Posted by on April 05, 2013 in Blog

As the demographics researcher at the Arab American Institute, I’m often asked about the total number of Arab Americans in the country. Though the question seems simple, the answer is anything but.

The problems of counting Arab Americans are many-fold, and range from social issues (including the contextual salience of national/ethnic identification, the subjectivity of categories of race, and the relationship between marginalized communities and the state) to methodological concerns (including sampling size methodology, historical undercounts, sampling models that fail to account for small urban population concentrations).

Despite these limitations, we do have two numbers we generally give out.

Census-estimated Arab American population: 1,967,219

AAI-adjusted population total: 3,664,789

Both of these are problematic in their own way, but the latter number is almost certainly a more accurate reflection of the total Arab American population than the former. We base our estimate on a fairly robust sampling model that accounts for documented cases of Census undercounts in the past, and tries to apply these metrics to larger population areas. Without getting too much into the statistical weeds, here are some of the basic issues that our demographics analysis is trying to overcome:

  1. Sampling Methodology
  2. Since the Census Bureau doesn’t consider “Arab” to be a race or ethnicity (“Arab” in classified as a subset of the “white” race, as per Office of Management and Budget classifications), there is no way to identify oneself as Arab American on the decennial Census form. Instead, information on Arab Americans and other “ancestry groups” is derived from the American Community Survey, a form sent by the Census Bureau to 3% of households every year, and extrapolated to determine national estimates on more nuanced demographic information. While the Census form only asks for very basic information, the ACS includes questions about income, employment, distance to work, health care, education levels, and many other topics. The information from the ACS is incredibly helpful in social service programs, academic research, government spending, and community organizing efforts, but since only 3% of the population is surveyed, the ACS estimates carry a significant margin of error for small groups such as Arab Americans.

    The problem is compounded by the fact that Arab Americans are heavily concentrated in a few urban areas, so a national sampling is very likely to undercount (or entirely omit) Arab population groups, which has led to a significant undercount, sometimes by as much as a factor of 3. Of course, adding Arab Americans (and other ancestry or origin groups) to the decennial census would go a long way in remedying this problem, and we have seen some movement in that direction for the 2020 Census. If this occurs, it would significantly improve the Bureau’s population estimate of the Arab American community, but still wouldn’t solve all the associated problems.


  3. Subjectivity of Arab American Identity
  4. What does it mean to be Arab American? Though AAI has a working definition of the term, it’s a hotly-debated issue.

    The first level of difficulty emerges from the very concept of “Arab” itself. Who is considered to be Arab? Though some cases are very clear (most Syrians, Egyptians, Moroccans, and others from Arab League states consider themselves as Arab), many problems arise at the margins. Groups that don’t speak Arabic (many Sudanese, Somalis, and Kurds), or claim a different cultural background (Copts, Chaldeans, and Assyrians for example), or a different historical genealogy (Berbers, Lebanese “Phoenicians,” etc) have all struggled with their relationship to the term “Arab,” and the issue of their inclusion as “Arab American” is no different. The Census Bureau doesn’t consider individuals who claim Sudanese, Somali, or Assyrian/Chaldean/Syriac ancestry to be “Arab,” but it does include Kurds.  

    In addition, the level of identification with Arab identity in America is deeply subjective. Since Arabs have been immigrating to the US in large numbers since the mid-19th century, many have assimilated quite fully into American society. A third- or fourth-generation American citizen with some ancestry in Lebanon may consider herself Arab American, or she may not, and there are no objective criteria to guide that choice. Identification can also depend on context. A man could identify as Tunisian among his family and friends, but prefer to identify simply as “White” on official paperwork and government documents. These Arab Americans are all but invisible to official sampling attempts.

    AAI sampling tries to take into include individuals who may consider themselves Arab in ways the Census Bureau and other demographic research attempts might not readily identify.


  5. Historical Community Distrust

The issues mentioned above are common across many ethnic groups, but Arab Americans have an additional level of complication in demographic polling, that we largely experience alone. Since 9/11 Arab Americans have lived under a pall of suspicion and distrust by many aspects of the government, including surveillance and profiling programs in the FBI, the NYPD, and countless other counterterrorism and countering-violent-extremism programs.

There have even been documented cases of the FBI using Census data on Arab Americans to plan its racial-mapping and surveillance program, despite assurances by the Census Bureau that demographic information would never be used against the respondents. This has created an unprecedented level of distrust in the Arab American community, and a general unwillingness to openly identify as Arab in public spaces. This has exponentially increased the difficulty of getting reliable information through Census Bureau outreach.


For more information on our demographics research, and to see detailed profiles of the Arab American community in your state, check out our Demographics page

comments powered by Disqus