From Nominal to Ordinal:
Reconceiving Racial and Ethnic Hierarchy in the United StatesJennifer Hochschild
For Christina Wolbrecht and Rodney Hero, eds., The Politics of Democratic Inclusion, Philadelphia: Temple University Press, 2005.
NOTE: not quite final version
There are many… variables that are not matters of degree. And it is these variables that define what it means to be black in America.… Police do not stop whites for ‘driving while black,’ but police do stop blacks, particularly wealthy blacks, for this offense. … Thus, it would be wiser to regard ‘driving while black’ and being black not as two variables but, instead, as part of the same condition. It is this second type of variable that forces one to conclude that by definition blacks and whites do not occupy the same social space.
–Samuel Lucas, 2000
I knew some [other black people] who not only had complexions ten shades lighter than that brown paper bag, and hair as straight as any ruler, but also had multiple generations of “good looks,” wealth, and accomplishment…. It was a color thing and a class thing. And for generations of black people, color and class have been inexorably tied together.
–Lawrence Graham, 2000
In the view of Samuel Lucas, the nominal category of blackness so outweighs the distinctions anatomized by Lawrence Graham that variation in skin color has no impact on the degree of blacks’ exclusion from American society. But Graham and many others have pointed to the profound effect that skin color has had within the black – and by extension, other – communities throughout American history. This paper argues that the United States is moving away from the situation that Lucas describes toward a more publicly-relevant version of the society depicted by Graham. That might imply greater inclusion of some Americans previously excluded from our society and polity – but perhaps at the cost of even firmer ostracism of others. How to weigh those equal but opposite trajectories remains a dilemma.
I begin with the observation that our conventional racial and ethnic categories no longer do as much work analytically, politically, or normatively as they used to. Analytically they are blunt, so that they often run roughshod over subtle but important distinctions or must be strung together in awkward locutions in an effort to signal those distinctions (i.e. non-Asian minority, non-Hispanic white, “mixed race black children” [McBride 1996]). Politically, the category “black,” for example, must encompass conservative insiders such as Condoleeza Rice or Clarence Thomas as well as radical outsiders like the followers of Al Sharpton or Sonny Carson, who probably agree with one another on almost nothing of political importance. The category “Hispanic” similarly ranges from Republican Cubans through Democratic Mexican Americans to radical dreamers about Aztlan. (The same, of course, holds for Asians and European Americans.) Normatively, we get tangled in the nuances of different connotations between black or African American or Afrikkan, Latino or Hispanic or Chicano, minorities or people of color, and so on. Economically, a substantial portion of blacks or Hispanics are wealthier and have better jobs than a substantial proportion of whites or Asians. Demographically, an increasing number of Americans simply do not belong in any one of these categories.
This paper begins by articulating more fully the arguments behind my assertion that “our conventional racial and ethnic categories no longer do enough work.” Given that motivation, it then proposes an alternative framework for thinking about the complexities of racial and ethnic hierarchy. If the alternative framework is persuasive, it will suggest a new way of understanding who is included or excluded from real participation in the American polity, and might therefore have implications for racial politics and policy. In short, this exercise seeks to capture a moment in which the United States might be moving to a new and possibly more liberating construction of race and ethnicity – or simply reifying in a more complicated way the age-old barriers among us.
Conventional Racial and Ethnic Categories Are Increasingly Problematic
My argument is predicated on the changing demography of the United States. Almost 20 percent of Americans are immigrants or children of immigrants– a figure that has not been reached since the early days of the twentieth century. Most are from Latin America and they are spreading across the nation from the original gateway cities and states. A substantial influx of immigrants from various nations in Asia and Africa is also transforming particular cities or regions. By 2020, if current immigration policies persist and if current racial and ethnic lines were to remain fixed (which they will not), over a third of the residents of the U.S. will be Asian, black, Latino, or Native American, up from about 15 percent in 1960. Sometime before 2100, if one continues with this exercise, “whites” will be a minority of Americans.
But those figures are artificial, for a variety of reasons. The one most relevant here is what is now virtually an article of faith among most social scientists: race is a social construction. The meaning of the very term, “race” -- as distinguished from ethnicity, culture, nationality, and religion -- has changed over the past few centuries (Sollors 1989; Fredrickson 2002; Hattam forthcoming 2004). Sometimes analysts or actors sharply distinguish a race from an ethnicity or religion; sometimes the two terms are used as synonyms. What groups are perceived to be separate races has similarly varied over the centuries. For example, a Harvard anthropologist published a textbook in 1939 on The Races of Europe which showed 18 races spread across the continent, including “Partially Mongoloid,” “Lappish,” “Brunn strain, Tronder etc., unreduced, only partly brachycephalized,” “Pleistocene Mediterranean Survivor,” “Neo-Danubian,” and so on (Coon 1939). That is, to put it mildly, no longer how we conceive of Europeans. Who is defined in and out of a given race varies over time and occurs for different reasons, including deliberate political choice; Matthew Frye Jacobson, for example, argues that the category of “white” expanded, then contracted, then expanded again during the nineteenth and twentieth centuries (Jacobson 1998; see also Omi and Winant 1994; Ignatiev 1996; Gerstle 1993). American courts struggled mightily for decades with the question of defining who is white, and eventually threw their hands up in despair (Jones 2000; Pascoe 1996; Haney Lopez 1996; Johnson 2003).
Most generally, human races have vastly more shared than distinct genes, and the mixtures of genetic pools over time and across space is extremely complex (Cavalli-Sforza et al. 1994). Partly as a consequence, the American Anthropological Association declared the concept of race to be “a worldview, a body of prejudgments that distorts our ideas about human differences and group behavior,” and recommended that the term be dropped from the census and scholarly writings (American Anthropological Association 1996, 2000). The editorial policy statement of the major medical journals similarly holds that “authors should avoid terms such as ‘race,’ which lacks precise biological meaning, and use alternative descriptors such as ‘ethnicity’ or ‘ethnic group’ instead” (International Committee of Medical Journal Editors 2001).
Many medical professionals and some scientists dissent, arguing that “race” remains a meaningful, even essential, descriptor. Doctors point out that for genetic reasons sickle cell anemia occurs most often among Africans, that northern Europeans are most likely to have lactose intolerance, that Swedes are most susceptible to the mutation that causes hemochromatosis. In their view these differences matter for medical diagnoses and treatment; “a ‘race-neutral’ or ‘color-blind’ approach to biomedical research is neither equitable nor advantageous, and would not lead to a reduction of disparities in disease risk or treatment efficacy between groups” (Risch et al. 2002; see also Satel 2001-2002; add Science Mag. Article 8/02). In 2003, medical researchers at Howard University pledged to collect DNA from 25,000 patients seen by the Howard College of Medicine over the next five years; most are expected to be African American. The dean of the medical school “said genetic information would increasingly find the causes of disease, predict susceptibility to an illness, and choose which drugs would work best for a particular patient” (Pollack 2003). But these are almost the only voices in mainstream academic or professional communities claiming that race is “real.”
The United States census provides another indicator of the complications of the concept of race. It is hard to see the census as an initiator of postmodern identity fragmentation, but that is what it has become. The instructions in 2000 permitted respondents to “mark one or more” of fifteen “races” (including “some other race”), and provided three lines inviting further differentiation. It also included a separate category asking if the respondent is Hispanic (with further subcategories). By combining some of these choices to generate only (!) six races,  and ignoring write-ins, the United States now officially recognizes 63 possible racial categories, 126 if one includes Hispanic or Latino as a “race.” Kenneth Prewitt, the former director of the census, points out that there is no reason to assume that we will not add “partly Hispanic” to the ethnicity question, and perhaps other “races,” such as Middle Eastern or South Asian to the race question in the next census. Any addition will increase the possibilities to 200, 300, or more – and the statistical and political mess now evident in the census will become a total quagmire (Hochschild 2002; Prewitt 2002).
Perhaps the census’s embrace of multiple identities will go the way of its earlier 70-year-long enumeration of mulatto (along with quadroon and octoroon in the 1890 census, Mexican and Hindu in 1930, and an array of Asian nationalities since the late nineteenth century) (Anderson and Fienberg 1999, 174-184). But that seems unlikely. Census respondents now choose their own racial identification rather than having it chosen by the enumerator, which makes a great deal of difference in the connotations of racial labeling. In 2000, fewer than 3 percent of Americans chose more than one race (although 823 people chose all six). Another 12.5 percent reported being Hispanic, along with a racial designation Bureau of the Census 2001). Depending on whether one treats being Hispanic and white as analogous to being black and white, a tiny or small minority of Americans have a mixed racial identity. There was, however, no advertising or advance information provided to alert Americans that they would have this new option on the 2000 census; within a few years, forms for all federal agencies will be required to permit people to “mark one or more,” and most universities, corporations, state and local government agencies, and other large organizations will probably follow suit. Thus the idea is becoming more familiar and may be a commonplace by the next census. Furthermore, the people who did choose more than one race in 2000 were disproportionately young and well-educated. All of this suggests that the proportions choosing multiple racial identities will grow; Prewitt speculates that by 2010, perhaps 10 percent of the population will mark more than one, and by 2020 perhaps a quarter of Americans .
