II. Racial Rules

For most of its conceptual history, race has been understood as being rooted in biology. Phenotypes and morphological characteristics such as skin color, eye shape and size, nose width, and hair texture were the telltale signs that distinguished races from one another. In this biological construction, race is mutable, essential, and determinative — one’s moral worth and human potential are a corollary of racial identity, and race can therefore account for every type of difference in attitude, ability, and achievement. The calamitous history of race is one of assignment and hierarchy. While the paradigm of biological racialism was far from static or coherent, the maintenance of impermeable racial boundaries was fundamental to ranking different population groups according to physical, intellectual, temperamental, behavioral, and moral traits. Needless to say, these racial orderings, believed to be permanent group hierarchies reflecting laws of nature or decrees of God,^{Footnote 2} justified and preserved the power, privilege, and domination of Europeans over non-Europeans, who were subjugated, exploited, excluded, and sometimes exterminated.

It is now generally recognized that race is a social construct, “a concept that signifies and symbolizes socio-political conflicts and interests in reference to different types of human bodies.”^{Footnote 3} Different physiological characteristics — skin color, eye shape and so on — do exist; the point is that these traits alone are part of a wider system of power relations that invest morphology with meaning.^{Footnote 4} Race operates at the crossroads of micro-level identity formation and macro-level social structures, and processes of racial signification and the meanings embedded therein are socially and historically contingent.^{Footnote 5} However, ample evidence points not simply to the social construction of race, but also to the instrumental role the state has played in the creation of racial identities. Though race exists outside of state institutions, racial categories remain heavily dependent on official recognition (or lack thereof) by the state. More to the point, the state has had a primary role in operationalizing race — not simply by creating discriminatory or emancipatory laws and policies, but also by making and manipulating the boundaries that divide supposedly distinct racial groups. Race is therefore more than an amorphous social construction: it is fundamentally a political one.

The process through which race became institutionalized in the census was once entirely state-driven. Enabled by the rise of statistical science during the nineteenth century, political elites and policy-makers broadened the use of the census from a mechanism primarily used to ensure democratic representation by population in legislatures to an instrument of knowledge, central to the building of state and nation.^{Footnote 6} Therein, racial knowledge was both given and taken. As James Scott argues in Seeing Like a State, it is through projects of simplification like the census that the modern state “attempts with varying success to create a terrain and a population with precisely those standardized characteristics that will be easiest to monitor, count, assess and manage.”^{Footnote 7} In this way, the state’s creation of racial classifications and the invocation of race in its policies and laws were used to dominate, banish, control, and segregate and expunge nonwhite populations from the social fabric of Western societies.

At the dawn of the twentieth century, the Canadian census explicitly enumerated the racial origins of the population. The census guide defined “race” as “a subgroup of the human species related by ties of physical kinship. Scientists have attempted to divide and subdivide the human species into groups on the basis of biological traits, such as shape of head, stature, colour of skin, etc.”^{Footnote 8} The Canadian state was particularly interested in counting its Aboriginal population. In 1901 enumerators were required to write the names of the tribes (e.g., Chippewa and Cree) and were instructed to carefully record white/Indian “mixes.” The instructions read:

Two distinct sets of information could be gleaned from this elaborate taxonomy of Aboriginal and French, English, Scotch, or Irish offspring. First, counting the number of persons with Aboriginal ancestry was necessary in order for the state to determine the progress of the assimilative goals of the Indian Act, a totalizing regime that paternalistically governed the lives of status Indians from cradle to grave. Since the Crown had a fiduciary obligation only to those with Indian status, the government had clear incentive to reduce the legal ability of Aboriginal people to claim either status or Aboriginal title to unceded territory. Second, the specific details about the composition of these racial amalgamations is related to the history of disputes and violent clashes between the Canadian government and the Métis population of the prairie provinces during the Red River Rebellion of 1869-70 and the Battle of Batoche in 1885. Legally designated as “half-breeds” by the federal government, these descendants of Aboriginal women and French fur traders were never seriously considered by the state to be either within the definition of Indian or a distinct segment of the Aboriginal population.^{Footnote 10} To acknowledge either of these scenarios would contradict the purpose of the Indian Act regime, which was to remove all legal distinctions between the Native population and other Canadians while maintaining hierarchical race relations and social stratification. This complex enumeration enabled the government to keep tabs on a population that, in its view, threatened the security of the still unstable, still largely unsettled, western provinces.

