The Consequences of a Middle Eastern or North African (MENA) Survey Question

Arab American organizations began lobbying the U.S. Census Bureau in the 1980s to create a new racial category for Arab or Middle Eastern Americans to increase their political influence and, perhaps, eventually benefit from affirmative action.³ After decades of hearing those lobbying efforts, the Census Bureau studied whether it should add the new racial category to the 2020 census, ultimately rejecting the addition.⁴ In 2022, the Office of the Chief Statistician started the process of revising Statistical Policy Directive No. 15: Standards for Maintaining, Collecting, and Presenting Federal Data on Race and Ethnicity (SPD 15), which is OMB’s statistical standard for collecting and reporting data across federal agencies.⁵ In January 2023, OMB published a notice and request for comments to inform the Federal Interagency Technical Working Group on Race and Ethnicity Standards as they seek to reform statistical standards by adding a MENA category, among other changes.⁶ OMB plans to complete revisions to SPD 15 no later than summer 2024.

Organizations that claim to represent MENA Americans are hoping the new racial category will allow for more accurate information to receive advantages from the government. Those advantages include more welfare benefits, benefit from affirmative action, addition of MENA as a protected class in employment and civil rights law, expanded sensitivity or awareness training to cover MENA Americans in workplaces, extended Voting Rights Act protections, more detailed data on hate crime victims, preference in government contracts, and other vague goals like respect, stronger identity, recognition, and alignment of government surveys with widespread sentiment.⁷

In practice, affirmative action is for Americans of ethnic or racial groups that are less successful than Whites on average. (This brief uppercases races and ethnicities to be consistent with U.S. government stylistic guidelines.) There are rare exceptions, such as Asian Americans getting preferences in government contracts.⁸ Affirmative action discriminates against successful groups by allocating fixed benefits such as school admissions. Thus, whether MENA Americans would gain through government largesse depends on whether MENA Americans are better off than Whites. If they are more educated and affluent than Whites, they could be discriminated against just as Asian Americans are.

Remarkably, there is little published research on MENA Americans’ economic and educational outcomes.⁹ Existing research focuses on immigrants from the Middle East or North Africa, a subsection of people who would fall into the MENA category that excludes native-born Americans.¹⁰ Unless there’s a trove of unpublished research that supporters of the new MENA category have seen, anyone promising benefits such as affirmative action is flying blind. A Cato brief recently filled that knowledge gap by providing the first estimates of the demographic, economic, and educational standing of MENA respondents relative to other race and ethnicity categories based on the 2021 American Community Survey.¹¹ Given the relative economic and educational success of Americans who would fall into the new MENA category, they would be net losers from affirmative action.

There were approximately 3.8 million MENA residents in the United States in 2021, which is about 1 percent of the U.S. population. MENA Americans had higher average incomes than Whites and lower average incomes than Asians (see Table 1). The mean MENA household income in 2021 was almost $115,000, 13 percent higher than the White household income at about $102,000. Mean Asian household income was about $134,000, substantially higher than MENA and White household incomes. Table 2 shows that more than 53 percent of MENA respondents aged 25 or older have a bachelor’s, graduate, or professional degree. That’s the second-highest level of education, just below 57 percent for Asians and well above Whites at 39 percent. If MENA had been a racial category in the 2021 ACS, it would have had the second-highest income and level of education, ahead of Whites and behind Asians.

Americans in a new MENA classification would have the second-highest incomes and education levels of all racial groups. The Supreme Court’s decision in SFFA v. Harvard exposed how some universities raised the standards for Asian applicants to free up slots for students from other racial groups with lower grades and test scores as part of their affirmative action programs.¹² Universities discriminated against Asians in admissions decisions because they were the most affluent and educated racial group in the United States. The Court ruled that the admissions programs at Harvard and the University of North Carolina violated the equal protection clause of the Fourteenth Amendment.¹³ Universities committed to using affirmative action will likely find ways to give preference to applicants from certain racial groups, which will require discriminating against others, but it will probably be a small fraction of the discrimination that existed before the Court’s ruling.

To the extent that discrimination against Asians persists in university admissions in the wake of SFFA v. Harvard, MENA applicants would also be discriminated against if they are lumped into a new racial category that exposes their high average academic success. Affirmative action will not harm MENA applicants in university admissions as much as it would harm Asian applicants, but it would penalize them more than it does White applicants. For example, affirmative action has been illegal in admissions in California public universities since 1996.¹⁴ That ban greatly diminished affirmative action admissions, but the University of California (UC) system also considers other achievements in light of opportunities available to the applicant, their demonstrated capacity to contribute to the university’s intellectual life, and the location of the applicant’s secondary school and residence.¹⁵ Those criteria allow UC admissions offices to practice some affirmative action under the guise of other goals.