The phenomenon of being multiracial is not new to Americans, of course (Hollinger 2003; Sollors 2000); neither is governmental identification of some Americans as racially mixed. What is new is that multiracial identity has become a point of public pride and assertiveness, and a social movement built around multiracial identity has shown surprising political strength and social cachet. I will return to the issue of multiracialism and democratic inclusion below; my immediate point is that the political and social legitimation of multiraciality blurs nominal racial categories in conception as well as in practice.
If larger proportions of Americans mark more than one race box on the next few censuses, it will not only be because they increasingly accept the concept and understand the procedure. There really will be more multiracial Americans since intermarriage rates and the number of multiracial children is rising. About 7 percent of African Americans marry outside their race (as of 1990), which means that about 13 percent of children with at least one black grandparent have a non-black grandparent as well. To assume that black outmarriage will rise to about 12 percent “is not a radical assumption. Yet if this rise does occur, a fifth to a quarter of children with a black grandparent being born in 2010 would also have a non-black grandparent. And from such levels,… the history of outmarriage in other groups suggests it might well soar within a generation after 2010” (Perlmann 2002, 15-16). African Americans have by far the lowest rate of outmarriage of any non-white group; as of 1990, for example, 36 percent of Asian Americans were married to a member of one of the other three ethnic or racial groups (Qian 1997, table 1). Depending on how multiraciality is defined, analysts estimate that up to 12 percent of youth can now be called multiracial (Goldstein and Morning 2000; Harris and Sim 2002; Hirschman, Alba, and Farley 2000). By 1999, over 20 percent of births to native-born mothers in California were multiracial or multiethnic (Tafoya 2002, figure 3.2); if California is the harbinger of the United States’ racial and ethnic future, as many argue, the number of multiracial Americans will grow rapidly.
Since families do not contain only parents and children, a single intermarriage or interracial child can have an impact well beyond the people most immediately involved. As of 1990, “one in seven whites, one in three blacks, four in five Asians, and more than 19 in 20 American Indians are closely related to someone of a different racial group. Despite an intermarriage rate of about 1 percent, about 20 percent of Americans count someone from a different racial group among their kin” (Goldstein 1999, 399). And those calculations do not include marriages or offspring involving a Latino and a non-Latino.
Individuals and families who are visibly and politically multiracial introduce fluidity into nominal racial categories at the aggregate level; the issue of whether one identifies as a multiracial person introduces another kind of fluidity, at the individual level. The fluid category of multiracial is itself fluid:
When more than 10,000 middle and high school students were asked to report their race on separate school and home surveys, about 12 percent failed to provide consistent responses. Seven percent reported being multiracial on only one of the surveys, and nearly 3 percent of the youth switched between single-race groups. Multiracial reports were almost twice as likely on the school survey… as on the home survey.
Having parents of two different races was neither necessary nor sufficient for a child to call him- or herself biracial on this survey (quotation in Harris 2001; evidence in Harris and Sim 2000; Harris and Sim 2002; Harris 2003).
Fluidity in individual categorization or self-identification is not limited to multiracials. People’s self-identified race may differ from that assigned to them by observers (Harris and Sim 2000; Telles and Lim 1998); people are sometimes assigned different races at birth and at death (Hahn et al. 1992); and the race assigned to them by observers may itself differ depending on the race of the observer (Harris 2002). Over half of young adult children of immigrants in San Diego reported having a different ethnicity, and over a quarter reported a different race, in 2002 than in 1995, barely seven years earlier (Rumbaut 2002, table 6).
If we look at the status of groups across time, we find once again that racial categories are increasingly problematic for understanding the boundaries around inclusion. One can no longer talk simply about “minority status” in the United States because of the astonishing trajectory of Asian Americans. Seventy-five years ago, they personified literal exclusion; almost all were denied the right to immigrate into the United States. Some Asian nationalities were denied the right to own certain types of property. Sixty years ago, most Japanese Americans were interned in World War II, although few German Americans or Italian Americans were. It was barely 40 years ago that members of all Asian nationalities were granted the right to become a naturalized citizen. Movies and cartoons about the yellow peril, Japs, and mysterious Chinese gangsters casually abounded until a few decades ago.
But now Asian Americans are described as the model minority, with very high rates of intermarriage with Anglos and higher average incomes and levels of educational attainment. The most select universities must use informal quotas to keep too many from beating out their nonAsian competitors; a Newsweek cover story promotes the strength and attractiveness of Asian men (Pan 2000); whites who live near a lot of Asians are more likely than those who don’t to endorse increased immigration (Hood and Morris 1997). All is not perfect for Asian Americans of course, and I discuss problems and persistent racial bias below. But we can no longer assume that the status of not being European American is sufficient for exclusion from American society and politics– an astonishing change.
The category of Hispanic or Latino is also decreasingly useful, for different reasons; the term is a conceptual mess. Is it a race, an ethnicity, or simply a bureaucratic convenience? Hispanics may identify as white, black, Native American, or even Asian; over 40 percent rejected all of these choices in favor of “other” on the 2000 census. When Latinos were a tiny fraction of the American population, this anomaly was statistically confusing but politically unimportant in most locations. Now that there are more Hispanics than African Americans, and given that there will be twice as many within our lifetimes, the conceptual confusion is empirically consequential. More generally, the degree to which Latinos are included in American society and politics varies enormously with nationality, location, and immigration status; it is often not helpful to discuss Cubans in Miami and Dominicans in New York with the same terminology.
Even the category of “black” is becoming less useful as a racial category, given increasing levels of immigration from Africa and the West Indies. Dark-skinned immigrants do not necessarily identify as black (or African American), nor are they always seen as black by other Americans (Waters 2001; Kasinitz 1992; Rogers 2001). The choice of identity (by oneself or others) and the degree of political inclusion are mutually causal; political parties and other groups may be more willing to include dark-skinned immigrants than blacks, and immigrants do or do not identify as black partly as a consequence of how included they perceive themselves to be. Thus even the quintessential racial category in American politics is blurring around the edges.
Ironically, the group which until recently received the least attention from social scientists is perhaps the one for which a nominal racial designation is least informative. Some analysts and political activists retain the traditional black-white binary by treating Hispanics like nonHispanic whites, while others make sharp differentiations. We lack settled conventions about whether non-Jewish immigrants from the Middle East or subcontinental Indians are white. In addition, if only because they are such a large share of the American population, whites are perhaps the most varied group in terms of income, political ideology, social practices, religion, and other measures of ethnicity. In simple comparisons across nominal groups, whites epitomize inclusion, but if other dimensions such as class or religion are permitted to intrude, the boundaries must be renegotiated.
An Alternative Conceptualization: Skin Color Hierarchy
Arguably, then, our traditional nominal racial categorization of Americans is not of much help in analyzing the politics of democratic inclusion. But to conclude from this that the United States is now colorblind or approaching the end of racial exclusion would be as bootless as to insist that races are biologically pure or politically insulated. Is there an alternative?
I propose a thought experiment; even if we reject it as a guide to action, we will learn a great deal about patterns of racial inclusion and exclusion in the United States by identifying why and how it fails empirically, emotionally, and politically. What would we gain and lose by moving from nominal categories of Anglo, African American, Latino, and Asian to ordinal measures of skin color as a way of giving order to our understanding of American racial hierarchy?
The basic and most radical hypothesis is simple: the darker a person’s skin color, the lower he or she is likely to be on any scale of whatever is broadly perceived to be desirable in the United States. Thus the basic racial hierarchy runs from European American, to Asian American, to Latino, to African American. More interestingly and controversially, if the hypothesis is correct light-skinned blacks are likely to be better off than dark-skinned Hispanics, light-skinned Hispanics are better off than dark-skinned Asians, and light-skinned Asians are likely to be better off than dark-skinned Anglos.
Qualifications and extensions follow immediately. First, the basic hypothesis will be complicated, though probably not deeply undermined, by gender differences. Second, this is a probabilistic, not a deterministic hypothesis: some dark-skinned people have more of the goods typically desired in our society than some light-skinned people . Most importantly few nationality groups are on average “too high” or “too low” on the skin tone hierarchy given their appearance. Third, neither the content nor the intensity of a person’s racial or ethnic identity tracks the skin-color hierarchy very well; the hierarchy mostly refers to socioeconomic status or treatment by others. In the vocabulary of this paper, there may be a disjunction between individuals’ commitment to a nominal category and their own inadvertent role in creating an ordinal continuum.
