Långsiktiga prognoser av platsbehovet på fängelser
The article discusses whether it is possible to make reliable predictions of prison populations five to ten years in advance. This period corresponds approximately to the planning period that the Swedish Prison and Probation Service considers necessary for building new prisons. After a brief descrip...
Saved in:
Main Author: | |
---|---|
Format: | Article |
Language: | Danish |
Published: |
De Nordiske Kriminalistforeninger
2010-08-01
|
Series: | Nordisk Tidsskrift for Kriminalvidenskab |
Subjects: | |
Online Access: | https://tidsskrift.dk/NTfK/article/view/137439 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | The article discusses whether it is possible to make reliable predictions of prison populations five to ten years in advance. This period corresponds approximately to the planning period that the Swedish Prison and Probation Service considers necessary for building new prisons. After a brief description of forecasting and its various methods, three quantitative forecasting methods are discussed: trend projection, extrapolation from causal models, and statistical modelling. Swedish and international prison population data (comprising years 1961 ta 2010) are used as illustrations. It is concluded that accurate long-term forecasts are unlikely, since prison population data tend to show strong random walk patterns. |
---|---|
ISSN: | 2446-3051 |