Sex Offender Recidivism: A Multistate Study Using Criminal History Records

1
Sex Offender Recidivism A Multistate Study
Using Criminal History Records

• Stan Orchowsky
• Justice Research and Statistics Assn.
• Presented at the NCJA Conference
• Louisville, KY
• August 5, 2008

2
Criminal Histories and NCHIP

• Criminal history records describe an arrest and
  all subsequent actions that are identifiable to
  an individual
• The National Criminal History Improvement Program
  (NCHIP)
• Administered by Bureau of Justice Statistics
• Over 500 million to states since 1995 to improve
  the quality of records

3
Project Goals

• help SACs obtain and analyze their states
  criminal history records (build capacity)
• coordinate SAC analyses of state criminal history
  records to address the same key criminal justice
  issue across a number of different states
• identify critical data quality issues in states
  criminal history records

4
Awards

• BJS awarded funds to JRSA to coordinate projects
• JRSA made subawards in two categories
• Capacity-building AK, IA, NM, TN, VA, WI
• Coordinated analysis AZ, DE, IL, SC, UT
• 9 SACs participated in sex offender recidivism
  analysis (all but VA and WI)

5
Criminal History Record Issues

• Arrest Records with no Dispositions
• Disposition never received
• Disposition misrecorded
• Disposition Records with no Arrests
• Less common
• Same explanations
• Changes in Criminal Codes Over Time

6
Related Issues

• Non-matches between criminal history files and
  prison release files
• Most matches by State Identification Number
• Identified as sex offender by DOC, but no sex
  offense in criminal history files

7
Examples of Data Issues

• Utah
• 17 of arrests had missing dispositions
• NM
• 77 of dispositions coded as unknown
• AZ
• no criminal history info found for 11 of all
  releasees
• Of releasees identified as sex offenders by DOC,
  57 had a sex offense in the CHx

8
Methodology

• Modeled after 2003 BJS study
• Released in 1994
• Violent male sex offenders
• Rape
• Sexual assault
• Released from prison in 2001
• Rearrest measured for 3 years after release

9
Demographic Characteristics

10
Sentence Length and Time Served (average, in
years)

11
Prior Criminal Record
12
Recidivism
13
Sex Offenders v. Other Offenders
14
Conclusion

• Criminal history records can be used to study
  violent/sex offender recidivism, but extensive
  data cleaning is required
• Descriptive studies using only criminal history
  records are easier than those that require
  matching to other datasets
• Periodic audits of criminal history records are a
  valuable tool