viewed by others or the way it has been viewed historically, the question of how people view (or viewed) themselves becomes very interesting. For example, Judy Wingerd observed that in her research, Haitians in Miami considered it almost an insult to be called black. Instead, the Haitians she interviewed had their own register of colors and resisted being confined to one color. In Creole, the term raz (race) is equivalent to “my people,” “the area of the country I’m from,” or “my history.” These concepts match Latinos’ explanations of why they checked “other race” and wrote in a Latino referent, such as Puerto Rican, Dominican, or Honduran in response to the census’s race question (Rodríguez 1992).

EDUCATION AND RACIALIZATION The role of education may be very important to constructing one’s concept of race. Most of the Haitian interviewees were both poorly educated and immigrants to the United States, and they could not understand why race and culture were different concepts. Although the younger people that Wingerd interviewed learned to call themselves black, some had different interpretations of the word. It may be that the more education that people have in any country, the more often they are exposed to either an alternative or the dominant view of race in the United States.

Summary In summary, for Latinos, questions concerning “race” are very much affected by contextual variables, which in turn affect their responses. The variation in Latinos’ responses (compared with the more

152

THE “OTHER RACE” OPTION

consistent responses of whites, African Americans, and Asians) reflects the influence of context. Who asks the question, who answers it, and how and where it is asked—that is, whether the interviewer is Anglo and a Hispanic category is a possible choice—the presence of other cultural groups as categories, and the question’s phrasing, structure, placement, format, and purpose all affect Latinos’ responses. But what also is evident from this review is that many Latinos who chose the “other race” category saw their “race” as equivalent to their nationality, culture, familial socialization, birthplace, skin color, ethnicity, or a combination of these. The respondents who answered “other race” to the race item were not necessarily “mixed” or mixed up, nor were they forced into the “other” response solely by context. Rather, they interpreted the question according to their own frame of reference, which differed from that generally used in the United States. Whether the tendency to choose “other race” represents (or incorporates) a denial of blackness or an alternative view of race needs further research. We also must ask whether this question itself reflects the hegemonic nature and pull of the United States’ bipolar racial structure. The U.S. Office of Management and Budget stated that it changed its data collection standards and policy because it needed to collect information reflecting “the increasing diversity of our Nation’s population stemming from growth in interracial marriages and immigration” (U.S. Office of Management and Budget 1999:3). Our review of Latinos’ “other race” responses sheds light on the complex dynamics underlying these changes; we will examine in the next chapter the political sources accompanying these changes.

8 Redefining Race in 2000

I N N O V E M B E R 1 9 9 3 , Congressman Tom Sawyer, chair of the House

Subcommittee on Census, Statistics, and Postal Personnel, completed a series of hearings on federal measurements of race and ethnicity. There was nothing particularly remarkable about a series of hearings conducted by a fairly junior congressman, especially when they received relatively little press attention. What made them extraordinary were their proposals.

THE PROPOSALS The hearings focused on four proposals to amend the race item on the U.S. Census: (1) the addition of a multiracial category, (2) the addition of a special category for Middle Easterners/Arab Americans, (3) the shift of Native Hawaiians from the “Asian and Pacific Islander” category to the “Native American Indian” category, and (4) the inclusion of “Hispanic” as a race category.1 Except for the proposal on Hispanics, each had been advanced by representatives of the affected constituencies, and government officials and community representatives commented on the proposals (U.S. House of Representatives 1994). Surprisingly, all four proposals challenged the status quo and the assumptions inherent in the government’s current racial and ethnic classification. The proposals also implicitly reinforced the social constructedness of race categories and their malleability and susceptibility to political, intellectual, and social redefinition. But no groups formed alliances to support any of the proposals.2 Rather, the one area in which there was agreement was on the need for more research. The U.S. Office of Management and Budget (OMB) also agreed at the hearings to undertake a comprehensive review of the race and ethnic categories used by government agencies, as it was responsible for 153

154

REDEFINING RACE IN 2000

Directive 15, which specified and defined the categories (U.S. House Committee 1994n). Directive 15 had been issued on May 12, 1977, to meet the needs created by legislation passed to protect civil rights monitoring and enforcement, as well as the requirements of Public Law 94311, passed one year earlier, which called for the collection, analysis, and publication of economic and social statistics on persons of Spanish origin and descent.

Multiracial Americans The multiracial proposal challenged the long-held assumption that racial categories were (or had to be) mutually exclusive.3 Whereas all the current racial categories assume one (predominant?) racial identity, that is, white, black, Asian or Pacific Islander, or Native American Indian, a multiracial category would acknowledge that a person could be more than one race. Even the “other” race category in the current census race question is mutually exclusive, for it is the choice to be checked when one is “none of the above.”4 As noted earlier, this mutually exclusive way of viewing race has enabled North Americans in the United States to think of racial categories as representing “pure” races (Lee 1993). The extent of mixing (miscegenation) between, for example, whites and blacks that has produced “mixed” children has thus been overlooked and the myth of “pure” races sustained. Consequently, the possibility that a person might have more than one racial identity (particularly at the same time) defied the conventional approach to race in the United States. This new approach questioned the rule of hypodescent—in which one drop of black blood makes a person racially black (see Davis 1941; Russell, Wilson, and Hall 1992; and Williamson 1984 for excellent analyses of the evolution of this racial construction). Despite the traditional, exclusivist way of viewing race, the concept of multiple identities, which was inherent in the multiracial proposal, is increasingly viewed as appropriate for people with various heritages. This concept is particularly relevant to Hispanics and to other groups as well, for example, the children of interethnic or interracial marriages, ethnically identified Jews, and second-generation immigrants from many European, Caribbean, and Asian countries.

REDEFINING RACE IN 2000

155

Middle Easterners/Arab Americans Although the multiracial proposal garnered the greatest media attention, it was the proposal for a “Middle Eastern” category that most challenged traditional and idealized assumptions about race and ethnicity in the United States by pointing to precedents already in place. The proposal was for a Middle Eastern category, and the Arab American representative who argued in its favor did so on behalf of the population of the entire Middle East. She argued for “an ethnic non-racial classification for persons from the Middle East”—whether or not Arab (U.S. House Committee 1994g:183).5 The Arab American representative contended that Middle Easterners/Arab Americans deserved their own category for many of the same reasons that Hispanics and other groups have their own categories. The arguments presented raised basic questions about the nature of race and ethnicity, and the way they are determined in the United States. One argument was about self-classification versus classification by others. Middle Easterners and Arab Americans have most recently been classified by the census as white, although a number of scholars have noted that the media regard Arabs as nonwhite (Naber 1998; Shaheen 1984; Shohat and Stam 1994). The Arab American representative also pointed out at the hearings that in their personal lives, many Arabs and Middle Easterners identified themselves as “people of color” and that this classification was increasingly influenced by current political and ideological disputes and representations (U.S. House Committee 1994g).6 According to one Arab American researcher, some Arab Americans identify as white, and others as nonwhite, with a broad range of phenotypical diversity—some Arab Americans have very dark skin and kinky hair, and others have blonde hair and blue eyes (Naber 1998). Moreover, some identify as nonwhite because they feel discriminated against because of their political views or their Muslim identity, which in mainstream American discourse is seen as different from and inferior to a white identity (Naber 1998). Thus, some were embracing a notwhite-American position at the same time that they were being classified by the census as white. Furthermore, because the census classifies Arab Americans as white, it is difficult to get a separate count for the group. The representative argued that such counts were necessary because of current

156

REDEFINING RACE IN 2000

trends, such as discrimination against Arab Americans in the United States. This discrimination is seen as related to the politicization of ethnicity, in which Arab Americans are often viewed negatively because of the United States’ changing political relations with some Arab nations. Another trend noted is the change in rates of immigration, with a more (physically and socially) diverse stream of Middle Easterners currently coming to the United States than in the past. Also mentioned was the current context of pluralism, which encouraged immigrant children to respect and be proud of their diversity, to view their native language as an asset, and to preserve their religious and cultural practices. The representative noted that the current context favoring diversity conflicted with the speed with which Middle Easterners/Arab Americans were becoming Americanized. Consequently, they would continue as unassimilated (or visible) and persecuted minorities and therefore should be counted separately. In addition, the Arab American representative pointed to “perhaps a demonstration of certain cultural disadvantages” that Arab Americans might experience and to the possibly greater affirmative-action benefits and protections to be gained as a result of identifying as a minority. These protections were seen as necessary because of the discrimination against them, which has been documented by the American Arab Anti-Discrimination Committee (Ekin and Gorchev 1992). The Arab American Institute contended as well that both the Hispanic and the Asian and Pacific Islander categories contained models relevant to the reclassification of Middle Easterners. “The rationale for the Hispanic classification was to measure a population sharing common geographic and linguistic roots that could distinguish them from the rest of the white population” (U.S. House Committee 1994g:188). The institute pointed out that the Asian and Pacific Islander race category was similar in that it transcended precise racial characteristics and covered a geographical region that represented many nationalities, languages, and even racial groups.7 The proposed Arab American category would include Arabs, Iranians, Turks, Afghans, and others, who, it was argued, had similar religious, linguistic, and cultural backgrounds; faced similar discrimination and exclusion; and were distinguishable from the European-based white majority. Arab Americans noted a number of fluid and contextually dependent race constructs in the classification of Hispanics and Asians. For example, the focus on the increase in politically related discrimination, the