Finally, the skin color continuum may be more contextual or historically bounded than the basic hypothesis suggests. In some periods of history, political environments, or social settings, darker-skinned people may be perceived as more authentic, closer to “the people,” or otherwise of higher status than lighter-skinned people. If there are a large number of such “exceptions” that can be contextually and not just idiosyncratically explained, then the basic hypothesis will need to be raised a level of abstraction, to read: “skin color matters much more than we have typically taken into account, but whether dark or light skin tone is associated with advantage will depend on the circumstances.” That will still be interesting and important; I read the evidence, however, to imply that light tone is almost always advantageous so the abstract hypothesis remains less informative than the concrete one.
A softer version of the basic hypothesis melds the nominal and ordinal. That is, it may be that skin tone has (increasing?) importance in determining patterns of inclusion and exclusion within each racial or ethnic category, and that since two of the three non-Anglo groups are gaining in salience in the United States, the whole issue is increasingly important as a public matter – but that the nominal categories persist in creating sharp exclusionary boundaries. The image here is, not one continuum arrayed by skin tone, but four separate continua (or five, if we add Native Americans), each in parallel fashion arrayed by skin tone but remaining distinct from one another. To put the point less abstractly, inclusion is still more likely for dark-skinned European Americans than for light-skinned Asians because Asians still seem ineluctably “foreign” to the majority and dominant population. Similarly, one is still better off as a dark-skinned Hispanic than as an African American, because one lacks the stigma of being black and can substitute the less-stigmatized identity as an ethnic American or immigrant.
The evidence described below lies almost entirely in the realm of the softer hypothesis; with rare exceptions, it compares outcomes by skin tone within a given nominal race or ethnicity. I am compiling evidence about the plausibility of the basic (one-continuum) hypothesis, but in this chapter I largely leave this question open.
A final preliminary point: the use of skin tone to determine who warrants inclusion in a polity is not new. The connection between lightness and virtue is at least as old as Shakespeare; in the United States, one need only point to phrases such as the black Irish or the blue vein test to identify long-standing color hierarchies within groups (Morgan 1975, e.g. 20; Drake and Cayton 1993 ; Lawrence-Lightfoot 1995; Graham 2000). Nor is the focus on skin tone unique to African or European Americans, or to Americans at all. A dark-skinned Dominican reported for a job interview, only to find the interviewer surprised into stammering, “I was looking for someone who looked different, I mean Hispanic, I mean…” (Foner 2000, 158-59; more generally, see Menchaca 2001). According to an ancient Japanese proverb, “white skin makes up for seven defects” (quoted in Wagatsuma 1968, 129).
Before evaluating what is in fact new about skin color hierarchy and whether it implies greater or lesser inclusion in a democratic polity, however, let us look at the evidence of its breadth and depth.
Skin Color Hierarchy within Racial Groups
It is uncontroversial to point out that the level of inclusion within the American polity, society, and economy varies by skin color, with the lightest-skinned nominal racial or ethnic groups most strongly included and the darkest-skinned most firmly excluded. The more interesting question is the degree to which the same skin tone hypothesis obtains within groups. The basic answer is, “a good deal.”
Differentiation in outcomes is one way to measure degrees of inclusion. Strictly speaking, of course, poor or unemployed people are “included” in the economy as much as affluent and fully employed ones. Similarly, in formal terms people who are disfavored as marital partners or lose in a campaign for electoral office are also included within the system, just not on very favorable terms. But the connotation of inclusion is that people are “able to seek and achieve status and influence” rather than “remain[ing] structurally…unequal,” in the words of the introduction to this volume. Thus it is a slide, but not too long a leap, to move from the concepts of inclusion and exclusion to the facts of affluence or poverty, popularity or disfavor, electoral victory or defeat.
By that route, we can conclude that dark-skinned African Americans suffer from even more socioeconomic exclusion than light-skinned ones. To begin with, both African and European Americans use skin tone as a cognitive organizing device (Maddox and Gray 2002), and those cognitions have real impact. Boys who were identified as mulatto in the 1920 census grew up to be more likely to attain a white collar job, at a statistically significant level, than similarly situated boys who were labeled black (Hill 2000). The 1968 Kerner Commission survey of 2,800 blacks in 15 cities showed increasing levels of income, college attendance, and white-collar occupations as respondents’ categorization moved from dark to medium to light skin tone (Edwards 1972). According to the 1979-80 National Survey of Black Americans, light-skinned blacks attain more years of education, higher status jobs, higher incomes, and higher family incomes, than do dark-skinned blacks. All of these differences are statistically significant, and in each case, the movement in five steps from “very dark” to “very light” exactly tracks the progressing levels of success (Keith and Herring 1991; Hunter 2002; Allen et al. 2000).
Three other recent surveys yield similar findings. In the 1982 General Social Survey (GSS), light-skinned blacks received more education, had higher occupational prestige, and enjoyed higher family incomes in their teens than dark-skinned blacks. All of these results are highly significant statistically. This survey showed no difference in current income between dark- and light-skinned African Americans, and minimal to no differences in political ideology, policy views, or levels of alienation (especially when the survey results were controlled for education) (Seltzer and Smith 1991). Analyses of the 1994 Los Angeles Study of Urban Inequality show that “being a dark-skinned African American male reduces the odds of working by 52 percent.” Highly-educated light-skinned black men had an unemployment rate similar to that of comparably-educated white men (10.3 percent and 9.5 percent respectively), whereas dark-skinned men with the same level of education were twice as likely to be unemployed (19.4 percent). The same pattern holds among men who have participated in job training programs. Being a high school dropout, having a criminal record, and living in a very poor neighborhood similarly penalized dark-skinned black men in the labor market even more than light-skinned black men (Johnson et al. 1998). Finally, a 1990s survey of young black adults also found that dark-skinned respondents were more likely to be in the working class than professionals, had lower family incomes, and were more likely to have dropped out of high school (Krieger et al. 1998).
Several of these studies move beyond bivariate relationships into more sophisticated multivariate analyses. The theoretical question here is whether “colorism” is primarily a direct response to the person standing before the potential employer or teacher, or an indirect effect of being able to take advantage of the higher social origins that have accrued over many generations to light-skinned African-Americans. The evidence shows consistently that both phenomena are at work, but that the direct effects of skin tone are more powerful than the indirect effects of being born into a light-skinned black family. As Mark Hill summarizes, “differences in social origin are responsible for only 10 to 20 percent of the color gap in adult attainment” (Hill 2000, 1454; see also Keith and Herring 1991; Hughes and Hertel 1990).
Other scholars, typically with smaller and more opportunistic samples, have shown how skin color affects interpersonal relations among African Americans. Blacks (especially men) perceive light skin to make people more attractive as a dating or marriage partner (Bond and Cash 1992; Ross 1997; Hill 2002; Hunter 2002), or adoptive child (McRoy and Grape 1999; Kennedy 2003). Both blacks and whites attribute more positive psychological traits to light-skinned African Americans (Maddox and Gray 2002).
I know of no systematic research on the relationship of skin color to political participation or electoral success (I am currently conducting such research.) Some analysts have suggested that “looking ‘more white’ in appearance might help Blacks initially as they try to cross the color line in statewide campaigns” by helping them to “blend more easily”(Strickland and Whicker 1992). Governor Douglas Wilder of Virginia is a frequent example. One laboratory experiment similarly showed that highly prejudiced whites with little tendency to monitor themselves for social desirability are indeed likely to prefer lighter-skinned to darker-skinned hypothetical black candidates, whereas whites with less prejudice or more monitoring expressed little or even the reverse preference (Terkildsen 1993). In some cases, in other words, light skin makes it more likely that black candidates will be included in the group of potential office-holders that white voters deem acceptable.
Occasionally the political dynamic is reversed; dark-skinned black candidates persuade (predominantly black) voters that they will include more members of their own race within the fold. The most spectacular recent use of dark skin color as a political tool came in the 2002 mayoral race in Newark, New Jersey, between two self-identified African Americans. Incumbent Sharpe James reportedly called his opponent, Cory Booker, many things – a tool of the Ku Klux Klan and/or Jewish interests, a Republican, and a “faggot white boy.” For whatever reason, James won the race.
Skin color appears to have the same impact among Latinos, although the academic literature is not as extensive or well-developed. With a few exceptions, light tones are associated with greater economic success, more social status, and probably more electoral victories – that is, more inclusion of a positive sort in the American polity.
