Common Sense, Common Census: Should America Follow France and Abolish Race Questions on the Census?
PlanHaut de page
“The truth is that there are no races: there is nothing in the world that can do all we ask race to do for us.”
Anthony Appiah, In My Father’s House
“We feel that if all the possible scientific questions be answered, the problems of life have still not been touched at all.”
Luwig Wittgenstein, Tractatus Logico-Philosophicus
1We, the People. The opening words of the American Constitution intimate a promise of democracy, a radical idea at odds with the structural features of the document. But the provisions for constitutional amendment and for free speech did establish the framework, however flawed, for progressively more democratic self-government. Another key mechanism for sustaining the momentum of democratic enlargement was the Article I mandate to the national government to find out, on a decennial basis, precisely who “we, the people” actually are :
- 1 U.S. Const., Art. I, section 2.
The actual Enumeration shall be made within three Years after the first Meeting of the Congress of the United States, and within every subsequent Term of ten years, in such Manner as they shall by Law direct1.
- 2 See Reynolds v. Sims, Wesberry v. Sanders. Of course, even when the “one person, one vote” doctri (...)
2This “actual Enumeration” census requirement in Article I anticipated by nearly two centuries the emergence of the “one person one vote” doctrine in the Warren Court, which interpreted Equal Protection to require that each and every adult American in the states have the right to vote in federal and state elections on an equal basis2. While the census clause commanded the society to hold a mirror up to itself, the Court, in Reynolds v. Sims and Wesberry v. Sanders, insisted that we become politically true to the image we see.
- 3 “Representatives and direct taxes shall be apportioned among the several States which may be incl (...)
- 4 73 N.Y.U.L. Rev. 333, 433.
3Yet, like the rest of the Constitution, even the census clause was tainted from the start by association with the Framers’ original fateful compromise with slavery. The sentence commanding a census every ten years followed directly upon the sentence announcing the infamous “three fifths” formula for counting slaves for the purpose of congressionalreapportionment3. This concession tothe slavemasters inflated the political power of the white oligarchs, who (needless to say) disenfranchised their slaves but happily counted 60 % of their bodies to increase their own representation in Congress and the presidential electoral college. The resulting “slave power” of the South, as explained by Garry Wills, not only provided the margin of victory in the presidential election of 1800, among others, but assured the dominance of pro-slavery politics in Congress for decades4.
4The interacting power of the three-fifths clause and the census clause gave rise and sanction to two critical institutional mechanisms of white supremacy in our history: the mass disenfranchisement of African-Americans and the development of a racialized census.
- 5 Shaw v. Reno , 509 U.S. 630 (1993).
- 6 See Jamin Raskin, 3 Election Law Journal 559 (2004).
- 7 See id.
- 8 14 Cardozo L. Rev. 1135.
5The first mechanism, racial disenfranchisement, has been gradually dismantled through the Civil War, the Thirteenth, Fourteenth and Fifteenth Amendments, and the passage of the Voting Rights Act of 1965. However, its doctrinal legacy survives in two ways. The first is in the Supreme Court’s Shaw v. Reno doctrine, which knocked the breath out of the Voting Rights Act by essentially constitutionalizing the right of white voters to be in a racial majority in single-member districts except where African-Americans and Hispanics can form a majority in a compact way that is aesthetically pleasing to the Court’s majority5. Second, despite the substantial gains we have made in black enfranchisement, we are still contending with the glaring absence of a constitutional right to vote, a reality invoked by the Supreme Court majority in Bush v. Gore that explains the continuing disenfranchisement of millions of American citizens, including people living in Washington, D.C. and in Puerto Rico and people convicted of crimes in a number of states6. Elsewhere I have argued that our recurring failure to protect the right to vote and to be represented in our fragmented federalist polity compels passage of a constitutional amendment guaranteeing all of the people the right to vote and to have their votes count7. This is a matter of democratic urgency. Similarly, the Supreme Court’s racial double standard with respect to redistricting forces those who take democracy seriously to consider more effective ways to guarantee the political voices of everyone, and here, as Lani Guinier has argued, it is essential to move away from single-member districting toward mechanisms of proportional representation that are used in most other democracies8.
6But it is also crucial to see that the racialized census of our early days has survived intact. The census carries into the present the ideologies of race that bolstered European colonialism, infused the institution of American slavery and rationalized white supremacy.
- 9 Melissa Nobles, “Racial Categorization and Censuses,” in Census and Identity : The Politics of Ra (...)
- 10 Id.
- 11 Id.
7Ironically, the constitutional census clause itself did not explicitly command a racial inventory of the population. Although it made reference to “Indians not taxed,” the actual text of the census clause was oblique and allusive on the subject of race. Slaves are never identified racially and only enter the picture in the semantic mist, as “three fifths of all other Persons.” As Melissa Nobles has observed, it would have been possible for the new government to comply with the census clause by simply dividing respondents into slaves and free citizens and not making explicit use of race. But in our first national census in 1790, the United States government quickly worked out racialized categories for our population focused on the baseline norm of whiteness : “Free White Males ; Free White Females ; All Other Free Persons ; Slaves9.” Our early censuses thus “registered race as it was then understood10.” Race was used, Nobles argues, because it “was a salient social and political category. Eighteenth-century political elites regarded race as a natural and self-evident component of human identity, in keeping with European Enlightenment thought11.”
- 12 Melissa Nobles, “Racial Categorization and Censuses,” in Census and Identity : The Politics of Ra (...)
- 13 Id. at 49.
8In this sense, race was used as an organizing ideology of the new nation, and so it has been, more or less, ever since. Since 1790, we have continually placed racial categories on the census and sent out enumerators – or, since 1960, invited self-identification – to fit everyone into the right category or (beginning in 2000) categories12. The official purpose of the census has been to take a kind of prose snapshot of the population. But it has also served as the active tuning rod and experimental journal of the nation’s racial ideologies, obsessions and neuroses. Nobles usefully divides the history of our census into four periods : 1790 to 1840, when census racial categories were “shaped by representational apportionment, slavery, and racial ideas” ; 1850 to 1920, when they were “used expressly to advance the racial theories of scientists,” most prominently certain bizarre speculations about the hybrid qualities of “mulattos” offered by the reigning polygenist school of thought ; 1930 to 1960, “when census definitions of racial categories were identical to those of southern race laws,” and 1970 to the present, a period in which racial categorization “has been shaped most profoundly by civil rights legislation, the implementation of Statistical Directive No. 15, and the lobbying efforts of organized groups13.” Statistical Directive No. 15 was released by the Office of Management and Budget in 1977 to define the five standard categories of racial and ethnic statistical reporting by the Census Bureau and other federal agencies : “American Indian or Alaskan Native,” “Asian or Pacific Islander,” “Black,” “Hispanic,” or “White”. These census categories have been used to assemble information useful for civil rights enforcement and affirmative action policies in the fields of housing, education, employment, health care, and voting rights.
