On April 11, JRSA held the second in its series of webinars on administrative records. The webinar focused on the use of administrative records by the Illinois and Vermont Statistical Analysis Centers (SACs).
Max Schlueter, the director of the Vermont SAC, discussed how the SAC obtains and uses court docket data for analysis and planning. He began by defining court docket data as consisting of all of the activity related to a court case: the identity of the defendant, the charges, dispositions of those charges, and sentences received. Max identified the ways in which the Vermont SAC uses these data, including analyzing how much time is typically given for a particular offense; analyses of sentencing practices throughout the state; policy and planning; throughput analysis; time analysis; program evaluation; and disproportionate minority contact. He then provided examples for each type of analysis; for example, looking at defendants sentenced to incarceration by race; looking at outcomes (dismissed, plea agreements, sentences) for DUI offenses; and evaluating Special Investigation Units for child sex crimes and rape.
Max then discussed strategies for obtaining court data. He suggested doing some homework to explore how the court system in your state is organized and what the court Management Information System looks like; determining whether any fees would be involved in receiving the data; assessing how much work the court would have to do to provide the data; setting yourself up as a "trusted partner" (giving analytical information back to the court in exchange for its data); establishing multiple users for the data; knowing what data fields you want, how you want them, and how often you want them; establishing how the data will be secured/protected; being flexible in what you are willing to accept; and developing a strong relationship with the court's information technology staff.
He also identified the type of data you might want to ask for, including: docket number and agency (police or prosecutor) case numbers; demographic information on the defendant; both arraignment and final disposition charges; any relevant dates; and detailed dispositional and sentencing data.
Mark Myrent, the Illinois SAC director, discussed how the SAC merges criminal history and records data to conduct both prospective and retrospective analyses. Mark began by noting some of the reasons the SAC has undertaken this work: the need to examine policy issues related to prison overcrowding and prisoner reentry; the lack of an offender-based tracking system; the amount of missing data in the criminal history records regarding prison admissions and exits; and the richness of the data collected by the Department of Corrections (DOC). He noted that the SAC begins with their corrections data and looks for a corresponding criminal history record in the hope of finding a one-to-one match between the two records systems. The search is built on "person-event date," the specific date that a specific person was admitted to, or released from, corrections.
The state identification number (SID) and the DOC number are the keys to making this match, although there are instances of multiple SID and DOC numbers, and the SID number is not included in the DOC record. The SAC does deterministic matching-that is, matching based on a set of rules. They look for an exact match on last name, first name, and date of birth. They also consider partial matches based on the first three letters of the names and an exact match on date of birth. For partial matches, further evaluation is necessary to determine if the match should be accepted. Mark described the matching process in further detail, including the use of other data fields, such as gender and prison admission and exit dates.
Once the matches are made, the data are aggregated so that they can be used for analyses. The first level of aggregation is for each person-event date. Two levels are created: one is criminal history (events that occurred prior to admission). The second level is post-release events, which are used for assessing recidivism. The next level of aggregation is by county and by cohort year, and this information is made available on the SAC's website.
Mark discussed the types of policy questions the SAC has addressed using the matched data files. One question relates to whether drug offenders have prior histories of other types of offenses, such as violent offenses. They have also addressed the question of whether violent sex offenders are more likely to commit a new sex offense after release from prison. They are hoping to use these data for cost-benefit analyses based on recidivism rates and costs.