Benefit vs. cost of Alaska criminal justice programs

A shorter version of this article appeared in the Winter 2018 print edition.

In October 2017, the Alaska Justice Information Center (AJiC) released its Alaska Results First (RF) report on Alaska’s adult criminal justice programs. The report found that approximately $20.58 million in state funds were invested annually in 19 programs whose effectiveness has been evaluated by academic studies and rigorous reviews. Using Alaska-specific inputs, including program costs, recidivism patterns, and criminal justice system costs, along with national criminal justice data from the Pew-MacArthur Results First Initiative, Alaska RF provides a benefit cost analysis of the state’s investment in evidence-based programs.

Alaska Results First Initiative: Compile — Cost — Compare. This graphic illustrates the three stages of the Alaska Results First benefit-cost process. The "Compile" stage involves compiling a program inventory. "Cost" includes analyzing the programmatic costs and programmatic benefits for each program. "Compare" involves using the benefit cost model to evaluate each program. Courtesy Alaska Justice Information CenterBenefits and costs

The benefit to cost ratio, or monetary return on the state’s investment in adult criminal justice programs, was calculated by comparing program costs with costs avoided by a program’s ability to reduce recidivism. Avoided costs — or benefits — include avoided future criminal justice costs and avoided future victimization costs.

In order to calculate a program’s ability to reduce recidivism, AJiC needed to understand Alaska’s patterns of recidivism without the programs. To do this, AJiC collected information on all convicted offenders released from Alaska Department of Corrections (DOC) institutional custody in 2007. Because of the date of release, these individuals had likely not participated in the evidence-based programs.

Calculating recidivism

AJiC developed nine cohorts from among the offenders released in 2007. These cohorts were made up of groups of offenders who would have been eligible to participate in the evidence-based programs. Table 1 shows the characteristics of offenders selected for each cohort.

Table 1. Cohorts in Alaska’s Results First Model. This table shows the nine cohorts used in the model, the number of offenders in each cohort, and the participant selection criteria for each cohort. [Note a. All cohorts were based on offenders discharged from DOC facilities during 2007, after an incarceration stay for an original criminal offense. Offenses associated with the incarceration stay were used to qualify the offender for a cohort. (See Valle, 2017, Appendix E.)] Cohort name: Prison (GT120) (N= 1,081); Participant selection criteria: Stay associated with a felony conviction; Incarcerated for more than 120 days. Cohort name: Probation (LTE120) (N=1,279); Participant selection criteria: Stay associated with a felony conviction; Incarcerated for less than or equal to 120 days. Cohort name: GT120 Prison Mix (N=1,200); Participant selection criteria: Incarcerated for more than 120 days; 900 (75%) randomly selected from offenders whose stay was associated with a felony conviction; 300 (25%) from those whose stay was associated only with misdemeanors. [Note b. The 75% felon and 25% misdemeanor mix was based on the distribution of offenders in the PsychEd program.] Cohort name: Sex Offender (N=197); Participant selection criteria: Stay associated with a sex offense (excluding failure to register as a sex offender; Male offender. Cohort name: Felony DUI (N=353); Participant selection criteria: Stay associated with a felony DUI conviction; Offender had at least one prior DUI conviction. Cohort name: Misdemeanor DUI (N=533); Participant selection criteria: Stay associated with a misdemeanor DUI conviction; No felony offense associated with this stay; Offender had at least one prior DUI conviction. Cohort name: Drug Court (N= 527); Participant selection criteria: Stay associated with a felony alcohol or drug offense; Stay NOT associated with an unclassified or A-level felony, a homicide or an offense involving drug distribution [Note c: Based on rules set for the Anchorage Wellness court.] Cohort name: Mental Health Proxy (N= 5,000); Participant selection criteria: Random sample drawn to match most serious offense distribution found among FY15 Mental Health Court participants. Cohort name: Domestic Violence Proxy (N= 2,325); Participant selection criteria: Stay associated with a DV-associated statute [Note d: Based on analysis of offenses with DPS DV conviction flag in a DPS 2012 arrest conviction data set. (See Valle, 2017, Appendix E.)]; Male incarcerated for less than or equal to 120 days.The cohorts were tracked for eight years following their release from DOC institutional custody in 2007. AJiC used information from the Department of Public Safety to determine when individuals in the cohorts had been arrested for a new crime that resulted in a conviction. This information made it possible to compute the percentage of those who recidivated from the release date through the end of each year during the follow-up period, or the cumulative recidivism rate.

AJiC used national data for evidence-based programs similar to those in Alaska to estimate the recidivism reduction rate that could be expected if individuals participated in Alaska’s programs. The criminal justice administration costs and costs to victims that would be avoided due to this recidivism reduction were also computed. This “benefit” was then weighed against the program’s costs to arrive at a benefit cost ratio.

New information from Alaska RF

The Alaska RF report provides a wealth of new information to policymakers including eight-year recidivism patterns for the nine cohorts of offenders, measures of how effective a program may be at reducing recidivism, and how changing cost structures and delivery methods may impact the benefit to cost ratio of programs. The RF model may also be used to assess the benefit to cost ratio of new programs — providing an estimate of how a new program would impact recidivism and its return on investment using Alaska criminal justice costs.

In the following article, Araceli Valle, author of the Alaska RF report, discusses how tracking offenders for eight years for the RF project is adding to our understanding of recidivism in Alaska.


Valle, Araceli. (2017). Alaska Results First Initiative: Adult Criminal Justice Program Benefit Cost Analysis. Anchorage, AK: Alaska Justice Information Center, University of Alaska Anchorage.

PDF of this article