If a future Supreme Court overturned SFFA v. Harvard and again allowed universities receiving government funds to practice affirmative action, MENA Americans would be penalized. That may seem to be a remote possibility, but many supporters of a federally recognized right to abortion thought they were on solid ground before Dobbs v. Jackson Women’s Health Organization overturned Roe v. Wade.¹⁶ No matter what, Americans who would fit into the proposed MENA classification would be harmed in university admissions, especially if SFFA v. Harvard is ever overturned. The prospect that a new racial category for MENA Americans would facilitate discrimination against their constituents ought to give pause to groups currently supporting the change in racial classification.

Of course, a new racial category for MENA applicants isn’t strictly necessary for discriminatory admissions policy to exist in higher education. Harvard used to discriminate against Jewish applicants even though federal government surveys have never asked about religion.¹⁷ However, a MENA racial category would inform university admissions departments, government bureaucrats, and others that MENA Americans have high incomes and are well educated. If the government creates the MENA racial category, many organizations would learn that MENA Americans are the second-most academically successful racial group in the United States, resulting in discrimination against them under race-based affirmative action.

Those who favor a MENA racial category promise other small potential benefits to members of that group in addition to affirmative action. In many cases, the purported benefits would fail to materialize, could backfire, and may even impose high costs on Americans of MENA ancestry.

The government sets aside contracts and grants for minority-owned businesses.¹⁸ Samer Khalaf, the former president of the American-Arab Anti-Discrimination Committee, said that the creation of a new MENA racial category would allow “[s]mall-business owners in the community … to take advantage of grants that we’re not entitled to, because we’re factored into the White category.”¹⁹ Some businesses owned by MENA Americans would likely gain preferential treatment in government contracts or small-business grants if the new category were created, but the legal standing of these types of programs is shaky, and the benefits are small in light of the economic and educational success of MENA Americans in the United States.²⁰

Some advocates in favor of a new MENA racial category claim it would facilitate an accurate population count.²¹ Currently there is no robust census counting of residents of MENA ancestry, but the ancestry and country of origin reporting in the ACS and census allow an excellent estimate of the number of MENA residents and vast descriptive data. Researchers curious about MENA residents could download and format current ACS or census data to get an accurate picture of MENA Americans. If the point of this new racial classification is to learn something about a subgroup, that goal is already met by currently available government data.

The ACS’s ancestry and country of origin data are imperfect, but census data also contain errors. Ancestry and country of origin data are likely subject to more errors than race data, but they are pretty good and better than any estimates of religion or social attitudes. It’s telling that nobody has bothered to create their own MENA estimates using existing data sources. If the demand for that data were great, surely somebody else would have compiled it before the Cato Institute.²² There is more research asking people of MENA ancestry whether they want a new MENA racial category than there is research using the data from existing government surveys to answer questions about that group.²³ Data concerns do not drive demand for a separate category.

Some argue that Americans with MENA ancestry are distinct from Americans of European extraction because of cultural, experiential, and other differences that are vast enough to justify another racial category.²⁴ However, there is enormous diversity among respondents in the MENA category itself. They are religiously diverse, with large numbers of Christians, Jews, Muslims, and believers of other religions. They are ethnically diverse, with Turks, Arabs, Kurds, Persians, and others. They are even linguistically diverse, with languages in the Altaic language group (Turkish); Afroasiatic language group (Arabic and Hebrew); and Indo-European language group (Farsi and Pashto). MENA would have the largest variance in many socioeconomic metrics. For instance, Americans of Iranian ancestry have the third-highest median household income, behind Indians and Pakistanis. Afghans, who would also count as MENA, have the second-lowest, just ahead of Appalachians. If reducing variance were the concern, government surveys would focus on the ancestry question rather than create a new race category. After all, Afghans likely have more in common with other Afghans than they do with Moroccans, even though they’d both be lumped into the same MENA category. A MENA racial category obscures those differences and glosses over the wide disparities in income and education among people from the MENA region, to say nothing of the vast religious, linguistic, cultural, social, and other differences.²⁵

Many MENA respondents in survey experiments don’t self-identify as MENA.²⁶ In one such experiment, 80 percent of MENA respondents self-identified as White, 5 percent as Asian, and 15 percent as “Some Other Race” in the control group. When the MENA race category was added, 60 percent identified as MENA only, 27 percent as MENA and White, 12 percent as White, and about 1 percent as MENA in combination with another race or races. Muslim respondents with MENA ancestry were most likely to identify as MENA.

There are a few takeaways from this experiment. First, many people of MENA descent treat it as an ethnicity rather than a separate race. If MENA is added to government surveys, it should be an ethnic category, similar to the Hispanic or Latino ethnicity. Second, when the option is available, most MENA Americans select MENA only, but large numbers do not. These data could be used to get a good subsample of Muslim Americans, who are likelier to identify as MENA only. Third, even though respondents from MENA countries are more likely to select MENA when given the option, they’d be more likely still to select their ancestral country of origin. If one of the goals of reforming government surveys is to produce more accurate data, then scrapping race questions entirely and relying on self-reported ancestry would yield more granular and useful results. Government surveys aren’t supposed to, nor can they, resolve or reconcile various personal identities that people possess.

