THE STATISTICAL ANALYSIS TESTS A MODEL THAT VIEWS THE DECISION TO ENGAGE IN PROPERTY CRIME AS A RATIONAL ONE BASED ON A PERSON'S SELF-PERCEIVED ABILITY TO ATTAIN PERSONAL GOALS THROUGH LEGITIMATE AND ILLEGITIMATE ACTIVITIES AND THE INDIVIDUAL'S RELATIVE PREFERENCES FOR THESE ACTIVITIES. IN THE CASE OF VIOLENT CRIMES, SUCH AS ASSAULT, THE PERSON IS UNABLE TO MAKE THIS DECISION, AND THE RESULTING FRUSTRATION MANIFESTS ITSELF IN AGGRESSIVE BEHAVIOR. UNEMPLOYMENT DATA (FOR ONTARIO PROVINCE ONLY) AND CRIME STATISTICS ON MALES ARRESTED OR SUMMONED MONTHLY IN TORONTO FOR PROPERTY AND VIOLENT CRIMES INDICATE A CLEAR RELATIONSHIP BETWEEN ECONOMIC CONDITIONS AND BOTH PROPERTY CRIMES AND CRIMES OF VIOLENCE, PARTICULARLY FOR MALES AGED 16 TO 20 YEARS. THE RELATIONSHIP BETWEEN UNEMPLOYMENT AND CRIMES OF VIOLENCE FOR YOUNGER MALES IS MUCH STRONGER THAN ANTICIPATED, A FINDING SAID TO SUPPORT THE MODEL WHILE CASTING DOUBT ON THE VALIDITY OF THE SUBCULTURAL THEORIES OF VIOLENCE. DIRECTIONS FOR FURTHER EMPIRICAL TESTING OF THE MODEL ARE SUGGESTED. TABULAR DATA ARE INCLUDED. (LKM)