How Might We Measure Race Without Reifying It?

It always intrigues me to be told that by thinking differently from the herd (a hallmark of progressivism) I must actually be a faux progressive. Recently, for challenging my government’s use of racial classifications on the census form, I have been called a selfish, irresponsible, childish bigot by readers who are (presumably) generally in favor of disagreeing with one’s government.

These comments fascinated and amused me, not because many disagreed with my point (I expected that of course), but because the disagreement was so uniform. I had expected a cacophony of different perspectives and some serious debate about a pretty important issue in our political culture. But with a few notable exceptions, most of the substantive critiques (to date) have instead fallen into one of the following categories:

1) “nice idea, but utopian and therefore dangerous”
2) “if we don’t measure race we can never get past racism”
3) “Charli only thinks this because she’s white and obviously not a genuine liberal.”

Let me briefly elaborate my actual views by explaining by responding to these three critiques, then conclude with some modest proposals for acknowledging race on the census forms without reproducing or imposing racial categories on the public.

Critique #1: Opposing the government’s racial categorization on the census is nice in theory, but dangerous in practice.

“It’s a nice sentiment, but still just a utopian one.”

“Also, it’s very, very easily twisted to right-wing purposes.”

Both these statements may be true, but I also don’t think they’re convincing reasons not to hold the position I do. It’s just as easy to twist census race data for nefarious purposes (like to throw people into internment camps) as to twist arguments that we should avoid forcing people into racial boxes. As Comply notes in his/her comment, the Census has recently been used for the former purpose. Of the two, I’d prefer to be accused of the latter.

As to the utopianism of Olmos’ argument, the idea that a black man could become president in this country was also considered “utopian” until very recently; that argument was in fact used to try to prevent Democrats from nominating him. On both counts – empirically and morally – I find it the idea that progressives should avoid suggesting a possibility simply because it’s utopian uncompelling.

Critique #2: Checking Racial Categories on the Census is Necessary In Order to Combat Racism.

“Unfortunately, we still need to be able to track racial inequality.”

“The Census isn’t a form of political expression, it’s necessary in order to get a handle on what the country is really like in 2010.”

Misconceived survey methods are unlikely to result either in accurately tracking or alleviating racial inequality. They may in fact perpetuate it. There is nothing “correct” about the racial demographic data that answers to this form will produce, as the boxes on the form exclude numerous racial and ethnic groups in this country and will likely result in many people checking a box that does not accurately describe them. For example, while the form invites people to identify themselves as “of Hispanic, Latino or Spanish origin” it does not include categories for those of Middle Eastern origin. Nor is there a box to check if you would like to identify yourself as multi-racial.

So, as I see it, the census is actually the ultimate form of political expression – a politicized expression of racial politics endorsed by government, one that structures not only how we should think about and study race but also how we think about discrimination in general. (Why is race data gathered but not religious data in a country where perceived discrimination against various religious minorities is a significant political issue? Why collect binary data on sex (while excluding those who don’t fall into either category) but exclude data on sexual preference?

In short, my fear is the data generated by this method will not reflect racial/ethnic divisions in this country accurately so much as construct misleading notions of what the country looks like in the minds of citizens and bureaucrats, while communicating to individuals who see themselves as members of groups excluded from the form that they are so marginal as to merit no formal consideration.

Critique #3: Resisting Government-Imposed Race Discourse is Something Only Illiberal White People Do

“Only white people pretend that being color-blind is even possible, let alone good.”

Now, statements like this imply you know what all people of color do or do not think or do. I’d encourage you to watch Edward Olmos’ speech again and consider the fact that he would normally be considered ‘Hispanic’ – though as far as I can tell from his public statements (and having not interviewed him in person on the matter), he himself rejects this label.

But for what it’s worth, neither Olmos nor I argue anyone should be “color-blind.” Rather his argument as I understand it, and certainly mine, is that our political institutions should be incentivizing individual behavior that renders race less salient as a means of categorizing people, rather than doing the opposite.

So What, Then?

We are back to the question in the header of this post: how does one acknowledge and capture the “social facts” of race and racial stratification in a country without reifying those social facts themselves?

Off the top of my head at 3:23 in the morning, I would suggest a simple answer, and a slightly more complicated one:

The Simple Answer: Remove all boxes from the ethnicity and race questions and retain only an open-ended space where individuals can write in their race. In short, treat every individual the way the government now treats all those individuals who choose the “some other race” category. The result will be a set of text data that can be easily coded using automated tools to produce what I hypothesize will be a somewhat similar result to what you get when you prompt people, but without any of the negative externalities in terms of feedback into social processes. (That is, many people will simply write “white” or “black” as easily as they would have checked the box, but without the government telling them that this is how they should think about themselves.) You will also get a very interesting and useful long tail of more complex answers that could help the government enormously in thinking about the complexity of race in this country. Methodologically, this approach would also provide a means to empirically measure over time the extent to which specific racial categories used by the general population are becoming more or less salient or how they are changing. And it will have the effect of forcing people to reflect on race (that is, not to be color-blind) without telling them what to look for or how to think. Most importantly, no one will be excluded – racial categories will emerge if at all from the collective speech acts of the citizens of the US, rather than be imposed on them by the government. Ultimately, this is a much more accurate way to measure “social facts” like ideas about race.*

The Slightly More Complicated Answer. In an ideal world, I’d want the Census Bureau to go a step further, and include two open-ended questions instead of one. The first would ask respondents to write in their race as they identify themselves. The second would ask the respondent to write in his/her race as they perceive they are identified by others. This would not only give the government much better tools for tracking changing conceptions of race and racial inequality in our country, it would encourage all US citizens to be much more reflective about race/ethnicity and what it means in our communities.

I remain concerned that the census, as currently structured, does the opposite, and I stand by my opinion that this is unfortunate.

* They might follow suit in gathering non-dichotomous information on gender and gender minorities, and capturing categories relevant to combating other forms of discrimination like religious belief.