The study found that felony recidivism was not reduced for the average offender in the study sample who was confined for a violation of conditions for community supervision. In fact, in all of the models tested, confinement for a parole/probation violation increased recidivism. One possible explanation offered by the researchers is that confinement for a violation made it more difficult to enter the labor market upon release. Another plausible explanation offered is that a community corrections officer (CCO) has the ability to observe an offender's risk for recidivism beyond what is actually measured in the Risk Level Classification. In this case, the larger effects found in the preferred model implies that some CCOs are routinely better at assessing higher risk offenders who are more likely to recidivate. The study used a non-experimental design based on observational data from the Department of Corrections. This report notes that this design increases the risk that selection bias and unobserved factors can undermine causal inference. The study population included all offenders in Washington State who were at-risk for recidivism in the community between July 1, 2001, and June 30, 2008. The study population was limited to those offenders with at least one violation during the 36-month follow-up period (N=70,398). The sample of violators was further restricted to those offenders who had a single CCO during community supervision and with a sufficient number of observations per CCO (n=1,273). 7 exhibits and a listing of publications used in the meta-analysis