As before, I begin with economic outcomes. The 1979-80 National Chicano Study showed that dark-skinned Mexican Americans earn less than their light-skinned counterparts, at a statistically significant level in some though not all analyses. They also receive significantly less schooling, are less proficient in English, and are less likely to be unionized workers. The latter results hold even after a long string of controls are added; in some analyses they are stronger for men than for women, and in some the reverse (Telles and Murguia 1990; Hunter 2002; Murguia and Telles 1996; Allen et al. 2000). As with African Americans, skin color appears to directly affect Chicano men’s position in the labor market and earnings more than it indirectly affects their life chances through the human capital acquired through their family of origin (Telles and Murguia 1990; for disagreement based on an alternative methodology, see Bohara and Davila 1992).
Other analyses based on different data find the same patterns. White Hispanic men earn more than black or dark-skinned Hispanic men; white Hispanics with little education or labor market experience succeed more in the job market than similarly situated black Hispanics. These results hold even with standard social and economic controls (Cotton 1997, Gomez 2000).
Spatial location suggests a different dimension of inclusion and exclusion, again to the general disadvantage of Latinos with dark skin. Census data from 1970 and 1980 show that black Hispanics are much more residentially separated from Anglos than are white Hispanics in all of the ten metropolitan areas studied. Black Hispanics are, conversely, much more residentially integrated with African Americans than are white Hispanics. White and black Hispanics are also residentially separated from one another. White Hispanics are just as likely to live near Anglos as near black Hispanics; black Hispanics are more likely to live near U.S. blacks than near white Hispanics. When Hispanics are divided into three groups rather than two (with a self-identified “Spanish race” occupying the space between white and black), levels of separation for the newly-added group fit in between the other two just as the skin color continuum would predict. Finally, the pattern of skin color hierarchy holds even with controls; as we have seen in the case of African Americans, skin color is not merely a stand-in for social class but rather generates a directly discriminatory response on its own (Denton and Massey 1989; see also Relethford et al. 1983).
Evidence on psychological and interpersonal correlates of skin color among Latinos is sparse and not clearly patterned. “Dark Indian-looking men who were born and raised in the United States had significantly higher depression scores” than other Mexican Americans on the National Chicano Study -- but so did lighter European-looking women who had emigrated from Mexico (Montalvo and Codina 2001, 331; see also Codina and Montalvo 1994). Other findings are qualitative or impressionistic or driven by the analyst’s own framework for analysis; this is an arena that awaits further serious research.
Asians and Asian Americans
There is even less systematic research on skin color within the Asian American population. Female South Asian-Canadian university students “wished to be significantly lighter in skin color than they perceived themselves to be, compared with European-Canadian participants” in one study. “The desire to become lighter became greater the further South Asian-Canadian participants perceived themselves to be from the White ideal” (Sahay and Piran 1997). These are not Americans, but arguably students at the University of Toronto are not dissimilar to students at universities in the United States.
Skin color gradations certainly occur in the nations from which Asian immigrants to the United States come. In Japan, members of the upper class have lighter skin tones than do members of the middle or lower classes; the author of this study concludes that “social selection for light skin color has had some genetic effect” in a nation that has been genetically isolated for almost 1500 years (Hulse 1967, quotation in abstract). Japanese and Indonesian university students consistently report that fair skin tones are more beautiful and that people with fair skin present a more likeable impression (Saito et al. 2001). In South Asia, “skin coloring tends to be a minutely calibrated index (of beauty, of Aryan ancestry, often of class)…. Lighter skin coloring is the overwhelmingly preferred shade” (George 1997, XX). Perhaps not surprisingly, therefore, a “preference for light-skinned spouses can be easily verified by the matrimonial columns of some of the Hindu dailies [in India]. Among the most desirable traits a bride can have are virginity and light skin, as evidenced by those placing ads” (Hall 1995, 179). None of this is close to dispositive – but it reinforces rather than contradicting the hypothesis of a skin color continuum within the Asian American population.
Even European Americans, the quintessentially “included” group, have experienced the same relationships between inclusion and skin-color gradations, at least in earlier generations. Compared with southern Europeans, northern immigrants and their descendents enjoyed higher status, greater wealth holdings, and more political power in America until well into the twentieth century. As late as 1950, nationalities of the “old” immigration (from England, Wales, Ireland, Norway, Sweden, and Germany) averaged well over eight years of education in the first generation of immigrants, whereas nationalities of the “new” immigration (from Poland, Czechoslovakia, Austria, Italy, and Russia or the USSR) averaged well under eight years. Except for the Irish, northern Europeans in the old wave of immigration were more likely to be in agricultural or professional occupations and less likely to be in domestic or personal service occupations than those in the new (Lieberson 1980, chapters 4 and 5).
After the first generation, however, white immigrants’ trajectory got more complicated. Occupationally, the pattern was one of continuity: “there is almost an exact correlation between the present-day ethnic participation in agriculture and the levels observed for immigrant groups 80 years earlier.” A similar, though weaker, relationship holds for domestic and personal service workers (Lieberson 1980: XXX). Social status has also been continuous; light-skinned northern Europeans have always had higher standing in the eyes of fellow (white) Americans than have darker-skinned newer immigrants. On the Bogardus ethnic distance scale, for example, all ten of the former were preferred to all eight of the latter from the 1920s through the 1970s. The single exception was Italians, who make it into the top five preferred groups by 1977 (Smith and Dempsey 1983, 588). Even so, more than four times as many GSS respondents in 2000 stated that the English had made “one of the most important positive contributions to this country” as said the same thing about Italians (Davis, Smith, and Marsden 2001, 758-761).
Nevertheless, with regard to educational attainment and the likelihood of holding a professional job, the chief pattern has been discontinuity across immigrant generations. Perhaps most importantly, “the old-new distinction does not operate with respect to distinguishing between the incomes earned by the groups. As a matter of fact,… [average income by 1980] is slightly higher for the six SCE European groups than for the ten northwestern European ones” (Lieberson 1980, 113-137 check pp).
In short, skin color hierarchy used to obtain among European Americans in the same way that it continues to matter among the other nominal racial or ethnic groups. After a century of intermarriage among Europeans, and perhaps as a consequence of the “whitening” of all who are not black, Hispanic, or Asian, skin tone is now a poor predictor of many measures of inclusion for whites. Whether this blending within what is now seen as a single racial group and its subsequent blurring of exclusionary boundaries signals the future for the other groups -- or whether it is a good predictor for Hispanics and Asians but not for blacks, or whether whites will continue to occupy a unique status and pursue a unique dynamic -- remains to be seen.
Evidence for the Thought Experiment: Skin Color Hierarchy across Racial Groups
So far I have demonstrated the exclusionary impact of dark skin tone within nominal racial or ethnic groups; let us turn briefly to the more radical hypothesis of a single skin color continuum across groups. If it holds, light-skinned African Americans do now or will soon have more education and income, higher status jobs, and more social standing than dark-skinned Hispanics, and so on up the chain of skin tone.
I know of no direct measures of this overlap, beyond the research I and my co-authors are currently conducting, and current sources of data will not permit a full comparison. Nevertheless, evidence suggests that the proposed overlap is plausible. First, one of the best studies of the effects of skin color finds that the income gap between light- and dark-skinned African Americans is almost as wide as the gap between African Americans and whites (Hughes and Hertel 1990). Second, census data show clearly that the bottom tails of the distributions of income and education among Latinos, Anglos, and even Asians fall well below the top tails of the distributions among African Americans, and so on up the skin tone hierarchy. Census data, of course, do not include measures of skin tone, but we have seen over and over that lighter-skinned members of each nominal category are better off than darker-skinned members.
Third, multiracial youth, who on average can be expected to have a skin tone somewhere between their darker and lighter parents, occupy socioeconomic positions between the average positions of their two ancestries. “In terms of annual family income and parents’ education and occupation, multiracial blacks are solidly between monoracial whites and monoracial blacks” (Harris 2003, 65; summarizing Kao 1999). Similarly, in suburban Chicago, “monoracial blacks live in neighborhoods with the highest percentage of poor residents (10.4 percent), followed by multiracial blacks (9.1 percent), multiracial whites (7.5 percent), and monoracial whites (5.3 percent)” (Harris 2003, 65; summarizing Corrin and Cook 1999). The families of multiracial children with one Asian parent are on average less wealthy than the families of children with two Asian parents, but not significantly different from the families of children with two white parents. The same pattern holds for parents’ educational levels (Kao 1999). An elaborate study compares adolescents who identify themselves as multiracial at school, home, or both, and disaggregates by the particular racial combination reported (it does not consider Hispanics). Its detailed conclusions can be summarized as following the same pattern; multiracial children fit between their two monoracial groups. That finding holds for family income, parents’ education, and residential separation. Multiracial children live in census tracts that are more racially diverse than is average for people like their lighter-skinned parent, but less racially diverse than is typical for people like their darker-skinned parent (Harris 2003). These results hold across various multiracial pairings.