Beginning in 1911, the census required that the names of “tribes” be recorded and that racial origin be traced through the mother’s side. This requirement of matrilineal descent is particularly interesting given that the progeny of Aboriginal women and non-Aboriginal men would be considered “Indian” in the census and yet denied Indian status in accordance with section 12.1(b) of the Indian Act. Why this inconsistency? On the one hand, a long process of institutional development may have prevented schematic standardization. During its formative years, the Dominion Bureau of Statistics was likely not overly concerned with promoting a standardized racial classification system throughout the federal government. The fact that two distinct arms of the state would categorize the same population in different ways is simply another example of the contradictory rationality of the legal construction of race. The Indian Act, whose unabashed goal was, as Duncan Campbell Scott, Deputy Superintendent for the Department of Indian Affairs, put it, to “get rid of the Indian problem . . . to continue until there is not a single Indian in Canada that has not be absorbed into the body politic and there is no Indian department,”^{Footnote 11} had a different function and purpose — and therefore, different policy and political consequences — than the census, which was designed to count the population.

On the other hand, however, both the census and the Indian Act were historically, and remain today, legal instruments through which racial categories are made and manipulated. They do not simply reflect racial reality, but are fundamental to its existence. As such, discourses of race and gender collide through the state’s schematizing impetus: there were legal sanctions for Aboriginal women who married non-Aboriginals through the Indian Act regime, once again affirming the stereotype that Aboriginal women were far more likely to “marry out” than the unfathomable circumstance that white women would “marry in.”^{Footnote 12} At the same time, the rationale of the census required Aboriginal ancestry to be traced through matrilineal descent since women were perceived as the vessels of (un)civilization, culture, and morality,^{Footnote 13} juxtaposed with the patrilineal tracing of white descent, since men were the conveyors of the privilege of property and citizenship rights.^{Footnote 14}

Patrilineal or matrilineal descent mattered not for mixed white/nonwhite progeny, as “the children begotten of marriages between white and black or yellow races will be classed as Negro or Mongolian (Chinese or Japanese) as the case may be.”^{Footnote 15} In the 1931 and 1941 censuses, eugenicist language permeated the wording of the question, referring to “the distant coloured stocks . . . involving differences in colour (i.e., the black, red, yellow, or brown races).”^{Footnote 16} In other words, any mixed-race person whose parental or ancestral lineage was comprised of European and nonwhite components would legally be counted by the census as nonwhite. This concern about the classification of mixed-race offspring reflects eugenicist anxiety not simply about nonwhite immigration to Canada, but also regarding the reproduction of “degenerates” and the so-called “unfit” already in the country.^{Footnote 17} Those with such inherently degenerate characteristics could never be considered potential members of the dominant race; in the words of Ann Laura Stoler, white is a color that is easily stained.^{Footnote 18}

As the edifice of biological racialism crumbled after the Second World War, the normative context surrounding dominant conceptions of race as irrefutably biological, determinative, and hierarchically ordered shifted. Canada’s Dominion Bureau of Statistics recommended that the word “race” be omitted from the census card because “‘[r]acial’ has some odious implications, stresses the biological aspect and is thoroughly unscientific. ‘Ethnic’ includes more to the cultural aspects, which is what we want, but it has been criticized as being too academic a term to be easily understood.”^{Footnote 19} Unlike human rights laws or immigration policies, which the government changed only in halting increments,^{Footnote 20} there was little at stake by altering the language of the origin question in the census.

The term “race” was dropped from the census, with the question in 1951 now requesting details on the “origin” of the population. The instructions to census enumerators relayed the importance of carefully distinguishing between citizenship/nationality and origin, admonishing that “[o]rigin refers to the cultural group, sometimes erroneously called ‘racial’ group, from which the person is descended; citizenship (nationality) refers to the country to which the person owes allegiance.”^{Footnote 21} Instead of assigning mixed-race offspring of white/nonwhite relationships to a nonwhite designation, these people were classified according to the same principles of patrilineal descent as other white ethnic “mixes.” This change in classification principles had statistical consequences, partially accounting for the decline in the Chinese and Japanese populations and largely explaining the large drop in the Black population, from 22,174 in 1941 to 18,020 in 1951.^{Footnote 22} The social construction of race certainly did not appear overnight, nor was the idea of race as biological fully abandoned in the social foundations of the census. The introduction to the origin question in the 1951 Census refers to the data as “partly cultural, partly biological, and partly geographical” in nature.^{Footnote 23}

