THE MODEL TRANSFORMS THE MULTIPLE SENTENCING OBJECTIVES FOUND IN SEVERAL SURVEYS OF JUDGES INTO A SINGLE, QUANTIFIABLE SENTENCING CRITERION WHICH CAN BE UNIFORMLY APPLIED TO ALL TYPES OF OFFENDERS IN A NONDISCRETIONARY MANNER. THE MODEL IS THEN USED TO STUDY THE EXTENT TO WHICH ACTUAL PATTERNS OF SENTENCING DISPARITY ACROSS JURISDICTIONS CAN BE EXPLAINED BY FACTORS WHICH ARE LEGITIMATE IN TERMS OF THIS SOCIAL INVESTMENT THEORY MODEL. FIVE SEPARATE REGRESSION STUDIES ARE PRESENTED: (1) A STUDY OF AVERAGE TIME SERVED BY ADULT FELONS IN THE UNITED STATES, BY STATE, FOR EACH OF SEVEN OFFENSES; (2) AVERAGE TIME SERVED IN JUVENILE CORRECTIONAL FACILITES, BY STATE; (3) AVERAGE SENTENCES OF AND TIME SERVED BY COUNTY JAIL AND CORRECTIONAL CAMP INMATES CONVICTED IN THE LOWER COURTS OF CALIFORNIA, DATA GIVEN BY COUNTY FOR EACH OF SEVEN OFFENSES; (4) AVERAGE SENTENCES OF INMATES OF NEW YORK COUNTY JAILS AND PENITENTIARIES, BY COUNTY, WITH SEPARATE REGRESSIONS FOR WHITES AND NONWHITES AND FOR MALES AND FEMALES; (5) A STUDY OF THE RELATIVE FREQUENCY OF INCARCERATION AS OPPOSED TO PROBATION IN THE LOWER CRIMINAL COURTS OF MASSACHUSETTS, BY INDIVIDUAL COURT. IT IS FOUND THAT SOME OF THE SENTENCE VARIATIONS CAN BE EXPLAINED BY THE MODEL. THE PAST CRIMINAL HISTORY IS HIGHLY SIGNIFICANT, AS ARE THE COSTS OF INCARCERATION IN THE STATE AND FEELINGS ABOUT THE DETERRENT VALUE OF THE SENTENCE. HOWEVER, THESE FACTORS DO NOT EXPLAIN ALL OF THE CROSS JURISDICTIONAL SENTENCING VARIATIONS. IT IS ARGUED THAT SOME SORT OF NORMATIVE MODEL WOULD BE ADVANTAGEOUS. THE DISCUSSION IS HIGHLY MATHEMATICAL. CHARTS AND TABLES PRESENT REGRESSON ANALYSIS DATA. APPENDIXES PRESENT FURTHER STUDY DATA AND REFERENCES. (GLR)