9If, in the past, the census has been a decennial occasion for American political society to interrogate its beliefs, fears and anxieties about race, the 21st century will apparently be no exception. In the lead-up to the 2000 census, there was great political and academic controversy focused on the proposal to add a “multiracial” category to the census choices offered to respondents. The underlying policy debate was between people who strongly identify themselves and/or their children as “multiracial” and leading civil rights advocates who feared that this proposed new category would undermine group political solidarity and weaken thenation’s ability to enforce civil rights law.
10The “multiracial category” debate captured vividly what could emerge as the new century’s agony around race in America. There is something close to scientific and academic consensus that biological race is not, and never was, a defensible scientific idea and that it is destructive to people in many ways. Yet, because the social and economic injuries of race are so profound and the idea of political organizing along received race lines has at times been used constructively in the struggle against white supremacy, many Americans of diverse backgrounds will be terrified of letting go of the racial idea for public purposes, however artificial and “constructed” they take it to be. The question then becomes whether there is any way that America can transcend both race (the proclaimed goal of conservatives today) and also racism (the traditional goal of progressives and liberals). Is there some means by which we can abolish race while continuing the emancipation of its countless victims in American society ? Can the Constitution play a constructive role in resolving this conundrum for us ?
11In Part II, I analyze the ferocious debate over the “multiracial” category before the 2000 census to demonstrate how both sides had important stories to tell but essentially spoke pasteach other : the advocates of the new category spoke from the standpoint of the individual presentlyfacingthe requirement of an existentially false racial self-identification while the opponents spoke from the historical vantage point of large numbers of people living with the cumulative injuries of racism and seeking redress. The striking aspect of the debate is that everyone involved acknowledges the scientific falsity of “race” but is still necessarily trapped in the race paradigm set by our original constitutional embrace of slavery and white supremacy. It is not clear that we the people, as a nation, can ever escape the awful power of this racial ideology. The most important and most valid reason we hang on to it today is found in the widely shared belief that it is necessary to maintain race on the census in order to advance civil rights enforcement and affirmative action.
12But in Part III, I closely interrogate this claim, using standard equal protection doctrine involving strict scrutiny of race-based classification. I argue that maintaining the essentially neurotic idea of race is not nearly so necessary to programs of social progress as we may think. Our most important civil rights agendas – effective enforcement of anti-discrimination law, affirmative action for diversity in higher education, and effective legislative representation for political minorities – in truth may not depend at all, or may depend far less than we imagine, on a racialized census. Thus, the idea that racial questions on the census narrowly serve compelling public ends in the least restrictive way, as required by the mechanism of strict scrutiny, will grow increasingly suspect and untenable. The basic norms of antidiscrimination law may not require much, if anything, from census data ; affirmative action is completely possible without it ; and better alternatives probably exist for any programs or policies that presently use racialized census data.
13In Part IV, I venture a guess as to why the debate over the “multiracial” category was so incendiary, arguing that, in conservative America, race has become a proxy, but an increasingly ineffectual and divisive one, for social progress towards missing universal rights and benefits. If we recognize that “political race” has taken the place of social solidarity, class consciousness, and wealth redistribution, we will see why progressives are so reluctant to let go of the concept that one progressive author defending a racialized census openly likened to “the devil”. But I will argue here that we should revive the central insight of every progressive movement we have had, from abolitionism and Populism to the Civil Rights movement : that interracial solidarity for universal social rights provides the greatest promise of the greatest benefit to the greatest number of people. Thus, it is time to liberate law and politics from our tiny race boxes and fashion ways to transcend racism and race at the same time. We need to struggle towards what will ultimately be a new constitutional arrangement : abandonment of race as a census category at the same time that we establish long-missing but essential rights of national citizenship, including the right to vote, the right to health care, and the right to equal spending on public education. We can think of this as the French solution since the French republic has never permitted race questions on the national census, has proclaimed a belief in universal republicanism since the French Revolution, and has provided health, education and suffrage as essential rights belonging to all citizens.
14The immediately striking facet of the ferocious debate that surrounded the suggestion to create a new “multi-racial” category in the run-up to the 2000 census is that observers on all sides agreed that, as a biological assertion or scientific category, “race” lacks all validity. All agreed, rather, that race is a pure political and social invention rooted in the history of colonialism and racism.
- 14 Lisa K. Pomeroy, “Restructuring Statistical Policy Directive No. 15 : Controversy Over Race Categ (...)
- 15 Christine B. Hickman, “The Devil and the One Drop Rule : Racial Categories, African Americans, an (...)
- 16 Anthony Appiah, In My Father’s House, 46.
15Thus, even the professors who defend the current practice of inviting respondents to sort themselves into racial boxes corresponding to White, African-American, Native American, Asian-American and Hispanic readily admit the fictional quality of distinct “race” categories that constitute the enterprise. For example, in the course of defending the current race boxes, Lisa K. Pomeroy, writing in the Toledo Law Review in the fall of 2000, records that, “Most anthropologists agree that there is no biological basis for race ; in fact, these experts contend thatrace is purely a socialandconstruct14.” Similarly, Christine Hickman, who also defends current practices, writes that the “proponents of [the multi-racial category] all correctly deny that there is any biological basis for race15.” This contradiction – denying the existence of race on scientific grounds but defending its meaning to uplift people – was remarked upon by Anthony Appiah in his brilliant analysis of W.E.B. DuBois’ tortured efforts to come to terms with race : “In his later writings, DuBois – whose life’s work was, in a sense, an attempt at just this impossible project – was unable to escape the notion of race he explicitly rejected16.”