REDEFINING RACE IN 2000

157

cultural context celebrating diversity, the greater retention of cultural differences, the greater diversity of the immigrating population, and the establishment of affirmative-action benefits all are context-dependent factors that influence how those affected view themselves and others. Moreover, the representative pointed out, “Just as self-definitions internal to racial minorities evolve and emerge, the lines between and around race and ethnicity as identifiers continue to blur, shift and intersect over time” (U.S. House Committee 1994g:188).8 What was perhaps most interesting in the representative’s presentation is that her request did not represent “a racial redefinition, but rather a recognition of new realities.” In other words, Arab Americans were not changing their “race”; rather, their position in American society had changed. In essence, they explained that in order to keep up with the changing realities, Arab Americans should be counted as a separate group, a new pan-ethnic group (Edmonston, Goldstein, and Tamayo Lott 1996:33). Some readers might view the Middle Eastern proposal as an attempt to capitalize on the benefits associated with the shift from an exclusionary to an inclusionary categorization. As a result of the civil rights movement, categories formerly used to exclude individuals now have been used to include individuals in affirmative-action programs, set-asides, and so forth (Fienberg 1994). This shift has introduced, however, a new tension into the issue of classification, with some groups wanting to be classified as protected minorities so that they can benefit from these programs or because they need the programs’ protection. But it is also possible that—regardless of the benefits to be gained— Middle Easterners/Arab Americans, like Hispanics and other groups or individuals, may simply want to be viewed in accordance with their own self-conceptions of race and identity. Their position reflects a different view of “race,” in which one’s (white or nonwhite) racial status is seen to bear no relationship to a group’s identity as a group. Indeed, at a later workshop, the Arab American Institute’s spokesperson insisted that Arab racial identity is ethnic and not racial (Samhan 1994). A Middle Eastern category was not established. In its final recommendations, the U.S. Office of Management and Budget noted that establishing a new category would require a “consensus building effort to arrive at appropriate terminology and a definition” (U.S. Office of Management and Budget 1997b:36934). Some of the issues in this group still requiring resolution were whether the term Arab American or Middle

158

REDEFINING RACE IN 2000

Eastern should be used. Should the category be defined as pertaining to persons whose “mother tongue” or culture was Arabic, or should the definition be more restrictive, and if so, which countries should be included?

Native Hawaiians The request by Native Hawaiians to be counted in the “Native American” category, instead of in the “Asian and Pacific Islanders” category, also challenged another tradition: the use of geographic origin (with its implied biological characteristics) to determine race. Instead, Native Hawaiians insisted that their history as a conquered and indigenous people be acknowledged, and not just their geographic location on the Asian and Pacific side of the globe. As Hawaii’s Senator Daniel Akaka explained, “Native Hawaiians have a unique historical and political relationship with the United States” (quoted in Omandam 1997), which differs from that of Native Americans, who claim certain federal benefits based on earlier treaty agreements in which they exchanged land for perpetual educational and health provisions. Before 1893, Hawaii’s treaties with the United States concerned friendship, commerce, and permission for U.S. ships to enter Hawaiian waters and dock in its ports. Between 1826 and 1893, the United States recognized Hawaii as a sovereign nation and extended it full diplomatic recognition. But in 1893, a U.S.-backed military coup overthrew the constitutional monarchy headed by Queen Liliuokalani and in 1898 ceded Hawaiian lands to the United States (U.S. House Committee 1994o:199). It was in 1920 with the passage of the Hawaiian Homes Commission Act that Native Hawaiians were first classified according to a blood quantum definition of 50 percent. The Hawaiian proposal displayed the U.S. government’s colonial and imperialist past. In so doing, Native Hawaiians placed themselves alongside groups like Native Americans, Puerto Ricans, and Spanish Americans in the Southwest who do not consider themselves immigrants to the United States but see themselves as part of the United States because the United States came to them and took over their land. Although the Hawaiians’ proposal was not supported, there has been some change. In the 2000 census, Native Hawaiians are not listed under the Asian and Pacific Islander category but have their own “Native Hawaiian” category along with other Pacific Islanders.

REDEFINING RACE IN 2000

159

Hispanics Finally, the Hispanic proposal also reflected a radical departure from current policy. An important difference, however, was that it was not advanced by the constituent group. The proposal called for the elimination of the “Hispanic” identifier and the addition of a “Hispanic” race category to the race question. The proposal challenged the Census Bureau’s official position that race and ethnicity were separate concepts, and it would reclassify what had been an “ethnic group”—in which Hispanics could be of any race—to a “race” group, in which all Hispanics were one race. The lack of constituent support for the proposal to include Hispanics as a race category was not noted during the hearings. Nor is it clear who first advanced this proposal. This lack of Hispanic involvement contrasted sharply with the Hispanics’ earlier involvement with the census. In 1970, “in response to demands by community groups for a comprehensive self-identification measure of Hispanic ethnicity,” such a question was included in the 1970 census long forms, which were sent to 5 percent of households. This question relied on self-identification and was not tied to parental birthplace or Spanish surname, as earlier questions had been (McKenney and Cresce 1993:175–176). According to Choldin, the census had resisted the demand for a question in which respondents would identify themselves as “Hispanic,” arguing that it was too late to test such an item and that “existing procedures for identifying Hispanic individuals were more valid” (1986:407). But the White House intervened and instructed the secretary of commerce to add a “Hispanic” self-identifier. Since millions of questionnaires had already been printed, the compromise reached was that the question would appear on the long form sent to 5 percent of households. (Only ten thousand copies of the long form had been printed.) The question was also tested in the 1969 Current Population Survey. The results of the 1970 mail-out, mail-back questionnaire were disputed and protested by Mexican American organizations who decided on a class-action suit. It called for a new category on all the questionnaires and for Mexican-American or Chicano face-to-face, Spanishspeaking enumerators, using Spanish-language questionnaires. Although the case never went to trial, the House subcommittee did hold a series of hearings on statistics for “Spanish-speaking Americans.”9 It

160

REDEFINING RACE IN 2000

was these political forces that contributed to the emergence of the Hispanic identifier on the 1980 census form. This sometimes contentious and antagonistic history was not repeated during the Sawyer hearings. Indeed, the Hispanic community’s silence and lack of involvement on the issue of reclassification generally and on the Hispanic proposal specifically were surprising. It is difficult to tell whether this resulted from the exclusion and obfuscation of the issues, a lack of awareness, a lack of Latino interest in the issue or in the complexity and perhaps perceived irrelevance of the discussions, an inherent aversion to discussions of race, a sense that it did not matter how Hispanics would be classified as long as they were counted, or a combination of all these and other factors. Although the major, requisite Hispanic organizations were present at the hearings, few representatives of the Hispanic community testified. Likewise, there was little coverage of the issue in the Spanish- or English-language media and few public discussions elsewhere. When the Mexican American Legal Defense and Education Fund (MALDEF) and the National Council of La Raza (NCLR) testified at these hearings, MALDEF indicated that a recommendation on “whether or how to change the Census’s Hispanic origin and race questions would be premature” (U.S. House Committee 1994b:179). Both groups felt that the current Hispanic item should be retained; neither endorsed the proposal as presented. Moreover, both recommended additional research before any change was made, and MALDEF added that any change contemplated should be targeted to reducing the differential undercount (U.S. House Committee 1994k:178–182).10 Only the National Council of La Raza made a statement at the hearings that was later cited as supporting this proposal (del Pinal 1994; Wright 1994). Yet a closer reading of the statement shows that by proposing that “Hispanic” be included as a category in the race item, the council was also requesting that the item be relabeled Race/Ethnicity. (Its suggested question has this label; see U.S. House Committee 1994p:178.) Moreover, the statement advocated retaining the current separate “Hispanic” identifier, whereas the proposal being considered would eliminate it. Perhaps because their position was misinterpreted by some, the National Council of La Raza decided to clarify its position to the Office of Management and Budget (OMB): The NCLR “would be inclined to support the combination of the race and Hispanic origin questions into a question re-labeled ‘Race/Ethnicity,’ if testing indicates

