The American government’s standard racial and ethnic categories make a certain amount of sense. They’re approximately continental groupings, and they capture how, over many centuries, circumstances have brought visually distinguishable people from different parts of the world to this country. Native Americans settled here long before the arrival of white Europeans, who brought black slaves from Africa, and subsequent immigration introduced populations from Asia and Latin America.
Thus the default options of white, black, Hispanic, Native American, and Asian. Beyond their relationship to geographic origins, these categories roughly correspond to the way Americans sort each other out socially. And they also correlate, to varying degrees, with skin color, cultural and linguistic heritage, and genetic ancestry. Given the names and pictures of 100 randomly chosen U.S. residents, the typical American could probably put the vast majority of the subjects into the same bins in which those subjects would place themselves.
But there’s a lot more going on just under the surface, which is the subject of Classified: The Untold Story of Racial Classification in America, the excellent new book from Antonin Scalia Law School professor David E. Bernstein. Classified traces the history of these categories, with a focus on their gray areas — especially cases in which ambiguities have proven highly consequential.
It’s a common belief that race in modern America is just a matter of self-identification — and often, it is. As Bernstein notes, “No one is going to police what box someone checks off on a census form or a mortgage application.”
But someone has to decide which boxes are available on such forms to begin with. And sometimes, racial classifications are more than just a matter of descriptive record-keeping: One’s race might determine one’s eligibility for numerous preferences or play a role in allegations of racial bias. In those cases, it’s an invitation for fraud to rely unquestioningly on self-identification. Some minority-owned business preferences, for example, require affidavits or other documentation to prove an applicant’s race.
Much of the system’s history traces back to the civil rights era and its policies against discrimination. The Civil Rights Act of 1964 banned racial discrimination in employment, for example, and tasked the Equal Employment Opportunity Commission with enforcing it. In collecting workforce statistics by race, the EEOC grappled with many tricky questions, such as which categories to include (verdict: “Negroes,” “Spanish-Americans,” “Orientals,” and “American Indians”) and how workers would be classified (visual identification by the employer).
The categories were already solidifying by the mid-1970s, when the Ad Hoc Committee on Racial and Ethnic Definitions recommended the schema of “American Indian or Alaskan Native, Asian or Pacific Islander, Black/Negro, Caucasian/White, and Hispanic” — with Hispanic considered an ethnicity rather than a race — and several federal agencies adopted these for a trial period. Then, in 1977, the Carter administration released Statistical Policy Directive No. 15, which imposed these categories, with some tweaks (such as dropping “Negro” and “Caucasian” as alternative names), as uniform data-collecting standards for federal agencies.
The directive noted the classifications were not “scientific or anthropological in nature” and also did not determine “eligibility for participation in any Federal program.” But they soon became the de facto standard for governments and researchers nationwide. And despite some further adjustments over the years, they more or less remain with us today.
Bernstein highlights numerous problems stemming from these classifications, which, in the spirit of the topic under discussion, we can crudely but helpfully group into four categories: general incoherence, outright fraud, legitimate disputes over how to classify an individual or subgroup, and debates over whether to divide the major groups into smaller ones.
Racial classifications tap into numerous features of humanity, ranging from geographical ancestry and skin color to language and culture. But as Bernstein explains, they don’t combine this information in a consistent way (and indeed, it would be hard to). Directive 15, for example, counted someone as Native American only if they maintained a “cultural identification through tribal affiliation or community recognition,” something not demanded of other groups. The official “white” category includes those who trace their ancestry not only to Europe but also (in Bernstein’s words) to “Asia west of Pakistan, the Asian parts of the former Soviet Union, and North Africa,” meaning that not everyone whose roots lie in Asia is Asian, and not everyone whose roots lie in Africa is black.
Further, it’s awkward to include the (ethnic and linguistic) “Hispanic” category alongside the other (racial) options, typically requiring a separate question for which the only responses are Hispanic and not-Hispanic — and in which a white European might find his shot at affirmative action benefits (if he’s from Spain). And Directive 15 did not allow multiracial individuals to check more than one box, though this has changed since.