- 17 Naomi Zack, “Border People and Antidiscrimination Law : American Mixed Race : The U.S. 2000 Censu (...)
- 18 Melissa Nobles, “Racial Categorization and Censuses,” in Census and Identity : The Politics of Ra (...)
16Needless to say, the advocates of the new multi-racial category vigorously rejected the notion of race as well. Naomi Zack spoke well for this pro-multiracial category camp : “Race is a social construction on all levels. Current scientific findings about biological race indicate that the main problem with ‘race’ in common sense is a failure to recognize that there is no biological basis for racial categories17.” In fact, there is a powerful “intellectual consensus” across disciplinary boundaries “that racial categories have no biological basis18.” Research in the Human Genome Project has established that human beings truly comprise a single species and that “race” carries virtually no explanatory or predictive power in analyzing human nature or behavior. As a way of describing the talents, aptitudes, limitations, gifts, virtues, vices, feelings, thoughts or attributes of human beings, racial concepts are practically worthless. Moreover, because the very idea of race is an ideological product of specific political and cultural systems, we have never been, and never will be, able to fix permanent racial categories that reflect some underlying natural reality. By the same token, we have never been, and never will be, able to classify everyone into our provisional and politically negotiated decennial race categories since individual people have their own ideas, experiences, and feelings involving “race” in America. People will often reject the deployment of inescapably artificial race divisions in the census as foreign to their self-identifications and betraying of their personal experience and feelings.
- 19 Patrick F. Linehan, “Thinking Outside the Box : The Multiracial Category and Its Implications for (...)
- 20 Id. at 50.
17The 2000 census debate in America began when proponents of the new census category asserted that, since racial categories are, by definition, artificial and provisional, we should adjust them to meet the subjective experiences and sentiments of hundreds of thousands of people with mixed-race parentage who feel that they do not fit into the current lineup of racial boxes. “The most commonly cited argument made by advocates of a separate multiracial category is that its absence constrains multiracial individuals’ ability to assert their identity as they subjectively perceive it to be”, writes Patrick F. Linehan. “Under the present categorization system, such individuals are faced with the dilemma of either ‘choosing a race” or throwing oneself into a category known as ‘Other19.’” Congressional hearings on the proposed change were filled with heartfelt testimony about the psychological dilemmas, emotional ambiguities and identity crises caused by forcing both children and adults to choose between or among the parents and grandparents identified with different races. Thus, champions of the new proposed category argued that “individuals should have the liberty to define themselves as they see fit, this decision should not be confined to the rigid racial categories offered under the current system20.”
- 21 Hickman, supra note 15 at 1264.
- 22 Tanya Kateri Hernandez, “’Multiracial’ Discourse : Racial Classifications in an Era of Color-Blin (...)
- 23 Id. at 105.
- 24 Id. at 167. The invitation to respondents to state “which single race and/or ethnicity you find m (...)
18Although they too acknowledged the scientific invalidity of racial taxonomies, opponents of the change insisted that the collection in the census of racial statistics remained essential to the meaningful enforcement of civil rights laws in the voting, employment and housing fields. Thus, Christine B. Hickman summed up the view nicely : “However imperfect the census may be, it is our main yardstick for measuring the progress we as a society have made toward ending racism. We tamper with it at our peril. Racism is still with us, and an accurate means for measuring its hold on our society is still vital21.” Similarly, Tanya Kateri Hernandez writes that, “although multiracial discourse may seem benign and appealing on a humanitarian level, its implementation will produce counter-egalitarian results in the struggle for racial equality22.” Hernandez argues for “a race-conscious classification system, which reflects the sociopolitical nature of race, to monitor racial discrimination more effectively and to dislodge the force of multiracial discourse23.” She advocates that the current list of racial choices be continued but accompanied by a lengthy explanation that begins like this: “Recognizing that there are no such things as scientifically pure races or ethnicities, and that a person’s individual identity can never be reduced to a single box, this form requests that you indicate which single race and/or ethnicity you find most politically or socially meaningful24.”
19Now, the two sides in this argument both made valid points Why should someone who is the daughter of an African-American father and a white mother have to identify herself for census and public purposes as being either African-American or White if she does not define herself as one or the other? Where does the government derive the power to compel her to make a self-definitional choice that is completely false to her ? As a matter of First Amendment law, this official racial coercion smacks strongly of the kind of “compelled speech” that was condemned in West Virginia v. Barnette and more recently in Boy Scouts of America v. Dale and Hurley v. Irish American – Gay, Lesbian, Bisexual Comm. Of Boston. Compelling a person to identify with a race in a way that is false to conscience or belief is far more of a speech imposition than asking the organizers of the St. Patrick’s day parade to admit a gay contingent or the Boy Scouts of America to tolerate gay scoutmasters.
20Furthermore, the Equal Protection problems are profound where government creates racial categories for the population that are not rooted in scientific concepts. When government draws lines among citizens that are not rooted in empirical knowledge and reason but superstition and fear, such policies fail minimal Equal Protection rational basis scrutiny. Thus, the Court invalidated such irrational anti-gay legislation in Romer v. Evans and Lawrence v. Texas, striking down laws that targeted groups of citizens for adverse treatment without any reasoned basis.
- 25 116 S. Ct. 1620 (1996).
- 26 For an excellent discussion of the origins and meanings of the “one drop” rule, see generally Hic (...)
21The denial to all citizens of an equal opportunity at racial self-definition cuts hard against the principles organizing these decisions. The basic idea of Equal Protection, as the Court put it in Romer, is “the principle that government and each of its parts remain open on impartial terms to all who seek its assistance25.” Why is there anything remotely defensible today about perpetuating the racial taxonomy erected by the racist 19th century rule of “hypodescent,” the “one drop rule” that “provides that one drop of Black blood makes a person Black26” ? Hundreds of thousands of people now resist the official hubris implied by compelling people to make choices of racial self-definition based on the underlying template of the “one drop” rule even when they reject it.
- 27 See Linehan, 44 How. L.J. 43, 66-67.
