On this page find:
Overview of the Logic Model
NIJ has developed a logic model for adult drug court programs that court administrators and their partners who want to examine the performance of their drug courts may find useful. The logic model can help clarify the best way to use resources and what long- and short-term outcomes drug court teams should consider measuring.
The logic model has six components:
- Inputs — financial, staff, equipment and other resources invested to support the program.
- Activities — structured services intended to deliver what is necessary to achieve objectives.
- Outputs — observable and measurable events resulting from program implementation.
- Short-term outcomes — immediate changes realized especially during program participation.
- Long-term outcomes — changes realized after program participation.
- External factors — conditions outside the program that affect implementation and outcomes.
The logic model components can be tied to program goals (e.g., promote public safety by treating drug-dependent convicted persons); program objectives (e.g., reduce recidivism by implementing risk/needs assessments, graduated sanctions and incentives and other drug court services); and performance measures, such as the following:
- Increase percentage of drug court participants who reduce substance use while in the program.
- Reduce percentage of drug court participants who reoffend while in the program.
- Increase the percentage of drug court participants who complete program requirements (excluding financial obligations necessary for graduation).
- Increase total number of drug court graduates.
- Reduce percentage of drug court participants who reoffend within one year after completing the program.
The logic model also can guide evaluations of drug court programs:
- A process evaluation documents a program's actual caseflow, service delivery and resources in relation to its planned target population, policies and procedures over time.
- An outcome evaluation measures the program's influence on graduation, criminal recidivism and relapse among cohorts of participants.
- An impact evaluation gauges the intervention's effect on the target population; using equivalent information available on comparable people outside the program, it contrasts participant outcomes while controlling for characteristics (e.g., criminal history) that may alone predict those outcomes (e.g., recidivism).
Cost-efficiency analysis indicates what impact the program has on public resources and expenditures and whether the investment yields savings over the status quo or an alternative.
- Read an NIJ Journal article on cost-benefit analysis that includes finding from an analysis of adult drug-courts.
- Read Cost-Benefit Analysis: A Guide for Drug Courts and Other Criminal Justice Programs, an NIJ Research in Brief.
Terms Used Frequently in Cost-Efficiency Analysis
Standing. A primary concern for evaluation is the population the program is designed to target. In cost-efficiency analysis, this is referred to as the standing — whose costs and benefits are being assessed?
Cost-efficiency: cost-effectiveness versus cost-benefit. Assuming that a program has the intended impact (e.g., it reduces recidivism), cost-efficiency analysis systematically assesses the value of costs and benefits to indicate how the program affects resources. Specifically:
- Cost-effectiveness analysis uses dollars to rank programs by their relative efficiency in producing the desired outcome. Suppose Program A costs $100 for a 10 percent reduction in recidivism, Program B costs $100 for a 15 percent reduction, and Program C costs $150 for a 20 percent reduction. A cost-effectiveness ratio is total cost divided by units of outcome effect, ranking the programs this way: B = 6.7, C = 7.5, and A = 10. (Note that the monetary value of the outcome is not estimated.)
- Cost-benefit analysis calculates the values of all costs and all benefits to determine the net benefits — total benefits minus total costs. Comparing net benefits for program participants with a group of nonparticipants can help determine whether the program investment yields a savings over the status quo. In addition to overall costs, cost categories or outcomes should be assessed separately; for example, higher treatment costs may be offset by lower recidivism costs.
Cost to taxpayers. Using taxpayer or government costs focuses the analysis on public resource allocations for the jurisdiction that will bear the brunt of the fiscal costs and receive the majority of the social benefits. This is a more narrow scope than a review of all tangible (e.g., job growth) and intangible (e.g., perceived safety) costs and benefits that accrue to taxpayers, neighbors, participants, competing organizations or others affected by the program.
Variable, fixed and step-fixed costs. Fixed costs, such as rent and utilities, remain the same for a period and are not easily affected by workload changes, whereas variable costs, such as overtime and supplies, relate directly to workload. Step-fixed costs remain constant up to a point and then change if the workload exceeds or falls below certain thresholds (e.g., increased staff salaries and benefits to manage higher caseload).
Average versus marginal costs. Marginal costs are incremental changes in total operating costs that result from practice or policy changes, such as adding case management services. Average costs include facility and other fixed costs that the program may not affect, and they are the same for the status quo.
Opportunity costs and resources. One cost to assess is loss of opportunity for alternate activities due to the program. For example, participants cannot work during periods when mandatory court hearings and service appointments are scheduled. Opportunity resources are the benefits realized as a result of the program, such as staff and resources normally spent on status quo activities (e.g., warrant service) that can be redirected to other activities (e.g., relapse prevention).
Cost avoidance. One benefit to assess is the cost savings that result from the program's preventing or reducing behaviors, such as costs avoided because fewer property crimes were committed than would normally occur in the program's absence.
Proxy measures, revealed preferences and shadow prices When true outcome measures are not easy to obtain, proxy measures with similar values may be acceptable substitutes. Self-report information may be collected to assess criminal behavior, but sometimes only data on rearrests and convictions are available. These data may underestimate actual behavior for some (undetected or unreported) crimes. When true costs are not easily estimated, researchers sometimes use jury trial awards that indicate through "revealed preference" how the public values different crimes given property and personal victimization. They may also use shadow prices or value assumptions based on similar events when actual events (e.g., arrests avoided) cannot be observed.
Discount rate Considering the time horizon over which the program and long-term outcomes occur, future costs and benefits need to be discounted to obtain a present value (in today's dollars). Because the value of a dollar decreases over time, future program benefit estimates are discounted to present value and can be compared with investment costs borne today. The higher the discount rate, the lower the present value of future benefits. Programs that require a multiyear funding commitment may also discount future costs to the present value; for example, they may use an annual inflation rate of 3 percent.
Top-down versus bottom-up estimation. Top-down estimation uses a gross total cost figure and assumes that program participants contribute equally to costs and benefits, at least on average. A bottom-up approach compiles detailed operations information that accounts for time spent by staff on different services and according to use by different program participants. It estimates costs at the individual level based on price-by-quantity calculations for each program output (e.g., drug tests, hearings, treatment), and for program outcomes that vary in severity (e.g., crimes involving victimization) and system consequences (e.g., probation, prison).