Agnosticism and the Search for EBP

We in the youth services field should never let our desire to be "evidence-based" turn us into a "faith-based" movement. When we're searching for the most effective ways to help young people avoid trouble with drugs and stay out of the justice system, we should be agnostics — even the most attractive new answer should never stop us from asking important questions.
The substance abuse treatment field has finally started to find a few treatment approaches that can be called "evidence based," meaning that sufficient research exists to conclude that a particular model will produce important effects on client behavior in a reliable manner and across varying settings. The same thing is happening in delinquency prevention. A number of researchers, especially Steve Aos and Peter Greenwood,  have described program models that produce positive returns on investment, including multisystemic therapy (MST), functional family therapy (FFT), the nurse-family partnership model (NFP), and multidimensional treatment foster care (MFTC). 
Studies show these programs to be cost-beneficial in terms of the changes they produce in youth behavior and their effects on reduced crime. In his 2005 book on the topic, in fact, Greenwood concluded that these programs were the only cost-effective interventions for juvenile offenders and he advised policymakers to concentrate their funding investments on these models. Greenwood's advice, however, should not be misinterpreted. Only a handful of intervention models to date have survived enough clinical trials to demonstrate their effectiveness, but this is not to say that they will always be the best and only interventions for young people. The absence of evidence is not proof of ineffectiveness. 
The search for effective programs is far from over. Policymakers and practitioners must continue to search for new program models and to test the effectiveness of existing models.  Some unproven and even untested programs may one day turn out to be highly effective.
Developing strong and consistent research findings about social programs is expensive.  Multiple investigators work for years or even decades to accumulate the high-quality evidence needed to establish a program’s effectiveness. This rarely happens, of course, and we have relatively few proven program models as a result. It would be foolish, however, to stop developing new programs now simply because a few models have finally been found effective – especially when our evaluation agenda is theoretically narrow and subject to the corrupting influence of funding politics and the self-interests of researchers. 
In an imaginary, “best of all possible worlds” where research investments followed intervention concepts naturally and without prejudice, it might make sense to use the existing research base to finalize a list of proven programs.  In the real world of competitive funding and politically-charged investment, however, relying exclusively upon today’s research to invest in tomorrow’s social programs would stifle innovation and limit the impact of policy and practice. Researchers must continue to pursue new models. 
Unfortunately, many of our most talented researchers are reluctant to invest their career capital in community-based programs that depend on non-professional staff and volunteers. Researchers need reliable research partners, and programs that deliver medical-style interventions are simply more adept at maintaining program fidelity and adhering to rigorous research designs. As a result, the most productive researchers are often unwilling to study interventions outside the reassuring environs of the medical model. 
Likewise, the resources necessary to generate high-quality evidence are limited and hard to secure. Funding agencies are hesitant to invest in community-based programs that rely on volunteers, that cannot control the context of service delivery, and that cannot always determine the sequence and timing of critical events. There are potentially valuable program models in juvenile justice that cannot attract the level of research investment required to generate sound evidence. This is a weakness of investment, not a lack of efficacy. 
For these reasons, there is still much work to be done in identifying effective programs for youthful offenders, especially non-clinical, non-medical, and community-based interventions. Practitioners and researchers must keep asking the tough questions. Is this program really the best we can offer? Does this approach really make sense for this client and this family? Is there another way we could approach this young person that might be a better fit with our concept of problem-solving? Are we collecting the information necessary to prove our effectiveness? Even if we are not yet considered an evidence-based program, are we taking steps today that will allow us to merit that designation tomorrow? 
If we are going to succeed in building better and better programs for drug-involved, youthful offenders, all of us working in this field must remember to nourish our inner agnostic.