group apparently most hostile to government accommodation, or an expansive sense of free-exercise rights, was not particularly irreligious. Contrary to the expectations of the "culture wars" thesis, such respondents appeared rather religious and religiously orthodox. Again, we suspect that this group is motivated by a strong sense of the autonomy of the sacred rather than disapproval of religion generally.
What all of this appears to mean is that, at the level of public opinion, questions of religious establishment and free exercise are empirically, as well as conceptually, distinct. This distinction further suggests that there will be internal divisions within a coalition of the "'religious." Doctrinally conservative Christians appear to hold highly majoritarian beliefs about the relationship between religion and politics or to regard the sacred and the secular as mutually incompatible spheres of human activity. As such, evangelicals are rather unlikely allies in political struggles for the free-exercise rights of religious minorities in an increasingly diverse religious mosaic. 5 Rather, Jose Casanova ( 1994) comes close to the mark when he characterizes evangelical Protestantism as the defacto civil religion of the United States. Similarly, religious minorities, be they Asian immigrants, Native Americans, or adherents of "cults" or "New Age" religions, seem unlikely to support the culturally hegemonic aspirations of some conservative Christians. A broad understanding of the free-exercise clause connotes an America characterized by a high degree of religious pluralism while the evangelical basis for religious accommodationism may be based on a belief in cultural consensus. Given such divisions in public opinion, we suspect that the types of religion practiced by Americans will continue to be as politically important as the extent of their religiosity.