THE AUTHOR FIRST DESCRIBES THE RESULTS OF TWO PREVIOUS STUDIES ON JUVENILE DELINQUENCY: 'THE MEASUREMENT OF DELINQUENCY' AND 'DELINQUENCY IN A BIRTH COHORT'. IN THE FIRST, A METHOD OF SCORING THE EXACT SERIOUSNESS OF A CRIME WAS DEVELOPED, WHILE IN THE SECOND, THE PATTERNS OF DELINQUENCY AND NON-DELINQUENCY IN A SAMPLE OF PHILADELPHIA YOUTHS WAS EXAMINED. THE AUTHOR THEN LINKS THE RESULTS OF THESE TWO STUDIES TO DEVELOP A NEW SOCIAL POLICY FOR THE JUVENILE JUSTICE SYSTEM. BASED ON THE RESULTS OF THE EARLIER STUDIES, HE ARGUES THAT LITTLE OR NOTHING IN THE FORM OF SOCIAL INTERVENTION SHOULD BE DONE UNTIL THE THIRD RECORDED OFFENSE OF THE JUVENILE, BUT THAT CUMULATIVE CRIME SERIOUSNESS SCORES SHOULD BE MAINTAINED IN THE FILES OF THE SYSTEM TO BE USED AS A BASIS FOR DETERMINING THE MOMENT AND GRAVITY OF SANCTIONING INTERVENTION. THIS WOULD ESSENTIALLY BE A 'POINT SYSTEM' OF JUVENILE JUSTICE SIMILAR TO THE MOTOR VEHICLE VIOLATION POINT SYSTEM USED IN SOME STATES. THE AUTHOR ARGUES THAT PENALTIES COULD BE APPLIED WITHOUT THE WIDESPREAD DISCRETION AND DISPARITY THAT HAS EXISTED TO DATE. HE MAINTAINS THAT SUCH A SYSTEM COULD INCREASE THE POTENTIAL FOR DETERRENCE BY INCREASING FOREKNOWLEDGE OF SOCIETY'S RESPONSE, AND COULD REDUCE THE INJUSTICE OF JUDICIAL DISPARITIES BY INCREASING CONSISTENCY OF SANCTIONING.