The census question on racial, then ethnic, origins occurred within the confines of established schematics, governmental priorities and dominant ideational paradigms about the relevance of race in Canada. Though Canadian multiculturalism, championed by Prime Minister Pierre Trudeau, was a celebrated national value by the 1980s, multicultural policies were premised on symbolic pluralism rather than substantive antiracist strategies.^{Footnote 24} In the early 1980s, organizations within and outside the state began to demand more accurate data on racial minorities, which would have most effectively been attained by adding a direct question on race to the census.^{Footnote 25} Government statisticians recognized the multiple problems with using the ethnic origin question to enumerate racial minorities; not only was the question on ethnic origin simply not designed to capture data on race, but the conflation of race and ethnicity produced inaccurate data. Haitians, for example, claimed a French ethnic origin as French, while many persons born in Jamaica identified as British.^{Footnote 26} Even after the Employment Equity Act of 1986 mandated accurate racial data in order to fulfill its policy goals, policy-makers chose to incrementally adjust the existing ethnic question. Estimates of the nonwhite population were derived from indirect measures such as ethnicity, language and place of birth, though it was understood that the use of racial proxies would be inadequate.

Change was driven, ironically enough, by a conservative backlash against multiculturalism in the 1990s. Shortly before census day in 1991, media outlets in Toronto and its surrounding areas, including the Toronto Sun, began a campaign entitled “Count me Canadian!”^{Footnote 27} Campaigners and their allies in the Reform Party of Canada decried the lack of a “Canadian” category on the ethnic origin question and urged followers to declare themselves “Canadian” on their census forms using the mark-in space. Facing a disastrous situation in which the ethnic origin question could produce useless information on Canada’s racial and ethnic diversity, Statistics Canada included instructions on the census form that read: “While most people of Canada view themselves as Canadian, information about their ancestral origins has been collected since the 1901 Census to reflect the changing composition of the Canadian population and is needed to ensure that everyone, regardless of his/her ethnic or cultural background, has equal opportunity to share fully in the economic, cultural, and political life of Canada.”^{Footnote 28} After the 1991 Census, however, “Canadian” became the fastest growing ethnic group. Whereas only 130,000 people gave “Canadian” responses in 1986, that number jumped to just over one million in 1991. “Canadian” became the fourth largest single response answer, following “French,” “British,” and “German.”

The substantial responses of “Canadian” to the ethnic origin question had path-dependent implications not only for 1991 census data but also for the future of the question itself. As Monica Boyd points out, because of the internal protocol at Statistics Canada to rank-order ethnic categories in terms of representative population size, “Canadian” would appear as one of the listed choices in the 1996 census. This automatically stimulated increased responses because respondents are more likely to check the box next to one of the listed options than write in their own response in the free-text field, and also led to increased responses because the French translation of “Canadian” is “Canadien,” which has a historic and symbolic importance in French Canada.^{Footnote 29} More importantly, the increased responses of “Canadian” on the 1991 Census made it impossible for Statistics Canada to determine who was and was not a racial minority using its standard approach of cross-tabulating racial proxies. With the need for a direct question on race to meet the legal obligations of the Employment Equity Act all the more acute, the government announced its intention to include a question on race in the summer of 1995 and a public debate over the nature and effectiveness of Canada’s multiculturalism and employment equity policies ensued. Some argued that the inclusion of the question was “a step backward,” based on ideas from the nineteenth century.^{Footnote 30} In response to these debates, Chief Statistician Ivan Fellegi published a statement on the Statistics Canada website and in major newspapers throughout the country, explaining that the question was designed to elicit information necessary for the implementation and evaluation of the Employment Equity Act. Fellegi fervently disassociated the question from the notion of race, stating that “this question is not designed to provide information on race or racial origins of the population of Canada.”^{Footnote 31}

