Officially, the Census Bureau considers Arabs to be white.

November 25, 2014 9:08 AM

While the U.S. 2020 Census is still more than five years away, planning has already begun in earnest. Among the chief issues [pdf] under discussion is how to rearrange options for racial/ethnic self-identification [NYT], particularly the (allegedly undercounted [pdf]) Arab/MENA (Middle Eastern and North African) population. The Arab-American Institute argues [pdf] that this proposal will have "positive impact on the treatment and services available to members of the Arab American community" but some have voiced concern that government agencies could use this data for less-savory ends (again) [NYT]. Via 538.
posted by psoas (37 comments total)

What the hell is race anyway? I do not believe the Canadian census has a field for race, but, rather, language. This is a question I have often thought about because my children have dual-nationality with Canada and Japan. In one country they are fully Canadian, in the other they are considered "half Japanese" which is very silly.
posted by Nevin at 9:19 AM on November 25, 2014 [1 favorite]

Kenneth Prewitt, a former director of the Census Bureau, a professor of public affairs at Columbia University and the author of “What Is Your Race? The Census and Our Flawed Efforts to Classify Americans,” said the problem with asking about race was that it forced people to think about skin color. Mr. Prewitt said asking people about their national origin rather than race would be a more accurate way to classify them. “ ‘What race are you?’ is a very complicated question,” he said.

Yeah, why the hell DON'T we do this? At the very least in addition to asking about 'race'?
posted by showbiz_liz at 9:21 AM on November 25, 2014 [1 favorite]

What the hell is race anyway?

A major factor in explaining wide disparities in social outcomes for children raised in societies with long histories of racial prejudice and systematic racial oppression?

It's all very well getting all high minded and dogmatically insisting that "there is no such thing as race!"--but you might want to think for a moment that if the US census (and other agencies) took that approach, there would no longer be any way of measuring, for example, racial disparities in educational outcomes. There would also be no way of gathering statistics about racial disparities in police arrests, or in judicial sentencing, or in housing etc. etc. etc.

We need to keep asking about "race" for as long as we live in a society where systemic racial discrimination is a genuine and pervasive ill. Ceasing to ask about "race" would provide aid and comfort precisely to those who perpetrate racist actions, not to those seeking to fight against them.
posted by yoink at 9:30 AM on November 25, 2014 [30 favorites]

But if this metric is so important, why is it based on self-identification? This determination should employ SCIENCE! Like a DNA test.
posted by Rash at 9:33 AM on November 25, 2014 [1 favorite]

There's as much if not more genetic variation within ethnic groups as between them. This is because race as we think of it is a social construct. Which doesn't mean it's "made up," but it does mean you can't really test for it.
posted by showbiz_liz at 9:36 AM on November 25, 2014 [3 favorites]

Officially, the Census Bureau considers Arabs to be white.

Well, duh. Jesus was from that part of the world. Ever seen a painting of him?
posted by ZenMasterThis at 9:38 AM on November 25, 2014 [6 favorites]

Mr. Prewitt said asking people about their national origin rather than race would be a more accurate way to classify them.

"National origin" seems a useful proxy only for first-generation immigrants. For everyone else, their "national origin" is "America."
posted by yoink at 9:39 AM on November 25, 2014 [2 favorites]

I'm in the UK and ethnicity forms include various types of African, Asian (what we call Indian in the US) and South Asian as well as sub-categories for Bangladeshi.

In the US, I am a Latino. In the UK I once asked a nurse what box I should tick. There is no box for Latin-American or even for Native American. So the conversation went like this:

"So...Latin-American. They just come from Europeans right?"
"Well, yeah, mostly. There is Native Americans in there too.."
"Yeah, just put 'White European'"
"Got it."

I do get it. I'm not part of the history of race relations in the UK. However, my skin is dark and I do get mistaken for being Indian and sometimes asked "Where are you from?"

Perhaps the forms should just ask you what your skin color is?
posted by vacapinta at 9:40 AM on November 25, 2014 [3 favorites]

I'm not sure it's useful to create a new racial category to talk about a group of people not united by a single race.

