THE MODEL WAS BASED ON THE CONSENSUS IN THE CURRENT LITERATURE THAT PARTICIPANTS IN CIVIL STRIFE ARE DISCONTENT. POLITICAL LEGITIMACY, OR ALLEGIANCE TO THE GOVERNMENT AND THE GOVERNMENT'S POTENTIAL FORCIBLE MEANS AND LEVEL OF PUNISHMENT FOR VIOLENCE, WAS ALSO FELT TO AFFECT THE NATURE AND LEVEL OF CIVIL VIOLENCE. NINE INDICES OF VIOLENCE (ASSASSINATIONS, STRIKES, GUERILLA WARFARE, GOVERNMENTAL CRISES, PURGES, RIOTS, ANTIGOVERNMENT DEMONSTRATIONS, REVOLUTIONS, AND THE NUMBER OF PEOPLE KILLED IN DOMESTIC GROUP VIOLENCE) WERE REDUCED THROUGH FACTOR ANALYSIS TO TWO INDEPENDENT CLUSTERS OF ACTIVITIES: SPONTANEOUS AND NONORGANIZED VIOLENCE AND ORGANIZED VIOLENCE. USING ZERO-ORDER CORRELATIONAL ANALYSIS, RELATIONSHIPS BETWEEN THE INDEPENDENT VARIABLES AND THESE TWO TYPES OF VIOLENCE WERE ESTIMATED. RESULTS SUPPORTED THE CONCLUSIONS THAT (1) SYSTEMIC SATISFACTION, AS MEASURED BY CHANGE IN PER CAPITA GROSS NATIONAL PRODUCT, WAS NEGATIVELY ASSOCIATED WITH POLITICAL INSTABILITY, ESPECIALLY ORGANIZED VIOLENCE; (2) GOVERNMENTS USING MINIMAL OR EXTREMELY LARGE AMOUNTS OF FORCE PREVENTED UNORGANIZED INSTABILITY, BUT INTERMEDIATE LEVELS OF FORCE WERE POSITIVELY ASSOCIATED WITH HIGH LEVELS OF UNORGANIZED VIOLENCE AND ORGANIZED VIOLENCE OCCURRED INDEPENDENTLY OF THE GOVERNMENT'S MILITARY STRENGTH; (3) LEGITIMACY, AS MEASURED BY INCREASING DEMOCRATIZATION, WAS NEGATIVELY CORRELATED WITH ORGANIZED VIOLENCE ONLY WEAKLY BUT ASSOCIATED WITH UNORGANIZED VIOLENCE. ORGANIZED VIOLENCE THEREFORE, APPEARED TO BE RELATED MOSTLY TO DISCONTENT AND TO DECLINING BELIEFS IN GOVERNMENTS' LEGITIMACY; UNORGANIZED VIOLENCE, TO MEDIUM LEVELS OF PUNISHMENT. TABLES, FIGURES, FOOTNOTES WHICH INCLUDE REFERENCES, AND A DESCRIPTION OF THE STUDY METHODOLOGY, ARE INCLUDED. FOR THE ENTIRE VOLUME AND FOR RELATED CHAPTERS, SEE NCJ-63568, 63569, AND 63571. (CFW)