Making Community Punishments more Effective

Sentencing research has been traditionally restricted by limitations on data availability. Researchers have had to rely on aggregate date from the Ministry of Justice, or one-off surveys of judges such as the Crown Court Sentencing Survey which is now outdated (data collection ceased in 2014). The new linked administrative databases released by ADR UK will permit researchers to explore a much wider range of sentencing questions and to greater depth.
This project will make an important contribution to the literature on Effective Sentencing. Specifically, the analyses will explore the relative effectiveness of different requirements attached to a community order. Courts have the discretion to impose a range of requirements on offenders serving community orders, yet almost nothing is known about which conditions or combinations of requirements, are most effective in reducing re-offending. Sentencers are effectively sentencing in the dark, relying on the individual experiences and intuitions of judges and magistrates.
Furthermore, this dataset will enable the analysis of hitherto under-explored aspects of sentencing that carry potential consequences for social inequalities. In particular, this project will examine whether the number, nature, and onerousness of requirements varies according to the ethnicity of the offender. If, for instance, the requirements imposed on BAME defendants differed from those imposed on White defendants convicted of the same offence, then this could, in turn, help us understand differences in re-offending rates. Given the great desire by central government to make sentencing more consistent, effective, cost-effective and fair, it is anticipated that this research will have an important impact on policy and practice in the field.
Given the absence of empirical research on this area, there is scope to produce impactful findings just from simple descriptive statistics. However, the PhD student will have the opportunity to be trained in two areas of advanced statistical analysis, data visualisation and longitudinal data analysis. Creating intuitive data visualisation tools will be key to be able to summarise and convey large amounts of findings in a way that is meaningful to non-statistically trained audiences. Similarly, the longitudinal dimension of the dataset and the research questions, will require the student to be trained in advanced statistical techniques like sequence analysis and growth curve and autoregressive models. Many of these techniques are currently covered in the MSc in Data Analytics organised by the CDT Data Analytics & Society. Those techniques that are not covered there will be learn through participation in specific short courses such as those delivered by LIDA Societies, the Consumer Data Research Centre, or external courses such as those advertised by the National Centre for Research Methods or the Royal Statistical Society.