The evidence on the simple, radical hypothesis is incomplete. Most strikingly, we lack any data on levels of discrimination and social status for light-skinned members of disfavored groups compared with dark-skinned members of more favored groups. So the question of whether patterns of inclusion and exclusion can be boiled down to skin color, tout court, remains open.
Exceptions to the Skin Color Hierarchy
Individuals can move in either direction along the continuum from where they “belong.” Some dark-skinned individuals from disfavored groups are more fully included in the American polity than are almost all light-skinned people of favored groups by virtually any conventional measure; think of Clarence Thomas, Colin Powell, Michael Jordan, Richard Parsons, Oprah Winfrey, Henry Cisneros, Jennifer Lopez, Jackie Chan, and Gary Locke. And some light-skinned individuals are more excluded than almost all dark-skinned people; deeply poor residents of eastern Kentucky, junior members of Asian gangs in Minneapolis, or white imprisoned felons are examples.
An accumulation of individual exceptions could at some point overturn the basic rule of skin color hierarchy within or across groups; I cannot say when that point might be reached or how we would recognize it, but our nation is not now close to it. One might also ask whether “excessively” included exceptions are most likely to be economic (wealthy individuals), political (national leaders), or cultural (sports heroes or stars of popular culture). I also set that question aside for another day.
More important here is the fact that whole nationalities can be located at a different point from where their racial or ethnic group “belongs” on the skin color continuum. For example, as of 1990, compared with 21 percent of Anglos, 60 percent of Asian Indians, 40 percent of Chinese and Filipinos, and 35 percent of Japanese and Koreans had completed college. But barely 5 percent of Cambodians, Laotians, or Hmong had done so – even fewer college graduates than among Hispanics, 9 percent of whom have graduated from college. The same pattern holds for family income, poverty status, and proportions holding a professional or managerial job. The best-off Asian nationalities are lighter-skinned than the worst-off, so the skin-tone hypothesis continues to hold within the Asian race. But Japanese are not usually considered lighter than European Americans or Hmong darker than African Americans – so the standard skin tone hierarchy is violated when Asian nationality groups are compared with whites or blacks. Subcontinental Indians present another complexity; the skin-tone hierarchy holds this group, but it is confounded in complicated ways when Indians are lumped together with other Asians.
The question again arises of when these sorts of nationality-based exceptions would cumulate enough to topple the basic rule that levels of inclusion parallel lightness of complexion. But in this case, as in the case of individuals, the rule is not yet in danger even from these important anomalies.
Do Racial Identity and Commitment Follow the Skin Color Hierarchy?
The main answer to that question is, we do not know. The evidence is sparse, inconsistent, and unsystematic. Some suggests that dark-skinned people correctly perceive that their complexion is associated with a high degree of exclusion, as the skin color hierarchy would predict. In 1968 and 1980, for example, dark-skinned African Americans reported more discrimination than their light-skinned counterparts (Edwards 1972; Keith and Herring 1991; see also Ransford 1970; Brown, et al. 1997). The 1968 survey also showed that dark-skinned blacks perceived whites to be more hostile and untrustworthy and “suggest[ed] a greater sense of black pride or identification with blackness among the darker population.” The differences from light-skinned African Americans were small but consistent, and they persisted when gender or (to a lesser degree) family income were controlled (Edwards 1972). In the 1990s, dark-skinned working class black men perceived more bias from police and courts (Krieger et al. 1998). Among Mexican American students in the 1990s, “dark-skinned subjects, who reported being oriented towards the Mexican culture, identified mostly with the ethnic community and least with the Anglo community” (quotation from Montalvo and Codina 2001, 330; citing Vasquez et al. 1997). Another study similarly found that light-skinned Mexican Americans perceived less discrimination (Arce et al. 1987).
But researchers have also found the reverse, which would work against the skin-color hypothesis. In the CARDIA survey, light-skinned men of all economic classes perceived more bias at school (Krieger et al. 1998). Similarly, in another study “the highest levels of interest in the Latino community occurred among those who self-reported having intermediate skin-color and regarded themselves as bicultural” (Montalvo and Codina 2001, 330; citing Vasquez et al. 1997). One analyst “described whiteness as a ‘bleaching’ agent that could rob her of her culture, language, and Chicana identity if she was not diligent about consistently reasserting it” (Hunter 2002, 179; describing Moraga 1983), and others found that English-speaking Mexican American women were much more likely to call themselves brown and less likely to call themselves white than were their Spanish-speaking counterparts. The color brown “is associated with mestizo heritage.… Choosing ‘brown’ is probably more related to increased awareness of the ethnic group than to racial identification with one’s individual appearance” (Montalvo and Codina 2001, 332; see also Mason and Martinez 1998).
Finally, some studies find little or no effect of skin color on racial or ethnic identity or perceptions of exclusion from mainstream society. In the CARDIA survey, skin color was not associated with reports of racial discrimination in five out of seven specified situations (Krieger et al. 1998; see also Hughes and Hertel 1990). Light-skinned African Americans have taken dark-skinned African Americans to court with claims of color discrimination— and the reverse; so far plaintiffs have won none of these cases (Hiskey 1990; and other cites on court cases). The growth of a politically active multiracial movement, and a proliferation of small groups who identify as multiracial, mulatto, mestizo, or hapa further complicates the picture (Williams 2003; DaCosta – get cite, and published cite for Williams). It is not clear how widespread this new identity is, or how consequential it will be. Nevertheless, defining oneself as a mestizo or hapa and joining a group of like-minded others has a very different connotation than identification as mulatto by a census enumerator a century ago. If these movements grow, they will confound conventional racial or ethnic identifications, and might change the perception and practice of racial exclusion.
Toward a Paradigm Shift
The evidence for a single continuum of skin color that cuts across nominal racial and ethnic categories, or even for multiple parallel skin-color continua within categories, is not tidy. But neither is the world. The politics of democratic inclusion or exclusion are shifting, and we simply cannot predict how far these changes will go and with what political or social consequences. These shifts include: challenges from both within and outside the black community to the one-drop rule for defining African Americans; the growth of multiraciality in fact, political standing, and social and emotional legitimacy; the growth and dissemination of the Latino population, such that most Americans will have to come to grips with the concept and practice of mestizaje; and the changing status of Asian Americans, with the implication that Americans need to reconsider the meaning of “minority” status. For these reasons and more, nominal racial and ethnic categories are blurring around the edges and their social meaning is becoming less and less settled. Skin color has always mattered within racial or ethnic groups, but its import for inclusion or exclusion from the polity may be growing as traditional group categorizations are loosening their hold.
“Skin color hierarchy” encapsulates these profound changes. Whether any of this represents an improvement in the sorry story of American racial domination and exclusion, or is simply another way to reify old racial or ethnic hierarchies remains to be seen. That will depend on demographic changes, the state of the economy and the practice of important institutions, political leadership, and choices by the citizenry not yet conceived of, never mind made. Instead of predictions, then, I conclude with some judgments about the meaning of this shift – if it persists-- for a liberal democratic polity as the United States purports to be.
From the perspective of the individual, moving from a strict and simple racial and ethnic hierarchy of nominal groups to a more fluid continuum within or cutting across groups might be liberating. Children with one black and one Hispanic parent, or one Asian and Anglo parent, need not feel forced to choose one heritage over the other; light-skinned blacks can elide the painful choice to or stigma of “passing,” and dark-skinned whites can avoid the appellation of “wigger” if barriers across races become more permeable. Degrees of inclusion would depend less on physical appearance, which one cannot control. From the vantage point of liberal individualism, this will be an advance.
However, a focus on skin tone rather than race is simultaneously deeply troubling for individual, especially if it does not generate new identities outside of the traditional racial or ethnic groups. Attention to phenotype has an intensely personal quality; categorization as "black" or "white" is cruder and therefore less directly relevant to a person's self-image than categorization as "light-skinned black" or "dark-skinned white." A discussion of skin tone always evokes uncomfortable anecdotes about variations within one’s family or community and its consequences for differential treatment and self-image,
More theoretically, focusing on skin color risks shifting our analytic and behavioral explanations for inclusion or exclusion from structures and institutions onto relationships among people. Political and economic institutions create and maintain the basic nominal hierarchy, reinforced by and reifying racial sentiments; those institutions clearly have a major impact on which individuals and groups are included in the polity. But variations within that racial and ethnic order, and boundary blurrings across its categories, cannot be blamed in any simple way on institutions or external oppression. The variations grow out of, but move beyond, the institutions that structure inclusion and exclusion, superiority and inferiority. It is intensely infuriating to be an “invisible man” in a white-dominated society, but at least that can be explained in terms of something outside oneself. It may be worse to feel oneself to be the object of “the gaze,” or to know oneself to be participating in such a gaze.