REDEFINING RACE IN 2000

161

that such a question solicits a greater and more accurate response rate” (1995:8, italics in original). The Census Bureau’s history concerning this issue also received scant attention in the hearings and other discussions of the proposal. Earlier, in 1984, the census formed and chaired interagency working groups (IWGs) to discuss the 1990 federal census data requirements. These groups were composed mainly of program specialists familiar with census data and their applications. The IWG on race and ethnicity supported retaining a separate question on Spanish/Hispanic origin and concluded that a combined race/Spanish origin question (i.e., to include “Hispanic” as a race group) “would not meet program needs and could result in an undercount of the Spanish origin population” (U.S. Bureau of the Census 1990:5). In addition, a proposal to count Hispanics as a race was introduced before the 1990 census but was so strongly opposed “through the most aggressive campaign ever seen by the bureau” that agency officials decided to abandon it, fearing it would lose needed community support (Hispanic Link Weekly Report, May 26, 1986, p. 3). Subsequent attempts by the census to institute such a proposal also were met with similar resistance (McKenney 1994). A study (done after the proposal was made) did find that a majority of Hispanics preferred the combined question (U.S. Dept. of Labor, Bureau of Labor Statistics 1995:table 3),11 but this may reflect a different understanding of the question. The study’s participants may have understood it as “Do you want to be included?” rather than “Do you want to be a ‘race’?” The preference for a combined question probably does not mean that Hispanics acknowledge or agree that they are a “race” in the same way that the census conceptualizes this term. Moreover, as Ruth McKay, a researcher at the Bureau of Labor Statistics, observed, “The respondents did not understand the consequences of combining the questions” (Torres 1996:4). Given the history of the proposal in the Hispanic community and its lack of apparent support or even involvement, we might ask why the proposal was presented, and continued to be presented, as a serious and legitimate proposal. A number of suppositions are possible. Making “Hispanics” a race would make life easier for the data gatherers because there would be one item on the census instead of two and all social races could be counted directly instead of subtracting Hispanics from the various race categories. Having a combined question would

162

REDEFINING RACE IN 2000

also be cheaper for the Census Bureau, as there would be one less item to tabulate. Indeed, one researcher at the hearings described getting detailed data on various race and ethnic groups as a “cumbersome and fallible process” (U.S. House Committee 1994m:54). The proposal would also make the counting of Hispanics consistent across government agencies. At present, the government counts Hispanics in two ways, the census’s and that specified in Directive 15—the executive order resulting from a federal interagency agreement in 1977. Directive 15 places all people into five major racial/ethnic groups: “white,” “black,” “Asian and Pacific Islander,” “Native American Indian,” and “Hispanic.” (The “other race” category is not included.)12 Those supporting the “Hispanic” race proposal may therefore have tried to adopt the Directive 15 model. But whereas the directive makes clear that it refers to both race and ethnicity, the census proposal referred just to race. In support of the Hispanic proposal, it could also be argued that the proposal was more like the Latino view of race in that it presented all the “social races” together. Apparently, some persons subsequently expressed support for a combined race/Hispanic-origin question because “many Hispanics do not identify as a race” and this would end the practice “of using the term race which they see as a social rather than a scientific construct” (U.S. Office of Management and Budget 1995:44678). The term used in the proposal, however, was “race,” not “social race” and not ethnicity; and the proposal did not refer to ethnic or cultural differences among groups, which are central to Hispanic views of race. In essence, Hispanics were included in the model, but at the cost of making them a race within a framework that privileged the white social race. Some people may have supported the proposal because they felt it was simply time to acknowledge a new nonwhite or “other” race in the census categories. According to their perspective, this new Hispanic race would span a color continuum from “almost white” to “black.” Still others (both Hispanic and non-Hispanic) argued that Hispanics were, for all intents and purposes, a race in the United States and should therefore be counted as such. The NCLR, for example, noted that despite the “technical” differences that might be found between “race” and “ethnicity,” the two terms were really used interchangeably by society and were synonymous for “Hispanics.” Furthermore, because Hispanics were treated as a “race,” it was important to be repre-

REDEFINING RACE IN 2000

163

sented in the race item (see chap. 1; Bendick 1992; de la Garza et al. 1992:94–95; Del Valle 1993; U.S. General Accounting Office 1990). Moreover, the lack of a Hispanic “race” category perpetuated the black/white paradigm, which consistently excluded Hispanics. (However, as noted earlier, what the NCLR envisioned as a “race” was somewhat different from what was spelled out in the Hispanic proposal.13) Such approaches appeared to argue that Hispanics be called a “race” in return for recognition, as this might make the counts more accurate, resolve the problem of Hispanic invisibility, alter the prevailing black/ white axis and paradigm, and be more in keeping with changing demographics.14 Although the proposal to make “Hispanics” a race received some attention by others testifying at the hearings, it was not enthusiastically endorsed by anyone. Indeed, some participants were explicitly against it. For example, Arthur Fletcher, chair of the U.S. Commission on Civil Rights, stated that the commission recommended “against reclassifying Hispanics as a racial group” because they were “a complex community of races bound by common cultural, linguistic and geographic origins” (U.S. House Committee 1994b:253, 260). Tony Gallegos, chair of the U.S. Equal Employment Opportunity Commission, believed that the census’s successful experience with five racial/ethnic groups precluded the need for such changes (U.S. House Committee 1994c:285–286).15 But the general consensus was that the proposal needed to be tested and its impact evaluated before it was put into operation (Morris 1994). In addition, it was clear in the hearings that a multiracial category would cause more problems than it would solve. Some groups felt such a category would jeopardize the numbers in their categories. It was not clear at the time, though, what a “Hispanic race” category would do to the counts of other minorities. The lack of support continued, and the Office of Management and Budget concluded that “most Federal agencies did not comment on whether race and Hispanic origin should be collected in one question or two questions. . . . Those few that commented were split on the issue” (1995:44678).

THE PERSISTENCE OF THE HISPANIC PROPOSAL Despite the lukewarm reception and earlier resistance, the proposal continued to be considered seriously. When the Office of Management

164

REDEFINING RACE IN 2000

and Budget requested comments on the proposals being reviewed for the “Standards for the Classification of Federal Data on Race and Ethnicity” in 1994, it mentioned having “Hispanic as a racial designation, rather than as a separate ethnic category,” adding that combining race and Hispanic origin has become one of “the more significant issues that have been identified for research and testing” (U.S. Office of Management and Budget 1995:44690). Hispanic input into the proposal continued to be minimal. Indeed, by the time the Hispanic Advisory Committee to the Census was established in 1994, the proposal had already been discussed and researched. Instead, the meetings of the committee in 1995/96 were to discuss the findings from the National Content Survey that had tested this combined question (U.S. General Accounting Office 1997:8). At subsequent hearings on this proposal, Hispanic involvement did not greatly increase. Nonetheless, the proposal persisted and became part of a massive research effort. Of the four proposals presented at the House hearings, only two were pursued, the “Hispanic” proposal and the multiracial proposal. The Office of Management and Budget created the Interagency Committee for the Review of the Racial and Ethnic Standards and, as part of this, an interagency research initiative. This research was to evaluate the proposals for revising racial and ethnic reporting categories and to determine the potential effect of any changes.

The Studies The first study in this research agenda was a supplement to the Current Population Survey (CPS) that collected information on several key issues, one of which was “the effect of adding ‘Hispanic’ to the list of racial categories” (U.S. Dept. of Labor, Bureau of Labor Statistics 1995:1). The Bureau of Labor Statistics designed the special supplement to the usual May 1995 Current Population Survey (CPS), so that it could evaluate how the “inclusion of an Hispanic category in the list of races” would affect racial and ethnic data (U.S. Office of Management and Budget 1995:44690).16 This special supplement surveyed by phone more than sixty thousand randomly selected households. The census also conducted cognitive research on this proposal, and by 1995, additional research plans were made to examine larger samples (U.S. Office of Management and Budget 1995:44691).

