STATISTICS ON TOTAL REPORTED OFFENSES, STATE PRISON POPULATION, STATE PRISON ADMISSIONS, AVERAGE SENTENCE LENGTH, AND PROBABILITY OF IMPRISONMENT AFTER CONVICTION ARE USED TO ANALYZE JUDICIAL BEHAVIOR IN THE STATES OF GEORGIA, TEXAS, AND MISSOURI. JUDGES IN BOTH TEXAS AND MISSOURI ARE LIKELY TO IMPOSE LONG PRISON TERMS WHICH KEEP THE PRISON POPULATION HIGH AND RESULT IN A STEADY STREAM OF INSTITUTIONAL ADMISSIONS. GEORGIA'S MORE FLEXIBLE JUDICIAL POLICIES RESULT IN FLUCTUATING PRISON ADMISSION STATISTICS AND VARYING PRISON POPULATION LEVELS. A MATHEMATICAL MODEL IS THEN USED TO FORECAST LONG-TERM BEHAVIOR OF THE CRIMINAL JUSTICE SYSTEM IN EACH STATE AND TO ESTIMATE THE SEPARATE INCAPACITATION AND DETERRENT EFFECTS OF EACH STATE'S SENTENCING POLICIES. THE POLICIES OF ALL THREE STATES SHOWED COMPARABLE PER CAPITA CRIME CONTROL POTENTIAL BUT GEORGIA SPENT LESS MONEY PER CAPITA ON INCARCERATION THAN EITHER TEXAS OR MISSOURI. THE SIMILARITY IN DETERRENCE INDICATES IT IS THE CERTAINTY OF PUNISHMENT, NOT THE SEVERITY, WHICH EXHIBITS THE GREATEST CRIME-CONTROL POTENTIAL. IN ADDITION, STATES WHICH HAVE HIGHLY SUBOPTIMAL JUDICIAL POLICIES HAVE THE GREATEST POTENTIAL FOR IMPROVED CRIME-CONTROL EFFECTIVENESS IF SENTENCING POLICIES ARE OPTIMIZED. THIS DISCUSSION INCLUDES CHARTS, TABLES, AND REFERENCES. (GLR)