- 28 Id. at 67
22The government’s ultimate resolution of the controversy in 2000 – allowing census respondents to check off more than one race category27 – is perceived as similarly unsatisfactory by many people because it “still fails to recognize that multiracial identity is a socially validated identity”, according to Patrick F. Linehan. “In the end, the government’s omission of a multiracial category leaves those who may not fall neatly into the existing race classification system carrying the ‘heavy emotional burden’ of finding validity in a multiracial identity outside of the rigid racial constructs imposed by the government and, ultimately, by society28.”
- 29 Hernandez at 105.
23On the other hand, the critics of the “multiracial” box have an equally powerful point to make. They approach the issue not from the standpoint of the individual who wants to escape the single-race census boxes but from the standpoint of large masses of people who have suffered under racial dogma and oppression in American history. While it may be true that “race” was a pseudoscientific ideological concoction of colonialism and racism, both law and society actually operated on this conceptual system for centuries, treating race as though it were a real scientific fact and thus dramatically affecting the lives and fortunes of millions of people. Those who were – and are – victims of racism in the real world and their descendants should not be abandoned now simply because we discovered that the ideology that oppressed them was based on superstition. All of the social progress we have mobilized in antidiscrimination law, voting rights law and affirmative action should not be discarded in a rush to the “Multiracial Category Movement”, which “inadvertently furthers the progression of color-blind jurisprudence in direct contravention of the MCM goal of promoting racial equality29.”
24And yet both sides also made telling, uncomfortable criticisms of the other side. The advocates of the change pointed out that the traditional civil rights leadership was now clinging to the discredited remnants of the racist “one drop rule” and thereby perpetuating an unhealthy social allegiance to racial ideas. Rather than search for new solutions to old problems, these leaders still prefer to compel people to adhere to a scientifically debunked racial ideology.
- 30 Id. at 110.
- 31 Id. at 113.
25Conversely, the opponents observed that, while the advocates typically begin their analysis by debunking race, the idea of designating “multiracial” as a category itself presumes the coherence of preexisting and independent monoracial categories. If some are “multiracial,” others must, by definition, be monoracial ; if not, then we are all multiracial and the category loses coherence. Thus, the logic of the proposal, ultimately, is either to reinforce racial thinking and create a kind of intermediate “mulatto” box on the census or else ultimately to destroy racial categories altogether, making it impossible to monitor and effectuate race-conscious change. Given these unpalatable choices, some of the critics of the proposal detected in the movement of “parents of biracial children” a sneaking “claim to having biracial children inherit all of the privileges of White status, which White parents logically would like to extend to their children as protection from racism against non-Whites30.” Hernandez argues that “the parentsof biracial children often are unconsciously deeply invested” in the “property value” that American history has assigned to “whiteness31.”
26This debate was an agonizing one for all involved except perhaps the racial conservatives promoting “color-blindness”. They clearly sought to use the emergence of the “multiracial” issue in precisely the way the opponents feared : to undermine the legitimacy of efforts to redress the injuries of racism and declare the end of not only race but the struggle against racism in American life. But the original proponents of the change, many of them people in interracial marriages and their children and grandchildren, suddenly saw themselves being linked rhetorically to Justices Scalia and Thomas and others who would close their eyes and turn their backs to the racial injuries permeating American history. The champions of the “multiracial” box were essentially being accused of betraying the very Civil Rights Movement that made their love, their marriages and their children legitimate and perhaps even possible. They were now being accused of heresy from a faith that they thought they embodied in their personal lives. Similarly, the civil rights leadership was suddenly thrown on the defensive for perpetuating the pseudoscientific, irrational and essentialist claims about race that the Civil Rights movement has always challenged, with the suggestion often thrown in that this unenlightened response reflected their determination to maintain their own power.
- 32 Grutter v. Bollinger, 539 U.S. 306, 322-3.
27Is there any way to escape this grim impasse that seems to trap everyone unhappily in the terms of racial thinking set centuries ago ? Can we expect the dilemma to just grow worse as the numbers of people identifying as “multiracial” increase dramatically in the future and as the 25-year clock on affirmative action suggested by Justice O’Connor in the Michigan affirmative action cases begins to tick away32 ?
28Let us take the problem in constitutional terms. Assume that race does not exist as a biological or scientific fact. By what right then does the government continue to propound race categories on the census ?
29Under now-settled Equal Protection doctrine, government use of racial categories is inherently suspect and must withstand strict scrutiny. Thus, the government would need to demonstrate, first, that it has a compelling interest in the project of racial categorization and classification on the census and, second, that these racial means are necessary to vindicate its interest.
30Of course, all of the pro-slavery, segregationist, and white supremacist interests served by the use of racial categories on the census from 1790 until 1960 fail strict scrutiny today. But today’s state interests are the reverse : to facilitate the enforcement of civil rights and antidiscrimination law, to enforce voting rights law and to allow for “diversity” based affirmative action policies to take place. Race is needed to identify and then rectify the damage visited on the victims of race.
31The opponents of the “multiracial” category invoke such an interest as a compelling reason for maintaining the current racial inventory. About this point, they surely must be right. Discrimination, disenfranchisement and educational apartheid are such powerful forces in our history that undoing them constitutes a social interest of the highest order.
32But are race-conscious means on the U.S. census truly necessary to implement the panoply of antidiscrimination laws ? To be more precise, do we need to ask people about their “race” on the census in order to enforce Title VII anti-discrimination law, the commands of the Fifteenth Amendment, Equal Protection and the Voting Rights Act, and educational affirmativeaction policies that promote “diversity” ? And if we really do need to in some cases, are there no less restrictive alternatives that will get the job done ?
- 33 Hickman at 1254.
- 34 See id.
- 35 Id.
- 36 Pomeroy at 87.
33There is far more to these questions than first meets the eye. The opponents of the “multiracial” category habitually assert that civil rights law would essentially collapse without race data compiled from the census. For example, Christine Hickman writes that “traditional civil rights and minority groups. depend on the census’s racial statistics in order to safeguard voting rights, job opportunities, and school integration plans, as well as to effectively enforce antidiscrimination laws33.” She quotes the congressional testimony of Steve Carbo of the Mexican American Legal Defense Fund, who argued that “the collection of race and ethnic data is fundamental34,” and another observer who stated that an inaccurate racial count “could have dire political consequences35.” Lisa K. Pomeroy argues that the “value of census data in the battle against race discrimination is beyond measure” and goes so far as to suggest that the census claims and beliefs of “persons of multiracial ancestry” will “perpetuate discrimination against traditionally defined minorities in the areas of housing, criminal sentencing, and employment discrimination36.”