REDEFINING RACE IN 2000

165

Other agencies were to carry out similar research. For example, the National Center for Health Statistics, using an approach similar to that of the Bureau of Labor Statistics (i.e., comparing combined and separate race and Hispanic-origin questions), was to examine the effects of racial classification changes on birth certificates. The Centers for Disease Control and Prevention were to evaluate the recording of racial classifications on death certificates. The Department of Health and Human Services (DHHS) was to conduct a literature search and make an inventory of DHHS minority health databases. Finally, the National Center for Education Statistics was to examine current issues, state legislation, and how schools currently collect, maintain, and report racial and ethnic data (U.S. Office of Management and Budget 1995:44690–44691). In 1996, the Bureau of the Census conducted two other major studies. Both sent self-administered, mail-back questionnaires to 90,000 households for the National Content Survey (NCS) and 112,000 households for the Race and Ethnic Targeted Test (RAETT). The National Center for Education Statistics and the Office for Civil Rights in the Department of Education conducted surveys in public schools to determine how they collect data. Finally, the National Center for Health Statistics, the Office of the Assistant Secretary for Health, and the Centers for Disease Control and Prevention studied the methods used to gather data in this area. The CPS, NCS, and RAETT tested a number of innovations, including the introduction of a multiracial category, a proposal to make Hispanics a race (subsequently called the combined format), and the reversal of the sequence of the race and Hispanic questions. Although several interesting and detailed results were produced, the net result of the CPS, NCS, and RAETT studies was that the combined format resulted in fewer Hispanics and whites being counted (U.S. Dept. of Labor, Bureau of Labor Statistics 1995:table 1).17 Consequently, the proposal to make “Hispanic” a race was abandoned.

The Purpose of the Hearings Why did the proposal to make “Hispanic” a race persist despite its lukewarm support? To answer this, we should ask, Why did these hearings take place at all? Why were they examining previously unexamined questions? What was their purpose?

166

REDEFINING RACE IN 2000

A later analysis of the revision process maintained that the OMB began to consider revising the federal standards for racial and ethnic classification because of the demographic and social changes taking place in the United States and because of the increasing dissatisfaction with the current standard among data users, data providers, and the public (Edmonston, Goldstein, and Tamayo Lott 1996:35).18 Thus the purpose of the hearings was to ascertain whether the way the federal government measured race and ethnicity was satisfactory. The fact that this question was being asked at all signaled a major adjustment in the way that racial and ethnic concepts—until now taken at face value— might be viewed in the future. The “concerns” expressed at the hearings hinted at some of the underlying issues that led to this reexamination.

The Concerns Leading to the Hearings Congressman Thomas Sawyer opened the hearings by citing three concerns: (1) the identification of multiracial persons, (2) Hispanics and Middle Easterners who do not identify with any of the four major racial categories, and (3) self-identification by foreign-born persons whose understanding of race is often shaped by different definitions and understandings in their countries or cultures of origin. Although only the second concern refers specifically to Hispanics, all three pertain to them because many are foreign born, many are seen to be multiracial, and many have different views of race. These concerns raise a number of questions, not the least of which is why these issues were being addressed in 1993. For example, “the identification of multiracial persons” implies that there is currently some interest in, or need to identify, multiracial persons. The last time that the census counted in a separate category those people whom it viewed as multiracial was in 1920, when it classified “mulattoes” as a race category. The third concern suggests an awareness of alternative views of race among the foreign born when it acknowledged that the understanding of race and “self-identification by foreign-born persons” is “often shaped by different definitions and understandings in their countries or cultures of origin.” The second concern, however, “Hispanics and Middle Easterners who do not identify with any of the four major racial categories,” raises the question of why it is acceptable for the foreign born, but not the groups called “Hispanics” and “Middle

REDEFINING RACE IN 2000

167

Easterners,” to have different understandings of race. Was the assumption that these groups were born in the United States and not abroad? If so, then the real concern was that different understandings of “race” were persisting among Hispanics and Middle Easterners born in the United States Preceding these hearings was a more general questioning and heightened awareness of the resurgence of racial and ethnic tensions on an international scale, for example, the Islamic fundamentalists’ conflicts. In addition, in the U.S. scientific community there was a major reexamination of race and ethnicity. The fact that this was receiving special attention in the major media added to the need to reconsider the meaning of race and ethnicity (see, e.g., Barringer 1993; Bernal 1987; Discover November 1994; Marks 1994; Newsweek February 13, 1995; Rosin 1994; Wood 1994; Wright 1994). Congressman Sawyer noted that the stated “concerns” had been voiced by “many people . . . during the 1990 census” (U.S. House Committee 1994a). These “many people” may have been those gathering the data or those constituencies interested in specific census issues. At the governmental level, the concern began to surface largely because of the results of the 1980 and 1990 censuses. With the shift in the 1980 decennial census from interviewer-identified race to self-identified race, unexpected issues and questions emerged. According to a former census official, the idea of “race and ethnicity” as a state of mind had not been accepted by the census earlier, but with the shift to self-identified race, the census recognized that this type of reporting raised other issues (Estrada 1994).

Data-Quality Problems Four key problems were discussed in a report submitted to Congressman Sawyer by the General Accounting Office (GAO) earlier that year:19 1. The growth of the “other race” category. 2. Problems in the consistency with which some groups reported their race and Hispanic origin. 3. A high allocation rate for the Hispanic item. 4. Some misreporting problems in both the race and the ethnicity items.

168

REDEFINING RACE IN 2000

Table 8.1 Percentage of Hispanics Choosing “Other Race” in Different Question Formats Race Question

Hispanic Question

Multiracial Category

Percentage Choosing “Other Race”

First First

Included Not included Included Not included

33.0 42.9 25.1 24.9

First First

Source: U.S. Bureau of the Census, National Content Survey 1996a:tables 11 and 12.

These were referred to as the “data-quality” issues that led to the hearings. Although these issues affected many groups, they particularly concerned the large and growing Latino population. Indeed, a close examination of the GAO report reveals that Hispanics were at the center of many of these issues, although this was not noted. For example, Hispanics make up the overwhelming majority (97.5%) of the “other race” category. Thus, its dramatic growth is due to the fact that Hispanics continue to choose it, and it is difficult to recode their national-origin responses into other race categories. With regard to the second problem, the extent to which individuals consistently give the same response to questions of race and Hispanic origin, many Hispanics changed their answers in reinterview studies and in response to a series of contextual factors, whereas other groups did not. The GAO’s report noted that only 36 percent of those who said on the census form that they were “other race” said that again when reinterviewed. The report did not specify whether most of these respondents were Hispanic, but they likely were because they made up 95 percent of this category (U.S. General Accounting Office 1993).20 Curiously, although “consistency” was a problem when answering “race” questions, responses to the “Hispanic” question were highly consistent, with 90 percent of Mexicans, 92 percent of Puerto Ricans, 86 percent of Cubans, and 100 percent of “those who said they were nonHispanic” responding the same way on both occasions. The exceptions here were those who said they were “other Hispanic,” with only 64 percent answering similarly in the reinterview study.

REDEFINING RACE IN 2000

169

On the third problem, it was also the “Hispanic” item that had the highest allocation rate.21 The Census Bureau allocates a particular response for questions left unanswered on the census questionnaire. This allocation procedure is based on a complicated series of steps for each item that best approximates the missing information. The allocation rate for the Hispanic-origin question was not only the highest of all the questions, but it also increased from 4.2 percent in 1980 to 10 percent.22 This was seen as particularly problematic: “The results from the 1990 census showed that the Hispanic-origin item continues to pose one of the more significant data quality challenges for the Bureau in terms of allocation rate” (U.S. General Accounting Office 1993:24). Why was the allocation rate so high for the Hispanic question in 1990? Why did 10 percent not answer the question about whether they were Hispanic? The GAO report saw two underlying problems: One was that many persons who were not Hispanic skipped the question altogether because they did not see it as relevant.23 Another problem was that some Hispanics “equate their ‘Hispanicity’ with race by responding ‘other race’ in the race item, indicating they are Hispanic in the space the race item provides, and then skipping over the Hispanic origin item because they see this item as superfluous” (U.S. General Accounting Office 1993:25). That is, “confusion” about the race item might have spilled over to problems with the Hispanic-origin item. The GAO report does not explain what it means by “confusion” with the race item, the implication is that it refers to Hispanics’ responding they are “other race.” Two years later, the OMB stated that most of those who did not respond to the “Hispanic” item were non-Hispanics (U.S. Office of Management and Budget 1995:44689). Consequently, the reason for the high nonresponse rate for the “Hispanic” item was more that nonHispanics saw the question as irrelevant to them than that Hispanics were “confused.” The last problem, misreporting, refers to several problems but is seen to be principally the result of mistakes or misinformation. An example is those who responded they were “other Hispanic” and later said in the reinterview studies that they were not Hispanic at all but wanted to indicate they were “other than Span/Hisp.”24 Examples of misreporting in the race question are those who checked the “other race” category and wrote in a response that the census reclassified as

170

REDEFINING RACE IN 2000

one of the other four race categories, so “German” in “other race” was reclassified as “white.” The report does not make clear to what extent Hispanics were involved in this problem. Indeed, it is these early reports’ lack of reference to Hispanics’ specific reporting behavior that is most puzzling, particularly because Hispanics were so central to many of the concerns raised at the hearing.25 These problems notwithstanding, the GAO report concluded that the 1990 data on race and Hispanic origin were “generally of high quality” (U.S. General Accounting Office 1993:28). The report did note a growing awareness of the population’s increasing diversity, but it was not Hispanics to whom it was alluding; rather, the report cited the more than 200,000 codes that had to be developed to accommodate all the write-ins in the “Asian and Pacific Islander,” “Native American Indian,” and “other race” categories (U.S. General Accounting Office 1993:28). The responses that produced the greatest number of codes because of write-in responses were the Ancestry and the Native American Indian items (Edmonston, Goldstein, and Tamayo Lott 1996:23).

