Competing Claims: Religious Affiliation and African Americans’ Intolerance of Homosexuals

ABSTRACT

Literature on religion and political intolerance indicates competing expectations about how Black Protestant church affiliation affects African Americans’ attitudes about civil liberties. On the one hand, Black Protestant theology emphasizes personal freedom and social justice, factors generally linked to more tolerant attitudes. On the other hand, Black Protestants tend to be conservative on family and social issues, factors often linked to intolerance of gays and lesbians. Data from the General Social Survey are used to examine the influence of religious group identification, as well as other relevant aspects of religiosity, on political intolerance among African Americans. Results indicate that although other aspects of religion (beliefs and behaviors) help explain variation in political intolerance, Black Protestant church affiliation has no relationship with attitudes about the civil liberties of homosexuals. However, additional tests show that Black Protestant church affiliation significantly predicts intolerance of other target groups (atheists and racists).

KEYWORDS:

• Black Protestants
• civil liberties
• homosexuality
• political intolerance
• public opinion
• religious affiliation
• religious behavior
• religious beliefs

Notes

^1. The phrase is from Sullivan, Piereson, and Marcus (1982, p. 4), but Gibson reasserts that “tolerance requires putting up with that with which one disagrees” (2006, p. 22).

^2. See Berlet and Chamberlain’s (2006) report for Political Research Associates provides an in-depth analysis of how “culture war” themes are used to mobilize conservative Christians. Also, Sherkat et al. (2010) examined religion as a source of African American’s opposition to same-sex marriage specifically but did not examine attitudes toward civil liberties more broadly defined.

^3. McKenzie and Rouse (2013) examined the effect of race’s interaction with religion on political attitudes, but did not measure intolerance directly as done in the analysis here.

^4. As is common in much of this literature, I use the words Black Protestant in reference to the distinctive faith tradition that developed apart from predominantly White mainline and evangelical churches. African Americans is used when speaking about the larger population.

^5. Davis’s (1995) study is noteworthy but does not include a wide range of controls for religiosity, and he was writing before a more precise denominational classification scheme was developed.

^6. Some research questions the direct nature of the link between religion and intolerance. For example, using a small phone sample from one Midwestern county, Eisenstein (2006) found only an indirect link between religious commitment and intolerance, as her measure of doctrinal orthodoxy (literal Bible) is associated with an increase in threat perception, which increases intolerance.

^7. Sherkat (2014) preferred to place Black Protestants in either one of several mainline or “sectarian” Protestant categories, but his approach downplayed the notion that Black Protestant theology is qualitatively distinct. Although this may be a direction in which scholarship is moving, it overlooks the argument of Steensland et al. (2000) and the more recent work of McKenzie and Rouse that showed African Americans, as a distinct social group, “hold more equitable opinions about disadvantaged individuals” (2013, p. 218).

^8. Critics may say that collinearity problems may arise between Black denominational influence and being Black, as noted by Steensland and associates (2000), but they argue that affiliation with a Black Protestant denomination has a separate effect than that of being Black. They also assert that in light of improved model fit and interpretive clarity, collinearity is not a problem.

^9. Cohen and Liebman (1997) used this technique to study liberal attitudes among American Jews. Also, given that there has been a great deal of highly visible pro- and anti-gay and lesbian rights political activity since 2000, conducting the analysis on a sample of years prior allows questions about religion and African Americans’ political tolerance to be answered more clearly. Although certainly needed, simply controlling for time’s passing cannot address the myriad ways in which time has interacted with the complete set of independent and control variables used here, but it is outside the scope of this articl to examine the ways in which the interaction between time and aspects of religiosity relate to attitudes about civil liberties.

^10. The GSS includes indicators of tolerance toward both left-wing and right-wing groups, but there is some confusion regarding where atheists should be placed. For example, Ellison and Musick (1993) included them as a left-wing group along with communists and homosexuals, but they are part of Davis’s (1995) right-wing groups along with militarists and racists. That distinction is avoided in this analysis, as I am concerned only with “threatening” groups for the reasons stated above.

^11. Responses are coded 1 = never, 2 = less than once a year, 3 = once or twice a year, 4 = several times a year, 5 = once a month, 6 = 2–3 times a month, 7 = nearly every week, 8 = every week, and 9 = several times a week.

^12. Psychological research has recognized Bible beliefs and church attendance as indicators of a “right wing authoritarian” personality, which is strongly linked to intolerant attitudes. See the work of Altemeyer (1988), Altemeyer and Hunsberger (1992), and Hunsberger (1996).

^13. A higher number on the morality indicator reflects more liberal moral attitudes and, although not specifically related to atheists and homosexuals, is left in the models to ensure comparability.

^14. The variable asks if respondents think of themselves as politically liberal or conservative, and the answers range from 1 = extremely liberal to 7 = extremely conservative.

^15. The GSS conducted these surveys in 1990, 1991, 1993, 1994, 1996, 1998, and 2000.

^16. Substantive findings do not change if ordered probit is used; OLS is used because its interpretation is more intuitive.

^17. In Model 3 the coefficient for the variable identifying Catholics is significant and indicates intolerance toward racists, but I hesitate to make a great deal about these findings. In general there is much animosity felt by Blacks regarding racists, and there are simply fewer Blacks belonging to the Catholic church.