In previous censuses, race was avoided because of its controversial nature, the threat it posed to data quality, and its potential for further fractiousness in an era of rampant identity politics. This time, the justification for its inclusion in 1996 relied on established normative principles of equality and social justice. The rationale provided on the census form relayed to Canadians that the information would support programs that promote “equal opportunity for everyone to share in the social, cultural, and economic life of Canada”.^{Footnote 32} Similarly noting that “much of the criticism of Question 19 appears to be directed more at the idea of employment equity than at the collection of statistics,” the Chief Statistician encouraged Canadians to accept that “employment-equity legislation has been the law of the land since 1986,” and recognize the value in using the census to collect this data, as “the census is the only possible source of the objective information which is needed to administer the act and to evaluate its impact.” Invoking the necessity of fair and informed debate on these issues, Fellegi argued that “it is in everyone’s interest that the debate on issues related to employment equity, and the many other issues illuminated by census data on the composition and characteristics of our population, be supported by objective, impartial and reliable data, rather than by impressions, unfounded opinion or stereotypes”.^{Footnote 33} Canadians may not have liked employment equity and were still uncomfortable with a question that walked and talked like race, but the invocation of these principles — equality, fairness, and full participation in Canada’s social, economic and cultural life — were at least familiar.

III. Decisions, Decisions

Deciphering the meaning of racial in racial equality requires a close examination of racial classifications. This investigation leads directly to the doorstep of the state, which has been largely responsible for the definition and manipulation of racial classification schema. In the classic Weberian definition, the modern state is premised on domination as it claims, monopolizes, and legitimates the use of force or violence in the territory over which it presides. However, the state does not simply wield one type of power based on either coercion or consent. Though practices and outcomes can reinforce the mythologized image of the monolithic state, the state is also a field of power and a site of struggle. Simply put, the state is not a unified entity; it is an arena where policy alternatives are contested and where the state, comprised of many different parts, each moving to its own drumbeat, participates among other actors. Addressing the question of who decides where racial boundaries lie implicates multiple layers of power and privilege at work within the state apparatus. The different government bodies, agencies, and departments in the census policy sphere have varying and potentially conflicting goals, agendas and interests. Rather than being an indicator of dysfunction in the ideal-type Weberian bureaucracy, intra-institutional conflict is both normal and central to the institutional organization and function of the state apparatus. The political institutions of the state are populated by actors who bargain, form coalitions, interpret ideas, and negotiate constraints. Political decisions about race are contested ground in multiple policy-making arenas.

Additionally, in an age of deliberative democracy and networked governance,^{Footnote 34} the boundaries of state and society are rarely clearly delineated. The technocratic change from ascribed classification to self-identification in the 1970s catalyzed a shift in the locus of power in designing census categories. Though the state still retains control over final classificatory decisions, substantial interactions between the state and advisory groups, public consultations, and targeted social engagement are now institutionalized procedures of census politics. But because racial classification is a type of decision-making most conveniently and most often done behind closed doors,^{Footnote 35} the idea of the state as a singular entity that exists autonomously from society works to both hide and enshrine these contractions, just as schematic outcomes make the state appear coherent, unabridged, and static.

Consider the review of federal classification standards in the United States as an illustrative example. In 1977, the Office of Management and Budget (OMB) implemented Statistical Directive 15, which mandated the use of four standardized racial categories in the official reporting of all statistical data pertaining to race: American Indian or Alaskan Native; Asian or Pacific Islander; Black, and White. In this formulation, Hispanic is considered an ethnic, not racial, category. Fifteen years later, amid changing demographics and patterns of racial intermarriage many felt these categories had become outdated,^{Footnote 36} and in 1993 the House Subcommittee on Census, Statistics, and Postal Personnel initiated hearings to review federal classification standards. In response to this exercise of congressional oversight, OMB proposed a comprehensive review of Directive 15. These two interrelated policy arenas each possessed a unique purpose, decisional bias, network structure, participants, and decision-making routines. OMB’s bureaucracy-led review had a mandate and desire to reach consensus and propose a viable alternative to the congressional committee. The purpose of the congressional hearings was to discuss and debate — and, it could be argued, sensationalize — alternatives and the review process.

At the bureaucratic level, the OMB enlisted the Committee on National Statistics of the National Research Council to determine the most important objectives of federal classification standards and the most feasible means of reaching them. To this end, the Committee identified eight major objectives for a federal standard of racial and ethnic classification: (1) fostering the exchange of statistical reports between agencies; (2) ensuring the availability of data for the monitoring of discrimination against minority groups; (3) designating the system for administrative and statistical records so that the data are reliable even when disaggregated by race and ethnicity; (4) ensuring that the categories are mutually exclusive and exhaustive and that the number of categories are manageable in size; (5) incorporating flexibility so that the standard can be adapted to the context of its use; (6) promoting longitudinal consistency for categories over time; (7) producing relevant and meaningful categories for federal policy purposes; and (8) producing categories that are relevant and applicable to individual respondents.^{Footnote 37} The Committee recognized that some of these objectives would undoubtedly conflict.