34These arguments proceed at an extremely high level of generality and do not descend to doctrinal particulars to explain where, within specific legal causes of action, racial census data are, in fact, required. Nor do these arguments ever explore whether, in cases where racial census data is in fact used, other alternatives are available that may be equally or even more effective for fighting racism.
35Thus, I want to explore here how necessary racial data on the census really is in several important areas of civil rights law. In some cases, it is simply unnecessary and irrelevant ; in other cases, where it is actually relevant and useful, there may be other alternatives available that will be equally – or more – effective at stopping discrimination and promoting integration. For analytical purposes, we can divide the positive use of race in “race law” into two types: antidiscrimination norms and positive integration or affirmative action norms.
- 37 426 U.S. 229 (1976).
36Consider, first, Equal Protection claims under the Fourteenth and Fifth Amendments. Since the Supreme Court’s 1976 decision in Washington v. Davis37, it is well-established that litigants alleging Equal Protection violations must demonstrate that an allegedly discriminatory government policy was motivated by a racially discriminatory purpose. Thus, it is never enough for a constitutional claim simply to show that there is a racially discriminatory impact on the population. As established in cases like Arlington Heights and McCleskey v. Kemp, plaintiffs must show that it was the purpose behind an enactment or policy to discriminate. Race data from the census is therefore irrelevant to such claims : the impact on the general population is irrelevant.
- 38 Title VII.
- 39 See Griggs v. Duke Power Co., 401 U.S. 424 (1971).
37Consider now the more complicated case of Title VII statutory employment discrimination law, which makes it unlawful for an employer to “discriminate against any individual with respect to his compensation, terms, conditions or privileges of employment, because of such individual’s race, color, religion, sex or national origin38.” The Court has read this prohibition to give rise to both intentional discrimination law suits based on “disparate treatment” and also suits that do not require a showing of intentional discrimination but arerather based on racially “disparate impact.”39 But when we take each in turn, we see that race data from the census is of little or no use.
- 40 See McDonnell Douglas v. Green, 411 U.S. 792, 802 (1973).
- 41 Perkins v. Lake County Department of Utilities, 860 F. Supp. 1262, 1278 (USDC, N.D. Ohio) (1994) (...)
- 42 Id at 1276.
38As with Equal Protection claims, census race data is clearly unnecessary to intentional discrimination claims that a person was fired, not hired or otherwise adversely treated in the employment setting based on his or her membership in a protected status group. This is because the individual need not refer to any general racial demographics in making out his or her complaint but simply allege, in order to establish a prima facie case, that he or she is a member of a protected group who was qualified for the job or promotion but was passed over40. At that point, the burden shifts to the employer to articulate a status-neutral explanation for its employment action, and the contest of reasons begins. Census race data is not used. Even if an employer tries to challenge a plaintiff’s allegation that he or she is a member of a protected class (perhaps by questioning his or her census identification), it appears that the crucial question becomes what was the employer’s perception of the plaintiff’s race or ethnicity during the course of the alleged discrimination, not what the plaintiff’s race actually is : “Objective appearance and employer perception are the basis for discrimination,” as one court put it, and are “in the opinion of this Court, the key factors relevant to enforcing rights granted members of a protected class”41. As this court also explained, “when bringing an action under Title VII African-Americans do not have to demonstrate that their relatives lived in Africa, or that they visit the site of their roots, or that they are involved in any kind of cultural activities associated with Africa. They only have to appear to be African-Americans to be deemed members of that protected class”42. Thus, white employers who fire a white employee because they think he is black are still guilty of practicing race discrimination.
- 43 401 U.S. 424 (1971).
39The vast majority of Title VII suits involve intentional discrimination claims. But some are disparate impact claims. Unable to show evidence of intentional discrimination, plaintiffs bringing these claims try to mobilize statistical evidence to demonstrate that seemingly neutral employment policies have racially discriminatory effects. In Griggs v. Duke Power Co., the Supreme Court explained that traditional employment practices, like imposing high school or college degrees as a job requirement, may be “neutral on their face, even neutral in terms of intent” but nonetheless still “operate to ‘freeze’ the status quo” in racially discriminatory ways43. Thus, the disparate impact cause of action seeks to displace “artificial, arbitrary and unnecessary barriers to employment” where they “operate to invidiously discriminate” on a forbidden basis. To make a prima facie case, the employee must offer a statistical showing that a particular practice is causing a statistically significant difference between the racial composition of the employer’s workforce and the racial composition of the pool of qualified available people in the specific labor market or the employer’s workforce. If he or she does so, then the burden shifts to the employer to demonstrate that there is a true “business necessity” for maintaining the effectively discriminatory practice or policy.
- 44 Michael Selmi, Was the Disparate Impact Theory A Mistake ?(unpublished manuscript on file with th (...)
- 45 Id. at 2.
40Unlike intentional discrimination claims, “disparate impact” claims clearly will rely on race data, often drawn from census tract information. Yet, despite a continuing “academic obsession” with the “disparate impact” theory after more than 40 years of its existence, it “has had a strikingly limited impact” on ending discrimination and “may have actually hindered the eradication of discrimination as much as, and perhaps more, than it has helped44.” After doing a comprehensive survey of Title VII cases involving “disparate impact” claims, Professor Selmi concludes that “outside of the realm of objective written examinations, the theory has had littleactual influence in litigation even within employment discrimination where it clearly had its greatest impact45.”
- 46 Id. at 10.
- 47 Id. at 11.
41Part of the reason for the ineffectiveness of the “disparate impact” theory was that it developed, in the context of courts confronting whites-favoring seniority systems, to deal with “the perpetuation of past explicit but lawful discrimination that would contravene the purposes” of Title VII46. It would have been far better to treat the conscious reproduction of racially loaded seniority systems as a form of intentional, rather than accidental, discrimination since employers could legitimately be charged with knowledge of the likely consequences of their actions. But, as Professor Selmi explains, “it became clear early on in the history of Title VII that negotiations with employers would be far easier to accomplish if they could move away from a focus on intentional discrimination, which carried with it an implicit label of blame, to something more amorphous47.”