FINDINGS FROM THE GOVERNMENT’S STUDIES • Regardless of the format used, a substantial proportion of the answers remain in the “other race” category. An important part of the “data-quality” issues addressed was the respondents’ tendency to choose the “other race” category. As we now know, almost all those (97%) in this category were Hispanic. The purpose of the proposal to make “Hispanic” a race was to reduce the number of persons choosing the “other race” category. I suspect that the reasoning was that if Hispanics saw their group represented with the others, they would choose “Hispanic” and not “other race.” In all three studies, when Hispanics were made a race; that is, when the combined question was used, the number of persons who chose the “other race” option dropped. But what the studies showed was that regardless of the context of the question, for example, whether or not a multiracial category was included or whether multiple responses were allowed, many people still chose the “other race” category.

REDEFINING RACE IN 2000

171

Sequencing was also thought to affect whether Hispanics chose the “other race” option; that is, if Hispanics were asked about their Hispanic origin first, they would not choose “other race.” In all three studies, placing the “Hispanic” question before the “race” question did reduce—but did not eliminate—the number of persons choosing the “other race” option. An example of this adherence to the “other race” response can be seen in the NCS study in which respondents were asked to choose their race under what might be considered—from a statistician’s perspective—fitting conditions for discouraging an “other race” response (U.S. Bureau of the Census 1996a). Ideal conditions meant including a “multiracial” category (so that those of mixed race could choose it) and placing the “Hispanic” question before the race question (so that those who saw “race” as “culture” and “national origin” would have already identified themselves as such and could now choose their race). Under these conditions, the proportion choosing “other race” did decline, but 25.17 percent of Hispanics still chose this option (see table 8.1). (Although this was a substantial proportion, it was not statistically significant.)26 • Hispanics choose more than one category even when instructed not to. Two other findings from these studies are relevant. One is that when the combined question was used in the RAETT test, a high percentage of respondents (18% to 19%) checked that they were Hispanic and also checked one of the other race categories. Indeed, the Hispanic respondents checked more than one category even when they were instructed not to do so (U.S. Bureau of the Census 1997:4, chart I). Although the results of the RAETT test can be generalized only to areas with relatively high concentrations of Hispanics and other targeted populations, it is interesting that the census’s Hispanic Advisory Committee recommended that respondents be allowed to choose “more than one” category on the “Hispanic” item as well. In other words, the committee thought it important that respondents be able to say they were both “Hispanic” and “not Hispanic” (those who might want to acknowledge a Hispanic component as well as a white, black, etc. component in their response). This recommendation was considered but not accepted because it had not been tested.

172

REDEFINING RACE IN 2000

• The responses of Hispanic-origin groups differed. Finally, some of the studies showed that the various Hispanic-origin groups responded differently to the questions. For example, in the CPS study, when having to choose the “Hispanic” or another category in the combined question, a minority of Cubans (39.92%) chose the “Hispanic” category, compared with a majority of Mexicans (85.15%), Puerto Ricans (71.51%), and Central or South Americans (77.67%). The introduction of a multiracial category increased the percentage of Cubans who chose the “Hispanic” category, but it was still only 46.40 percent. In the other Hispanic groups, the percentage choosing the “Hispanic” category also increased slightly or stayed about the same (U.S. Office of Management and Budget, 1997a:36916, table 4.4). These results are consistent with those cited in chapter 7 regarding who chooses the “other race” option. • Why Did Hispanics Choose “Other Race”? As noted earlier, the reasons that some Hispanics continued to choose the “other race” category are complex and require further research. To some degree, the context and format of the question influence the choice. But the choice also reflects different conceptions of “race” and perhaps a resistance to the racial structure as articulated in the United States. This resistance may be traced to Hispanics’ objections to being classified as a uniform, subordinate, not-white race. Or it may irritate Hispanics who see themselves as physically diverse and defined by national origin or culture. In either case, making “Hispanic” a race may have been seen as a perpetuation and extension of the racialist thinking of the past. The OMB’s final recommendations cite findings that both supported and did not support separate race and Hispanic-origin questions. Those findings that did not support a combined question were that the concepts of race and ethnicity were difficult to separate; that Hispanics want to identify their race in addition to their Hispanic origin; that some Hispanics, including the Census Hispanic Advisory Committee and most Hispanic organizations, opposed a single, combined question; that “Hispanic” was not considered a race by some respondents and users; and, finally, that a combined question would increase the need for additional tabulations because people would choose

REDEFINING RACE IN 2000

173

more than one category. Those findings that did support a single, combined question indicated that it would eliminate redundancy, thereby acknowledging that for many Hispanics, race, culture, and national origin are the same.

Race in Formation The OMB finally decided to retain the two-question format, but it also decided to allow individuals to choose more than one category.27 Moreover, it recommended that when self-identification was not feasible or appropriate, a combined question could be used (U.S. Office of Management and Budget 1997a:36930, 36939). The recommendation that a combined question be used when self-identification was not possible suggested that attempts be made “to obtain proxy responses (from family or friends) as opposed to using observer identification” in order to ensure accurate data.

Unresolved Issues According to the Office of Management and Budget, government research shows that less than 2 percent of persons are expected to choose more than one race category (U.S. Office of Management and Budget 1999:4; Tucker et al. 1996). The preliminary Census 2000 Dress Rehearsal Results, although not representative of the country as a whole, also do not show many persons choosing more than one category (del Pinal 1999). Therefore, the OMB does not anticipate any significant impact on redistricting decisions or on total population counts used for apportionment or for compliance with one-person, one-vote requirements because of the “choose more than one” option (U.S. Office of Management and Budget 1999:33 ff). But other researchers estimate that the impact will be larger, that this shift may be greater than the net size of the undercount (Goldstein and Morning 1999). Moreover, they estimate that this shift will have different effects on the single-race groups, with whites declining between 3 and 6 percent, blacks between 3 and 7 percent, Native Americans between 15 and 25 percent, and Asian and Pacific Islanders between 4 and 9 percent. This change in practice and policy has been put into effect. But at this writing, there still are a number of unknowns. Unknown (and not included in the preceding estimates) is the role of the media in

174

REDEFINING RACE IN 2000

influencing individuals to “choose more than one.” Also unknown are the implications for race-based public policies. As Goldstein and Morning (1999) asked, Will people who in the past said they were white and now claim Native American Indian ancestors in the race question be eligible for minority small business loans? Will those who previously said they were only black and now say they are white and black no longer be eligible? Should some individuals (or groups) of more than one race be protected classes and others not? For example, if those of Japanesewhite ancestry are economically more advantaged than those of Vietnamese-black ancestry, should the latter be protected but not the former? Last, it is not known how the data should be tabulated. A number of possibilities are under discussion, but no firm decision has been made (U.S. Office of Management and Budget 1999).

