BECAUSE ONE OF THE CENTRAL INTERESTS IN CRIMINAL STATISTICS IS THEIR RELATIONSHIP WITH REAL EVENTS, IT IS NECESSARY TO INVESTIGATE BOTH POLICE ACTIVITIES AND THEORIES REGARDING THE GENERATION OF BOTH CRIMINAL STATISTICS AND CRIME. IN THE BROAD SENSE, THERE ARE THREE GROUPS OF THEORIES CONCERNED WITH DETERMINING THE LEVEL OF RECORDED OFFENSES: (1) DETERRENCE THEORIES FOCUSING ON CONSEQUENCES, POSSIBLE PAYOFFS, AND PENALTIES OF ACTS; (2) THEORIES OF TRADITIONAL CRIMINOLOGY THAT SPECIFY CERTAIN GROUPS AS BEING MORE PRONE TO OFFENDING; AND (3) THEORIES OF SOME SOCIOLOGISTS THAT CONCENTRATE ON THE RECORDING OF OFFENSES AND THE LABELING OF OFFENDERS. DATA USED IN THE EXAMINATION OF ALL THREE GROUPS ARE BASED ON A CROSS SECTION OF POLICE DISTRICTS IN ENGLAND AND WALES FOR THE YEARS 1961, 1966, AND 1971. CONSIDERATION IS GIVEN TO FACTORS INFLUENCING THE LEVEL OF CRIMINAL ACTIVITY, SOCIAL REACTIONS TO CRIMINALITY, RECORDING AND POLICY ASPECTS OF OFFICIAL STATISTICS, AND THE ECONOMIC APPROACH TO CRIME AND PUNISHMENT. THE SIMULTANEOUS EQUATIONS TECHNIQUE IS APPLIED TO EVALUATE SIMULTANEOUS CAUSATION AND UNOBSERVED VARIABLES, AND THEN INTERRELATIONSHIPS AMONG CRIME, POLICE, AND CRIMINAL STATISTICS ARE EXPRESSED IN A MATHEMATICAL MODEL. FINDINGS INDICATE THAT ALL THE THEORIES EXAMINED PLAYED IMPORTANT ROLES IN UNDERSTANDING RESULTS. THUS, ONE OR TWO OF THE THEORIES SHOULD NOT BE EXAMINED IN ISOLATION. IN ADDITION, CRIMINAL STATISTICS HAVE BEEN CONSTRUCTED BY AGENCIES WHICH SHOULD THEMSELVES, TOGETHER WITH THEIR RELATIONS WITH THE COMMUNITY, FORM PART OF THE SUBJECT UNDER STUDY. SUPPORTING DATA AND EQUATIONS ARE GIVEN. A DATA APPENDIX CONTAINS DEFINITIONS AND SOURCES. REFERENCES, A NAME INDEX, AND A SUBJECT INDEX ARE INCLUDED. (DEP)