What accounts for the success on this side of the Atlantic of referenda strengthening state laws that recognize marriage as only the union of one man and one woman? A study published by three American political scientists, finding that religion more than anything else predicts attitudes about sanctioning same-sex relationships, suggests that the stronger religious character of the United States relative to Europe may be responsible.
The researchers examined data from a PBS-commissioned telephone poll of more than 1,600 American adults (which included oversamples of white evangelical Protestants, African-Americans, and Hispanics) conducted in the spring of 2004 just after the City of San Francisco began issuing marriage licenses to parties of the same sex. The respondents were asked to what degree they favored 1) allowing "gays and lesbians to marry legally" and 2) the creation of "civil unions." Those who expressed disapproval to the first question were also asked whether they support an amendment to the Constitution to "ban gay marriage."
In multivariate tests, the researchers found that measures of religious affiliation, religious practices, and moral attitudes yielded more statistically significant correlations with the answers to the first two questions than did demographic variables (age, marital status, gender, and education). While respondents who were not Protestant (especially Jews, secularists, and members of less conventional religions) were much more likely than Protestants to support gay marriage and to lesser extent, civil unions, respondents across all religious traditions who attend church regularly expressed greater disapproval of same-sex relationships than did those who did not attend church regularly.
The study also found that respondents who prioritized concern about "moral values" above matters like the economy and terrorism-as well as those who expressed traditional attitudes on morality and secularism-were also more likely to oppose legal recognition of same-sex marriage and civil unions. These religious and moral measures played a weaker role, however, in predicting attitudes toward the federal marriage amendment, where political conservatism was the strongest indicator of support. But as with the first two questions, religious activity and involvement more than religious affiliation contributed to expressing support for federal action.
While the study might have found broader support for natural marriage had the survey questions been less biased in favor of same-sex constructs, these findings explain what the researchers seem to lament: how religion uniquely informs public opinion and policy on marriage.