Developing methods to measure sentence severity and assess disparity in sentencing in the criminal courts of England and Wales

Wallace, Stephanie and Francis, Brian and Humphreys, Leslie (2015) Developing methods to measure sentence severity and assess disparity in sentencing in the criminal courts of England and Wales. PhD thesis, Lancaster University.

Text (2015wallacephd)
2015wallacephd.pdf - Published Version
Available under License Creative Commons Attribution-NonCommercial-NoDerivs.
Download (1MB)

Abstract

This doctoral research focuses on sentencing, and more specifically, disparity in sentencing in the courts of England and Wales. Previous research in this area has introduced bias into the samples by focusing solely on custodial sentences and therefore failing to consider the more common non-custodial sentences. This thesis overcomes this methodological problem. There are two parts to this research: the first part develops a sentence severity scale based on the principle of proportionality, and includes the full range of sentences meted out by the courts in England and Wales. The second part of this research then uses the new scale to assess disparity based on the extra-legal factors of the case. Data from the Offender Assessment System (OASys) was merged with data from the Police National Computer (PNC), which provided information on the offence and resultant sentence, as well as information relating to the socio-demographic characteristics of the offender. Goodman Row Column Association analysis was used to model the association between the offences and sentence categories. This model was then extended to control for three legal factors - offence plea, previous history of offending, and the number of offences the offender was sentenced for - which must all be taken into consideration when sentencing offenders. This analysis provided a series of scores for each of the sentence categories allowing them to be arranged by magnitude, ranging from the least to most severe sentence. To anchor the scale and make the scores less arbitrary, linear interpolation was used. This estimated the equivalent number of days in custody for the non-custodial sentences using the modal days in custody for each of the custodial sentence categories. The new sentence severity scale was treated as the dependent variable in which to model sentencing disparity using multilevel modelling. Here a number of legal and extra-legal variables were considered to try and explain the large amount of sentencing variation between offenders. After controlling for the legal factors of the offence this research found there to be significant sentencing disparity. Consistent with previous research, female offenders were sentenced less severely than male offenders, and White British and Irish offenders were also sentenced less severely than ‘like-situated’ Black offenders. However, offenders in the White other category were sentenced less severely than the White British and Irish group. This research also found that unemployed offenders were sentenced more severely than those in employment prior to sentencing, although the research did find that those on benefits were sentenced less severely. Finally, offenders who do not live in permanent accommodation were also sentenced more severely. The results of this research suggest significant disparities in the sentencing of offenders in England and Wales.