A chronic problem in many kinds of research is that we cannot directly observe the variable we are interested in. We have to observe something else that we think/hope is close enough to the one we care about that we can "operationally define" the observable quantity to be the important quantity.

How do we define "success" in rehabilitation efforts? True success is that the offender does not commit any more offenses. The observable quantity is whether the offender has been caught committing more offenses. On March 9, 1999, the U.S. Parole Commission issued the following certificate to a parolee:

"You are hereby discharged from parole," the March 9, 1999, certificate read.

"After a thorough review of your case, the Commission has decided that you are deserving of an early discharge," said the document signed by administrator Raymond E. Essex. "You are commended for having responded positively to supervision and for the personal accomplishment(s) you have made.

"The Commission trusts that you will continue to be a productive citizen and obey the laws of society."

The parolee was Phillip Garrido. He had already kept Jaycee Dugard in captivity for 8 years at that point, and he would keep her for 10 more. Sam Stanton and Denny Walsh have this story in the Sacramento Bee.

Let us keep that in mind every time we hear about recidivism rates. The one thing we know for certain is that the true rate is higher than the observed rate. The only question is how much higher.