The truth, Mr. Johnson and many other social scientists say, is that there is little reliable research proving the effectiveness of religious programs. They also add that there is scant evidence showing which religious programs show the best results and how they stack up against secular programs.
''From the left to the right, everyone assumes that faith-based programs work,'' Mr. Johnson said. ''Even the critics of DiIulio and his office haven't denied that. We hear that and just sit back and laugh. In terms of empirical evidence that they work, it's pretty much nonexistent.
''We've created an office out of anecdotes.''
In the history of grand presidential initiatives, this would not be the first to take the stage without a script. But this one is different. A body of research is essential to the project's success for the simple reason that it would be unconstitutional for the government to decide which religious programs to finance based on theology or favoritism or familiarity. President Bush and Mr. DiIulio have frequently said that a record of effectiveness is the only viable measure.
No one denies that religious organizations and volunteers do indispensable work caring for people in need. For example, Mr. Johnson's colleague, Ram A. Cnaan, a professor of social work at the University of Pennsylvania, found that more than 90 percent of Philadelphia's congregations provided community services. There is also extensive research showing the benefits of faith: religious people cope better with old age, sickness and hardship; they are healthier; they drink less alcohol; they volunteer more.
Mr. Johnson pulls reports off his shelves with more evidence: religious youths are less likely to use drugs, or be involved in crime. But these studies, he and other researchers say, do not prove if someone can get more help from a religious program than from a secular one.
One program that has opened itself to scrutiny is Teen Challenge, which treats nearly 3,000 drug and alcohol addicts annually in 150 centers around the country. The group says the secret to its success is what it calls the Jesus factor.
In 1995, Teen Challenge helped Aaron Todd Bicknese, a doctoral student at Northwestern University, track down 59 people one to two years after they had completed Teen Challenge's yearlong residential program. Mr. Bicknese compared them with a similar group of addicts who had spent one or two months in a hospital rehabilitation program.
The results were favorable to Teen Challenge, which posted a simplified summary of the dissertation on its Web site concluding that it had an 86 percent success rate. In recent months, politicians and evangelical leaders have used that figure to assert that religious programs are superior to secular ones.
Mr. Bicknese found that Teen Challenge graduates reported returning to drug use less often than the hospital program graduates, but not less than the hospital program graduates who continued attending Alcoholics Anonymous support groups (which some also consider to be religious because of their reference to a ''higher power'').
He also found that Teen Challenge graduates were far more likely to be employed (18 of the 59 worked at Teen Challenge itself, which relies on former clients to run the program).
Social scientists have pointed out that the 86 percent success rate of Teen Challenge is misleading. It does not count the people who dropped out during the program. And like many religious and private charities, Teen Challenge picks its clients.
Before they are accepted, most of the addicts have already been through detoxification programs, said the Rev. John D. Castellani, president of Teen Challenge International U.S.A. In the program's first four-month phase, Mr. Castellani said, 25 to 30 percent drop out, and in the next eight months, 10 percent more leave.
This raises questions for David Reingold, a researcher at the Indiana University School of Public and Environmental Affairs. A study Mr. Reingold has just completed of social services in Indiana found that religious programs are more likely than their secular counterparts to limit the clientele they serve. As a result, Mr. Reingold said, ''It's an extreme exaggeration to say that religious organizations are more effective.''
In other words, Teen Challenge's claims should be taken with a large grain of salt.