Shaping Data Ethics: A Young People’s Deliberative Forum
11 January 2023
Timothy Lowe Research Lead, Thames Valley Police Violence Reduction Unit ; Police Data Ethics Fellow, Ethox Centre
The police are the public
If you observe any prolonged lengthy twitter conversation involving the police, you are likely to see the phrase ‘the police are the public and the public are the police’ arise at some point. Such a phrase acknowledges a fairly simple truth: that police officers are members of the public first before they take their oath and remain so after taking it.
Fundamentally, the phrase connects UK policing to its roots in policing by consent, where the police acknowledge that ‘the power of the police to fulfil their functions and duties is dependent on public approval of their existence’. In essence, whilst individual members of the public can’t consent or not consent to being policed, there is a fundamental recognition that the activity of the police is only possible with the co-operation of the public.
Ethics and the use of data
The model of policing by consent presents its own questions and challenges, but these challenges perhaps become starker in the context of data use. At its core, as with many other areas, the growth, sheer availability of data, and the potential to exploit that data presents the possibility of a radical change in the way that policing gets done. With this comes many new ethical problems and questions that require new methods of engagement beyond the remit of your standard ethics committee or police public advisory group.
There are a limited number of Data Ethics Committees within policing, and the Thames Valley Violence Reduction Unit (VRU) houses a unique one. As part of its process, the VRU Committee has the option to co-opt wider members to support its decision-making. It is possible for the Committee to open up its doors, in this instance to members of the public, to add a wider array of scrutiny to projects being taken to it. It seeks to build into its process the possibility of public deliberation as part of the process of making decisions and recommendations.
Involving young people to support decisions making
As such, on the 15th November a programme of work run by the Thames Valley Violence Reduction Unit and the Ethox Centre culminated in an event entitled ‘Shaping Data Ethics: A Young People’s Deliberative Forum’. This forum was a pilot of a process where we access wider voices from members of the public in the Committee’s decision-making process. The event took place at the Big Data Institute and involved bringing together over 60 students from Henry Box School, Wheatley Park School, and Cherwell School to discuss a live data use-case currently being presented to the VRU’s Data Ethics Committee.
The specific project we wanted the students to think about was presented to the Committee a couple of months ago and concerns the social contagion of violence. In essence, we wanted to understand whether the students thought the VRU and the Police should use analytical methods to understand how violence might spread through networks of individuals (for which there is already some evidence in America), whether we should intervene if it is found to spread, and what those interventions should look like.
A complex set of problems
The programme of work undertaken with the students recognised that thinking through data ethical questions and analytical models and so on would be a novel, modern, and complex set of problems to engage with. As a result, we wanted to deliver some workshop sessions with the students first, but within their school environment where they felt comfortable to help them to develop the skills to think about an unusual set of issues.
We went in to each school to deliver two workshop sessions. The first workshop involved getting the students to think through issues in police practice as a warm up exercise. They were presented with a BBC article on a stop and search in Liverpool, and an article from the VRU website on Operation Paramount, a scheme that uses data to identify children with parents who have gone to prison and provide them with early support. In the same workshop we provided them with another activity where the students were asked to rank different kinds of data as to how personal they thought it was to get them thinking through data related issues and how they thought about those issues and categories of data together.
The second workshop sought to be more practically oriented and involved just the one activity, which tried to put the students in the position of a public sector professional who is considering whether or not to share information. The students were split into three groups (one acting as the police, one acting as a GP, and the other as the school). They were given a fictional story about a young person’s life and provided with some general information that all of them knew, and specific information that only that agency knew. They had to debate whether they wanted to share their information with the other tables.
Once this decision was made, we then helped facilitate any sharing and they were asked to review their decision. If they felt they wanted to get information from a specific table that they didn’t have they had to come up with arguments to persuade the other table to share their information. At each point of the exercise they were asked to revisit their decisions and whether they were happy with them. Finally, they were given some consequences for their choices (some good, some bad) to discuss. Whilst these were hypothetical consequences, they mirror the often complex outcomes that present themselves to professionals.