OMB also established an interagency committee in 1994, co-chaired by the Bureau of the Census and the Bureau of Labor Statistics and comprising representatives from thirty government agencies, to review the racial and ethnic standards and to make recommendations to OMB for any revisions. The different agencies within the state had different priorities and preferences that influenced their participation in the committee. Agencies with statutory responsibility for policy development, program evaluation and civil rights monitoring were concerned about the historical continuity of the data in light of the proposed changes. Data collection agencies, which use racial and ethnic data for federal programming, redistricting, and for the analysis of social, economic and health trends of the population, were also concerned with the historical continuity of data, but felt that the categories used in Directive 15 were confusing to respondents, overly broad, and inconsistent. The data collection agencies were also concerned about the significant technical, operational, and cost issues implied if a “mark one or more” approach was to be adopted.^{Footnote 38} Though divided initially, the interagency committee had clear monetary and professional incentives to reach consensus on these issues by the 1997 deadline. In addition, OMB sought public comment through notices published in the Federal Register and at public hearings, and initiated a comprehensive testing program. Throughout the review process, the bureaucratic arms of the state prioritized administrative needs such as statistical exchange within the state apparatus, data consistency, program and monitoring requirements, and the production of categories that were meaningful and relevant to the community and for policy purposes — in other words, the legibility, transportability, and standardization of the American racial schematic.

In the congressional arena, subcommittee hearings were held in 1993 and 1997. Though a number of issues were under consideration regarding the federal classification of race and ethnicity, the proposed addition of a multiracial category soon rose to the front of the battle lines. Project RACE (Reclassify All Children Equally) and the Association of Multiethnic Americans (AMEA) advocated for multiracial inclusion in the census. While the positions of both groups evolved over time through coalition agreements and political decisions, the major distinction was that Project RACE was unrepentant in advocating a stand-alone multiracial category, whereas AMEA endorsed a multiracial category that was followed by a listing of racial and/or ethnic groups duplicated from the main classifications.^{Footnote 39}

The multiracial movement changed the discourse of the census. In contrast to state concerns about the enforcement of civil rights legislation, funding allocations, and congressional redistricting, multiracial organizations introduced the idea that the census was a vehicle for the recognition of identities, separate from the material incentives associated with identifying as one race or another. Contending that appropriate labels are crucial for the positive self-development of children, Susan Graham of Project RACE testified:

Carlos Fernandez, the President of AMEA, testified that his organization sought “a government-wide reform to accommodate and acknowledge the particular identity of people whose racial or ethnic identification encompasses more than one of the designated classifications currently in use.”^{Footnote 41} The multiracial movement premised its arguments on the language of recognition, arguing that the right of self-identification and recognition of what people “really are” is not just a matter of civil rights, but is something more fundamental to the ethos of liberalism and democracy. Using very individualistic terms to make this point, the mixed-race movement argued that the interests of the state, schools, or other officially recognized racial groups were secondary to those of the individual checking the box.^{Footnote 42}

On the other side of the debate, civil rights organizations were strongly opposed to the addition of a multiracial category. Henry Der, of the National Coalition for an Accurate Count of Asians and Pacific Islanders, questioned the point of having a multiracial category, contending that “unless there is adequate testing or sufficient evidence is provided about the experiences of biracial or multiracial persons that are unique to their being biracial or multiracial, the National Coalition asks the Census Bureau not to create a biracial or multiracial category at this time. It is not clear at this time what is the salience of knowing how many biracial or multiracial persons there are.”^{Footnote 43} Representatives from the National Urban League, the National Congress of American Indians, and the United States Commission on Civil Rights also came out against the proposal to add a multiracial category. In 1994, the Lawyers’ Committee for Civil Rights Under Law, the NAACP, the National Urban League and the Joint Center for Political and Economic Studies issued a coalition statement on the issue:

Civil rights organizations invoked color-consciousness to emphasize the importance of maintaining existing racial categories in order to protect and maintain the legislative gains of the civil rights movement. Here, the value of counting by race in the census was connected to the liberal goals of promoting racial harmony and diversity through the monitoring of circumstances of racial inequality and disadvantage. Civil rights organizations took the analogy one step further, emphasizing that attaining substantive equality would require constant vigilance; the NAACP noted its concern with the proposals to add a multiracial category stemmed from its belief that “segregation and discrimination in this country has to be battled with deeds, not just with words,” and invited “all multiracial people who so identify themselves, to join us in that struggle.”^{Footnote 45}

