Kommentar till Peter Lindström
2008 (Swedish)In: Nordisk Tidsskrift for Kriminalvidenskab, ISSN 0029-1528, Vol. 95, no 2, 173-178 p.Article in journal (Other academic) Published
In a rejoinder to a paper (Lindström, 2008, in Swedish) that claims that a10 percent increase in the number of prisoners in Sweden decreases the number of homicides by 5 percent during the subsequent year, the authorshows that this claim is a methodological and conceptual artefact. Using data from Swedish conviction statistics (1981-2006), it is shown that the temporal distribution of the incidence of homicide is identical for perpetrators sentenced to custodial and non-custodial sanctions, respectively. Since the average stay in Swedish prisons is short (ca. 5 months), lasting incapacitative effects are impossible to obtain given the temporal distribution of Swedish homicide events. The author joins the ranks of scholars criticizingthe use of regression modelling as a superior research strategy.
Place, publisher, year, edition, pages
2008. Vol. 95, no 2, 173-178 p.
fångtal, brott, mord, dråp, kriminalpolitik
Law and Society
Research subject Criminology
IdentifiersURN: urn:nbn:se:su:diva-30211OAI: oai:DiVA.org:su-30211DiVA: diva2:242097