Why Study American Muslims?
For many years, American Muslims have experienced heightened scrutiny due to widespread societal apprehensions about Islam. Research has documented the discrimination against Muslims by the public, elites, and masses (Lajevardi Reference Lajevardi2020; Oskooii, Dana, and Barreto Reference Oskooii, Dana and Barreto2021) as well as the impact of this hostile environment on Muslims themselves (Dana et al. Reference Dana, Lajevardi, Oskooii and Walker2019; Sediqe Reference Sediqe2020). Overall, findings indicate that whereas American Muslims face challenges akin to other racial and immigrant groups, they also grapple with gaining acceptance and being recognized for their positive contributions to America’s diverse society.
Within this context, Muslims face unique challenges that raise particular questions about their democratic inclusion. Moreover, the diversity and heterogeneity of the US Muslim population across numerous markers—including race, denomination, nativity, age, socioeconomic status, sexual and gender identity, and religiosity—all matter for shaping whether and how intensely they have encountered such challenges, as well as their responses. A comprehensive understanding of their experiences, and how these compare to those of other American groups, requires extensive samples of both Muslim and non-Muslim individuals, along with detailed questioning on a broad range of sociopolitical issues.
Fortunately, the 2020 Collaborative Multiracial Postelection Survey (CMPS)—a national survey of voters and non-voters that oversamples racial and ethnic minorities on political and social issues—provides a unique opportunity to delve deeper into the experiences of US Muslims compared to other minoritized groups. The 2020 CMPS consisted of large samples (∼4,000) of white, Latino, Black, and Asian respondents. Importantly, given the salience of other socially stigmatized communities, the 2020 CMPS also included oversamples of LGBTQ+ Americans, Native Americans, Native Hawaiians, Afro Latinos, African Immigrants, and Muslim Americans. With the CMPS, for the first time, researchers can thoroughly investigate Muslim experiences in the areas of politics, identity, housing, employment, law enforcement, immigration, media, education, and environmental concerns.
Complications of Collecting Survey Data on Muslims
Studying US Muslims presents unique challenges, which distinguishes it from research on other marginalized groups. Typically, researchers use demographic details such as group size, composition, and location when they study marginalized populations (e.g., racial and ethnic minorities). This rich dataset is notably lacking for Muslims, making study design difficult. The following three main issues hinder comprehensive survey data collection about US Muslims:
-
1. The US Census does not query religious identification. Although certain groups, such as Hispanics and Asian Americans, may be inferred from census-recorded surnames, there is no analogous method for Muslims due to the absence of a reference list (Barreto and Dana Reference Barreto, Dana, Calfano and Lajevardi2019). This data gap limits accurate estimations of religious populations nationwide (Barreto and Dana Reference Barreto, Dana, Calfano and Lajevardi2019).
-
2. Muslims are incredibly diverse in their racial backgrounds. Although a substantial proportion have roots in the Middle Eastern/North African (MENA) region, using MENA populations for rough estimations is problematic because the US Census currently groups MENA Americans under the “White” category, which obfuscates specific national origins (d’Urso Reference d’Urso2022).
-
3. Alarmingly, some US Muslims perceive survey-recruitment efforts as a potential surveillance method, which heightens their anxiety, lowers response rates, and possibly biases responses due to social desirability (Calfano, Lajevardi, and Michelson Reference Calfano, Lajevardi and Michelson2019).
Findings From The 2020 CMPS
This section describes our sampling challenges and provides a preview of our sample.
Sampling Challenges
Achieving a fully representative Muslim sample is challenging due to inherent difficulties. However, we used benchmarks from the Pew Research Center’s 2017 survey on US Muslims to establish race and nativity quotas, which assisted in the weighting of our Muslim sample.Footnote 1
Our primary limitation in developing the Muslim oversample was that we did not pose a screener question about religion at the beginning of the survey. Respondents first answered questions related to race and were not asked about their religious identification until Question 58. Moreover, when they were queried about their religious affiliation, respondents were permitted to “check all that apply.”
Ultimately, the 2020 CMPS included 579 self-identified Muslims and 292 MENA respondents.Footnote 2 The CMPS directors combined these two groups for two reasons: (1) low response rates in the Muslim sample; and (2) a burgeoning interest in the MENA American population, especially in light of the likelihood of including a new racial category for MENA Americans in the 2030 US Census. Nevertheless, as other research argues (d’Urso Reference d’Urso2022) and as shown in table 1, it is crucial for users of these data to recognize that not all Muslims are MENA and vice versa. Although a single oversample indicator for a combined MENA/Muslim sample appears in the 2020 CMPS dataset and in the public documents, we recommend that users treat MENA and Muslim as distinct categories, disaggregating these two groups into their own analyses. Relatedly, the provided oversample weights by the 2020 CMPS are a rough estimate, complicated by the fact that there is no true underlying population to which we can weight. We advise researchers (of Muslims) to formulate their own weights relying on the Pew Research Center’s Muslim surveys, considering variables such as race, nativity, and age.