The outcome of these political battles was not determined by the power or prominence of its various participants. The multiracial movement was relatively new, tiny, unrepresentative, and operated on shoe-string budgets,^{Footnote 46} but nevertheless held its own against long-established, better organized and more powerful organizations such as the NAACP and the National Urban League while testifying before Congress. Civil rights organizations were anxious about the tangible impact that stand-alone mixed-race categories would have on their constituent numbers and the ability of the state to adhere to its obligations under civil rights legislation; they also had a vested interest in preserving the boundaries of their racial categories, which they had been using for decades to converse with the state. The implication is that these organizations represented well-defined, recognizable constituencies, able and prepared to parley with the state on its own terms. Though multiracial groups argued that mixed-race people faced discrimination by virtue of being mixed-race, they more often clothed their arguments in a language of “recognition,” contending that the absence of a census category was a denial of the validity and worth of their identities. This strategy, however, was not well aligned with the purpose of the policy, which was not about recognition but rather concerned the collection of accurate data in order to fulfill the state’s legal obligations and justify positive action.

Ultimately, OMB’s review process was the most decisive in the change to a “mark one or more” approach on the U.S. Census in 2000. After completing its analysis in 1997, the Interagency Committee unanimously recommended that the method for enumerating the mixed-race population take the form of multiple responses to a single question and not a stand-alone “multiracial” category. This course of action allowed the state to measure the growing mixed-race population without significantly disrupting the consistency of longitudinal data, using established classification schematics that were already well understood. Rather than being the lesser of all evils, the Interagency Committee clearly recognized the necessity of counting the mixed-race population and sought the most effective means for doing so, given its priorities involving satisfying statutory and program needs, dealing with voting rights issues, and taking into account data continuity concerns, financial costs, and public sentiment.

IV. Normalizing Racial Categorizations

All boundary-making processes are fundamentally, unavoidably normative. The use of racial classifications in public policy, for example, frequently treads on moral ground. The ethical value of colorblindness, once used as a weapon by NAACP counsel Thurgood Marshall in arguing for the end of racial segregation in Brown v. Board of Education, is now used by opponents of affirmative action to dismantle what they see as a regime of special treatment. “The way to stop discrimination on the basis of race,” Justice Roberts argued in Parents Involved v. Seattle, “is to stop discriminating on the basis of race.” Roberts’s logic in this case, widely criticized in critical race theory and a clear example of what Ian Haney Lopez calls “racial jujitsu,” effectively coopts the moral force of the civil rights movement, deploys that power to attack attempts at racial redress by defining racism as any and all use of race, and simultaneously defends racial hierarchies by defining nonracism as all interactions not explicitly predicated on race — no matter how racialized their consequences may be.^{Footnote 47} Because racial categories have been used to dominate and subjugate in the past, colorblind advocates seek to avoid the terminology altogether, thereby conflating the distinction between racialism (distinctions between supposedly discrete racial groups) and racism (the use of those distinctions to deny access to power, privilege, property, protection, and privacy).

The use of racial classification for good and ill is often discussed; less often explored is the fact that the making of racial classifications is itself a normative process. That is, insofar as all boundaries make distinctions for inclusion and exclusion, there are underlying norms that inform racial categorizations and express assumptions about who really, authentically, authoritatively belongs. All census categorizations legitimize and reify distinctions that become meaningful in the body politic: between those who can access the rights of citizenship and foreigners; between those who earn enough to provide a certain standard of living and those who require government assistance; between those whose domestic partnerships are legally recognized and those whose partnerships are not. The schematic created through and employed by the census embeds realms of meaning as it orders and binds. Contestation over rules of racial classification implies a continuing struggle over the normative implications of race and the meaning-making function of the census.

In Great Britain, as elsewhere, the meaning of race is linked to discourses of citizenship and belonging. As the nonwhite population grew in the decades following the Second World War, Britain imposed racially specific immigration controls and created new measures to prevent racial discrimination toward the nonwhite population already residing in Britain. The Race Relations Acts of 1965 and 1968 sought to end discrimination based on race and were wrought with an air of prevention; at the time it was believed that without political institutions to address the social problems of immigrants, Britain would soon be facing the prospect of American-style racial tension and violence.^{Footnote 48} The Race Relations Acts created special bodies to deal with problems faced by immigrants in relation to discrimination, social welfare, and integration, and to educate the population as a whole about race relations in an attempt to minimize the potential for racial conflict.