Issues Raised The “Hispanic” proposal, as well as the other proposals discussed at the initial hearings, raise a number of issues. They—and the events that followed these hearings—also revealed the dynamics of racial formation as we approach the next millennium. All the proposals made clear the extent to which race and the construction of racial categories are influenced by nonbiological factors, although this was seldom recognized or expressed. On a theoretical level, the proposal to make “Hispanic” a race raised the issue of how Hispanics should be counted. Should they be treated as a European ethnic group (albeit multiracial) or as a separate race? The first approach was (and is now) the one in effect in the census: Hispanics could be of any race. The second approach implied that Hispanics were seen (and saw themselves?) as a distinct social group—a race—regardless of phenotype. Hispanics, as well as many other groups, challenge the U.S. system of racial classification because they do not fit neatly into the given categories. They are neither a race nor a racially homogenous ethnic group. Rather, they are a diverse array of multiracial ethnic groups, bound together by language, cultural ancestry, and discrimination in the United States. They can best be understood in a paradigm acknowledging that the social constructions popularly called “race” are really all social groupings that convey political, social, and cultural differentials. Within such a paradigm, Latinos and other “races” are clustered eth-

REDEFINING RACE IN 2000

175

nicities in a hierarchy of power growing out of the history of whites and other social races in the United States. Hispanics contributed significantly, albeit silently, to the concerns voiced and to the “data-quality” issues raised that prompted the hearings and subsequent research. Indeed, it might be said that Hispanics have come to redefine everyone else, as in the use of terms such as “nonHispanic whites” and “non-Hispanic blacks.” An interesting irony here is that at the same time that the influx of Hispanics led to the redefinition of all other groups, the government attempted to redefine Hispanics as a race. Just as in 1930 when the government introduced a “Mexican” race category, as we end the twentieth century, the government is proposing to create a race category for all Hispanics. The increase in numbers at both times contributed strongly to these racial classification projects. The “Hispanic” proposal also highlighted the role played by “race” or color in the United States. Making “Hispanic” a race and eliminating a separate “Hispanic” identifier would not allow individuals to respond that they were, for example, both Hispanic and black (U.S. House Committee 1994k:179). As noted previously, research has shown that Hispanics who classify themselves as white or who are classified as white by others fare better economically than those classified in other racial categories (Arce, Murguía, and Frisbie 1987; Rodríguez 1990, 1991a, 1992; U.S. Office of Management and Budget 1995:44678; Telles and Murguía 1990). Using one combined question might make it more difficult to determine which Hispanics are more likely to be victims of discrimination. In addition, making “Hispanic” a race would make it difficult to compare the data with past censuses. Finally, a combined question might also dilute the counts of other race groups; for example, Hispanics who in the past might have reported that they were “black” or “white” might indicate instead that they were “Hispanic” (U.S. Office of Management and Budget 1995:44678). As Rachel A. Joseph argued, the counts and comparability of counts over time of Native Americans would be affected adversely by making “Hispanic” a race (U.S. House Committee 1994l). In 1990, 165,000 (or 8.5%) of the 2 million who said they were “Native American Indian” also reported that they were of Hispanic origin.28 The hearings and the subsequent process showed “race” in formation. They showed, and sometimes acknowledged, the difficulties and contradictions of the current racial classification structure. While in the

176

REDEFINING RACE IN 2000

past and in public discourse, racial and ethnic concepts had often been projected as fairly immutable and not subject to diverse interpretations, it was now being publicly acknowledged on a national level that these concepts were not mutually exclusive but were fluid and “dynamic” (U.S. House Committee 1994f).29 Thus, on the one (conceptual) hand, the intermingling of the concepts was recognized, but on the other (practical or applied) hand, the concepts were treated as separate. In stating the “lessons” from the hearings, Congressman Sawyer stated that “the categories had to be relevant to those responding if cooperation was to be secured.” This revealed the growing official concern that current categories might not be relevant to some respondents.30 This lesson also raises questions about the government’s ability to identify individuals correctly and clearly. In addition, it suggests an emerging awareness of procrustean census tactics on the part of government officials. Last, it poses the question of whether there is a conflict between providing recognizable categories that are relevant to respondents and needing to gather uniform, comparative data. Although the stated concerns and the final formulations raised a number of questions, taken in concert they suggest the extent to which “race” and “ethnicity” are being reassessed in the public sphere. They also reveal a changing demographic picture as well as a serious reexamination of race by academics and policymakers that may be having a significant influence on public discourse. The hearings also demonstrated that the government is beginning to question its former views on race and ethnicity and to explore alternative views. The final determination of how race and ethnicity will be measured or viewed in the next century will depend on several factors. Demographic diversity will continue. Individuals and groups will continue to have their own particular and changing views on race. The incidence of intermarriage and the number of interracial individuals will also continue to grow. Consequently, if we are to understand the growing diversity of this country, we must improve our understanding of how people view themselves.

Appendix A Data Limitations and the Undercount

When we examine data on the racial self-classification of individuals, we assume that the data reflect individual choice. But we do not know who fills out the census form and how the “race” of each person in a household is determined. Generally, one person in the household fills out the census forms, but which person that is, the mother, father, eldest son, or whoever, may affect the racial classifications recorded. The data on Hispanics do not include Brazilians, but they do include persons from Spain. Although Brazilians are not considered Hispanic because they do not speak Spanish, many Brazilians consider themselves Latinos, though not usually Hispanics.

THE UNDOCUMENTED AND THE UNDERCOUNT We do not know how many undocumented persons are included in contemporary census data, but we know that Latinos make up a large proportion of the growing numbers of both undocumented and documented immigrants. Different methods yield different estimates of the undocumented, with the total number ranging from 2 million to 5 million (Passel and Woodrow 1984; Woodrow-Lafield 1992, 1993). But because we do not have estimates of how many undocumented are counted in the census data, we must assume that they underestimate the numbers of Latinos, though we do not know by how much. Despite this underestimation, the numbers of Latinos have increased dramatically. As we have noted, in 1997 the U.S. population “officially” contained 29.7 million Latinos, or 11 percent of the total (Reed and Ramirez 1998:table 1). In 1999, the official figure was 31.365 million. The Latino population is also expected to continue to grow substantially. It ac177

178

APPENDIX A

counts for almost half of contemporary immigration (Passel 1993:1076), and it comprises 42.5 percent of the United States’ total foreign-born population (U.S. Bureau of the Census 1993c). This also is a population with a very high birthrate and a youthful median age, so that it would continue to grow even if all immigration were to cease. Somewhat separate from, but also related to, the undocumented issue is the undercount issue. Even though the Latino communities are large, growing, and diverse, Latino undercount rates have not been studied extensively. Hispanics are often not included in analyses because the data used are limited to race. For example, a historical analysis of undercount rates from 1940 to 1990 by Robinson and colleagues (1993) did not include Hispanic undercount rates. The official 1990 undercount estimate for Hispanics is 5.2 percent (Hispanic Link, August 1994, p.2), but this seems conservative, given the large number of estimated undocumented and documented Latino immigrants (Passel 1993:1076; Passel and Woodrow 1984; Woodrow-Lafield 1992, 1993). Given the diversity and rapid growth of the Hispanic population, it is surprising that the Hispanic undercount has not received more attention. In general, undercount rates overall have been found to vary widely by age, race, geography, and homeownership status, with black males having consistently higher rates over time than the total population (Hogan 1993; Robinson et al. 1993). Hogan’s (1993) analysis of 1990 undercount rates indicates that the rates for Hispanics also varied widely. For example, Latinos who did not own property but lived in large urbanized areas of the Northeast were undercounted by an estimated 6.72 percent, but if they lived in similar areas of the Midwest and owned property, they were overcounted by 4.33 percent.

DIFFERENT METHODS, DIFFERENT ESTIMATES Why the undercount rates differ is very complicated, particularly for Latinos because of how they identify or are identified racially. Both may affect the undercount estimates (Passel 1993:1076). If Hispanics are reported as “white” in vital statistics data or if they report themselves as “other” or “black” in census data, they will be counted in these categories and not as “Hispanics.” If Hispanics are reassigned to a “Hispanic” category, estimates of undercount or any other counts will be un-

APPENDIX A

179

derstated (see U.S. Bureau of the Census 1992 for estimates of these effects).1 How Hispanics identify household units also affects undercounts. Examining the census’s Current Population Survey, McKay (1993) found that Hispanics often omitted boarders as part of the household. Or the census takers would not recognize that a housing unit might have more than one household living in it. As do members of many African American communities, respondents distinguished between people who were “living with” and those who were “staying with” a family unit. Thus, boarders or people who were “staying with” them were often not counted. There are many reasons that individuals are not counted in the census. One is that they are missed because they live in irregular and complex household arrangements. Some households were not counted because their living quarters were not visible to postal clerks because they were, for example, behind and/or above a commercial establishment. Other reasons are that people move, fear government and outsiders, and speak English poorly (de la Puente 1993).