In short, from the perspective of the individual, focusing on skin tone, or phenotype more generally, does not make the institutions that foster democratic inclusion or exclusion disappear, but it does make them recede into the background – for better in some cases and for worse in others.
From the perspective of most actors concerned with civil rights and voting rights laws, attention to a skin color continuum is deeply problematic (Williams forthcoming 2004). Representatives of the NAACP, La Raza, and other racial advocacy groups testified against creating a multiracial option on the 2000 census on the grounds that it would make it very difficult to enforce the Voting Rights Act and similar laws. The former chief of racial statistics at the Census bureau concurs, arguing that the new census categorization
leaves the entire statistical mechanism for monitoring and enforcing civil rights vulnerable to a successful challenge in courts…. The nation should certainly determine… whether it still wishes to monitor and enforce equal opportunity legislation using the current methods and systems of assessing disparate impact…. However, the system should not crumble because federal statisticians…[did not] understand… the full ramifications of their revisions…. They are placing that system [of civil rights monitoring and enforcement] in jeopardy (Harrison 2002, 159).
In this view, in other words, moving from a legal and political focus on nominal racial groups to a focus on racial combinations and permutations will reinforce the old and persistent forms of exclusion.
Some proponents of multiracialism do in fact endorse the blurring of racial categories through the census and other means precisely because they anticipate that it will make it more difficult to pursue race-based policies. Prominent examples include former speaker of the House of Representatives Newt Gingrich and the libertarian Center for Equal Opportunity. In their view, racial and ethnic exclusion has been abolished, legally and behaviorally; it is time to move beyond concerns about group exclusion to a focus on individual self-definition, hopefully with little racial content at all.
Just to confuse matters, other proponents of multiracialism are deeply committed to traditional civil rights enforcement, but add that multiracial individuals are themselves excluded from full participation in the American polity and therefore merit their own attention. As Ramona Douglass, former president of the Association of MultiEthnic Americans, put it, the change in the 2000 census “allow[s] for the celebration of diverse heritages, support[s] the continued monitoring of existing civil rights legislation…, and provide[s] the most information for the accurate collection of racial/ethnic data for medical diagnosis and research” (Douglass 2002, 1-2).
So the debate over “mark one or more” involves almost all possible positions: full inclusion requires continued focus on nominal racial or ethnic groups; full inclusion requires us to stop focusing on nominal groups; full inclusion requires continued focus on nominal groups and attention to newly emergent, crosscutting groups. (There is also a small segment of the multiracial community arguing that full inclusion is best attained by focusing only on multiracial groups, which in this view are a large segment of the American population.)
Multiracialism is not the same thing as a skin color hierarchy, although its increased visibility will be necessary for the hierarchy to move from parallel continua within each racial or ethnic group to the radical hypothesis of a single continuum across groups. Multiracials represent the positive possibilities of a single continuum—a nation in which racial identity and identification are part of what matters about individuals and groups, but not dispositive and not confining. Just as fluidity in racial lines can be liberating for individuals, it could possibly become liberating for the nation, moving us closer to David Hollinger’s postethnic America in which race continues to matter, but only as much as or in ways that people want it to (Hollinger 2000). No one can think that we are there yet, but arguably a move from nominal groups to an ordinal continuum is a step in the right direction, to the degree that it suggests that institutions and hierarchies are permeable rather than solid.
A skin color hierarchy, however, can exclude as much as it can include – or possibly it can only include because it simultaneously excludes. One must not forget that a skin color hierarchy is based on, requires, traditional –even atavistic—associations of darkness with low status and lightness with high status. If a skin color hierarchy connotes shades of whiteness rather than multiracial fluidity or cultural mestizaje, then giving flexibility to the racial and ethnic order by allowing people maneuvering room at the top margins of each group simply strengthens that order, just as brittle iron becomes stronger by adding alloys to give it the flexibility of steel. That is, if a continuum (or multiple continua) of skin color simply allows the best-off members of a group to escape any sense of connection with their group or any commitment to abolishing the racial and ethnic hierarchy that permits their inclusion, then individual liberation comes at the considerable cost of group and national reification. In that case, the institutions that have changed enough to make skin color publicly salient are engaged simultaneously in generating more democratic inclusion (for the lucky) and deeper exclusion (for the unlucky). That is not an appealing image.
Allen, W., et al. (2000). “Skin Color, Income and Education: A Comparison of African Americans and Mexican Americans.” National Journal of Sociology 12(1): 129-80.
Anderson, M. and S. Fienberg (1999). Who Counts? The Politics of Census-Taking in Contemporary America. New York: Russell Sage Foundation.
Arce, C., et al. (1987). “Phenotype and Life Chances among Chicanos.” Hispanic Journal of Behavioral Sciences 9(xx): 19-32.
Bohara, A. and A. Davila (1992). “A Reassessment of the Phenotypic Discrimination and Income Differences among Mexican Americans.” Social Science Quarterly 73(1): 114-22.
Bond, S. and T. Cash (1992). “Black Beauty: Skin Color and Body Images among African American College Women.” Journal of Applied Social Psychology 22(11): 874-88.
Brackman, H. and S. Erie (1995). "Beyond 'Politics by Other Means'?: Empowerment Strategies for Los Angeles Asian Pacific Community." in The Bubbling Cauldron: Race, Ethnicity, and the Urban Crisis. ed. Smith, M. P. and J. Feagin Minneapolis: University of Minnesota Press.
Bureau of the Census (2001). The Two or More Races Population: 2000. Washington D.C.: U.S. Department of Commerce,
--- (2002). "Historical Poverty Tables." www.census.gov/hhes/poverty/histpov/hstpov2.html.
Cavalli-Sforza, L. L., et al. (1994). The History and Geography of Human Genes. Princeton NJ: Princeton University Press.
Clegg, R. (2002). personal communication. Hochschild, J. November 19-20.
Codina, G. E. and F. Montalvo (1994). “Chicano Phenotype and Depression.” Hispanic Journal of Behavioral Sciences 16(3): 296-306.
Coon, C. (1939). The Races of Europe. New York: MacMillan Co.
Corrin, W. and T. Cook (1999). Spanning Racial Boundaries: Multiracial Adolescents and Their Families, Peers, Schools, and Neighborhoods. Evanston Illinois: Northwestern University, Institute for Policy Research, Working Paper 99-20.
Cotton, J. (1997). "Color or Culture? Wage Differences among Non-Hispanic Black Males, Hispanic Black Males, and Hispanic White Males." in African Americans and Post-Industrial Labor Markets. ed. Stewart, J. New Brunswick, NJ: Transaction. 61-75.
Denton, N. and D. Massey (1989). “Racial Identity among Caribbean Hispanics: The Effect of Double Minority Status on Residential Segregation.” American Sociological Review 54(5): 790-808.
Douglass, R. (2002). "Shaping the Multiracial Experience: AMEA Past, Present, and Future". AMEA's Multiracial Child Conference, Tucson, Arizona.
Drake, S. C. and H. Cayton (1993 ). Black Metropolis: A Study of Negro Life in a Northern City. Chicago: University of Chicago Press.
Edwards, O. (1972). “Skin Color as a Variable in Racial Attitudes of Black Urbanites.” Journal of Black Studies 3(4): 473-83.
Foner, N. (2000). From Ellis Island to JFK : New York's Two Great Waves of Immigration. New Haven
New York: Yale University Press ;
Russell Sage Foundation.
Fredrickson, G. (2002). Racism: A Short History. Princeton NJ: Princeton University Press.
Frey, W. and E. Fielding (1995). “Changing Urban Populations: Regional Restructuring, Racial Polarization, and Poverty Concentration.” Cityscapes 1(1-65.
George, R. (1997). “"From Expatriate Aristocrat to Immigrant Nobody": South Asian Racial Strategies in the Southern California Context.” Diaspora 6(1): 31-60.
Gerstle, G. (1993). “The Working Class Goes to War.” Mid-America 75(3): 303-22.
Goldstein, J. (1999). “Kinship Networks That Cross Racial Lines: The Exception or the Rule?” Demography 36(3): 399-407.
Goldstein, J. and A. Morning (2000). “The Multiple-Race Population of the United States: Issues and Estimates.” Proceedings of the National Academy of Sciences 97(11): 6230-35.
