Recidivism: A Time Series Analysis of Navy Releases, 1997-2003
Nimon, R. Wesley, Purcell, Timothy E., Corrections Today
Historically, Navy inmates have had an overall recidivism rate that has generally been below that of civilian levels. Although updated civilian rates are not available for comparison, the current analysis suggests that Navy recidivism rates are still relatively low. However, there was a slow increase in the Navy's recidivism rate between 1997 and 2003. Analysis based on 2003 releases provides the most recent three-year recidivism rates, and the 1997 releases provide the longest-term recidivism rates available. This article disaggregates the trend by brig, confining offense and post-release crimes contributing to the Navy's aggregate recidivism rate. The analysis draws two conclusions. First, the number of military offenders (i.e., those confined for a military-specific crime) increased between 1997 and 2003. Second, the recidivism rate of military offenders increased during this time.
While it is challenging to rehabilitate inmates confined for military offenses because the motivations for their crimes are varied and their sentences generally short, military offenders present a potentially fruitful, yet relatively underserved, population to target with rehabilitation programs. Currently, there are few programs that target them, and their recidivism rates are high relative to those of other Navy offender types and they are increasing. Upon release, the percentage of individuals arrested for public order crimes, which include driving under the influence, has increased substantially since 2000. It is certainly possible that this is simply the result of increased enforcement without any implications for the characteristics of the inmates being released. Even so, if there are ways to integrate curricula into all rehabilitation programs that discourage future public order offenses, doing so offers an approach to target a relatively high and increasing source of Navy recidivism.
The data include all 13,997 individuals released from a Navy confinement facility in the seven years from 1997 through 2003. Demographic information maintained by the U.S. Navy was combined with criminal history information provided by the FBI. The FBI compiles arrest and court data from local municipalities, and the data were used to calculate Navy recidivism rates. Unless otherwise stated, the recidivism rate is defined as the percentage of releases who were arrested within three years after release. While the arrest data do not fully represent the actual number of arrests (due to underreporting), the FBI's arrest data are relatively robust compared with the court data. Thus, this analysis focuses on arrest data.
The FBI provided data on 9,905 of the 13,997 releases from a Navy brig between 1997 and 2003. This turns out to be slightly more than 70 percent of the releases. This difference could be because they had no prior or subsequent arrests, but it could also be that the FBI data are incomplete. The recidivism rates undoubtedly reflect some downward bias, but there is no reason to suspect systematic bias by crime category or demographics.
Two FBI data pulls are reflected in this analysis. Criminal history information on the 2002 and 2003 releases were drawn in February 2007, and data on those released from 1997 through 2001 were drawn in September 2005.
The Timing of Re-offense Decisions
One question often asked is how much time without an arrest must transpire before it becomes unlikely that a released inmate will re-offend. To assess this, the cumulative recidivism rate for the 1997 releases, the cohort tracked for the longest period, is plotted over time in Figure 1. The eight-year recidivism rate is an estimate because it reflects only the individuals who were eight years post-release when the FBI recidivism data were drawn on Sept. 6, 2005. This includes about 65 percent of the individuals released in 1997. Inferences drawn from that data are indicative but not definitive. …