NCJ Number
15371
Date Published
1969
Length
55 pages
Annotation
THIS REPORT DISCUSSES THE CONCEPTUAL FRAMEWORK FOR COST-BENEFIT EVALUATION, PROBLEMS IN THE MEASUREMENT OF COSTS AND BENEFITS, AND THE RELATIONSHIP OF PROGRAM EVALUATION TO THE PLANNING PROCESS.
Abstract
BENEFIT-COST ANALYSIS SHOULD BE CONCERNED LESS WITH AVERAGE BENEFIT-COST RATIOS THAN WITH MARGINAL BENEFITCOST RATIOS, SINCE THE ISSUE TO BE ADDRESSED IS WHAT THE EFFECTS WILL BE OF AN INCREASE OR DECREASE IN FUNDING LEVELS. MEASUREMENT OF BENEFITS AND COSTS MUST TAKE INTO ACCOUNT THEIR DISTRIBUTION AMONG DIFFERENT ECONOMIC AND SOCIAL CLASSES AND THE SOCIAL VALUE OF BENEFITS TO DIFFERENT GROUPS. THE QUANTIFICATION OF NONMONETARY BENEFITS IS IMPORTANT. CONTROL GROUPS ARE DIFFICULT TO OBTAIN. THE MOST RELIABLE CONTROLS WOULD BE A SAMPLE OF INDIVIDUALS QUALIFIED FOR A PROGRAM WHO FOR SOME REASON DID NOT ENROLL. LONGITUDINAL STUDIES ARE VALUABLE FOR OBTAINING CURRENT, RATHER THAN RETROSPECTIVE, DATA ON BOTH BENEFITS AND EXPENSES OVER TIME. CROSS-PROGRAM COMPARISONS ARE EXTREMELY USEFUL, BUT THEIR VALUE AS NATURAL EXPERIMENTS IS LIMITED BY PROBLEMS OF MULTIPLE CAUSALITY AND LACK OF THEORY FOR RELATING PSYCHOLOGICAL VARIABLES TO PERFORMANCE. EVALUATIONS OF TOTAL PROGRAM IMPACT DO NOT PROVIDE THE KIND OF INFORMATION THAT CAN AID IN PROGRAM PLANNING. ALTHOUGH GOVERNMENT FUNDING BODIES ARE CONCERNED WITH TOTAL IMPACT AS A GUIDE TO ALLOCATING RESOURCES AMONG PROGRAMS, PROGRAM MANAGERS WANT RICH INFORMATION ON DETAILS OF PROGRAM OPERATION. IN THE EARLY STAGES OF A PROGRAM, STUDIES SHOULD LOOK AT PROGRAM COMPONENTS AND PROCEDURES, IN THE LATER STAGES EVALUATION OF IMPACT IS SUITABLE. EVALUATION OF INDIVIDUAL LOCAL PROJECTS IS VERY EXPENSIVE AND IS NOT LIKELY TO BE WORTH THE INVESTMENT. (AUTHOR ABSTRACT)