What factors account for black–white differences in anti-Muslim sentiment in the contemporary USA?

Abstract

I use data from the 2004 General Social Survey (N=719) and multivariate analyses to: explore the effects of race on attitudes toward Muslims; evaluate the extent to which the racial differences were mediated by psychological and religious factors; and assess whether the race effects differed significantly by gender. The findings show that blacks report significantly more favourable feelings toward Muslims than whites. Those respondents who are female, more educated and Catholic also hold significantly higher scores on the 100-point scale assessing feelings toward Muslims. After controlling for religious and psychological factors, I find that the racial difference in feelings toward Muslims is increased, indicating that the race effect is suppressed by these factors. Moderation analysis reveals that white men hold the highest level of negative feelings toward Muslims, compared to women and black men. The findings suggest challenging the misconstrued perceptions of Muslims through education and endorsement of positive images.

Keywords:

• Islam
• Muslims
• religion
• race
• gender
• attitudes

Acknowledgements

I am grateful to Deborah Carr, Julie Phillips and Randall Smith for their encouragement and generous help throughout the multiple versions of this article and to anonymous referees for their useful suggestions. Research for this article was supported in part by the Open Society Foundations Doctoral Fellows Program. An earlier version was presented at the 81st Annual Meeting of the Eastern Sociological Society, Philadelphia, PA, 24–27 February 2011.

Notes

1. Responses on the dependent variable were missing for 1,939 cases who were not in the random subsample scheduled to receive the Muslim attitudes question, and eighty persons who did not answer the question when asked. Prior analyses show that age was positively associated and education inversely associated with the odds of answering ‘don't know’ or ‘refuse’ to the Muslim question.

2. A paired sample test showed that the mean differences of all four religious groups are significant, meaning that the ratings of Muslims are significantly lower than ratings for the other religious groups.

3. The GSS distinguishes urban and non-urban residents between Standard Metropolitan Statistical Area (SMSA) and non-SMSA residence. See the GSS codebook (http://publicdata.norc.org:41000/gss/documents//BOOK/GSS_Codebook.pdf) for definitions and more information.

4. I conducted a series of sensitivity analyses through transformation of the respondent's age into quadratic, cubic, multi-category and cohort variables, to determine the existence of non-linear association between age and attitudes. None produced significant results, confirming that anti-Muslim attitudes appeared to be unrelated to age.

5. Other potential predictor variables considered include trait optimism, political identity and a more fine-grained measure of geographic region. None was a significant predictor, nor did any improve the overall model fit.

Additional information

Notes on contributors

Hakim Zainiddinov

HAKIM ZAINIDDINOV is a PhD student in the Department of Sociology at Rutgers University.