If the mountain is getting scientists to talk to policy makers and practitioners, then Delbert Elliott is somewhere near the summit. Several times this week he has spoken to an audience of 1,000 people representing all interests in children's services. All have one goal in mind – better outcomes for children, and all agree on one mechanism for achieving that goal – evidence based programs.

But the climb has been steep and and difficult. When Elliott started out he was surrounded by scientists who delighted in pointing out that “nothing works”.

It is certainly true that there was scarcely any reliable evidence to show the impact of children's services on the child outcomes that most matter to policy makers, such as criminal behavior, drug use and violence.

He came to the conclusion that the problem was at least as much the responsibility of the scientific community as it was policy and practice makers. It was all too easy for them to criticize. But what about using evidence to design effective programs? And what about rigorously evaluating the result to generate better evidence? And what about working with policy makers and practitioners to increase the use of effective models?

These challenges have more or less defined his career.

At first the situation got worse. A US violence “epidemic” peaked in the early 1990s but in some respects, progress continues to be undermined by ill-conceived policies. Elliott gives the example of the harm caused by young offenders being transferred to the adult justice system, where the options are fewer, treatment is restricted and re-offending rates are higher.

But 40 years on, and from a situation in which nothing was working, at least 50 programs meet the most stringent criteria for effectiveness.

Talking to Elliott in Denver this week, three aspects of his approach stand out. Although undoubtedly a leader, he sees himself as just one figure in a sizable community finding a path to the crest of the mountain. Second, Elliott is not looking back with any great sense of satisfaction; the higher he goes the harder the climb. And, third, despite the collective achievement, he still questions whether he is on the right heading.

The community to which Elliott belongs represents a remarkable cross-section of individuals and interests. The Blueprints project, which he initiated in the mid-1990s, brings together two dozen or more program developers, plus hundreds of prevention scientists, policy makers and practitioners from a dozen or so countries.

The breadth of their interests is reflected in the range of the conversations to be heard at the Denver meeting: from sophisticated economic modeling to intimations about the impact on individual cases. There are workshops by trainers demonstrating how children experience programs, and by scientists discussing how to measure program fidelity. It is rare to find a conference where so many disparate interests are concentrated on a single cause.

As for the challenges to come, the primary objective continues to be to get effective programs embedded in mainstream practice.

“Life Skills Training is one of our most successful programs,” Elliott says. “But it is in place in just 12-15% of US schools. Delivery still depends on the corporation that owns it. That’s cumbersome for schools and school districts. And why should a school principal in Mississippi trust a study by a professor at Cornell that took place in upstate New York? The goal is getting teachers trained in a program like LST, so that it becomes integral to the lives of all schools."

Another target is cajoling what he calls the “old boys and old girls network” into operating on behalf of evidence based programs, instead of joining hands to make an obstacle to be repeatedly hurdled. “We need more people to push and enable. We need an eagerness to constantly improve the quality of existing programs and to elevate new ones.”

Then there are the emerging problems. It’s one thing to work out whether a program meets a threshold of effectiveness; it’s another to acknowledge that a program once considered to be effective is no longer supported by the weight of fresh evidence. So far this has only happened once in the life of the Blueprints project. Prevention Action will report next week on the Quantum Opportunities Program, and why it was removed from the Blueprints lists.

“In the future,” Elliott observes, “as we get more evidence, there are likely to be other examples of programs that we thought were effective – but now we’re not so sure. We can learn as much from programs that don't work as from those that do.”

And looking back – or down – is he sure he took the right path? “We’ve gone about this work program by program. We’ve looked in some way at over 1,000 now. My colleague Mark Lipsey takes a different route. He looks at lots of programs that are doing the same kinds of things and subjects them to a meta-analysis that gives an average effect size. He can tell people whether a broad strategy like mentoring or skilling-up parenting makes a difference. We can only say if program X or program Y makes a difference."

It’s the difference, he says, between understanding evidence based practices and evidence based programs. Both are important, but we don't yet know how the two fit together.

Thirty years hence these uncertainties should be much closer to being resolved. It was Dennis Romig who three decades back wrote a book on juvenile delinquent rehabilitation programs and concluded that “nothing works”. Now we know that plenty works. Delbert Elliott has played a crucial role in the community of people who are still searching for the right way.

Reference
Romig D. Justice for Our Children: An Examination of Juvenile Delinquent Rehabilitation Programs, Lexington. Books/D.C. Heath, Lexington, 1978, 1982.