THE ANALYSIS SHOWS THAT POSTCOMMITMENT DESTINY IS NOT THE CONSEQUENCE OF ANY SINGLE FEATURE OF WHAT YOUTHS BRING WITH THEM TO THE INSTITUTION OR OF WHAT THEY FIND IN THE INSTITUTION'S PROGRAMS. IF THE NEWCOMER BRINGS A SERIOUS OFFENSE HISTORY TO A LARGE INSTITUTION THAT HOUSES MANY LONGTIME RESIDENTS, THE PROSPECTS FOR REHABILITATION ARE NOT PROMISING. IF THE HISTORY INCLUDES ONLY MINOR OFFENSES AND THE OFFENDER ENTERS A SMALLER CORRECTIONAL PROGRAM, THE PROSPECTS ARE BETTER. OTHER COMBINATIONS FALL BETWEEN THESE TWO PROGNOSES. A KIND OF PEER INFLUENCE APPEARS TO BE AT WORK. THIS INFLUENCE IS NOT JUST THAT OF ONE INDIVIDUAL UPON ANOTHER, BUT RATHER IS PROGRAMWIDE AND DISTINCTIVE FOR EACH MAJOR TYPE OF PROGRAM. PROGRAM CULTURES ARE TRANSMITTED FROM EACH GENERATION OF YOUTHS TO THE NEXT. THE PROGRAM RETAINS ITS IDENTITY IN THAT THE BASIC CONDITIONS THAT DETERMINE PROGNOSIS -- SIZE, PREVIOUS OFFENSE HISTORY OF ITS YOUTH, AND NUMBER OF VETERANS -REMAIN FAIRLY CONSTANT. THE NATURE OF PEER INFLUENCE APPEARS TO BE DETERMINED LARGELY BY THESE BASIC CONDITIONS, AND PEER INFLUENCE IS ONE DETERMINANT OF PROGNOSIS. SUPPORTING DATA ARE INCLUDED. (AUTHOR ABSTRACT MODIFIED--LKM)