I appreciate that it's meaningful for the census to collect data about people of Middle Eastern descent. I'm just not sure "race" is the best way to do that, nor am I sure that the current approach to "race" on the census is really functional at all, for any group.

Also, Rash -- the problem is there's no such thing as racial DNA, or any meaningful scientific way to ascertain someone's race. Much less so if you're talking about groups like Hispanic people, people of Middle Eastern origin, or Asian/Pacific Islanders as a group. Because those groups describe an incredible amount of genetic diversity and are very clearly artificial social constructs.

So if you're going to have these imprecise "racial" categories, you're going to have to base it on self-identity, since otherwise it's some weird totalitarian grouping where there's an authority determining that a given individual is a certain race. Which is suboptimal for obvious reasons.
posted by Sara C. at 9:40 AM on November 25, 2014 [2 favorites]

"Well, duh. Jesus was from that part of the world. Ever seen a painting of him?"

Iconysterical!
posted by clavdivs at 9:50 AM on November 25, 2014 [4 favorites]

Ah, the perpetual question of just what box Middle Eastern people should tick. I for one would really appreciate a better box to tick than "white" because white does not feel entirely accurate when you're Middle Eastern. I think I actually almost picked "Other Asian," once, after spending way too long overthinking whether Afghan counts as Central Asian and thus "Other Asian," and then I got in an argument with myself over how Afghan isn't exactly descriptive either given the wide range of ethnic groups, and anyway, I'm pretty pale and pass as white so maybe I should just suck it up and pick white...

Anyway, I still have no fucking idea what box to tick really. All I know is there is an appreciable difference between the way I and my darker skinned relatives are treated. There's a reason a lot of Middle Eastern people feel ambivalent and confused about whether they count as white. It can feel pretty galling to have to tick that "white" box when you know you don't have all the privileges of whiteness, but it can feel equally galling to tick that "white" box when you do have some of those privileges by virtue of how you look but your name and a bunch of other cultural/ethnic markers still mark you as Other.
posted by yasaman at 9:53 AM on November 25, 2014 [11 favorites]

But if this metric is so important, why is it based on self-identification? This determination should employ SCIENCE! Like a DNA test.

You do realize that there is no "scientific" test for the production of meaning, right?

*

As an academic with demographic training (though I would hesitate to call myself a demographer proper), I'd note a few oddities in the post. For instance, the "alleged" undercount of the MENA population in the post--sampling is widely considered more accurate for large populations than enumeration.

Further, I'm not familiar with any demographer who does *not* argue that enumerations undercount non-hegemonic groups. If you want to look to Latinos as a potential analogue, Princeton's Mexican Migration Project and Latin American Migration Project show pretty indubitably through mixed survey methods, including in communities of origin and in US receiving neighborhoods (among others), that people of Latino origins are undercounted in the US census.

That's why we sample. People often have really good reasons not to want to be found, and also we have to be careful about endangering them if/when they do get their info incorporated into official documents and data sets.

As for the "less-savory," it's worth noting that for all the US Census's claims to confidentiality, and in addition to the link, it was the mechanism for internment of Japanese-Americans in WWII.
posted by migrantology at 10:37 AM on November 25, 2014 [4 favorites]

It's important to note that the Census is considering moving away from the loaded, dated, imprecise term "race" for these questions. One of the challenges of a census is there is strong value in being able to track changes over time, which means that questions and terms sometimes change more slowly than society does so that there is consistency with previous censuses. (They're planning on dropping the term "negro" from the 2020 census for example, after testing that it removing it did not reduce responses from the African-American population.)

It's sort of obvious that the Census grew from a "race" question by adding on a second "ethnicity" question to identify Hispanics, but this is muddled. The proposed combined single question makes a lot more sense to me, and I suspect to most people. And while the Middle Eastern / North African group is not a clearly united single group, neither are any of the other top level groups - "Asians" come from a pretty big place, and there are a lot of different ethnic groups going into "White" and many nations fit under "Native", etc.