The students showed an enormous amount of enthusiasm for the subject matter. In the second workshop in particular, the students became incredibly animated by the exercise, with some making impassioned speeches to other tables on the importance of sharing information.
The deliberative forum
The workshops prepared the students really well to be able to engage with the social contagion use case at the day-long forum on the 15th November and to answer the posed question: Should interventions based on peer network data be implemented in the Thames Valley area?
The day was split into 5 sections to help gain an answer to this question:
- Introduction-students were given a general overview of the day and what is at stake within the social contagion use-case (what data it uses, general principles etc) and asked to develop a general set of values (concerns, things they like etc)
- Facts- students were presented with the relevant facts, in this instance the possible interventions that might be deployed off the back of the research, and provided opportunities for clarification
- Framework- the students considered what they thought mattered to answer the particular question they were being asked
- Recommendations- students developed as tables what they thought they would recommended, and then deliberated as a room discussing their views as a body of people to come to a recommendation
- Panel- the students presented their recommendation to a panel of figures, including Supt Dave Horsburgh, South & Vale Local Police Area Commander, Matthew Barber, Police and Crime Commissioner for the Thames Valley, and Supt Stan Gilmour, VRU Director.
There will be a full report on the event so those who wish to delve into the considerable detail can read it there when published. Suffice to say, however, that the students engaged at an exceptional level with the ideas that they were presented with. They effectively considered pros and cons around things like: who provides the intervention? At what point should we intervene? What data informs the model and does it risk bias? What role does consent play? And so on.
A model that supports community engagement
I think what came out of the event and the workshop sessions for me was the passion with which the students considered the issues. They were able to put themselves in the shoes of public sector professionals tackling really difficult problems and understood why data might be able to help tackle some of those issues. On the other hand, they were also able to debate the nuances of the issues in great detail, pulling out that, whilst they understood the potential benefits of use, there were a number of things that needed to be considered and thought through first and that they were concerned about. Listening to, and talking through these views, is immensely important to understanding the ethical questions we are working through, especially in what can be an immensely charged and emotive area.
Ultimately, these discussions need to be two way, where members of the police service are able to discuss and debate issues with members of the public. The sheer value of such engagement was certainly felt by all on the day.
The public are the police
Ultimately, the work we carried out is an attempt to take the whole phrase ‘the police are the public and the public are the police’ seriously. You could interpret this phrase as a simple statement that the police are drawn from the public. Or, we can take it as a statement that not only are police members of the public, but that the public have a vested interest in the way policing gets done, particularly as data becomes more important. The issues for policing here are potentially starker than other areas, mainly due to the balance of power that exists in the relationship between an officer and members of the public.
Evidently, there are always going to be limitations even to the approach we have taken. We can’t access every voice in the Thames Valley and there will always be questions around representativeness, for example. However, to me, ‘the public are the police’ contains within it the possibility to be proactive; to engage with communities early and to incorporate their thoughts into the way decisions are made around data. That is what we have sought to do here by piloting a model for a committee that supports community engagement as one of its core functions. Yes,’ the police are the public’, but, using deliberative models like this as part of decision-making has the potential to make the public a part of the police.
Shaping Data Ethics – a young person’s view
Hannah Abbas Student, Cherwell School
During this event, I gained an insight into the ethics behind sharing data and policy making. The thing that made this event unique and successful was the element of interaction and communication which I believe should be at the forefront of policy-making to increase efficiency and effectiveness.
Overall, I came to the conclusion that data should be shared between institutions such as schools, hospitals and the police to create the framework for the peer network identification suggested in the proposals presented to us. These social links can be strengthened and made more accurate with the use of social media - which was an idea proposed by students. The sharing of data with the police is important since they are more than just a form of law enforcement, they have a legal obligation to protect society. Therefore, to do this best, maximum information is required to enable the police to tailor their intervention plans.