OMISSIONS Besides not being counted by the census, respondents may not supply all the information asked for on the census forms. Fein (1990) examined the causes of census omissions among different racial and ethnic groups and found some factors that were common “sources of omission” for different groups and other factors that were unique to particular groups.2 He concluded that because 40 percent of the differences among groups could not be explained by the available indicators, “idiosyncratic social, cultural, and economic aspects of ethnicity constitute major sources of census omission” (Fein 1990:297). Certain groups— blacks, Mexican Americans, and Puerto Ricans—were less motivated to fill out forms than other groups were (non-Hispanic whites, other Hispanics, and Asian and Native Americans).3 Puerto Ricans had the highest omission rate, followed by blacks, Mexicans, other Hispanics, Asians, and Native Americans. Most interesting was Fein’s conclusion that for Puerto Ricans, education seemed to measure something other than the respondent’s ability (1990:296). In other words, those who most often omitted

180

APPENDIX A

information were those with the most education. Thus, although most of the reasons for the undercount are straightforward, such as household structure and fear of government, apparently other, less tangible factors influence it as well.

IMPLICATIONS One result of the scanty research on the Latino undercount is that we do not have a clear view of the parameters of this population and its future impact on the United States. This affects the “visibility”—or the lack of it—of the Latino population at all levels, but particularly at the policy level. More generally, given the size and rapid growth of this population, such data are important to general analyses of the population. Other implications pertain to current practice. Both Passel (1993: 1076) and Vobejda (1991:9) observed that individuals who checked “other race” on the census forms were reclassified to determine undercount estimates. The Census Bureau estimates undercounts by comparing their population counts with birth, death, and immigration records. In order to match the census numbers to those administrative records that do not use an “other race” category, the bureau had to reassign racial categories to those who designated “other race” on the census. As noted earlier, about 43 percent of Hispanics in the nation (or approximately 9 million persons) chose the “other race” option, making up more than 97 percent of the “other race” category. These assignments to standard race categories are increasingly problematic (Passel 1993: 1076), and they also present a picture different from that given by the census without these reassignments. Also not well studied is how this practice affects the socioeconomic and health profiles of the resulting categories. Would, for example, whites’ median income be higher without this reassignment? This may not be a problem at the national level at which the impact is not great, but what about at the local level or in geographic areas where substantial numbers of Latinos live? When just racial data are examined and Hispanics have not been selected out, the question of how the presence of Hispanics in all those categories influences results is seldom addressed or considered. A final implication again involves an undercount. Despite issuing an official undercount estimate, the census figures were not revised to

APPENDIX A

181

include it, thus negatively affecting the representation of Latinos. According to an article in Hispanic Link (August 1994:2), including undercount figures could have benefited districts with large Latino populations. Areas such as southern California, Texas, and New York would likely have made adjustments that could have increased the representation of Latinos. Also, changing the figures might have created an extra congressional seat for both California and Arizona, though at the expense of Wisconsin and Pennsylvania. Moreover, it could have generated more income for Los Angeles and New York City, the latter the primary plaintiff in the undercount suit. (For comprehensive analyses of the litigation of the 1990 undercount suit, see Anderson and Fienberg 1999.)

Appendix B The Biological Concept of Race in the United States

The biological concepts of race and its implicit assumptions have been challenged on a number of grounds.

IT IS ILLOGICAL The concept of “pure races,” which underlies the biological concept of race in the United States, has been shown to be illogical and to be disregarded by many today. For example, in the past when people from two U.S. race categories had a child, the child was placed in the same category as the nonwhite parent; that is, the “hypodescent rule” was applied. The child was considered—as the prefix hypo- implies— less than white; thus he was nonwhite.1 Even though the child was half white and half black, he was placed in the “black” category and was viewed as solely black and not in any way white. In essence, and contrary to logic, even though the child was not purely “black,” he was classified as being so. In the case of a child of two nonwhites, other considerations entered the picture, such as physical type and socialization, but generally black ancestry predominated in classification, and the child was classified as “black.” In neither instance was “purity” a factor, but the classification as only one race contributed to the myth of racial purity. Race was a master status, and considerations of ethnicity or culture were relegated to issues of color and race. Contemporary practice, which allows people to choose their own race, shows that the majority of Asian/white couples state that the race of their children is “white.”2 In the case of Native American/white unions, the race reported for children tends to be consistent with that of 182

APPENDIX B

183

the mother, regardless of whether she is Native American or white. Only in the case of black/white unions are the majority of children given the race of the black parent, regardless of whether this is the father or the mother. Yet even in this case, the majority is just 66 percent (Bennett, McKenney, and Harrison 1995:18, 35–36, 22). Thus, presentday racial self-classifications deviate from earlier approaches, making racial categories even less “pure.”

GROUPS ARE NOT REALLY JUST ONE COLOR Second, the concept of race, particularly as it has been constructed and understood in the United States, has been criticized on the grounds that people are not really one color. For example, people “who are called ‘white’ are really pinkish, light grayish, creamy, very light brown, ruddy reddish and so on. Few are truly white-white. The term is symbolic and non-specific. The same is true of the term ‘black’” (Forbes 1988:95). Lee observed that some “white”-looking people are classified as “black” and that some whites are darker in color than are some nonwhites, for example, Asians; that not all “blacks” are black; and that most whites are really “pinkish-yellowish.” The theory here (i.e., the color categories) differs from the reality. Or as Lee phrased it, there is a variance between the ideology of beliefs and the reality of observations (1993:93).

OTHER VARIABLES AND CONTEXT ARE IMPORTANT According to Forbes, the classification of “many dark, curly-haired Italians, Egyptians and Middle Eastern-South Asian groups” as white has been based more on the recognition of their political and cultural accomplishments than on their physical resemblance to blond Scandinavians. It seems that many of these “so-called Caucasoid groups could just as well be regarded as being non-white or intermediate (mixed) types of people” (1990:6–7). Although color terms are used, cultural, political, and economic variables enter the racial calculus as well, and race often depends on context (Haney López 1996:xiii). Contexts affecting racial classification include the political/legal, social, and economic frameworks that

184

APPENDIX B

encompass individuals and groups. These contexts vary over time, and implicit in these frameworks are variables that influence racial and ethnic classifications, for example, national origin, religion, language, minority status, tribal membership (Bates et al. 1995:433–435), or the variables mentioned in the introduction to this book (Leets, Clement, and Giles 1996:2). In short, considerations other than color or biological ancestry are important to constructing these categories and classifying individuals. For example, many persons are put into the “red” group because they are seen, or they see themselves, as members of a political unit called a tribe or a nation. They are not “red” in color. Others are “black” because their ancestors were classified as “black” or because currently they identify culturally, politically, or spiritually with this group. Some people alter their classification as “white” and become members of other groups (e.g., the “red” group) when they develop a sense of cultural pride or, more crassly, so that they can be considered affirmative-action candidates. Thus, “race” in the United States is not just “color.”

IT IS DISREGARDED BY SCIENTISTS AND ITS SCIENTIFIC UNDERPINNINGS ARE QUESTIONED Third, the assumptions underlying the simple but classical conception of U.S. race as a scientific or biological construct have been intensely questioned. For example, the assumptions that biological race is a given; that hypodescent is a natural law; that some geographic regions are homogeneous; and that ethnicity is less important than color and race have been questioned in both the academic literature and the popular press. Earlier questioning in the academic anthropological community preceded the inquiry in the popular press. A 1989 survey indicated that about 70 percent of cultural anthropologists and 50 percent of physical anthropologists rejected race as a biological category (Begley 1995:67; see also Sanjek 1994). Common to the relatively recent, critical literature on race is a distrust of the validity of race as a biological concept and an increasing awareness of the illogic of the United States’ racial constructions. In much of this literature, experts argue that the differences within groups are often greater than the differences between groups (Begley 1995:67; Gregory and Sanjek 1994:6–7; Gutin 1994:73; Marks 1994,

APPENDIX B

185

1995; Rosin 1994; Shreeve 1994:60; Washburn 1963; Wills 1994:81). As Marks put it, fieldwork revealed, and genetics later quantified, that there is far more biological diversity within any group than between groups (1994:34).3 Moreover, specialists contend that what generally varies from one population to the next is the proportion of people in these groups displaying a particular trait or gene. For example, Marks (1994) pointed out that hair color varies greatly among Europeans and native Australians, but not among other peoples. He also noted that Africans are more biologically diverse than Europeans. At present, both academic and popular research is concerned not with whether there are rigid biological distinctions between races but with when our most recent common ancestors lived and when subpopulations branched off (Goldstein and Morning 1999:5; Lemonick and Dorfman 1999). The sense in this literature is that although there appear to be clearly visible differences among groups, in fact these are just population clusters. In other words, what we use to distinguish groups, for example, pigmentation, eye form, and body build, are anatomical properties, but they are not restricted to particular groups. Rather, they are distributed along geographical gradients, as are nearly all the genetically determined variants detectable in the human gene pool. In essence, the physical characteristics that appear to mark these groups (and that some people see as equivalent to distinct categories or races of people) are distributed along gradients. This variation is gradual, and these gradients span populations. Three factors account for the variation of populations: (1) natural selection, (2) genetic drift, and (3) gene flow. Natural selection occurs as people, that is, populations, adapt to their surroundings. Those who are best able to survive in a particular environment live to reproduce others who will carry their genetic heritage. Through natural selection, populations become differentiated from other populations. Genetic drift also contributes to the variation of groups. Random fluctuations in a gene pool occur as a result of genetic drift. These random fluctuations increase the uniqueness of populations, differentiating them from other populations, but in nonadaptive ways. Finally, as humans have migrated, developed trade networks, and engaged in the political conquests of other populations, intermarriage and “other child-producing unions” have resulted. This has increased the gene flow between populations, making neighboring populations more similar (Marks 1994:34).