If we want to know how people of differing national origin and appearance are doing under the system, we need to ask the question. And it's good to ask the question in the clearest way possible that the most people can feel confident about answering correctly.
posted by Homeboy Trouble at 10:39 AM on November 25, 2014 [1 favorite]

The appropriate question to ask would be something like "What is your ethnic/racial identification?" with a fill-in-the-blank next to it. For me, a middle-eastern person, there are different identities within my own nuclear family. Even if "Middle-Eastern/North African" were an option I still think my parents would answer "White" while I wouldn't. Race might not be real but identities are. Does anyone else here identify differently from other members of their own family?
posted by BuddhaInABucket at 10:53 AM on November 25, 2014 [2 favorites]

The appropriate question to ask would be something like "What is your ethnic/racial identification?" with a fill-in-the-blank next to it.

This is free-form, go nuts.
posted by overeducated_alligator at 10:58 AM on November 25, 2014 [4 favorites]

As an Indian-American, I usually just tick "Asian / Pacific Islander", but honestly that feels weird to me. Indians are Asians, but the subcontinent has a very different culture as compared with the more traditionally-considered Asian nations (which are themselves fairly disparate). It seems weird to be counted among that group, along with Hmong, Polynesians, and native Hawaiians. I don't think our racial experiences are really all that similar. One of my coworkers is native Hawaiian, and while we both share a general sense of being Asian, we're sufficiently different as to not really recognize each other's cultures.

I hate the "race is fake" argument anyway. Of course race is an artificial social construct. So is democracy and capitalism and the corporate ladder. We make up stuff all the time, and that stuff consequently has social force and imperative. Money isn't "real" either, but try telling your landlord that.
posted by Errant at 11:12 AM on November 25, 2014 [10 favorites]

So, wait let me get this straight.

In the history of U.S. census-taking, Middle Easterners have ALWAYS been categorized as "White"?

Are there any other "races" that are lumped in? (how many)

I'm interested to know whether the percentage of "other" is classified as white SOLELY on the basis of lack of creativity, or historic inertia, on the part of the Census Bureau?

Or: Is it organized to fluff the "White" count deliberately? (unlikely/paranoid)

Or: Was economic status the conjoining factor: Jack Donaghy to Kenneth Parcel ~ (“Socioeconomically speaking, you are more like an inner-city Latina.”)

I see some answers are coming in already..v cool
posted by bird internet at 11:53 AM on November 25, 2014 [1 favorite]

US category of "Asian" is kinda crazy. Even more so than the govt use, the way it is commonly used by Asian American activists, academics etc., it includes all peoples from Turkey in the west to Japan in the east, from Siberia in the north to Indonesia in the south plus more and then they add on the Micronesians/Pacific Islanders too for good measure

Generally true, but the actual census disaggregates: In 2010, the categories were "Asian Indian, Chinese, Filipino, Korean, Japanese, Vietnamese, Other Asian." Note that no other racial category is disaggregated this way.

Here's a pdf.
posted by migrantology at 11:58 AM on November 25, 2014

Yeah, US v. Thind happened because this country doesn't know how to deal with brown-skinned Caucasians. They didn't just classify Indians as "non-whites" though, they classified Indians as Asians, and so Indians became subject to the Asian Exclusion Act and were barred from the country for 40 years, which is an excellent way of dealing with their confusion. Then, once they weren't allowed to prevent Indian immigration in toto anymore, they decided that us being white was fine, because we weren't black and therefore historically aggrieved or threatening, and also because no one was ever really going to treat us as white in practice. So we got to be an invisible model minority for about 30 years. Then 9/11 happened and white America can't tell whether or not we're actually a fifth column for Muslim extremism, but it's better to be safe than sorry, so if you see something, say something. Like, if you see an Indian, tell them jokingly that they're a terrorist. If they don't laugh, call Homeland Security, cause by gum you nabbed one, good job patriot.
posted by Errant at 12:03 PM on November 25, 2014 [2 favorites]

It's important to remember that some ethnic groups actively campaigned to be classified as white as a way to get around immigration restrictions. See United States v. Bhagat Singh Thind, and this article linked in the 538 post.
posted by yasaman at 12:05 PM on November 25, 2014 [1 favorite]

bird internet, it comes from the fact that, at some point (19th century? Early 20th?), "scientists" thought race was an actually verifiable category, and that there were three races. Which, to avoid slurs, I'll call White, Black, and Asian. Everyone theoretically belonged to one of those three. Under this scheme, people from the Middle East, Central Asia, and the Indian Subcontinent were considered White. Because "science". Sure.

