Skip navigation.
Home | About Us | Contact Us
Rick Scott, Governor
Florida Department of Corrections, Secretary Michael D. Crews

Florida Department of Corrections
Michael D. Crews, Secretary

Executive Summary

This document reports inmate recidivism rates (reoffense and reimprisonment) and shows how certain offender characteristics and other factors affect these rates for Florida state prison inmates. This information provides context for using recidivism rates as performance measures. The rates reported provide descriptive information; they are not the results of an evaluation of the Department of Corrections, its activities, or its programs.

Conclusions

  1. The reoffense and reimprisonment rates reported here are consistent with results from large-scale, national studies of state prison inmate recidivism by the U.S. Justice Department Bureau of Justice Statistics.

  2. Consistent with other large studies by BJS, this report finds that certain factors influence recidivism rates.

  3. Evaluations of prison inmates that use reoffense or reimprisonment rates as outcome measures should take certain (or similar) factors into account by controlling for their statistical effects on recidivism rates.

  4. The data described and analyzed for this report provide a basis for analyzing the performance of Department activities, functions, and programs that are designed or expected to reduce recidivism.

The following provides specific information from the report about recidivism rates; factors that influence recidivism: which are most important, the size and direction of their influences, and their combined effect; and the report methodology.

Recidivism Rates

The rates are the percentages of inmates released from July 1995 through June 2001 who committed a new offense (reoffense rate) or returned to prison for a new offense (reimprisonment rate) within the specified time following their release:

Follow-up Period
(Months Since Release)
Reoffense Reimprisonment
Male Female Male Female
6 12.5 % 8.4 % 1.3% 0.6%
12 21.5 % 15.2 % 6.4% 3.0%
18 28.2 % 21.1 % 12.5% 6.3%
24 33.2 % 25.7 % 18.0% 9.1%
36 40.5 % 33.3 % 26.7% 15.5%
48 45.4 % 39.1 % 33.2% 20.2%
60 48.7 % 42.8 % 38.1% 24.6%

Rates by Year Released

  • Recidivism (reoffense or reimprisonment) rates do not vary much for inmates released during different fiscal years (e.g., 2-year reoffense rate—high: 36.8% for FY95-96 releases, low: 33.0% for FY 98-99 releases).

  • No upward or downward trend in rates appears for inmates released during each of the last six fiscal years. The data presented here does not mean rates are declining or increasing by year. A separate analysis is required to determine whether annual rates are changing, one that compares data collected each year with the prior year.

  • Rates may have limited value as performance measures of progress on a year-to-year basis, although in proper context they may serve as useful benchmarks for program evaluation.

  • The latest reoffense rates for annual release cohorts that have the full follow-up period (males only):

    12 - month reoffense rate 19.4 % FY 1999-00 male releases
    24 - month reoffense rate 33.0 % FY 1998-99 male releases
    36 - month reoffense rate 42.7 % FY 1997-98 male releases

Factors Affecting Recidivism Rates

  • Factors analyzed are those that many scientific studies have shown to influence rates.

  • Only factors the Department has reliable, readily available data on at time of prison release were considered.

  • The factors included are the most influential from a wide variety of potential measures analyzed.

  • Five factors raise both reoffense and recidivism for males and females: prior recidivism, more disciplinary reports, worst crime burglary, more property crimes, and more drug crimes.

  • Seven factors lower both reoffense and recidivism for males and females: higher age, Hispanic ethnicity, low custody, more time in prison, higher education level, worst crime homicide, and worst crime sex/lewdness.

  • Six factors have mixed effects on reoffending and reimprisonment or between males and females: black race, high custody, supervision after release, worst crime robbery, worst crime other violent, and more weapons crimes.

Relative Effect of Factors on Recidivism Rates

  • For both males and females, the two most influential factors on reoffending and reimprisonment are prior recidivism and age at release, in that order.

  • For males, the last tested education grade level is the third most influential factor on reoffending and reimprisonment.

  • For males, 16 of these 18 factors significantly influence reoffending and reimprisonment; whereas for females, reoffending is significantly influenced by only 12 factors and reimprisonment by 10 factors.

  • Several factors have roughly similar relative importance for reoffense and reimprisonment:

    Males: prior recidivism, age, education level, disciplinary reports, property crimes, drug crimes, homicide (worst crime), high custody, other violent (worst crime).
    Females: prior recidivism, age, drug crimes, ethnicity, sex/lewdness (worst crime), low custody.

  • Some factors influence reoffending more than reimprisonment:

    Males: supervision after release, sex/lewdness (worst crime), ethnicity, time in prison.
    Females: supervision after release, time in prison, disciplinary reports, homicide (worst crime), high custody, property crimes.

  • Some factors influence reimprisonment more than reoffending:

    Males: robbery (worst crime), burglary (worst crime), low custody, race.
    Females: robbery (worst crime), education level, race, other violent (worst crime), burglary (worst crime), weapons crimes.