- 48 Id. at 22.
42In any event, Professor Selmi has found that, in 130 reported Title VII cases reaching appellate courts over a period of years, only 25 (or less than 20 %) won on a “disparate impact” theory and 15 of these won only remands. Given that it was most effective at confronting the perpetuation of practices from the days of segregation, the disparate impact theory has had decreasing appeal and utility over time. As a theory, it “is inconsistent with our broadly held idea of justice, an idea that is steeped in concerns of remedying intentional acts of discrimination48.”
43Given the declining importance of “disparate impact” claims, which were never a significant force compared to intentional ones, we might make the social judgment that the whole category should be reconsidered and collapsed back into intentional discrimination claims. Yet, even if we are to maintain this category, is census data necessary for a plaintiff to make such an initial prima facie showing of a racially discriminatory effect of an employment policy ? It is very unlikely. The key information relates to the demographics of the employer’s own workforce and applicant pool, information obviously not available from the census, and to the specific demographics of the available labor market, which also is an incredibly tiny slice of life requiring independent discovery and research. Plaintiffs and defendants need to conduct their own fine-tuned research relating to the specific institutions and markets relevant in the case. The census is a broad brush that might help but is hardly sufficient. The decision to drop racial data on the census would not interfere with the development of anyone’s employment discrimination case.
44Sections 2 and 5 of the Voting Rights Act invite a similar analysis for antidiscrimination norms. An African-American or Hispanic-American who is actually being denied the right to vote by a jurisdiction need only allege under Section 2 that he or she is a member of a minority group being disenfranchised by a state policy or practice for that reason. Census data never comes into play.
- 49 42 U.S.C. 1973.
- 50 Id.
45However, under the 1982 amendment to the Voting Rights Act, members of racial minority groups may bring suit against any voting procedure or standard that results in racial minorities having “less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice. The extent to which members of a protected class have been elected to office in the State or political subdivision is one circumstance which may be considered49.” This section would appear to give minorities representational rights that depend on racial census data to be made effective, but the amendments also have a proviso that reads : “Provided, that nothing in the section establishes a right to have members of a protected class elected in numbers equal to their proportion in thepopulation50.” This provision greatly undermines the utility of racialized census data. If racial groups had a right to proportional representation, we would obviously need such information ; without this right, the need dissipates.
46Still, under Section 5 “preclearance” procedures, jurisdictions seeking approval of proposed changes in voting practices and procedures must show there is no “retrogression” in the ability of racial minorities to participate and influence election outcomes. This mechanism implicates the use of racial data when it comes to the task of redrawing congressional and state legislative district lines.
- 51 509 U.S. 630 (1993).
- 52 115 S. Ct. 2475 (1995).
- 53 Id. at 2488.
47Yet, when we examine the complete picture of our voting rights jurisprudence, it becomes clear that racial data gathered from the census may not be used by state legislatures to deliberately draw majority African-American or Hispanic or Asian-American congressional or state legislative districts. Under the Supreme Court’s decisions in Shaw v. Reno51 and Miller v. Johnson52, the Court has found that state legislatures violate Equal Protection whenever they “subordinate traditional race-neutral districting principles, including but not limited to compactness, contiguity, respect for political subdivisions or defined by actual shared interests, to racial considerations53.” The Court has accordingly not hesitated to wipe out majority-minority districts created by states. Therefore, while one might argue that race data on the census should be gathered to facilitate the creation of majority-minority districts long suppressed by racism, the Supreme Court has already made it impossible to use census data in that way. Indeed, there is probably now far more of a chance to successfully create such districts without census data since having the data – and the knowledge it imports – becomes prima facie grounds for a court overturning a majority-minority district. While so-called “influence districts,” in which racial minority groups can influence the selection of leaders, are lawful, they can, and will, be created even without race data from the census.
48In any event, even if census data were helping to build majority African-American or Hispanic districts – a possibility that the Supreme Court has foreclosed – there are superior alternatives for expanding the possibilities of traditionally marginalized communities electing political leaders of their choice. I refer to electoral regimes based on proportional representation such as preference voting and cumulative voting, which give political minorities, however defined, the opportunity to organize cohesively and elect candidates they support even without majority backing. Such methods have been used in Illinois legislative elections, Cambridge city council elections and dozens of other jurisdictions, not to mention most corporate shareholder proxy elections. These race-free systems promoting full representation essentially allow voters to district themselves on election day and assure that substantial political minorities will gain some representation along with majorities.
49To be sure, it will be said that the non-retrogression principle also operates occasionally to prevent other kinds of adverse voting changes, such as the shift from district voting to at-large voting or the movement of polling places. It is possible that the lack of race data could require more discovery on relevant issues than otherwise, but most of these changes will require specific fact-finding anyway. Again, the stakes seem very small on this issue, and the costs of a racialized census much higher than any conceivable benefit under the new post-Shaw v. Reno regime.
50Defenders of racial categories on the census fear that losing racial census data will undermine educational and occupational affirmative action efforts. But it is hard to see why this would be the case.
- 54 See Grutter v. Bollinger, 539 U.S. 306, Gratz v. Bollinger, 539 U.S. 244.
51Take the case of universities practicing affirmative action to promote diversity, a policy just upheld by the Supreme Court in the University of Michigan decisions54. First of all, the universities ask student applicants to self-identify their race and ethnicity and this practice would continue unaffected. No census information is necessary on student applications or to enable individual applicants to define their own racial or ethnic background. Assuming their policies are constitutional, the universities take an individual applicant’s race or ethnicity into account in the admissions process as only one factor among many for the purposes of advancing diversity in the student community. Census data again does not enter the admissions equation.
52One might argue that census information needs to be available in order to bring the percentage of particular underrepresented groups up to their percentage level in the population, but this would be a racial quota, something explicitly rejected in Bakke v. Regents of the University of California Davis in 1976 and rejected again by the Court in the Michigan cases. It simply does not make any difference to the shaping of the diversity rationale in a particular university’s admissions process how many racial minorities live in the United States.
