An investigation into racial bias in court case outcomes in England and Wales

在英格兰和威尔士，亚裔黑人和少数族裔(BAME)公民在刑事司法系统的几乎每个阶段都有过多的代表，对黑人公民来说尤其如此。拟议研究的主要目的是利用新获得的治安法官和刑事法院数据集，增进对刑事司法系统内种族差异的了解。了解这些差异，并找出可能的解释，是解决这些问题的关键一步。拉米审查显示，BAME在刑事司法系统中的过度代表每年花费约3.09亿美元。因此，在经济上和道德上都有必要更好地理解并最终减少BAME在系统内的过度代表。Lammy强调了BAME社区对刑事司法系统缺乏信任。最近在英国举行的黑人生命也是种族不公正的抗议活动，是由长期存在的种族不平等助长的。在一定程度上，这是BAME公民对警察和刑事司法系统缺乏信任的结果。缺乏信任意味着BAME被告更有可能提出无罪抗辩，如果被判有罪，这将增加他们的刑期，进一步加剧不平等。与此同时，BAME被告的定罪率较低。这可能意味着BAME嫌疑人被指控的证据标准较低。这也可能是因为BAME被告更有可能在刑事法院受审，与地方法院相比，刑事法院的定罪率较低。拉米的建议之一是“要么解释，要么改革”的原则；如果对明显的差异没有基于证据的解释，那么应该引入改革来解决这些问题。拉米指出，人们仍然倾向于忽视种族群体之间的差异，认为这些差异可能是由其他因素造成的，比如一个群体的年龄特征，而不是种族本身。目前关于刑事司法系统内潜在偏见的现有统计数据还不够充分。他们往往不会控制其他因素。拟议的研究项目将解决这些问题，在统计上控制治安法官和刑事法院数据中可获得的其他因素，如年龄和犯罪行为，但遗憾的是，无法控制无法获得的变量，如宗教。该项目将包括以下内容：将使用不同的治安法官和刑事法院的数据集进行适当的回归分析，以审查种族与案件结果之间的关系，并控制数据集中现有的其他因素。此外，该项目将审查种族和性别是否与病例结果相互影响。由于相同的人将多次出现在每个数据集中，而且被告嵌套在法院内，因此可能会有多层次建模的情况。将使用SQL编程来生成每个被告一行的数据集，其中包含其他变量，例如每个被告的起诉次数。这一数据集将提供给其他研究人员，将用于进一步调查BAME被告较低的定罪率。该项目将审查种族问题：同一被告在一段时间内多次被起诉的程度(重复使用)；以及每个被告被定罪或不定罪的案件比例。这项计划亦会考虑到其他因素，例如罪行种类等，并考虑到其他因素，例如罪行类型，以及考虑到被告更可能在刑事法院受审的情况下，比较两者的结果。

Black Asian and Minority Ethnic (BAME) citizens in England and Wales are overrepresented at almost every stage within the criminal justice system, this is particularly so for black citizens. The primary aim of the proposed research is to use the newly available magistrates' and Crown Court datasets to increase understanding of racial disparities within the criminal justice system. Understanding these disparities, and identifying potential explanations, is a crucial step in addressing them.The Lammy Review revealed that BAME overrepresentation within the criminal justice system costs around 309 million per year. Hence, there is an economic as well as a moral imperative to better understand and ultimately reduce BAME overrepresentation within the system.Lammy highlights the lack of trust in the criminal justice system amongst BAME communities. The recent Black Lives Matter racial injustice protests in Britain, were fuelled by long-standing racial inequalities. They were, in part, a consequence of lack of trust in the police and criminal justice system amongst BAME citizens. Lack of trust means BAME defendants are more likely to plead not guilty, if then found guilty this will increase the severity of their sentence, further fuelling inequality. At the same time, conviction ratios are lower for BAME defendants. This may mean that BAME suspects are being charged on a lower standard of evidence. It may also be due to BAME defendants being more likely to be tried in the Crown Court, which has a lower conviction rate compared to the magistrates' court.One of Lammy's recommendations is the principle of 'explain or reform'; if there is not an evidence-based explanation for apparent disparities, then reforms should be introduced to address them. Lammy points out that there continues to be a tendency to dismiss disparities between racial groups by suggesting they may be due to other factors, such as a group's age profile, rather than ethnicity itself. Currently available statistics relating to potential bias within the criminal justice system are not sufficiently developed. They tend not to control for other factors. The proposed research project will address these issues, statistically controlling for other factors that are available within the magistrates' and Crown Court data, such as age and offence, unfortunately it will not be possible to control for variables that are not available, such as religion. The project will include the following:Appropriate regression analyses will be conducted using the separate magistrates' and Crown Court datasets to examine relationships between race and outcomes of cases, controlling for other factors available in the datasets. In addition, the project will examine whether race and gender interact in relation to case outcomes. As the same people will appear in each dataset on a number of occasions and as defendants are nested within courts, there may be a case for multilevel modelling. This will be used if appropriate.SQL programming will be used to produce a dataset with one row per defendant with additional variables, such as the number of prosecutions per defendant. This dataset, which will be made available to other researchers, will be used to further investigate lower conviction ratios for BAME defendants. The project will examine race in relation to: the extent to which the same defendant is prosecuted on multiple occasions over a time period (repeat use); and the proportion of cases for each individual defendant which result in a conviction or otherwise. This would be broken down for the magistrates' and Crown Court, throwing further light on the lower conviction rates for Crown Courts.The project will also compare outcomes of triable either way cases in the magistrates' and Crown Court, taking account of other factors, such as offence type, and consider the effect of this in relation to BAME defendants being more likely to be tried in the Crown Court.