Though the census was the most obvious vehicle to gather information on both the extent of racial discrimination and the effectiveness of the Race Relations Acts, it was not considered feasible to ask a question on race in the 1971 or 1981 Census. In the late 1960s, political elites recorded concern about the “political difficulties” of including a question on race on the census,^{Footnote 49} which could potentially provide more fodder for conservative politicians to make the case for more-restrictive immigration policies. For example, between 1967 and 1972 Enoch Powell made frequent speeches in the House of Commons and elsewhere on the subject of the number of colored immigrants in the United Kingdom, often using the 1967 statistical prediction (invalidated shortly thereafter) that the nonwhite population of Great Britain would reach 3.5 million by 1986.^{Footnote 50} By the late 1970s, the “political difficulties” of counting by race referred to the nonwhite population rather than conservative elites. Racial minorities, suspicious of Thatcher’s proposed Nationality Act, refused to cooperate with the 1979 Census Test, which included a question on race. A local campaign in Haringey urged racial minorities to condemn the question, claiming that the census data were linked to nationality laws that “would make nationality dependent on your parents’ nationality, not where you were born. . . . If we say now who is and who is not of British descent, we may one day be asked to ‘go home’ whether we were born here or not.”^{Footnote 51}

Throughout the 1970s, 1980s, and 1990s, racial minorities sought some means of demonstrating that they were second- or even third- generation British. Unlike the white majority, racial minorities eventually had little objection to a public testament that Britain had become a diverse society, but “what they did find deeply offensive — understandably enough —was any indication that their distinctiveness might be read as an indication that they were in some sense non-British.”^{Footnote 52} In the late 1970s, the government found the label “Black British” unacceptable because it placed too much emphasis on differences of race and/or color; Cabinet Ministers instead instructed the Office of Population, Censuses and Surveys (OPCS) to find an alternative system of classification couched exclusively in ethnic terms, avoiding the use of the words “black” and “white.” This unease with the concept of race is not an indication of its nonexistence, but rather its omnipresence. In Britain, discourses of race and nationality are articulated with the same breath, meaning that “statements about nationality are invariably also statements about ‘race’.”^{Footnote 53} Attempts to promote racial equality are often perceived as threatening to national unity, and British identity remains strongly raced and classed.^{Footnote 54}

As the 1980s and 1990s wore on, Great Britain experienced a more prominent disconnect between discourses of race relations and immigration as second- and third-generation blacks and Asians lay claim to being just as British as anyone else. In previous decades, blackness was perceived by the majority population as being synonymous with “immigrant”^{Footnote 55} but the growth of a politically active generation of British born and bred racial minorities not willing to settle for anything but full citizenship (complete with a sense of belonging as part of the nation) helped to challenge the unspoken, but dominant, paradigm that Britishness was equated with whiteness.^{Footnote 56}

After the inaugural appearance of the ethnic question in the 1991 Census was heralded as a resounding success,^{Footnote 57} the Office for National Statistics (ONS) made several important changes to the format of the question. The proposal to include the “Black British” and “Asian British” as headings rather than categories in the 2001 Census came from ONS. The working group charged with revising the ethnic question had struggled with the issue of retaining high-quality and comparable data that detailed ancestry while allowing respondents to identify as British, since there was still such demand from the public. On this topic, one working group member commented:

The inclusion of some way of recognizing Black British or Asian British identities, according to this working group member “had been a sticking point for a long time.” This solution, which acknowledges both race and British nationality and/or citizenship, suggests a symbolic function of the census beyond the task of counting the population. The more intangible elements of census politics, where the census collides with ideas of citizenship and belonging, demand that policymakers negotiate between technical requirements and the politics of recognition.

