When Stephen V. Monsma—formerly a political scientist at Pepperdine University and now a research fellow at the Paul B. Henry Institute for the Study of Christianity and Politics at Calvin College—completed his four-year study of 500 welfare-to-work programs in four cities (Chicago, Los Angeles, Philadelphia, and Dallas), he concluded that "much of conventional wisdom was wrong" about religious organizations. He talked about his stereotype-defying findings with senior associate editor Agnieszka Tennant.

What fallacy does your study correct?

It's that mainline Protestants are more active in social service programs than evangelicals—and that evangelicals merely serve their own congregations and are more concerned with evangelism than with social welfare programs. I found the opposite to be true: Of the welfare-to-work programs in the four cities that I studied, there were more evangelical programs than mainline Protestant programs.

How many programs were faith-based, both mainline and evangelical?

Of the 500 programs we studied, 117 were faith-based. Of the 96 Protestant programs, 61 were evangelical and 35 were mainline.

When did you classify programs as evangelical?

I considered them evangelical if they identified themselves as interdenominational, evangelical, or Pentecostal, or as one of the denominations traditionally considered evangelical, such as Assemblies of God or Baptist or Salvation Army.

What are these programs like?

I made the distinction between evangelical and mainline programs on the basis of their theological orientation. This meant that a number of the programs that I classified as evangelical were predominantly African American in terms of their church sponsorship, though some [African American ...