Are you saying you can statistically differentiate the effect of conditional release from the effect of monitoring the offender? Can you somehow measure the effectiveness of their counselling programs or their reintegration into society separately from the monitoring?
Perhaps I'm oversimplifying. Say you take two populations. One is subject to conditional release and electronic monitoring, and the other is not subject to electronic monitoring and is incarcerated. If you have a higher degree of recidivism in that population, it would appear to me that the conditional release system is not as effective in reducing recidivism as incarceration. I'm oversimplifying matters, because in my scenario, I've completely discounted counselling and other aspects of conditional release.
My question is whether you can statistically measure the effect of those other aspects.