The 2001 Census also disaggregated the “white” category and provided the options of “British,” “Irish,” and “Any other White background” with a write-in space. These options stem from the efforts of a surprisingly vocal lobby that persuaded members of the Working Group and ONS to add an Irish category. The Irish lobby used the notion of disadvantage to make their case in multiple institutional access points, targeting both the working group and other parts of the state, finding a particularly powerful ally in the Department of Health.^{Footnote 59} This was a necessary step; as one ONS representative noted, lobby pressure from other white ethnic groups, such as the Cypriots, the Greek Cypriots, the Cornish, and the Welsh were not persuasive because they could not demonstrate that their groups had experienced disadvantages in health, education, and the social realm.^{Footnote 60} When the ethnic question working group recommended the inclusion of an Irish category on the 2001 Census, the ONS was hesitant: “ONS was reluctant to have it at all . . . maybe because it wasn’t driven by color. ONS didn’t really have a strong appreciation of the nature and scale of the disadvantage. Although evidence was beginning to be published then. . . . The Irish group was a tougher sell — it was about 1997 when they came around. There was quite a lot of resistance.”^{Footnote 61} Though the lobby eventually achieved their goal, ONS’s unease with counting Irish was partially because the category differed from the state’s conceptualization of what racial disadvantage is, and therefore, what the ethnic question was designed to measure.

The most recent census round in Great Britain featured a new question on national identity, which preceded the question on ethnicity. In England and Wales, respondents chose from six options: “English,” “Welsh,” “Scottish,” “Northern Irish,” “British,” or “Other” with a write-in space. This new question is not a response to the demands for a means of allowing racial minorities to identify themselves without feeling as though such an identification would detract from their sense of belonging in the national community (i.e., Black British or Asian British), but rather is a byproduct of processes of devolution in Great Britain, a growing sense of national identity in Scotland and Wales, the thrust to keep British nationalism intact through policies emphasizing “community cohesion,” and a domestic concern about increasing immigration from Eastern European countries of the EU. The census is a reflection of a national community and when states use ethnic enumeration to promote multiculturalism and integration, public recognition through the census becomes a means for groups to seek and attain a sense of belonging.

V. Conclusion

Racial classification is a process, and a highly political one at that. Classification systems require consistent and unique principles that underpin the method of creating order out of chaos, categories that are mutually exclusive, and an organizational infrastructure that is complete and all-encompassing.^{Footnote 62} Categories, classifications, and standards are often entirely taken-for-granted; racial categories seem so obvious, so institutionalized, so very normal that they appear ahistorical. However, there are normative stakes in classifying and counting. Categories presuppose distinctive symmetry — white and black are both equally racial, but are indeed separate races. The act of classifying creates a connection between phenomena judged to be similar, and therefore each classification abides by criteria that determine which items, people, or groups belong and which do not. Thus, while black and white are considered racial categories, Hispanic, Irish, or Cree are not. Moreover, once a racial schematic is established, it renders the work, conflict, discord, uncertainty, and normativity undertaken in the process of its creation invisible.

This work — conceptual, normative, pragmatic — is fraught with tension, ambiguity, and power. As evidence from the development, abandonment, and recreation of the race question on the Canadian census demonstrates, the rules of racial classification can change substantially over time, driven by dominant ideological paradigms that dictate the role of race in the national imagination. The state plays an integral role in these processes, as the example of the American review of federal racial classifications makes clear. However, the state is neither a unified entity nor the only stakeholder in debates over where the boundaries of race lie. The census is not simply a means of enumerating the population or the categorical basis of ameliorative laws and policies; it is also a vehicle of public recognition. Though racial minorities in Britain first avoided the state’s gaze, they later sought to use this means to lay claim to substantive belonging in the nation.

The design and designation of racial boundaries is particularly pertinent to the study of equality and public policy. Boundaries are made meaningful when they are challenged by those outside or on the periphery of an established category. Similarly, the ideal of equality pertains not simply to equal opportunity or consequence, but also equality in the process of becoming. In Fisher, the Supreme Court confirmed once again that any use of racial classification must meet strict scrutiny, for when government decisions “touch upon an individual’s race or ethnic background, he is entitled to a judicial determination that the burden he is asked to bear on that basis is precisely tailored to serve a compelling governmental interest.” The Supreme Court may or may not have erred in its decision in this case; an indictment either way is far beyond the reach of this essay. Where the majority did falter, however, is in its assumption that racial classifications and governmental interest are distinct entities, wholly separate and constituted independently of one another. The state is highly involved in the process of making and manipulating racial taxonomies, which can be used for good or ill. Claims to race-consciousness, colorblindness, or post-racialism, must therefore first wrestle with whether or not the methods of making race that lie beneath them are fair, transparent, and equitable.