53Perhaps the claim is that stripping race from the census will weaken our belief in the social relevance of racism and therefore our commitment to affirmative action. But this is a non sequitur. Abolishing race should be seen as part of the effort to destroy the racism that has injured millions of our people over the centuries. Indeed, abolition of race on the census might improve affirmative action by encouraging universities to tailor their policies much more closely to the diversity rationale itself. Universities could simply ask all applicants to identify personal characteristics and experiences that will enrich the diversity of the class and the community. This will allow applicants to discuss race and ethnicity, along with other economic, social and personal experiences, if they find these factors relevant to explaining their characters, ambitions and talents. Under this approach, universities could actually collect far more meaningful information about the kind of diversity different applicants would introduce to their classes. Similarly, the applicants would not feel as if the mere formal invocation of their “race” or “ethnicity” earns bonus points but rather that they have had the opportunity to think seriously about how these social forces, along with others, have actually shaped their personal development and outlook.
54In principle, this same analysis should apply to professional and occupational affirmative action programs. Taking race off the census will not hamper affirmative action in any way. The loss of racialized census data would not threaten any efforts to integrate with the possible exception of racial quotas, which are illegal anyway under Title VII.
55The one area in which ending race questions on the census could theoretically interfere with integrationist public policy would be in public education, where antidiscrimination and affirmative action norms come together. Of course, in Brown v. Board of Education, the Court categorically struck down de jure segregation, a holding that bans official apartheid and discrimination and does not depend in any way on the gathering of demographic statistics. However, to the extent that formally desegregated school systems continue in de facto fashion the “vestiges” of segregation, they have an affirmative obligation to desegregate. This principle led to our national experience with judicial supervision of school assignment policies, busing and other active desegregation efforts. But, today, fifty years later, these efforts are almost completely gone and the nation’s school districts have overwhelmingly been declared “unitary” by the courts – that is, free from the taint of de jure segregation. This does not mean, of course, that they are necessarily integrated but simply that the legal causation chain for finding judicially sanctionable segregation has been broken. Given that this is the case, the practical effects of ceasing the collection of racial data will be very slender indeed.
56In general, it seems clear that the public interest in collection of race data on the census depends is, at best, negligible. Antidiscrimination, Equal Protection and affirmative action norms do not require it at all. The project of public school integration theoretically involves continued monitoring of the racial dynamics of student populations, but since the vast majority of school districts have long since been declared unitary, the actual relevance of this information is slight. There is no compelling interest left in maintaining a racialized system on the census, and far better means available to achieve any interests that are served by it.
- 55 Orlando Patterson, “The Race Trap”, New York Times, July 11, 1997.
- 56 See Jamin Raskin, “From Color-Blind White Supremacy to American Multiculturalism”, Harvard Journa (...)
57Even if the above argument is correct as an analytical proposition, it will undoubtedly make many of us nervous to contemplate taking race off of the census. To be sure, some distinguished sociologists have already come to this conclusion, such as Harvard professor Orlando Patterson, who argued forthrightly in 1997 that “the Census Bureau should drop its attempt to classify Americans by race55.” But for many people, this idea will strike fear that our civil rights commitments have come to an end and that the reactionary see-no-evil politics of color-blindness have triumphed56. This fear undoubtedly animated much of the explosive reaction to the “multiracial category” proposal.
- 57 Christine B. Hickman, “The Devil and the One Drop Rule : Racial Categories, African Americans and (...)
58I want to argue that this nervousness and fear about losing race as a social belief system reflect the enormous weight that we place on the idea of race to aid and measure social progress. In conservative America, race has become a proxy, but an increasingly ineffectual and divisive one, for chronicling our social progress towards missing universal rights and benefits, such as health care, income equity and college education. What Lani Guinier and Gerald Torres have called “political race,” or identity politics, has generally taken the place of social solidarity, anti-racism, class consciousness, and an explicit politics of wealth redistribution, which were the defining principles of progressive politics in prior historical periods. This is why progressives today are so reluctant to let go of the concept that one progressive author defending a racialized census openly likened to “the devil57.”
59But we have almost certainly reached the point of diminishing returns on what we can draw from the irrational, pseudoscientific and historically poisonous idea of race. It is time to revive the central insight of every progressive movement we have had in the United States, from abolitionism and Populism to the Civil Rights movement : that solidarity for universal social rights – across perceived lines of race, gender, and ethnicity – provides the greatest promise of the greatest benefit to the greatest number of people. Thus, it is time to liberate ourselves from our tiny race boxes on the census and fashion ways to transcend racism and race at the same time.
60Our large problems began with the fateful compromises embodied in our Constitution, which was tainted from the start with the political barbarism of slavery. In the 21st century, we should struggle towards a new constitutional covenant that at once abandons race as a census category and simultaneously establishes long-missing but essential rights of national citizenship, including the right to vote and to be represented for all citizens, the right to health care, and theright to equal spending on public education. Imagine just the effects of genuinely establishing this first right : millions ofAmerican citizens living in Washington, D.C., Puerto Rico, American Samoa and Guam would be represented in the U.S. House and the still segregated U.S. Senate. Millions of others presently disenfranchised because of criminal felony convictions would get their voting rights back. Imagine what it would mean for our poorest citizens to enjoy a national right to health care. Imagine what it would mean to have every state paying for equal educations for all schoolchildren regardless of the wealth of their school districts.
61If we inscribe national political and social rights of citizenship for all, the terror associated with dropping our racial taxonomies should vanish. Those who have been exploited and oppressed the most by race will have the most to gain by exchanging the illusory solace of race for the enactment of real social and political rights. It is not that the race data offers us nothing at all in the interstices of civil rights enforcement, but the benefit is small and vanishing. We should exchange the false pride and vanity of racial assignment with the development of universal social, educational and political rights.
- 58 See generally Alain Blum, “Resistance to Identity Categorization in France” chapter 5 in David Ke (...)
- 59 Id. at 143.
62We can think of this as the French solution since the French republic has never permitted race or ethnicity questions on the national census58, has always maintained a passionate belief in universal republicanism, and has provided, in the last half-century, all of its citizens with health, education and suffrage as essential rights. In France, no one would be surprised to see a linkage between strong antidiscrimination norms and a republican-style refusal to entertain superstitious notions about race and ethnicity that have great currency in the United States. For as Alain Blum observes of French census politics, “It is important to bear in mind that the fight against discrimination currently underway is often linked to a firm opposition to identifying individuals through categories that could be used to ascribe identity based upon ethnicity59.” Here is the ticket : combat race discrimination, drive the idea of race out of society and create universal rights of respect, participation and belonging for all.