Size and Direction of Factor Influences on Recidivism Rates

  • Statistical analysis of the factors indicates the direction and size of each factor's effect when controlling for all others, as follows (non-significant are shaded):

    Prior Recidivism - for each prior event recidivism probability changes:
      Reoffense: males +21.9 % females +27.2 %
      Reimprisonment: males +28.4 % females +34.3 %
    Age - for each year older at release recidivism probability changes:
      Reoffense: males -3.2 % females -2.9 %
      Reimprisonment: males -3.3 % females -3.4 %
    Disciplinary Reports - for each report received recidivism probability changes:
      Reoffense: males +1.8 % females +2.3 %
      Reimprisonment: males +1.6 % females +1.4 %
    Education - for each higher education grade tested recidivism probability changes:
      Reoffense: males -2.4 % females -2.5 %
      Reimprisonment: males -3.5 % females -4.0 %
    Time Served - for each additional month served recidivism probability changes:
      Reoffense: males -0.7 % females -1.5 %
      Reimprisonment: males -0.4 % females -0.9 %
    Burglary - inmates with this as their worst crime recidivate relative to others:
      Reoffense: males +11.6 % females +27.1 %
      Reimprisonment: males +28.7 % females +58.6 %
    Race - black inmates recidivate relative to non-blacks:
      Reoffense: males +27.1 % females -13.5 %
      Reimprisonment: males +30.8 % females -21.7 %
    Drug Crimes - for each offense recidivism probability changes:
      Reoffense: males +2.6 % females +4.8 %
      Reimprisonment: males +2.1 % females +5.7 %
    Property Crimes - for each offense recidivism probability changes:
      Reoffense: males +2.8 % females +1.5 %
      Reimprisonment: males +2.7 % females +1.6 %
    Supervision - inmates with supervision after release recidivate relative to those not supervised:
      Reoffense: males -14.2 % females -17.3 %
      Reimprisonment: males +6.8 % females +2.0 %
    Low Custody - inmates released at low custody (community and low) recidivate relative to medium custody:
      Reoffense: males -14.3 % females -9.1 %
      Reimprisonment: males -17.1 % females -9.2 %
    Homicide - inmates with this as their worst crime recidivate relative to others:
      Reoffense: males -28.3 % females -42.3 %
      Reimprisonment: males -21.6 % females -37.4 %
    Sex / Lewdness - inmates with this as their worst crime recidivate relative to others:
      Reoffense: males -31.5 % females -53.1 %
      Reimprisonment: males -19.8 % females -57.0 %
    Robbery - inmates with this as their worst crime recidivate relative to others:
      Reoffense: males -1.5 % females +1.7 %
      Reimprisonment: males +13.7 % females +33.7 %
    Ethnicity - Hispanic inmates recidivate relative to non-Hispanics:
      Reoffense: males -6.2 % females -26.7 %
      Reimprisonment: males -6.2 % females -28.8 %
    High Custody - released at high custody recidivate relative to medium custody:
      Reoffense: males +3.9 % females -24.1 %
      Reimprisonment: males +6.7 % females -11.1 %
    Other Violent - inmates with this (e.g., aggravated battery, aggravated assault, etc.) as their worst crime recidivate relative to others:
      Reoffense: males -2.8 % females -4.6 %
      Reimprisonment: males +4.0 % females +15.1 %
    Weapons Crimes - for each offense recidivism probability changes:
      Reoffense: males -1.8 % females -1.3 %
      Reimprisonment: males -3.4 % females +5.7 %

Combined Effect of Factors on Recidivism Rates

  • The factors as a group improve prediction for reoffense and reimprisonment somewhat better for males than for females, except for reoffending at 60 months.

  • For males, the predictive power of the factors is greater for reoffense at 18 months than for reimprisonment, about the same at 36 months, and greater for reimprisonment at 60 months.

  • For females, the predictive power of the factors is greater for reoffense at 18 months than for reimprisonment, but greater for reimprisonment at 36 and 60 months.

  • The amount of combined influence of these factors on recidivism is consistent with the explanatory power of similar variables found in other recidivism studies.

  • These factors predict recidivism from 6.3 to 15.1 points better than chance does, depending on the cohort, recidivism measure, and follow-up period of each model, as follows.

Methodology

  • New: Reimprisonment rates have been added as a recidivism measure in addition to reoffense (reconviction) rates.

  • New: Male and female inmates have been analyzed separately.

  • New: A thorough analysis of criminal history measures has improved our measurement of those factors that influence recidivism rates.

  • New: Time "not at risk" for recidivism—when an inmate is back in prison for a technical violation—has been accounted for to avoid inflating time to recidivism.

  • A standard statistical method (survival analysis) is used to generate rates, allowing maximum use of available data on inmate releases, new offenses, and returns to prison, and providing estimates of short- and long-term rates (up to 60 months).

  • Statistical techniques are used to establish how much the combined factors influence reoffense rates (logistic regression) and to specify the relative contribution of a factor while controlling for other factors' influence (proportional hazards regression).

  • Only releases from original prison commitments are used; releases following technical violation returns are not included to avoid inflating "successful" releases.