NCJ Number
213498
Journal
Journal of Criminal Justice Volume: 34 Issue: 1 Dated: January/February 2006 Pages: 57-72
Date Published
January 2006
Length
16 pages
Annotation
This case study used a local jail forecast method to explore the inherent difficulties involved in the long-term forecasting of local jail bed needs.
Abstract
Overall, forecasting jail bed space needs emerged as feasible over the short-term and inherently flawed over the long-term. Forecasting of correctional populations was revealed as empirically valid for, at most, 1 to 2 years. The long-term forecast of the bed space needs of a local jail revealed four inherent difficulties in long-term forecasting of correctional bed needs: (1) the inherent volatility of local jail populations; (2) a general lack of jail forecasting data; (3) a lack of reliable long-term lead indicators of jail populations; and (4) jail population forecasts are often self-destructing prophesies as lawmakers and practitioners scramble to ensure the forecast does not come to fruition. The authors concede, however, that long-term forecasting of jail bed space needs is not useless as it can help inform policy as long as it is recognized as one possible scenario and not as a forgone conclusion. The case study involved a forecasting demonstration of municipal correctional bed space need for the Orange County Florida Corrections Department, conducted in late 2000. Two common empirical forecasting approaches--ARIMA and autoregression, were employed to construct four July 2000 forecast models with population projections through December 2009. Figures, tables, notes, references