IMPROVEMENT IS DEFINED AS SOMETHING WHICH MUST BE MEASURED AGAINST CLEARLY DEFINED OBJECTIVES, AND THEREFORE, INNOVATIONS SUCH AS TEAM POLICING OR RADIO COMMUNICATION SYSTEMS DO NOT NECESSARILY IMPLY POLICE DEPARTMENT IMPROVEMENTS. IT IS DIFFICULT TO MEASURE POLICE IMPROVEMENT IN THE AREAS OF PEACE AND PROTECTION SINCE THE GOALS ARE VAGUE. IF POLICE IMPROVEMENT IS MEASURED IN RELATION TO CERTAIN TASKS ASSIGNED TO POLICE AGENCIES, THE MEASUREMENT IS HAMPERED BY THE FACTORS OF LEGALISM AND TRADITION. LEGALISM REQUIRES THAT POLICE DATA BE COLLECTED IN A MANNER THAT IS USABLE FOR LEGAL RATHER THAN MANAGEMENT PURPOSES, AND POLICE TRADITION DOES NOT INCLUDE PROFESSIONAL MANAGEMENT TECHNIQUES. POLICE IMPROVEMENT HAS BEEN HINDERED BY A LACK OF IMPROVEMENT IN OTHER COMPONENTS OF THE CRIMINAL JUSTICE SYSTEM. USE OF NEW RESOURCES HAS OFTEN TAKEN THE FORM OF NEW-PROGRAM APPENDAGES ATTACHED TO A TRADITIONAL OPERATIONAL PROGRAMS, RATHER THAN ACTUAL CHANGES IN THE POLICE SYSTEM. IMPETUS FOR POLICE CHANGE IS COMING FROM OUTSIDE THE AGENCIES, AND POLICE DEPARTMENTS ARE BEING MADE ACCOUNTABLE FOR FUNDS SPENT. CONSTRAINTS ON POLICE IMPROVEMENT INCLUDE POOR LEADERSHIP IN THE LOCAL POLICE AGENCIES AND AT THE NATIONAL LEVEL; POOR COMMUNICATION WITHIN AND BETWEEN POLICE DEPARTMENTS AND THE CONSEQUENT LACK OF CONSTRUCTIVE CRITICISM AND DEBATE; AND UNREASONABLY HIGH EXPECTATIONS THAT ARE ESTABLISHED FOR POLICE BY THE POLICE THEMSELVES, THE COMMUNITY, AND GOVERNMENT LEADERS. THESE CONSTRAINTS SHOULD BE REMOVED AND RESEARCH SHOULD FOCUS ON HARD, BASIC ISSUES RATHER THAN ON SUPERFICIAL DEMONSTRATION PROJECTS. EMPHASIS MUST BE PLACED ON ALTERATIONS IN THE BASIC OPERATING PROCEDURES WITHIN AGENCIES; I.E., BASIC ORGANIZATION AND THE FUNCTIONS OF THE VARIOUS PERSONNEL. BIBLIOGRAPHIC REFERENCES ARE PROVIDED. (AUTHOR ABSTRACT MODIFIED--DAG)