1 U.S. Const., Art. I, section 2.
2 See Reynolds v. Sims, Wesberry v. Sanders. Of course, even when the “one person, one vote” doctrine was being enunciated, there were classes of people implicitly left out of the ringing language of equal suffrage and representation : citizens living in the District of Columbia, citizens living in Puerto Rico and the Territories, people serving time in prison for felony convictions and released felons in many states.
3 “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.” U.S. Const. Art. I, section 2.
4 73 N.Y.U.L. Rev. 333, 433.
6 See Jamin Raskin, 3 Election Law Journal 559 (2004).
7 See id.
8 14 Cardozo L. Rev. 1135.
9 Melissa Nobles, “Racial Categorization and Censuses,” in Census and Identity : The Politics of Race, Ethnicity and Language in National Censuses 2002, 67.
12 Melissa Nobles, “Racial Categorization and Censuses,” in Census and Identity : The Politics of Race, Ethnicity and Language in National Censuses 2002, 49.
13 Id. at 49.
14 Lisa K. Pomeroy, “Restructuring Statistical Policy Directive No. 15 : Controversy Over Race Categorization and the 2000 Census”, 32 U. Tol. L. Rev. 67, 68 (2000).
15 Christine B. Hickman, “The Devil and the One Drop Rule : Racial Categories, African Americans, and the U.S. Census”, 95 Mich. L. Rev. 1161, 1168 (1997).
16 Anthony Appiah, In My Father’s House, 46.
17 Naomi Zack, “Border People and Antidiscrimination Law : American Mixed Race : The U.S. 2000 Census and Related Issues”, 17 Harv. Blackletter J. 33 (2001).
18 Melissa Nobles, “Racial Categorization and Censuses,” in Census and Identity : The Politics of Race, Ethnicity and Language in National Censuses 2002, 47.
19 Patrick F. Linehan, “Thinking Outside the Box : The Multiracial Category and Its Implications for RaceIdentity Development”, 44 How. L.J. 43, 46 (2000).
20 Id. at 50.
21 Hickman, supra note 15 at 1264.
22 Tanya Kateri Hernandez, “’Multiracial’ Discourse : Racial Classifications in an Era of Color-Blind Jurisprudence,” 57 Md. L. Rev. 97, 98 (1998).
23 Id. at 105.
24 Id. at 167. The invitation to respondents to state “which single race and/or ethnicity you find most politically or socially meaningful” would be terribly vague since this inquiry is different from asking which group the person belongs to. It is also a bit puzzling since it is not clear what it means for a race or ethnicity to be “meaningful.” More to the point, from the standpoint of champions of the change, what if they do not accept the terms of the question that one race or ethnicity is most meaningful or even exists ?
25 116 S. Ct. 1620 (1996).
26 For an excellent discussion of the origins and meanings of the “one drop” rule, see generally Hickman, supra note 15. I obviously depart from some of her normative conclusions about the continuing utility of maintaining the rule.
27 See Linehan, 44 How. L.J. 43, 66-67.
28 Id. at 67
29 Hernandez at 105.
30 Id. at 110.
31 Id. at 113.
32 Grutter v. Bollinger, 539 U.S. 306, 322-3.
33 Hickman at 1254.
34 See id.
36 Pomeroy at 87.
37 426 U.S. 229 (1976).
38 Title VII.
39 See Griggs v. Duke Power Co., 401 U.S. 424 (1971).
40 See McDonnell Douglas v. Green, 411 U.S. 792, 802 (1973).
41 Perkins v. Lake County Department of Utilities, 860 F. Supp. 1262, 1278 (USDC, N.D. Ohio) (1994) (finding that, regardless whether or not a Title VII plaintiff was actually American Indian, he was entitled to the benefit of the statute where employer had some reason to believe he was).
42 Id at 1276.
43 401 U.S. 424 (1971).
44 Michael Selmi, Was the Disparate Impact Theory A Mistake ?(unpublished manuscript on file with the author) 3, 1.
45 Id. at 2.
46 Id. at 10.
47 Id. at 11.
48 Id. at 22.
49 42 U.S.C. 1973.
51 509 U.S. 630 (1993).
52 115 S. Ct. 2475 (1995).
53 Id. at 2488.
54 See Grutter v. Bollinger, 539 U.S. 306, Gratz v. Bollinger, 539 U.S. 244.
55 Orlando Patterson, “The Race Trap”, New York Times, July 11, 1997.
56 See Jamin Raskin, “From Color-Blind White Supremacy to American Multiculturalism”, Harvard Journal of Law and Public Policy.
57 Christine B. Hickman, “The Devil and the One Drop Rule : Racial Categories, African Americans and the U.S. Census”, 95 Mich. L. Rev. 1161 (1997). Hickman writes that, “The Devil fashioned [the one drop rule] out of racism, malice, greed, lust and ignorance, but in so doing he also accomplished good : His rule created the African-American race as we know it today, and while this race has its origins in the peoples of three continents and its members can look very different from one another, over the centuries the Devil’s one drop rule united this race as a people in the fight against slavery, segregation, and racial injustice.” Id. at 1166.
58 See generally Alain Blum, “Resistance to Identity Categorization in France” chapter 5 in David Kertzer and Dominique Arel, Census and Identity : The Politics of Race, Ethnicity, and Language in National Censuses.
59 Id. at 143.Haut de page
Pour citer cet article
Jamin B. Raskin, « Common Sense, Common Census: Should America Follow France and Abolish Race Questions on the Census? », Droit et cultures, 49 | 2005, 21-40.
Jamin B. Raskin, « Common Sense, Common Census: Should America Follow France and Abolish Race Questions on the Census? », Droit et cultures [En ligne], 49 | 2005-1, mis en ligne le 02 mars 2010, consulté le 27 avril 2017. URL : http://droitcultures.revues.org/1481Haut de page
Haut de page
Droits et Culture est mis à disposition selon les termes de la licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.