Revisiting anomalous outcome data from the "Breaking the Cycle" program in Jacksonville

Michael D. White, Michael Hallett

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


Results from the national evaluation of the Breaking the Cycle program indicated that the participants in the program re-offended less frequently than comparison group defendants in Birmingham, Alabama, and Tacoma, Washington, but not in Jacksonville, Florida. This paper seeks to re-examine the same BTC program in Jacksonville, using different samples, methodology, data and analysis. The study employs T-tests, logistic regression and CHAID to examine differences in re-offending among contemporaneous random samples of program participants and comparable untreated defendants (n = 100 for each group). Using State Attorney and BTC data, both samples were drawn from the population of program-eligible cases from October 1999-April 2001, the complete period of operation for BTC Jacksonville. Similar to the national evaluation, results indicate that BTC did not produce significantly lower rates of re-offending among its participants. The authors explore possible explanations for the no-difference finding including Jacksonville's decision to link program participation to gaining pretrial release, the jurisdiction's heavy reliance on cash bail, and the differential impact of individual risk factors. The analysis produces more questions than answers and highlights the need for additional research to fully understand the Jacksonville BTC experience.

Original languageEnglish (US)
Pages (from-to)1-22
Number of pages22
JournalJournal of Offender Rehabilitation
Issue number1
StatePublished - Jan 26 2005
Externally publishedYes


  • Alternatives to incarceration
  • Breaking the Cycle
  • Drug treatment

ASJC Scopus subject areas

  • Rehabilitation
  • Law


Dive into the research topics of 'Revisiting anomalous outcome data from the "Breaking the Cycle" program in Jacksonville'. Together they form a unique fingerprint.

Cite this