No matter how you interpret the separation of church and state, it is clear that the president's new faith-based initiatives cross an old, well-established line. The question is not whether religious groups can get government money: they can and have for years. The question is whether they must keep their state-funded practices separate from their "sectarian" ones. This administration not only allows the two to mix, it encourages the overlap.

At an October conference of the Roundtable on Religion and Social Welfare Policy [http://www.religionandsocialpolicy.org/publications/publication.cfm?id=20], James Towey, head of the Office of Faith-Based and Community Initiatives, said he could not understand opposition to the new initiatives. Who could be against building partnerships? But Rabbi David Saperstein, at the same conference, put his finger on it when he said that most people agree with about 95 percent of religion and government partnerships, but the new initiatives insist on instituting the 5 percent that threaten to "tear us apart."

Since 1996, faith-based organizations have been able to receive certain government funding without diminishing their "religious character." They can display religious symbols and hire co-religionists exclusively. They cannot use government funds for worship or evangelization, but they can use private funds to do those things alongside their publicly-funded services. The Bush administration did not create this change, but it is testing the limits of its application.

The question now is how much "religious character" is too much when it comes to public money? The old model featured separate organizations and separate books, which is why denominations and congregations founded faith-based 501-3c organizations. Underlying expectations were clear, even if participants sometimes fudged. Now, we must establish a new set of expectations and boundaries.

This won't be easy. Disagreement around the limits of "religious character" and government support exposes a deep American ambivalence about the role of government in our lives. Must government be guided only by broadly public values (e.g., equal justice under the law)? Or can government support private values such as religious character development, if those values create desirable behaviors and eliminate undesirable ones?

As a society, we probably agree that sobriety, hard-work, and responsible parenting are desirable character traits. We know that religious groups provide services to the community that support and promote these behaviors. But we also know that religious groups do not limit themselves to just teaching values on weekdays, saving religious instruction for Friday night through Sunday morning. The argument for funding sectarian religious groups is that their services are delivered in the context of a total life change; character development and the religious instruction behind it are part of the package -- seven days a week.

Public policy analysts are now studying the effectiveness of faith-based programs, but that misses the point. Our national question is not whether religion has the power to change people's lives fundamentally. It does. Nor are we asking whether faith-based organizations do a good job. Some do, some don't. The question is whether we, as a society, are willing to pay for sectarian religious content in social service provision.

When it comes down to it, does government neutrality toward religion mean that government cannot support programs with sectarian values, even if they work, or does it mean that government should not care what values are involved, so long as they work? If the goal is to help prisoners and drug addicts become responsible citizens, does it matter whether Catholics, Methodists, Mormons, Sunni Muslims, or the Nation of Islam supply the content? Does it matter if they use my money?

Additionally, on a practical level, our expectations for faith-based partnerships must be realistic. Enthusiastic reformers have especially misrepresented congregations. Congregations have, on average, less excess capacity -- money, time, volunteers -- than reformers seem to think. Many congregations are not indigenous neighborhood organizations. Effective partnerships are hindered by the uneven flow of information among congregations, local government, and secular service groups. In fact, even to portray congregations as social service organizations is to mischaracterize their mission, despite the amazing amount of social services they provide.

Considering such practical issues will help us build better partnerships, but the question still remains: do we want government to fund "sectarian" social service programs that distinguish right and wrong and good and evil on religious grounds? Will we pay for behavioral change that is rooted in specific religious beliefs? No amount of analysis on outcomes and efficacy will spare us the burden of this decision. Where do we draw the line?

Author, Arthur E. Farnsley II, is a Senior Research Associate for The Polis Center's Project on Religion and Urban Culture at Indiana University Purdue University Indianapolis. His new book is Rising Expectations: Urban Congregations, Welfare Reform, and Civic Life (Indiana University Press).