Gomez, C. (2000). “The Continual Significance of Skin Color: An Exploratory Study of Latinos in the Northeast.” Hispanic Journal of Behavioral Sciences 22(1): 94-103.
Graham, L. (2000). Our Kind of People: Inside America's Black Upper Class. New York: Harper.
Hahn, R., et al. (1992). “Inconsistencies in Coding of Race and Ethnicity Between Birth and Death in U.S. Infants: A New Look at Infant Mortality, 1983 through 1985".” Journal of the American Medical Association 267(2): 259-63.
Hall, R. (1995). “The Bleaching Syndrome: African Americans' Response to Cultural Domination Vis-a-Vis Skin Color.” Journal of Black Studies 26(2): 172-84.
Haney Lopez, I. (1996). White by Law: The Legal Construction of Race. New York: New York University.
Harris, D. (2002). In the Eye of the Beholder: Observed Race and Observer Characteristics. Ann Arbor MI: University of Michigan, Institute for Social Research,
--- (2003). "Does It Matter How We Measure? Racial Classification and the Characteristics of Multiracial Youth." in The New Race Question: How the Census Counts Multiracial Individuals. ed. Perlmann, J. and M. Waters New York: Russell Sage Foundation Press. 62-101.
Harris, D. and J. Sim (2000). An Empirical Look at the Social Construction of Race: The Case of Mixed-Race Adolescents. Ann Arbor MI: University of Michigan, Population Studies Center, Research Report 00-452.
--- (2002). “Who Is Multiracial? Assessing the Complexity of Lived Race.” American Sociological Review 67(4): 614-27.
Harrison, R. (2002). "Inadequacies of Multiple-Response Race Data in the Federal Statistical System." in The New Race Question: How the Census Counts Multiracial Individuals. ed. Perlmann, J. and M. Waters New York: Russell Sage Foundation. 137-60.
Hattam, V. (forthcoming 2004). “ Whiteness: Theorizing Race, Eliding Ethnicity.” International Labor and Working-Class History xx(xx): xx.
Hill, M. (2000). “Color Differences in the Socioeconomic Status of African American Men: Results of a Longitudinal Study.” Social Forces 78(4): 1437-60.
--- (2002). “Skin Color and the Perception of Attractiveness among African Americans: Does Gender Make a Difference?” Social Psychology Quarterly 65(1): 77-91.
Hochschild, J. (2002). "Multiple Racial Identifiers in the 2000 Census, and Then What?" in The New Race Question: How the Census Counts Multiracial Individuals. ed. Perlmann, J. and M. Waters New York: Russell Sage Foundation. 340-53.
Hollinger, D. (2000). Postethnic America: Beyond Multiculturalism. New York: Basic Books.
--- (2003). “Amalgamation and Hypodescent: The Question of Ethnoracial Mixture in the History of the United States.” American Historical Review 108(5): 1363-90.
Hood, M. V., III and I. Morris (1997). “Amigo o Enemigo? [reversed ? before Amigo]: Context, Attitudes, and Anglo Public Opinion toward Immigration.” Social Science Quarterly 78(2): 309-23.
Hughes, M. and B. Hertel (1990). “The Significance of Color Remains: A Study of Life Chances, Mate Selection, and Ethnic Consciousness Among Black Americans.” Social Forces 68(4): 1105-20.
Hulse, F. (1967). “Selection for Skin Color among the Japanese.” American Journal of Physical Anthropology 27(xx): 143-56.
Hunter, M. (2002). “"If You're Light You're Alright:" Light Skin Color as Social Capital for Women of Color.” Gender & Society 16(2): 175-93.
Ignatiev, N. (1996). How the Irish Became White. New York: Routledge.
Jacobson, M. (1998). Whiteness of a Different Color : European Immigrants and the Alchemy of Race. Cambridge MA: Harvard University Press.
Johnson, J. J., et al. (1998). “The Declining Social and Economic Fortunes of African American Males: A Critical Assessment of Four Perspectives.” Review of Black Political Economy 25(4): 17-40.
Johnson, K., Ed. (2003). Mixed Race America and the Law. New York New York University.
Jones, T. (2000). “Shades of Brown: The Law of Skin Color.” Duke Law Journal 49(xx): 1487-557.
Kao, G. (1999). “Racial Identity and Academic Performance: An Examination of Biracial Asian and African Youth.” Journal of Asian American Studies 2(3): 223-49.
Kasinitz, P. (1992). Caribbean New York: Black Immigrants and the Politics of Race. Ithaca NY: Cornell University Press.
Keith, V. and C. Herring (1991). “Skin Tone and Stratification in the Black Community.” American Journal of Sociology 97(3): 760-78.
Kennedy, R. (2003). Interracial Intimacies: Sex, Marriage, Identity, and Adoption. New York: Pantheon Books.
Krieger, N., et al. (1998). “Racial Discrimination and Skin Color in the CARDIA Study: Implications for Public Health Research.” American Journal of Public Health 88(9): 1308-13.
Lawrence-Lightfoot, S. (1995). Balm in Gilead: Journey of a Healer. New York: Penguin.
Lieberson, S. (1980). A Piece of the Pie: Blacks and White Immigrants Since 1880. Berkeley: University of California Press.
Lien, P.-t. (1997). “Does Under-Participation Matter? An Examination of Policy Opinions Among Asian Americans.” Asian American Policy Review 7(xx): 38-54.
Lucas, S. (2000). “Hope, Anguish, and the Problem of Our Time: An Essay on Publication of The Black-White Test Score Gap.” Teachers College Record xx(xx): xx.
Maddox, K. and S. Gray (2002). “Cognitive Representations of Black Americans: Reexploring the Role of Skin Tone.” Personality and Social Psychology Bulletin 28(2): 250-59.
Mason, P. and P. Martinez (1998). "Is Race Endogenous? Some Preliminary Evidence on Latinos." www.nd.edu/~pmason/paperMasnMart.pdf.
McBride, J. (1996). The Color of Water: A Black Man's Tribute to His White Mother. New York: Riverhead Books, Putnam's.
McRoy, R. and H. Grape (1999). “Skin Color in Transracial and Inracial Adoptive Placements: Implications for Special Needs Adoptions.” Child Welfare 78(5): 673-92.
Menchaca, M. (2001). Recovering History, Constructing Race: The Indian, Black, and White Roots of Mexican Americans. Austin TX: University of Texas Press.
Montalvo, F. and G. E. Codina (2001). “Skin Color and Latinos in the United States.” Ethnicities 1(3): 321-41.
Moraga, C. (1983). "La Guera." in This Bridge Called My Back: Radical Writings by Women of Color. ed. Anzaldua, G. and C. Moraga New York: Kitchen Table Press. xxx.
Morgan, E. (1975). American Slavery, American Freedom: The Ordeal of Colonial Virginia. New York: Norton.
Murguia, E. and E. Telles (1996). “Phenotype and Schooling among Mexican Americans.” Sociology of Education 69(4): 276-89.
Omi, M. and H. Winant (1994). Racial Formation in the United States. New York: Routledge.
Pascoe, P. (1996). “Miscegenation Law, Court Cases, and Ideologies of 'Race' in Twentieth-Century America.” Journal of American History 83(1): 44-69.
Perlmann, J. (2002). The Intermingling of Peoples in the United States: Intermarriage and the Population History of Ethnic and Racial Groups since 1880Levy Economics Institute of Bard College: Proposal to Russell Sage Foundation.
Perlmann, J. and M. Waters (2002). "Introduction." in The New Race Question: How the Census Counts Multiracial Individuals. ed. Perlmann, J. and M. Waters New York: Russell Sage Foundation. 1-30.
Pollack, A. (2003). "DNA of Blacks To Be Gathered to Fight Illness." New York Times, May 27: A1.
Prewitt, K. (2002). "Race in the 2000 Census: A Turning Point." in The New Race Question: How the Census Counts Multiracial Individuals. ed. Perlmann, J. and M. Waters New York: Russell Sage Foundation. 354-61.
Qian, Z. (1997). “Breaking the Racial Barriers: Variations in Interracial Marriage Between 1980 and 1990.” Demography 34(2): 263-76.
Ransford, H. E. (1970). “Skin Color, Life Chances, and Anti-White Attitudes.” Social Problems 18(2): 164-79.
Relethford, J., et al. (1983). “Social Class, Admixture, and Skin Color Variation in Mexican Americans and Anglo-Americans Living in San Antonio, Texas.” American Journal of Physical Anthropology 61(xx): 97-102.
Risch, N., et al. (2002). “Categorization of Humans in Biomedical Research: Genes, Race, and Disease.” Genome Biology 3(7): 1-12.
