Document Type


Publication Date



This article seeks to provide an alternative to the polarization that so often characterizes debates about church and state. In Part I, the author suggests that there are good policy reasons for supporting faith-based initiatives, and that these reasons ought to be attractive to liberals and progressives, many of whom have opposed faith-based initiatives. Faith-based social services are, after all, social services, and are often the very types of welfare services that liberals and progressives tend to support. Core religious values--in particular, concern about the less fortunate, a belief in human dignity, and a commitment to the possibility of redemption--reinforce liberal values that appear to have lost ground in modern America. Religious institutions are an integral element of a vital civic society and have an independent normative authority that may permit them to succeed where secular institutions have not. The case for supporting faith-based services does not require proof that faith-based services are better for all; it only requires that they may be better for some. That seems likely, even though solid empirical evidence is not yet available.

Several important caveats should be noted. The faith-based initiative may simply be a cover for privatizing public services, and for reducing public support for the most needy. Faith-based institutions and their proponents often attribute problems of poverty to personal moral failings, minimizing the systemic and structural features of those problems. Religious institutions can be a fount of intolerance and prejudice toward nonbelievers generally and women and gays and lesbians in particular. And government support of religious institutions may itself undermine and dilute the effectiveness of religious institutions. Thus, the faith-based initiative is not without significant dangers. But its potential benefits nonetheless provide strong reasons for supporting the concept in principle.

In Part II, the author argues that the current constitutional polarization on public aid to religion is also unwarranted. Both the assimilationist and the separationist camps paint with too broad'a brush. He suggests an expressivist approach to the Establishment Clause, built upon Justice Sandra Day O'Connor's endorsement test, as a mediating principle. The endorsement test asks whether a reasonable observer would interpret challenged government conduct as approving or disapproving religion. It provides that the government must avoid messages that make adherence to religion relevant to political standing in the community. While initially developed to assess government messages and displays, the endorsement test need not be limited to those confines, and in particular provides a critical tool for assessing the constitutionality of government funding. This is because once one acknowledges that some government support of religion is permissible under, and indeed required by, the Constitution, it is not enough to ask whether the government has aided religion. Nor is it sufficient to ask whether the government has been formally neutral, as the assimilationists do. The endorsement test focuses not on the mere fact of funding, but on what the funding expresses. Funding schemes should be invalidated only where they express official approval or disapproval of religion, and thereby send the message that religious adherence is relevant to citizens' standing vis-a-vis their government. The expressivist approach proposed here avoids formalism, pays attention to effects, and takes cognizance of the special status of religion.

Publication Citation

75 S. Cal. L. Rev. 559-603 (2002)