NCJ Number
67412
Date Published
1977
Length
38 pages
Annotation
VARIOUS METHODOLOGIES FOR PREDICTING FUTURE INMATE POPULATION IN THE STATE OF FLORIDA CORRECTIONAL SYSTEM ARE ANALYZED AND COMPARED FOR ADVANTAGES AND LIMITATIONS IN THE FACE OF RECENT UNPREDICTABILITY TRENDS.
Abstract
IN THIS 1977 RESEARCH STUDY, THE FLORIDA DEPARTMENT OF OFFENDER REHABILITATION DESCRIBES ITS EFFORTS IN TRYING TO DEVELOP MORE EFFECTIVE METHODOLOGIES FOR THE PREDICTION OF THE FUTURE INMATE POPULATION. THE EARLIEST ATTEMPTS TO MAKE POPULATION PREDICTIONS WERE THE PAST PERCENTAGE INCREASE PROJECTIONS AND THE LINEAR REGRESSION METHOD. LINEAR REGRESSION HAD THE ADVANTAGES OF AN EASILY-GATHERED AND AVAILABLE DATA BASE AS WELL AS REASONABLE ACCURACY, GIVEN A STABLE AND LINEAR HISTORICAL TREND. THE LATTER WAS SUDDENLY BROKEN, HOWEVER, WHEN THE INMATE POPULATION STARTED GROWING AT AN UNUSUALLY RAPID PACE AFTER 1974. OTHER METHODS WRE CONSIDERED BY THE DEPARTMENT, INCLUDING MODELS THAT SIMULATED SOME OF THE WORKINGS OF THE CRIMINAL JUSTICE SYSTEM. THESE MODELS, HOWEVER, ENABLED PREDICTIONS OVER SHORT PERIODS OF TIME ONLY AND WERE THUS UNACCEPTABLE FOR THE DEPARTMENT'S LONG-RANGE PLANNING OF NEW FACILITY AND CAPITAL OUTLAY. NEW METHODS, SUCH AS THE SIMULATED LOSSES/ADMISSIONS MODEL, WHICH BREAKS THE PREDICTION DOWN INTO RELEASES AND ADMISSIONS, AND USES A MULTIPLE REGRESSION EQUATION FOR PREDICTING ADMISSIONS, ARE CURRENTLY BEING USED BECAUSE THEY MAKE PREDICTIONS POSSIBLE FOR A LONGER TIME SPAN. A SIMULATION OF THE WORKINGS OF THE ENTIRE CRIMINAL JUSTICE SYSTEM (PATTERNED AFTER THE MARYLAND MODEL) IS CURRENTLY BEING CONSIDERED BY THE DEPARTMENT IN ORDER TO IMPROVE STILL FURTHER THE ABILITY TO PREDICT ADMISSIONS.UNFORTUNATELY, THE PAUCITY AND FREQUENT UNRELIABILITY OF FLORIDA COURT DATA PRECLUDE THE USE OF THIS METHODOLOGY AT THE PRESENT TIME. UNTIL THIS OBSTACLE IS REMOVED, THE CURRENT METHODS USED TO PREDICT THE FUTURE INMATE POPULATION OF THE FLORIDA CORRECTIONAL SYSTEM CAN BE REGARDED AS A GREAT IMPROVEMENT OVER PAST METHODS.