186

APPENDIX B

Geographic barriers are also cited as important to creating different development paths for different population groups (Diamond 1997). Consequently, “the racial categories with which we have become so familiar are the result of our imposing arbitrary cultural boundaries in order to partition gradual biological variation” (Marks 1994:34). These culturally constructed categories that we develop and call “races” are discrete and are unlike biological, graduated distinctions. Moreover, our conception of “pure” racial types is also a construction. Thus, for example, very light skinned groups in cold, northern latitudes are not pure types or populations; rather, they are those best adapted to these environments. Despite the recent, extensive criticism of the concept of race and even though “scientists have been broadly unable to come up with any significant set of differences that distinguishes one racial group from another, the controversy over racial differences persists” (Morganthau 1995:64). For example, the recent controversy over racial differences in intelligence was revived—but not settled—by books like the best-selling The Bell Curve by Richard Hernstein and Charles Murray.

Appendix C A Technical Oversight or Racial Flux?

Given that the “color” term and concept endured in the U.S. census for most of the nineteenth century and a large part of the twentieth century, an intriguing question is why the first three censuses did not contain a color term. Related to this question is why the original “all other free persons” category was replaced by the “free colored” category.1

CHANGE IN TERMINOLOGY, NOT DEFINITION One explanation is that the shift to a color category was inconsequential. The category then represented only a very small proportion of the population—1.5 percent (or 59,557) in 1790 (Wright 1956:49). It could be argued that the shift reflected a mere change in terminology but not in definition (Bill Creech, National Archives, conversation, November 1995). In other words, the people in this category were the free colored, even though it was not so named. (This was the position that later censuses took with regard to the meaning of this category. See, e.g., U.S. House of Representatives 1895:xcv–xcvi; U.S. Secretary of the Interior 1853a:table xxxvii; 1853a:xxxvi, 6, 926, tables ix and lxvii.)

Missing or Miscellaneous Information The “all other free persons” category may have included errors and the “don’t knows.” It may have reflected the most generic category possible for what was then a very unstandardized process of data gathering. No specific instructions were issued to census takers until 1820 (U.S. Bureau of the Census 1989:ii), and before 1830, the states designed their own forms and categories. Thus, the information they collected 187

188

APPENDIX C

varied. For the early censuses, marshals submitted their returns to the federal government “in whatever form they found convenient and sometimes with added information” (U.S. Bureau of the Census 1978:7). In addition, many states changed the categories they used from one decennial census to another. For example, the Massachusetts census form of 1810 counted “free white males,” “free white females,” “slaves,” and “all other free persons, except Indians, not taxed.” In its 1820 census, when it was no longer a slave state, Massachusetts kept all these categories except “slaves” and added “free colored persons” and “foreigners, not naturalized.” Massachusetts’s 1820 census also asked for the age and gender of the “free colored persons” (U.S. Bureau of the Census 1978). In other states, of course, the category of “slaves” remained.

Change in Political Leadership Another view is that the shift simply reflected political changes in leadership. From the beginning of the census, some of the founding fathers had been interested in counting the “colors” of the free population. Indeed, all the legislative acts mandating the 1790, 1800, and 1810 censuses specified that the “colors of free persons” be reported (Jackson and Teeples 1976; U.S. Bureau of the Census 1967:43).2 But the census categories did not reflect this requirement, and it is not clear why.3 The 1820 legislation was similar to the earlier legislative acts but also included a substantially different schedule form (U.S. Statutes at Large 1963:548). On this schedule, foreigners were “white,” and the third category (“all other free persons”) was placed under the “colored” category. The emphasis on color in this census is also seen in Secretary of State John Quincy Adams’s instructions to the marshals, in which he specifically noted the need to count the population according to “sex, color, age, condition of life” (U.S. Bureau of the Census 1967:133, italics in original). This letter was published at the front of the 1820 census.4 President James Madison had earlier proposed a more comprehensive first census to Congress, arguing that the census should count the numbers engaged in the different “professions and arts.” He also suggested that Congress count “free blacks” along with “free white men,”

APPENDIX C

189

“free white women,” and “slaves.” (Wright 1956:132–133) Some in Congress opposed a census that did more than count individuals,5 but it may be that President Madison and Secretary of State Adams finally had their way in the 1820 census.

RACIAL FLUX Alternatively, the absence of a color term before 1820 may also have reflected the influence of an earlier period of racial formation in the United States in which there was a great deal of flux and classifications may have been more fluid and differently determined. A number of scholars note that in the very early colonial period, an individual’s status was not “determined solely on the basis of race” (Franklin 1967:225 ff; Higginbotham 1978:22). One’s status as a free person may have been more important than color. Color may also have been viewed differently at that time. It is interesting that of twenty-six reprinted advertisements from the Pennsylvania Gazette for “white indentured servants who fled their masters” (1728–1790), more than a quarter of the runaways—or seven of the twenty-six ads—were described as “brown,” “swarthy,” “dark,” or “of a brownish complexion” (Smith and Wojtowicz 1989:5, 161–172). In addition, Meaders discovered in his more extensive review of runaway indentured servants between 1729 and 1760 in Pennsylvania that there were “a few black and Indian indentured fugitive servants” (1993:xi). In his sample of 1,036 runaway servants and apprentices, he found that 5.79 percent were of unknown or mixed race, Indian, or black (p. 505). Thus the group referred to today as “white” included some who were described then as “not-white” or “dark.” Some colonial figures suggest that there was more heterogeneity during this period than is generally acknowledged today. According to Parrillo’s analysis of colonial data, in 1776, Africans accounted for 39.2 percent of the population of the southern states and 20.5 percent of the total; and African Americans and Native Americans together were more numerous than the English in the South (1994:530). The English also made up only 46.9 percent (less than half) of the total U.S. population of 2.587 million. In addition, Galenson found that whereas in the northern colonies, blacks constituted less than 10 percent of the total

HEADS OF

FA~IILIES-,J\'EW

.i!,:Ir.~ r""""'"''"' "' 1m~ m ,~, '~1 1~f , ;~,! ~

Hi ''''"'I'~ I: ]m··, p1i~J

.N.H!E " ' f[£AU Or F ...!!L>

,~

",~~~z;;;:"~:_::,~· "~:\~-

:i

""j

f~!;;:~::::'c~l, fJ;";.~ "']'~;,;,_--:

''

~{:i~;~i£,~~-::t~'-'''_

·_: --::

phlH· .•. , . • Johmton, J~hn.

CD."'-"" ;,•fo!:U-,U!J.

i

ll,

~w.n!'l"'''"'----------'

C•>J>p. ',\'llliarn • B.,-:mnlnn~. u.-,orgc C"nroilt. l'ctrr..

i::

'i

-~ll&r' Homes_ lotuL Jumec•. Hc,i.,rt

Hr,~~-. :;gn":o"C'g;,;o.:t.

~;~~~~:·,io~;~';;',;l;;.--

Rut\, J•,bn.---

8-oon,

lo, frunel• .. Kc-tth, PB>Id, l'rl••r.J~"'""··

..

bee•~ ..

hell_)·, \;ccrgo .. _ :\Gan .....

Cnrnel:u.< ..

I'Birr~·r.

f.'

(',\lY ••

Ogol\'le. Joh"

Booo~

'

;t~~;~~f~~.~~:~,:;t.,~t:

And