T A R R A N T C O U N T Y
Ex- Offe n de r Re e n t r y I n itiat ive
PAROLEE RECIDIVISM IN TARRANT COUNTY
ANGEL ILARRAZA, PH.D., Reentry Program Coordinator
JULIE H. MARTIN, MS., Research Associate
Tarrant County Administrator’s Office
1 0 0 E . We a t h e r f o r d , S u i t e 4 0 4 F o r t Wo r t h , T X 7 6 1 9 6 - 0 6 0 9 • t e l e p h o n e : 8 1 7 . 8 8 4 . 1 5 7 5 • f a x : 8 1 7 . 8 8 4 . 1 7 0 2 •
w w w. t a r r a n t c o u n t y. c o m / e t c r i
Parolee Recidivism in Terminology
Releases - Offenders leaving Texas Department of Criminal
Tarrant County Justice, (TDCJ) due to the discharge of their sentence or to another
form of supervision.
(2000-2006) Parole - The conditional release of an offender from prison, by a
Board of Pardons and Paroles decision, to serve the remainder of
Introduction his/her sentence under supervision in the community.
This research report presents the findings of a Discharge Releases - Prison offenders who are released from
recidivism study conducted by the Office of TDCJ either by full expiration of their sentence or by court order.
Criminal Justice within the County Years Post-Release - Calculated in increments of 365 days from
Administrator’s Office in Tarrant County, Texas. the date of release
The purpose of this study is to establish a baseline Release Cohort - All individuals released within a calendar year.
understanding in regard to the amount and type of
Year to Date Recidivism - Recidivism rate for each annual cohort
recidivism among the population that is released
to parole supervision in Tarrant County.
Recidivism, in this study, is defined as any re-arrest by a law enforcement agency, any re-conviction by a judicial court
and/or any re-incarceration occurring after the initial parole release date to Tarrant County. For county government, a
better understanding of the nature, extent, and trends in parolee recidivism is critically important for evaluating the
efficacy of any program(s) put in place to reduce recidivism.
This analysis presented in this report, examines the criminal history data for individuals released under parole
supervision in Tarrant County by TDCJ. The data used in this study was obtained through an open records request
submitted to the Parole Division of TDCJ for all persons placed on parole within Tarrant County between 2000 and 2006.
State identification numbers were obtained and subsequently used to identify the respective offense history records from
the criminal history repository maintained by the Texas Department of Public Safety (DPS). These efforts yielded the
complete criminal history records for 14,851 individuals paroled to Tarrant County between 2000 and 2006.
Year to Date Recidivism
The first part of this analysis describes global recidivism rates for each of seven annual release cohorts. Table 1 presents
Table 1. Ye ar t o Dat e Re c id iv is m
YEAR 2000 2001 2002 2003 2004 2005 2006
Number Paroled 1,850 2,166 2,092 2,372 2,145 2,099 2,127
Years of data 8 7 6 5 4 3 2
re-arrested 63.7% 62.6% 60.4% 58.9% 57.2% 51.4% 41.3%
re-convicted ≥42.6% ≥41.8% ≥39.9% ≥40.3% ≥36.8% ≥36.4% ≥28.3%
re-incarcerated ≥41.0% ≥40.4% ≥38.5% ≥38.5% ≥35.6% ≥34.7% ≥27.2%
the year-to-date recidivism for each annual release cohort (as of November 2008). It is noted that the data in Table 1 are
not comparable across release cohorts because the number of years of available post-release data are different for each
cohort. For example, the 2000 release cohort has eight complete years of post-release data, while the 2006 release cohort
only has two years of complete post-release data. As reported, 1,850 individuals were paroled in 2000, and after eight
years post-release, 63.7% were re-arrested, 42.6% were re-convicted, and, 41.0% were re-incarcerated. Similarly, after only
five years post-release, 58.9% of the 2,375 individuals in the 2003 release cohort were re-arrested, 40.3% re-convicted, and
38.5% re-incarcerated. Of the 2,127 individuals in the 2006 release cohort, 41.3% were re-arrested within the first two
years post-release, 28.3% were re-convicted, and 27.2% re-incarcerated. The greater than or equal to sign (≥) appearing
with the re-conviction and re-incarceration rates, acknowledges that missing conviction and incarceration data in the
criminal history record resulted in the underestimation of re-convictions and re-incarcerated.
