The custodial remand population has reached a 50-year high. However, we lack detailed knowledge on the reasons why it is considered necessary to remand defendants into custody. There are four broad exceptions to the presumption of bail: it is believed the defendant may not turn up to their trial, may commit another offence while on bail, may obstruct the course of justice (e.g., witness intimidation), or may pose a risk to their own welfare (Bail Act 1976). Decisions on remand and bail are made in the magistrates’ court, yet previous research suggests that magistrates rely significantly on information and recommendations from the police. Little is known about how the police interpret the exceptions under the Bail Act (1976) and make those recommendations. In particular, we do not know how these exceptions are interpreted by practitioners, how the risks are calculated, which exceptions are most relied upon, the impact of other factors, and how this relates to the defendants’ later custodial journey and related outcomes.

The Data First project links administrative records on defendants’ journeys from court into custody, including which exception under the Bail Act was relied upon when someone is remanded into custody. This pilot study will analyse synthetic datasets from Data First in order to demonstrate and refine the methodology before the real data can be accessed, enabling identification of large-scale trends in the use of the different exceptions to the right to bail. Additionally, this study will conduct interviews with police officers to bring insight into the rationale behind the officially recorded reasons.