Junk data in census makes us a nation of ‘others’

We are becoming a nation of write-ins. So found a report released last month by the U.S. Census Bureau. When filling out census forms in 2010 (the year of the last national population tally), more people than ever before did not choose one of the race options provided; they chose “some other race.”

The report, part of a years-long project to re-examine the census’ racial and ethnic categories, underscores the extent to which demographic changes in the U.S. have outpaced our methods of documenting them.

The bureau’s concerns about the unrepresentative nature of its census categories appear to be well-founded. Approximately one-third of the 47.4 million respondents who self-identified as ethnically Hispanic also self-identified as “some other race.” A full 96.8 percent of all people claiming to be “some other race” were Hispanic.

The Pew Research Center’s summary of the report noted, “The ‘some other race’ option … was never intended to be a category selected by so many respondents. The category was added to the 1980 census form to capture the small numbers of people who did not select one of the official race categories. But since then, it has grown to become the third-largest race category in the census.”

To understand why this may be so, let’s take a look at the census form (PDF). The race question is preceded by the ethnicity question, which asks if you are “of Hispanic, Latino or Spanish origin.” (The broad-stroke distinction that is usually made between ethnicity and race is that race is biologically determined while ethnicity takes more cultural factors into account. For example, I might be racially classified as black but ethnically identify as Arab or Cuban.) The questions are accompanied by a note: “For this census, Hispanic origins are not races.”

One would think, of course, that a country as diverse as the U.S. would provide more ethnic options than Hispanic/Latino/Spanish or not-Hispanic/Latino/Spanish, but these are — inexplicably — the only two. (Persons who select the first category are given a space to further identity themselves by country.) The racial identity options are similarly puzzling in their apparent arbitrariness; they include “white,” “black, African Am., or Negro,” “American Indian or Alaska Native,” “Filipino,” “Vietnamese,” “Guamanian or Chamorro” and “Samoan,” along with several others. The “other Asian” category comes with instructions to “print race, for example, Hmong, Laotian, Thai, Pakistani, Cambodian and so on,” and self-identified “other Pacific Islanders” are told to “print race, for example, Fijian, Tongan and so on.”

What the census questions suggest is that race may coincide with national origin if you’re Laotian or Fijian but not if you hail from one of the 21 Spanish-speaking countries in the world. In other cases, skin color designates race — but not if you’re brown.

The U.S. government “periodically alters race and ethnicity questions to keep up with shifts in the social fabric of the nation,” according to the optimistic wording of a 2013 USA Today article. How well it actually keeps up is debatable. It was not until 2000 that census respondents were permitted to check more than one box for race. The article speculates that the racial category “‘Negro’ may finally be dropped in 2020.”

Survey methodology, too, is far from cutting-edge. Kenneth Prewitt, a former Census Bureau director, argues in his book, “What Is Your Race? The Census and Our Flawed Efforts to Classify Americans,” that “21st century statistics should not be governed by race thinking that is two and a half centuries out of date”— referring to racial divisions proposed in 1776 by German scientist Johann Blumenbach. The U.S. is, Prewitt writes, “the only country in the world firmly wedded to an 18th century racial taxonomy” in which were embedded “theories of a racial hierarchy: There were not just different races but superior and inferior races.”

Statistics play a significant role in shaping how people see themselves and one another. In a nod to the work of political scientist Benedict Anderson, Prewitt explains that during the colonial era, “powerful nations categorized and counted their subjects in order to rule them, and … those subjects gradually took on the identities imprinted on them by colonial census taking.”

Anthropologist Adrienne Pine, whose areas of expertise include U.S. imperialism in Latin America, contends that the census is inherently problematic because its categories propagate the erroneous notion that race is “something real, fixed and therefore quantifiable” rather than a social construct. Race is a product of racism and colonialism, she wrote in an email, and its boundaries are in constant flux thanks to “changing societal dynamics of power and exclusion.” (Irish-Americans and Italian-Americans were previously not considered white, for example.)

According to Pine, the census form is “patently unequipped to reflect the complexity of racism and race as a process” because it obscures the extent to which racial categories are wrapped up in racist-imperialist history. But, she notes, this “vicious cycle of bad science” gets perpetuated in part because members of historically maligned or underrepresented groups often depend on financial resources that are pegged to census data.

The Pew Research Center confirms that race and ethnicity tallies “help determine how congressional districts are drawn” and “how $400 billion in federal aid is distributed.” The Census Bureau, meanwhile, claims that Hispanic-specific data collection is necessary in order to implement proper education, housing and health programs and to enforce laws such as the Civil Rights Act. But one is hard pressed to determine how such a sloppy survey might helpfully influence policy.

Within the next three decades, non-Hispanic white Americans will constitute a minority in the U.S. It is quite possible, therefore, that the country will soon be populated by a majority that self-identifies as “some other race.” (Hispanics have already surpassed African-Americans as the largest minority group, making up 17 percent of the population in 2012.)

Perhaps with these near-future demographic changes in mind, the Census Bureau is considering combining the two questions on Hispanic ethnicity and race. Hispanics, in other words, may be granted race status in future censuses. It’s not so simple, though; the Pew Center has found that most Hispanics prefer to self-identify by their country of origin first, raising questions of how much such a change would improve community representation.

The fact that I myself have continued to employ racial and ethnic terms while alleging the defectiveness of divisive labeling systems underscores the obstacles to a radical rethinking of demographic categories. But one option for thinking outside the box in future censuses might be to focus more on socioeconomic factors — which, unlike race, are quantifiable.

In the end, though, maybe the census form should be left as is. After all, a nation of “others” is a pretty good label for such a persistently racist system — and one not at risk of being confused with a nation of equals.