Table 1 MENA/Muslim Sample in the 2020 CMPS
Multiple approaches were taken to recruit Muslims for the sample and from these myriad approaches, we have learned several beneficial lessons. First, flagging religion initially led to a higher proportion of Muslim respondents in the survey. Second, the 2020 CMPS turned to multiple sources to recruit respondents for the survey, including survey panels, the random-recruit-to-web (RRW) method,Footnote 3 community samples cultivated by sample directors, and non-voter lists. Analyses indicated that the approximately 10 survey panels relied on to recruit respondents for the instrument all produced similar—and rather poor—results for recruiting Muslims into the 2020 CMPS. Third, of all the sources used, the non-voter list that the CMPS purchased for Muslims yielded significant results (i.e., 29% of all of those who started the questionnaire and who made it to Question 58 for the purchased non-voter list identified as Muslim), and most efforts tried to maximize completions from that source. Fourth, the RRW sample also performed well: of those who identified as Muslim, 39% identified as Muslim at Question 58. Fifth, in the limited community sample that we were able to cultivate (N=20), our contact rate for Muslim respondents was 100%.
Sample Preview
Overall, using all of the approaches taken to recruit Muslims for the sample, a total of 782 respondents took the survey and identified as either Muslim or MENA in the 2020 CMPS. Of these respondents, 579 identified as Muslim, 89 as MENA/Muslim, and 203 as MENA/non-Muslim (see table 1). Although we did not recruit 1,000 Muslims as originally planned, the robust and wide-ranging questions about which respondents were queried makes this sample an invaluable resource to learn more than ever before about the US Muslim population (Dana and Lajevardi Reference Lajevardi and Dana2024).
Discussions with the 2020 CMPS principal investigators throughout the recruitment process piqued our interest in those flagged Muslim respondents in the RRW and non-voter lists who ultimately did not subsequently identify as Muslim in Question 58. Fortunately, in addition to asking respondents about their own religious affiliation, the 2020 CMPS asked which primary religion, if any, that respondents’ families practiced or followed when they were being raised. As shown in table 2, 552 respondents indicated that they were raised in a Muslim household, with most still identifying as Muslim (N=462, 82.94%). Nevertheless, 95 respondents (i.e., 17.06% of those who indicated they were raised in a Muslim household) did not identify currently as Muslim. We refer to these individuals as “Muslim by Heritage” and, moving forward, we argue that one best-practice lesson is to query respondents early in the survey about the faith group in which they were raised and their current beliefs. Table 2 also reveals that 117 respondents were not raised in a Muslim household but currently identify as Muslim. Most likely, these are Muslim converts; descriptive analyses reveal that the majority identify as Black (∼51%) or Latino (∼22%). This reinforces findings that racial and ethnic minorities are the fastest-growing segments of the US Muslim population.
Table 2 Finding Respondents Who Are “Muslim by Heritage” in the 2020 CMPS
Table 3 lists basic descriptive statistics comparing the Muslim, MENA, and Muslim by Heritage samples to the primary oversample groups. The table displays means across the samples for indicators including US Born; LGBTQ+; Voted in the 2020 Presidential Election; 2020 Trump Support; and Perceived Discrimination with Respect to Race, Skin Color, Gender, Sexuality, Immigration Status, Religion, and Accent. Three noteworthy trends are illustrated in the table. First, self-identified Muslims reported the greatest levels of religious discrimination; however, Muslims by Heritage and non-Muslim MENA respondents also were more likely to report higher levels of religious discrimination than non-Muslim Black, Asian, Latino, and white respondents. Second, Muslim, MENA/non-Muslim, and Muslim by Heritage respondents also reported high levels of racial, skin color, and gender discrimination, similar to other non-Muslim minorities. Third, the Muslim, MENA/non-Muslim, and Muslim by Heritage categories had low turnout rates in the 2020 election (hovering around 30%)—lower than any of the other oversample populations. When they were queried, however, respondents reported low levels of Trump support.
Table 3 Descriptive Statistics
Note: These figures represent weighted averages.
Lessons Learned and Future Avenues for 2024
Given how Muslims have been central to discourse in American politics, and given how much discrimination they face along numerous markers, the data reconfirm the need for a Muslim survey in 2024 and beyond. This data-collection effort also yielded several lessons in best practices, as follows:
Given how Muslims have been central to discourse in American politics, and given how much discrimination they face along numerous markers, the data reconfirm the need for a Muslim survey in 2024 and beyond.
-
• Respondents’ religion should be queried as a screener sooner in the survey.
-
• Respondents should be asked about the faith group in which they were raised. Future surveys should recognize that there are many whose attachment to Islam may be cultural rather than religious in nature. Therefore, those individuals should be asked whether they consider themselves “cultural” or “secular” Muslims as researchers consider how to treat Muslims by Heritage.
-
• The most successful means of recruiting Muslim respondents for the 2020 CMPS was using RRW off the non-voter list and by developing a community sample through “snowball” sampling.
-
• Users should disaggregate the MENA/Muslim subgroups.
-
• The 2020 CMPS survey weights for the MENA/Muslim sample were rough estimates; therefore, users should create their own weights.
ACKNOWLEDGMENTS
We thank the editor and the anonymous reviewers for their thoughtful and constructive feedback. We also thank participants at the 2023 CMPS National Advisory Board Meeting and the Michigan State University Muslim Studies Program 2024 Annual Conference, “Researching Muslim America,” who together provided exceptional suggestions and insights. Finally, we thank Lorrie Frasure, Janelle Wong, Edward Vargas, and Matt Barreto for including Muslim and MENA Americans in the 2020 CMPS, and for helping us to think through sampling at every stage of the process. We are forever grateful.
DATA AVAILABILITY STATEMENT
Research documentation and data that support the findings of this study are openly available at the PS: Political Science & Politics Harvard Dataverse at https://doi.org/10.7910/DVN/SZBQJN.
CONFLICTS OF INTEREST
The authors declare that there are no ethical issues or conflicts of interest in this research.