Of course, nowadays we know that there is no biologically based black and white scientifically classifiable notion of race. And for the most part, we use the concept of race to talk about all sorts of things, sometimes in ways that are useful and (somewhat?) logical, but often in ways that are not, at all.

And, yes, of course, as Errant mentions, all this Super Scientific Racial Science went right out the window as soon as nativist white Americans realized that pesky brown people could use "science" to argue for open immigration to the US.
posted by Sara C. at 12:12 PM on November 25, 2014 [2 favorites]

Soooo... Are Jews with dual US/Israeli citizenship MENA? How about Sephardic Jews - Latin American? MENA? Ashkenazi Jews from India - MENA? How about spanish+yiddish speaking Ashkenazi Jews from South America - Latin American? MENA? White? How about my cousin, a "non-aboriginal South Sea Islander" of Jewish descent (API)?

Oyyyyyyyy.
posted by Dreidl at 12:19 PM on November 25, 2014 [1 favorite]

Sociologist Eduardo Bonilla-Silva has talked about the trajectory of racism in the US heading toward a tri-racial system with categories of "white, honorary white, and collective black."
posted by audi alteram partem at 12:23 PM on November 25, 2014 [2 favorites]

In the history of U.S. census-taking, Middle Easterners have ALWAYS been categorized as "White"?

It would be more accurate to say that in the recent history of US census-taking, there hasn't been an option of "Middle eastern or north African," so people of those ethnicities have either had to choose from several options that don't fit their self-perception very well or check off SOME OTHER RACE to identify as Arab or Lebanese or Persian or Sephardic Jew or whatever.

Soooo... Are Jews with dual US/Israeli citizenship MENA? How about Sephardic Jews - Latin American? MENA? Ashkenazi Jews from India - MENA? How about spanish+yiddish speaking Ashkenazi Jews from South America - Latin American? MENA? White?

In all those cases, you'd answer as you self-identify. It's not an examination with right and wrong answers to give. You pick the answer that best summarizes how you feel about yourself, with the caveat that none of the check-off choices might be very good answers for you.
posted by ROU_Xenophobe at 12:31 PM on November 25, 2014 [3 favorites]

I worked as a census taker one year here in Canada, and it was made very clear to us that we were to never question someone's answers, just to confirm that they put an answer down. Your racial classification is up to you. We wouldn't ever question it. Heck, it was drilled into us if someone put their birth year down as 1776 we were to accept it.
posted by thecjm at 1:56 PM on November 25, 2014 [1 favorite]

The question of race and the census is really interesting - how to classify people into categories, when the categories are socially constructed, have fuzzy edges, and no agreed-up definitions - yet the the categories do have real-world impact.

I like BuddhaInABucket's suggestion of asking something like What is your ethnic/racial identification?, but I think something like Do you identify as any of the following? would work really well practically - there could be a list of options (check as many as you'd like!) with various descriptions. For example, someone could choose to check both "White/Caucasian" and "Arab/MENA" or just one of those. Someone could mark only the "Hispanic/Latino" box, or also check that they identified as black (or white, another category). It would be difficult to come up with a list, and the list wouldn't be perfect, but it seems doable and better than most of the alternatives.
posted by insectosaurus at 2:35 PM on November 25, 2014

'Do you identify as any of the following?' would work really well practically - there could be a list of options (check as many as you'd like!) with various descriptions. For example, someone could choose to check both "White/Caucasian" and "Arab/MENA" or just one of those.