Rogers, R. (2001). "Black Like Who? Afro-Caribbean Immigrants, African-Americans, and the Politics of Group Identity." in Islands in the City: West Indian Migration to New York. ed. Foner, N. Berkeley, CA: University of California Press. 163-92.
Ross, L. (1997). “Mate Selection Preferences among African American College Students.” Journal of Black Studies 27(4): 554-69.
Rumbaut, R. n. (2002). Handout on "Inequality and Achievement Paths among Children of Immigrants in Early Adulthood: A Decade-Long Panel Study". December 2Cambridge MA: Harvard University, Kennedy School of Government.
Sahay, S. and N. Piran (1997). “Skin-Color Preferences and Body Satisfaction Among South Asian-Canadian and European-Canadian Female University Students.” Journal of Social Psychology 137(2): 161-71.
Saito, M., et al. (2001). “Beyond the Pale.” SPC Asia xx(25): 19-xx.
Satel, S. (2001-2002). “Medicine's Race Problem.” Policy Review: 49-58.
Seltzer, R. and R. Smith (1991). “Color Differences in the Afro-American Community and the Differences They Make.” Journal of Black Studies 21(3): 279-86.
Smith, T. and G. Dempsey (1983). “Ethnic Social Distance and Prejudice.” Public Opinion Quarterly 47(4): 584-600.
Sollors, W., Ed. (1989). The Invention of Ethnicity. New York Oxford University Press.
--- (2000). Interracialism : Black-white Intermarriage in American History, Literature, and Law. New York: Oxford University Press.
Strickland, R. A. and M. L. Whicker (1992). “Comparing the Wilder and Gantt Campaigns: A Model for Black Candidate Success in Statewide Elections.” PS: Political Science and Politics xx(June): 204-12.
Tafoya, S. (2002). "Mixed Race and Ethnicity in California." in The New Race Question: How the Census Counts Multiracial Individuals. ed. Perlmann, J. and M. Waters New York: Russell Sage Foundation. 102-15.
Telles, E. and N. Lim (1998). “Does It Matter Who Answers the Race Question? Racial Classification and Income Inequality in Brazil".” Demography 35(4): 465-74.
Telles, E. and E. Murguia (1990). “Phenotypic Discrimination and Income Differences among Mexican Americans.” Social Science Quarterly 71(4): 682-96.
Terkildsen, N. (1993). “When White Voters Evaluate Black Candidates: The Processing Implications of Candidate Skin Color, Prejudice, and Self-Monitoring.” American Journal of Political Science 37(4): 1932-053.
U.S. House of Representatives (1997). Hearings on "Federal Measures of Race and Ethnicity and the Implications for the 2000 Census". Subcommittee on Government Management, Information, and Technology of the Committee on Gov't Reform and Oversight. April 23, May 22, July 25.
Vasquez, L., et al. (1997). “Skin Color, Acculturation, and Community Interest in Mexican American Students: A Research Note.” Hispanic Journal of Behavioral Sciences 19(3): 377-86.
Wagatsuma, H. (1968). "The Social Perception of Skin Color in Japan." in Color and Race. ed. Franklin, J. H. Boston MA: Houghton Mifflin Co. 129-65.
Waters, M. (2001). Black Identities: West Indian Immigrant Dreams and American Realities. Cambridge MA: Harvard University Press.
Williams, K. (2003). The Next Step in Civil Rights? The American Multiracial MovementCambridge MA: Harvard University, John F. Kennedy School of Government.
--- (forthcoming 2004). Race Counts: American Multiracialism & Civil Rights Politics. Ann Arbor MI: University of Michigan Press.
Wyly, E. (1997). Social Trends and the State of the Nation's Cities. New Brunswick NJ: Rutgers University, Center for Urban Policy Research,
 My thanks to participants in the conference on “The Politics of Democratic Inclusion,” University of Notre Dame, October 18-19, 2002; participants in the Cluster on Immigrant Incorporation at the Radcliffe Institute for Advanced Studies; Victoria Hattam, Vesla Weaver, Traci Burch, and two anonymous reviewers for very helpful comments and suggestions.
 Lucas 2000: XX; Graham 2000, 4.
 Other reasons for mistrusting these projections include the fact that they depend on a continuation of current immigration laws, they pay little or no attention to differing and growing levels of intermarriage, and they necessarily make heroic assumptions about birth and death rates.
 There are no new ideas; in the first paragraph of the introduction, Coon writes, “if there is one consistent theme in this book, it is that physical anthropology cannot be divorced from cultural and historical associations, and that there is no such thing as ‘pure’ biology, at least in reference to human beings” (p. vii).
 The census bureau grouped seven categories into “Asian,” and another four into “Native Hawaiian and Other Pacific Islander.”
 “The Mexican government responded with an official protest to the effect that all Mexicans are white and the category was dropped for the next few censuses” (Perlmann and Waters 2002: 5).
 Almost half of Hispanics identified as white, over 40 percent chose “some other race,” 4 percent chose black or one of the other named races, and about 6 percent reported two or more races.
 This analysis did not include Native Americans, well over half of whom marry non-Native Americans.
 These analyses did not treat Hispanics as a distinct category; doing so would presumably raise all of the numbers here.
 There is a complicated relationship between skin color and broader phenotypical qualities that make a person look more or less European. For much of the following discussion, it would be appropriate to substitute “a more European appearance” for “lighter skin color.” That substitution does not hold, however, for Asian Americans, who may be as light as or lighter than Europeans but do not have Anglo features. By focusing on skin color per se, I will ignore this complexity in this chapter.
 This point raises the question of how to determine when changes stop being exceptions to a rule and become instead an indication that the rule no longer holds. This whole paper is an assertion that the paradigm of nominal racial or ethnic groups is so riddled by exceptions and changes that we should consider substituting a new one, or that a new paradigm is in the making through people’s behavior regardless of our constructions. But I cannot identify in the abstract the point at which the weight of more and more exceptions topples a standing rule.
 Gold, says poor King Timon of Athens, “will make black white, foul fair, wrong right, base noble, old young, coward valiant” (Timon of Athens, act IV, scene iii).
 Asians’ preference for light skin may be partly due to the western colonial influence over the past few hundred years, but that is not all of the story. Centuries before westerners were known, Japanese literature depicted white skin as the most beautiful; upon first meeting westerners, in fact, Japanese in the 1960s often denigrated them as insufficiently light (Wagatsuma 1968).
 Vesla Weaver and Traci Burch, both of Harvard University, and I are examining patterns of skin color overlap among groups, to the degree that it is possible with extant survey data. We will also analyze various status distributions for people who identify with more than one race.
 For example, 49% of nonHispanic white adults have a high school education or less, as do 41% of Asians and 73% of Latinos. Yet 40% of African American adults have at least some college, and 14% have a college degree or more. CITE
 In recent Gallup polls, Powell was the most popular political figure in the United States (Moore 2002).
 From 1990 through 2001, roughly 9 percent of Anglos were poor; comparable rates for blacks averaged 28 percent, for Latinos 29 percent, and for Asian Americans, 13 percent (Bureau of the Census 2002: Table 2). Fully 30 percent of Samoans and Vietnamese, however, were poor during the 1990s (Frey and Fielding 1995; Lien 1997; table 1-2; see also Brackman and Erie 1995; Wyly 1997; table 3). Obviously, these outcomes are totally confounded with home country institutions and recency of immigration, so I am not making any causal claims here.
 See testimony or written statements by Rep. Eleanor Holmes Norton, Harold McDougall (director of the Washington bureau, NAACP), Eric Rodriguez (policy analyst for National Council of La Raza), JoAnn Chase (executive director of National Congress of American Indians), Jacinta Ma (legal fellow of National Asian Pacific American Legal Consortium), in U.S. House of Representatives 1997.
 Clegg 2002 [general counsel for Center for Equal Opportunity]. Speaker Gingrich testified in the census hearings on the multiracial option that “it is wrong for some Americans to begin creating subgroups to which they have a higher loyalty than to America at large…. America is too big and too diverse to categorize each and every one of us into four (sic) rigid racial categories…. Ideally, I believe we should have one box on federal forms that simply reads, ‘American.’ But if that is not possible at this point,… allow[ing] them [i.e. Americans] the option of selecting the category ‘multiracial”,… will be an important step toward transcending racial division and reflecting the melting pot which is America” (U.S. House of Representatives 1997, 661-662).
 Thinking in terms of a skin color hierarchy, or even parallel continua within racial and ethnic groups, also has the advantage of allowing us to analyze the different trajectories of racial groups within the same theoretical framework. That might be an advantage that only an academic could love, but it is not one lightly to be thrown away absent a better broad framework.