Next, members of nonpreferred groups can gain access to benefits by checking the wrong box on purpose. Or at least, they can give it a shot. Paul and Philip Malone provide one of the most famous, and funniest, examples of this. The two brothers became Boston firefighters in the 1970s. They initially failed the civil service exam as whites but then passed by identifying as black, which gave them access to a lower cutoff on the test. When Philip again listed his race as black on a promotion application, the fire commissioner noticed, became confused, and questioned the brothers separately. “Paul told [the commissioner] that his father was black,” Bernstein writes; “Philip reported that someone in his family was black, but he did not remember who.”
After a disciplinary hearing, the brothers lost their jobs. They couldn’t demonstrate a “good faith belief that they could claim black status,” didn’t look black to the hearing examiner, couldn’t produce documents listing them as black, and couldn’t show that they presented themselves or were regarded by others as black socially.
Other times, there are more legitimate arguments over whether a given individual or population “counts” as part of a group. Categories’ definitional limits are often determined in part by lobbying: The Association of Indians in America, for example, succeeded in getting Indian Americans classified as Asian in Directive 15 when they’d been considered white previously.
At the individual level, disputes over eligibility for preferred categories often play out in unseemly legal proceedings. Bernstein highlights two cases in which Small Business Administration loans depended on whether the applicant was “really” Hispanic. Steve Lynn “was a Sephardic Jew whose ancestors had fled Spain centuries earlier,” and that counted. But Christine Combs didn’t have the same luck, even though her “maternal grandparents were born in Spain” and “she grew up in a bilingual family, was fluent in Spanish, and acted as an interpreter for Mexican and Spanish customers.” She hadn’t shown that she faced discrimination for being Hispanic, and the officer handling her case “noted that neither Combs’s maiden name nor her married name was recognizably Spanish, and her blond hair and blue eyes did not give her a noticeably Hispanic appearance.”
There are also debates over whether the existing categories are too big. Some of these arguments simply note the immense diversity that can be found within each group. Bernstein discusses how Pakistanis and Filipinos are lumped together as Asians, while Icelanders and Moroccans are both considered white, and Native Americans come from “dozens of distinct tribes with different languages, cultures, problems, and opportunities.”
Other arguments in this vein come from activists who want to see certain groups in particular broken out. There have long been calls for a “MENA” category for Middle Easterners and North Africans, for instance, or for splitting up the Asian category to account for the fact that some Asian subgroups, such as the Hmong, have not been as socioeconomically prosperous as others and thus would be more likely to receive preferential treatment if considered separately. Bernstein further notes the efforts of white ethnics, including Italian and Polish Americans, to win status as official minorities.
At a minimum, Bernstein’s work here is informative and interesting. But he’s also trying to prompt a discussion about fundamentally reworking these categories or abandoning them altogether, albeit with full knowledge the chances are slim.
This wouldn’t merely be a semantic matter. How the government defines racial categories has an obvious relationship to which groups are able to receive preferential treatment, and therefore, any changes are likely to spur massive resistance from activist groups. Yet it’s worth considering the options Bernstein discusses.
At one extreme is the French solution, in which the government simply doesn’t classify citizens by race or ethnicity at all. Or the government could group people, when needed, in a more context-sensitive way. In medical research, that might involve genetics. In affirmative action, it might mean focusing efforts on specific populations to whom the United States owes a special debt, such as Native Americans and descendants of slaves, rather than granting preferences to a hodgepodge of various ill-defined racial and ethnic groups who are often present here mainly through voluntary immigration. Well beyond cleaning up definitional problems with our current racial categories, that might prove a workable compromise for the affirmative action debate in a rapidly diversifying country (though the landscape of that debate will surely change when the Supreme Court weighs in next term).
A less aggressive approach is just to tweak the categories we have — which, Bernstein admits, are “generally good enough” for some purposes, such as monitoring the most salient types of racial discrimination — to address the most egregious problems with them. Bernstein would, for instance, divide the Asian category in two (South Asians and East Asians) to better reflect the immense variety of places “Asians” can come from. That, in turn, raises the interesting question of what role a group’s size should play in our willingness to subdivide it: The Asian American population is drawn from an enormous and populous continent overseas, but it also constitutes just 6% of the U.S. population.
The Directive 15 categories are hardly the best system imaginable. But they’ve been in use for decades, and they’ve ingrained themselves deeply into the American psyche. A total overhaul may not be in the cards, but Classified helps readers understand where these categories came from and the many disputes they’ve spawned over time.
Robert VerBruggen is a fellow at the Manhattan Institute.
