A RESEARCH PAPER BY ROSS INVOLVING THE EXAMINATION OF EXTENSIVE TIME SERIES DATA FOR BOTH NORWAY AND SWEDEN CONCLUDES THAT THE EFFECTIVENESS OF LAW ENFORCEMENT MEASURES IN CONTROLLING DRUNKEN DRIVING CAN BE DISMISSED AS THE 'SCANDIVANIAN MYTH.' THIS CONCLUSION IS EVALUATED IN LIGHT OF BEHAVIORAL AND CAUSAL INTERRELATIONSHIPS USING AN ANALYTICAL MODEL THAT CONSIDERS THE LEVEL OF MOTOR VEHICLE ACCIDENTS, THE LEVEL OF DRUNKEN DRIVING, AND THE INCIDENCE OF NONALCOHOL-RELATED ACCIDENT FACTORS FOR A GIVEN POPULATION. THE MODEL ASSUMES THAT ACCIDENTS WILL BE POSITIVELY RELATED TO DRUNKEN DRIVING AND NONALCOHOL-RELATED ACCIDENT FACTORS AND THAT THE PRODUCTIVITY OF LAW ENFORCEMENT AS MEASURED BY CONVICTIONS FOR DRIVING INFRACTIONS WILL BE POSITIVELY RELATED TO THE DRUNKEN DRIVING LOAD AND RESOURCES DEVOTED TO LAW ENFORCEMENT. TWO DATA SETS (NORWEGIAN TIME SERIES AND SWEDISH CROSS-SECTON DATA) ARE EMPLOYED TO EVALUATE THE GENERAL HYPOTHESIS THAT DRUNKEN DRIVING IS CONTROLLABLE, I.E., THAT ADDITIONAL RESOURCES DEVOTED TO THE ENFORCEMENT OF DRIVING LAWS WILL DETER DRUNKEN DRIVING AND HENCE REDUCE ACCIDENT LEVELS. IT APEARS THAT VARIATIONS IN ALCOHOL CONSUMPTION HAVE AN APPRECIABLE EFFECT ON ACCIDENT LEVELS IN NORWAY BUT THAT DRIVERS CAN CURB THEIR DESIRE TO DRIVE WHILE UNDER THE INFLUENCE OF ALCOHOL. THE STRONG EFFECT ASSOCIATED WITH CONVICTIONS FOR MOTOR VEHICLE VIOLATIONS SUGGESTS THAT THE ABILITY TO CURB THE DESIRE TO DRIVE IS REINFORCED BY DRUNK DRIVING REGULATIONS AND THEIR ENFORCEMENT. LAW ENFORCEMENT DISPOSITIONS FOR TRAFFIC CASES REPRESENT A BETTER PREDICTOR OF ACCIDENTS THAN CONVICTIONS FOR ALL MOTORING OFFENSES, CONVICTIONS FOR DRIVING UNDER THE INFLUENCE, AND THE PROBABILITY OF LICENSE REVOCATION. CONTROLS ON DRINKING IN SWEDEN, SHORT OF AN ABSOLUTE PROHIBITION OF ALCOHOL CONSUMPTION, DO NOT REDUCE DRUNKENNESS ON THE ROAD. STATISTICS INDICATE THAT SMALL CHANGES IN THE LEVEL OF PER CAPITA ALCOHOL CONSUMPTION HAVE NO MEASURABLE IMPACT ON DRUNKEN DRIVING AND ACCIDENTS, AND THE RELAXATION OF LAWS REGULATING THE SALE AND CONSUMPTION OF ALCOHOLIC BEVERAGES SEEMS TO HAVE LITTLE IMPACT ON ACCIDENT LEVELS. FOR BOTH COUNTRIES, THERE IS SUBSTANTIAL EVIDENCE THAT PROSECUTION FOR DRIVING OFFENSES IS AN IMPORTANT FACTOR IN THE DETERMINATION OF ACCIDENT LEVELS AND THAT LAW ENFORCEMENT DOES SERVE A DETERRENCE FUNCTION. THIS EVIDENCE CONTRASTS WITH THE RESULTS OBTAINED BY ROSS. SUPPORTING EQUATIONS AND DATA AND A LIST OF REFERENCES ARE INCLUDED. DATA SOURCES FOR THE RESEARCH ARE LISTED IN AN APPENDIX. (DEP)