Risk Assessments in the Juvenile Justice System: a Relevant Critique

One of the more important developments in the juvenile justice system (as well as the criminal justice system) of the last few decades has been the adoption of assessment tools to classify youth by their risk to reoffend.
This is critical, since research has shown that teens (and adults) are more likely to return to the justice system the more contact they have with it; as a result, interventions need to be different depending upon whether youth are classified as low-, medium-, or high-risk. (In its most simplest form, this means keeping lower-risk kids out of detention. While this can be difficult for communities to do, it seems to result in lower incidence of crime.)

But a critique of risk assessment tools used throughout the justice system issued by the National Council on Crime and Delinquency argues that risk assessment tools have gotten less effective.
For one thing, these tools sometimes confuse risk assessment with needs assessment, and many of the items measured have no actual relationship to recidivism.
Furthermore, shorter is better, as "[a]ssessment scales containing 25–125 variables introduce significant noise and create potential problems with reliability." A 2003 study of a criminal justice risk assessment instrument, for example, showed that reducing its 54 items to a mere eight improved its reliability in a statistically significant way.
Which risk assessment tool is used in your juvenile court? Is it doing the job it's supposed to?
*Photo by Flickr user mag3737.