Relatedly, some people whose ancestors are from the Middle East or North Africa argue that they are not White, so they need a new race category to accurately represent their identity. The White category in the census is very broad and currently includes people of Icelandic and Afghan descent, but a MENA category would be only slightly narrower and still open to the same criticism by including widely different ancestries. The government created the Asian racial category, which includes people of Japanese and Pakistani ancestry, based on geography, and the Hispanic or Latino ethnicity, which includes people from Spanish-speaking countries as different as Bolivia, Argentina, and Spain, based on linguistic commonalities. Those government-created categories created an identity for some people, but the government classification came first. A person whose ancestors came from the Middle East or North Africa doesn’t need a new government-designed racial category to validate their ethnic origins, cultural practices, or ancestry any more than they need a special religious category in a government survey to validate their Islamic, Christian, Jewish, Yazidi, or other religious beliefs. The government cannot validate every person’s identity, and it should not attempt to supply such validation. Other survey questions about ancestry capture that well, and freedom to determine one’s own identity will lead to better results that are more consistent with people’s actual beliefs, cultural practices, and ancestry than a new government-created racial group.

Some who support the new racial categorization do so to extend civil rights and employment protections to MENA Americans.²⁷ However, Title VII of the Civil Rights Act of 1964 already outlaws employment discrimination based on race, color, religion, gender, pregnancy, or national origin. According to an article published by the Society for Human Resource Management, “the U.S. Equal Employment Opportunity Commission has filed or settled various lawsuits alleging discrimination based on national origin and religion against Muslim, Sikh, Arab and Middle Eastern workers” since 2001.²⁸ A new racial category for government surveys doesn’t affect laws against discrimination, but it may increase the focus on MENA Americans in training sessions conducted by human resources departments.²⁹ In that way, a new MENA race category could be a boon for human resources employees, lawyers, and others who would directly benefit from increasing the saliency of a small minority and services targeted at them; MENA workers, who are already protected under the law, would receive no extra benefits.

Another argument for a new MENA racial classification is that it would reduce “othering,” a sociological term for treating someone as though they are not part of a group and are different.³⁰ It’s hard to see how creating a new racial category in government surveys would do anything other than increase othering of that group.

Slides from a 2022 Census Bureau meeting discussing issues and comments on the creation of the MENA racial category posited that it would aid in the reporting of hate crimes.³¹ However, hate crimes against Muslims, Arabs, and others who would be included in a new MENA category are already identified in national hate crime data.³² Furthermore, future administrations could use more detailed census microdata to better surveil MENA Americans, as happened to Muslim Americans after 9/11, when the Census Bureau shared tabulations of ZIP-code-level breakdowns of Arab Americans by country of origin with Customs and Border Protection.³³ Advocates for MENA Americans rightly point out that they have been the victims of targeted surveillance programs and watchlisting, but then claim they have no way to study those programs without the more definitive government data a MENA question would provide.³⁴ Of course, more definitive data on government surveys would lower the cost of targeted surveillance programs or watchlisting of MENA Americans. If the results from survey experiments are to be believed, such a question would be especially useful in identifying Muslims of MENA ancestry.³⁵ If government surveillance harms MENA Americans, then it seems perverse to argue that the government should have more information on them because it might also help researchers.

During World War II, the Census Bureau released block-by-block data and, possibly, microdata on the locations of individual Japanese Americans to the military and the U.S. Secret Service.³⁶ The government would have undoubtedly identified Japanese Americans and others for internment during World War II without the aid of census data, but the data provided lowered the cost of internment. Plenty of data are available to aid the federal government in identifying MENA Americans for future law enforcement actions, but not the detailed microdata that would be produced by adding a new MENA racial category to future censuses. Current laws prohibit the census from sharing data with other government agencies. Still, Congress could change those laws, and agencies could ignore the laws, so existing statutes are no guarantee against future abuses.³⁷ At the very minimum, not collecting MENA data in the first place would raise the costs of similar discriminatory actions.

The creation of a new MENA racial category in government surveys would backfire in several possible ways: by resulting in discrimination against Americans of MENA ancestry in university admissions, by producing many fewer benefits than reported, by benefiting non-MENA lawyers and bureaucrats, by making current data-collection practices redundant, and by aiding surveillance of MENA Americans.

As a social scientist, I find a new government racial category tempting. It would be better to have access to that category with fewer mouse clicks, all else being equal. But all else is not equal. As a libertarian concerned about the scope and power of the government, I believe a new MENA racial category raises many more concerns that outweigh my interest as a social scientist. As an American libertarian, I oppose the creation of a MENA category for other principled, ethical, and universalistic reasons. As an American with Middle Eastern ancestry, I also oppose it out of self-interest for myself, my family, and my children.