Year-to-date recidivism reported in Table 1
F igu re 1. Re c id ivis m ac ros s A n n u al Cohor t s : Pe rc e n t Re - ar re s t e d
depicts global recidivism for each cohort
given the number of years of available data.
In order to asses recidivism comparatively
(across cohorts), it is necessary to examine
the recidivism rates within a specific post- 41.25%
release time period.
Comparing the Rates of Recidivism Across
Release Cohorts by Post-Release Year 27.50%
Figures 1, 2, and 3 depict annual post-release
comparisons of the re-arrest, re-conviction,
and re-incarceration recidivism rates for six
annual release cohorts (2000-2005). With
very little exception, the annual post- release
comparisons (across release cohorts) reveal
an upward trend in recidivism, with Release Year 1 Year 2 Year 3
2000 2001 2002 2003 2004 2005
steadily increasing recidivism for each
subsequent release cohort at any given year post-
release. In other words, at any given point in time after release, proportionately more persons (within a given release
cohort) were re-arrested, re-convicted and re-incarcerated than in all previous release cohort(s).
Figure 1, for example, denotes that 16.2% of
F igu re 2. Re c id ivis m ac ros s A n n u al Cohor t s : Pe rc e n t Re - c on vic t e d the 2000 release cohort was re-arrested
40% within 12-months of release, as compared to
21.4 % of the 2003 release cohort, and 24.8% of
the 2005 release cohort. At 24-months post-
release, the data show that 30.9% of the cohort
released from confinement in year 2000 was
re-arrested. By comparison, annual re-arrest
data for the 2003 release cohort show that
20% 38.1% of the cohort was re-arrested, and that
41.7 % of the 2005 release cohort was re-
arrested within the 24-month time period post-
10% release .
The general upward trend evident in the re-
arrest data remain present at re-conviction (see
Figure 2). During the first year post-release,
Release Year 1 Year 2 Year 3 10.5% of the 2000 release cohort was re-
2000 2001 2002 2003 2004 2005
convicted compared to 18.2% of the 2005 release cohort. At three years post-release, 27.9% of the 2000 release cohort and
36.4% of the 2005 release cohort were re-convicted. The steady increase in the proportion of re-convictions continues
across all release cohorts.
F igu re 3. Re c id ivis m ac ros s A n n u al Cohor t s : Pe rc e n t Re - in c arc e r at ion . Figure 3 presents data depicting
the re-incarceration rates for the
release cohorts 2000-2005. Consistent with
upward trends reported for re-arrest and
re-conviction, the data for re-incarceration
26.25% also show the presence of a steady increase
in the proportion of persons re-
incarcerated subsequent an initial release
17.50% to parole supervision. As depicted in
Figure 3, the proportion of persons re-
incarcerated at one year post-release
increases from 10.2% in 2000 to 17.3% in
2005. After two years post-release, 20.1%
of the 2000 release cohort and 29.3% of the
2005 release cohort were re-incarcerated.
Release Year 1 Year 2 Year 3 Comparing re-arrest, re-conviction, and
2000 2001 2002 2003 2004 2005 re-incarceration recidivism (across release
cohorts) at yearly intervals subsequent to
the initial release supports the finding that recidivism rates in Tarrant County have increased. In other words, the data
show that proportionately more and more persons within an annual release cohort have been re-arrested, re-convicted
and re-incarcerated. Mann-Whitney U tests were conducted to determine the statistical significance of the increase in the
proportional recidivism rate between the 2000 and 2005 release cohorts. Results confirm statistically significant
differences in the re-arrest, re-conviction, and re-incarceration rates at year one, year two, and year three post-release,
between the 2000 and 2005 release cohorts (p ≤ .001 for all nine analyses).
Recidivism at Four Years Post Release F igu re 4. Re c id ivis m at ye ar s fou r pos t - re le as e .
Figure 4 reports recidivism at four years post- 60%
release for release cohorts 2000-2004. At least 50%
of each respective cohort were re-arrested. Over a
third of all persons annually released were re- 45%
convicted. One-third of the 2000 and 2001 release
cohorts were re-incarcerated. For the remaining
cohorts, 2002-2004, slightly over one-third of the 30%
individuals released were re-incarcerated within
2000 2001 2002 2003 2004
Re-arrest Re-conviction Re-Incarceration