This is pretty much how it was in the 2000 and 2010 censuses. Above the list of races, the instructions say "Mark one or more boxes". So multi-racial identity was totally allowed, and in 2010 about 9 million people chose to identify as more than one race. Also, there's an "other" box, with a text entry area, so you can fill in whichever race you choose.
posted by mr_roboto at 2:54 PM on November 25, 2014 [1 favorite]

This is pretty much how it was in the 2000 and 2010 censuses

Yes, I realize that. I'm suggesting that it would work really well to eliminate the "Hispanic origin" question and roll it into the race question, plus add several more categories/descriptors (including Arab/MENA).
posted by insectosaurus at 3:43 PM on November 25, 2014 [1 favorite]

This determination should employ SCIENCE! Like a DNA test.

This would be 1. horrendously expensive and 2. very unlikely to actually be a better measurement of race than self-report, because while race is certainly related to ancestry, it really has more to do with how you are perceived in society and those things are often not strongly correlated (e.g., Native American admixture and skin color had only a subtle correlation in this sample of Mexican-Americans from San Antonio).
posted by en forme de poire at 4:57 PM on November 25, 2014 [1 favorite]

Due to some recent experience with the medical system, I had the privilege of learning that there are only 2 ethnicities in this world: Hispanic and Non-Hispanic. Go figure.
posted by I-Write-Essays at 6:50 AM on November 26, 2014 [1 favorite]

Culturally and genetically Arabs are closest to Europeans. A recent study found that Europeans likely originated from three main populations - an original hunter-gatherer population, a later migration of farmers from the Near East, and a third population from "North Eurasia," probably related to modern Hungarians. Skin color is not very reliable here. I think it's better to look at bone structure, amount of body hair, eye color and things of that nature. Arabs and Persians, by those criteria, would best fit into the "Caucasoid" grouping in terms of the classical division of races into Asiatic, African, Caucasian and Native American, if we're going the route of population genetics and biological distinctions. Their DNA is closer to Europeans than people in the Far East. Of course, you can argue race is socially constructed as well, but even in that case, those from the Middle East and North Africa are arguably "white." There's also the fact that quite a lot of cultural and social mixing has gone on between the Near East and the Mediterranean over the millennia, though the degree of that will be played up or down depending on the degree of Eurocentrism in the scholarship you happen to be reading.
posted by ChuckRamone at 7:03 AM on November 26, 2014

I did read something recently though that said Arabs and other Middle Easterners in America started checking "Asian" in censuses and such for political reasons, starting sometime in the mid-20th century. It supposedly gave them a leg up socially. They could just as easily go the other route if it proves advantageous later.
posted by ChuckRamone at 7:12 AM on November 26, 2014

I think it's better to look at bone structure, amount of body hair, eye color and things of that nature.

This is what 19th century pseudoscientists were looking at when they lumped Middle Easterners and Central Asians in with Caucasians, but again none of that is really scientific. At best, it's a sort of "people from here are more likely to have this hair texture, while people from there are more likely to have that hair texture". At worst, it's the lead-up to Naziism.
posted by Sara C. at 8:07 AM on November 26, 2014 [1 favorite]

Which is why I mentioned the classical division of races. It's as outdated as phrenology. My main point is that skin color is not reliable for this kind of thing. Skin color is mostly a function of the latitude at which a population has existed, and is an adaptation to the amount of UV in the atmosphere. East Asians, for example, are often pale but would not be considered "white" socially or genetically.

But there are population genetic markers as well linking Middle Easterners and Europeans (see the NYTimes link in my post). Sorry, I linked it incorrectly above. Link
posted by ChuckRamone at 8:12 AM on November 26, 2014

But not in a way that in any way correlates to modern ideas about race, which I assure you do not have any basis in what Eurasian hunter gatherers were doing thousands of years ago.
posted by Sara C. at 8:15 AM on November 26, 2014

8,500 years ago is not that long ago. Anatomically modern humans have been around roughly 200,000 years.
posted by ChuckRamone at 8:17 AM on November 26, 2014 [1 favorite]

Anyway, as I mentioned before, if you want to look at this in terms of what's socially constructed about race, and totally disregard genetic markers, a later generation Middle Eastern person could easily assimilate into white American or European culture because of their physical attributes. Someone whose ancestors are African, Asian, Native American, etc. could not.
posted by ChuckRamone at 8:24 AM on November 26, 2014