THE MODELS PARTS CONSIST OF THREE ALGORITHMS. THE FIRST, ASSESSOR, ASSESSES THE PREFERENCES OF CITIZEN, POLICE (OR OTHER SERVICE PERSONNEL), AND ADMINISTRATOR INTEREST GROUPS FOR EFFICIENCY AND QUALITY OF SERVICE USING MULTIATTRIBUTE UTILITY FUNCTIONS. THE EVALUATOR COMBINES UTILITY FUNCTIONS WITH PERFORMANCE ESTIMATES FROM THE HYPERCUBE QUEING MODEL (OR OTHER ANALYTIC MODELS) TO EVALUATE ALTERNATIVE DISPATCH RULES (RULES ASSIGNING SERVICE UNITS TO CALLS FOR SERVICE). THE IMPROVER ALGORITHM GENERATES IMPROVED SERVICE AREAS AND ACCOMPANYING DISPATCH RULES. THE MODEL TAKES INTO ACCOUNT A VARIETY OF PERFORMANCE MEASURES, MOST NOTABLY TRAVEL TIME, WORKLOAD, AND NUMBER OF INTERSECTOR DISPATCHES, AND THEIR TRADEOFFS FOR EFFICIENCY AND EQUALITY. DATA FOR INTEREST-GROUP PREFERENCE WERE COLLECTED IN INTERVIEWS WITH REPRESENTATIVES OF POLICE, CITIZENS, GRADUATE STUDENTS COGNIZANT OF PRACTICAL AND ANALYTICAL ISSUES OF EMERGENCY SERVICE SYSTEM DESIGN, AND LOCAL GOVERNMENT. EACH PART OF THE DECISION MODEL IS ILLUSTRATED BY APPLICATION TO THE POLICE SECTOR DESIGN PROBLEM. THE MODEL WAS DEVELOPED AS PART OF THE NATIONAL SCIENCE FOUNDATION INNOVATIVE RESOURCE PLANNING IN URBAN SAFETY SYSTEMS RESEARCH PROJECT WHICH FOCUSED ON EVALUATION CRITERIA, ANALYTICAL TOOLS, AND IMPACTS UPON TRADITIONAL METHODS, STANDARDS, ROLES, AND OPERATING PROCEDURES. APPENDED ARE THE QUESTIONNAIRES USED FOR DATA GATHERING AND A BIBLIOGRAPHY AND REFERENCES. THE ALGORITHMS ALSO ARE INCLUDED. (AUTHOR ABSTRACT MODIFIED--DAG)