As part of Involve’s work with the Liverpool Civic Data Co-operative (CDC), we asked Reema Patel, formerly of the Ada Lovelace Institute and now Head of Deliberative Engagement at Ipsos, to bring her expertise in participation in a data cooperative environment. We are pleased to publish this blog post sharing some of her thoughts and conclusions, as well as a longer provocation paper she wrote to help stimulate our thinking. Involve’s final recommendations to the CDC on the next steps to take in building a participative approach to data stewardship will be available shortly.

Reimagining the Civic Data Cooperative

In January 2020, the UK’s Liverpool City Region’s combined authority announced new plans for a civic data cooperative (CDC), led by the University of Liverpool, to enable the analysis of anonymised health and social care data. The aim of the CDC is to benefit both society and the economy by aligning this work with public and societal expectations. It will do this through innovative citizen participation and the principles of cooperativism. 

The need for effective data stewardship has been brought into sharp relief by both the pandemic and several key events which highlight the need for critical reflection about the trustworthy use of health data. Key among these events are GPDPR and the liberalisation of the sharing of more confidential patient data across the NHS and partner organisations

However, the field of data stewardship is new and emerging; Liverpool is leading the way in exploring how to ensure effective data stewardship for large populations. It was with this in mind that I worked with Involve and the University of Liverpool to think through what the role of participation might be in the delivery of a ‘civic data cooperative’

This blog post draws on a provocation paper I developed for that purpose. It focuses primarily on how best to embed participation and engagement, rather than the narrower question of technical design in trustworthy data stewardship models.

1. Why move beyond the technical and legal towards the participatory?

The primary focus on data stewardship models to date has been legal and technical - but, as GPDPR shows, public questions and concerns include who should have access to data, for what purposes and who will make such decisions. Different publics have different concerns, hopes and fears in relation to GPDPR; there is no single, settled public view. As a result, if data stewardship models are to be successful, they need to ensure that such concerns are properly understood and acted on. Implementing civic engagement at different levels will be central to securing trustworthy data models. 

Data cooperatives are often defined legally, primarily through the lens of control over data; a recent Ada Lovelace Institute report describes it as ‘a legal mechanism that gives members more control over their own data’. Whilst this may be accurate, this generates further questions - who are members, and how broadly should they be defined? And what does control practically constitute in this respect? Answering these questions will help us develop a better understanding of good participation and engagement in data stewardship.

2. Broadening the notion of membership

Who is a member? Whilst almost all cooperatives have members, it is entirely feasible that they benefit those who have not become a member as well as those who have. Indeed, a quick scan of existing case studies of data cooperatives that take a member-centric approach highlights how requiring membership of a cooperative risks excluding many of the potential benefits to wider society. 

In the health context this is particularly significant - for instance, patients or residents in the Liverpool City Region should not have to be members to benefit. With this in mind, the broader concept of "beneficiary" is suggested to encompass those who may not have the time or resources to participate, but may be deeply impacted by the use of data. Their views will need to be taken into account just as much as those of formal members of the data trust.

3. Common and shared purpose - enabling cooperation and reciprocity in the use of data 

Trustworthy data governance requires ensuring that people affected by the use of data have active agency when it comes to governing their data rather than being ‘done to’ by the members of the data trust. 

The idea of trustworthy data governance also alludes to data having value that is societal and collective, rather than simply individual or commercial. When Sidewalks Lab proposed its unsuccessful and failed ‘civic data trust’ in Toronto, it was met with criticism precisely because the design of its proposed model largely remained proprietary and ambiguous; and had not adequately engaged city residents and stakeholders in the vision or the model. History teaches us therefore that models that work need to generate a shared consensus about the appropriate use of data. 

One dictionary definition of cooperation offers us a route forwards, turning towards the principle of mutuality and shared common goals and endeavours with reciprocity at its basis:

adjective

involving mutual assistance in working towards a common goal.
"every member has clearly defined tasks in a cooperative enterprise"

The notion of mutuality is understood by Guerini in relation to the Italian approach to social cooperatives, rooted in norms of social solidarity as ‘expanding the concept of mutuality, offering services and benefits for people who were not necessarily associated with the cooperatives.’ Here, reciprocal benefits aren’t just by members towards other members, but rather, by the cooperative towards wider society’. Effective public engagement will be the mechanism by which the cooperative and its members can better understand how wider society conceives of these benefits. 

4. Different points for participation in data governance - data stewards can engage in different ways

There are three key points at which co production and public engagement can be particularly valuable in the life cycle:

  1. Before: To help shape and inform the design of the cooperative, working closely with technical and policy stakeholders around scoped and potential options and working models. There must be sufficient information to permit ‘intelligent consideration’, as well as adequate time to incorporate reflections into the design of the initiative.
  2. During: Institutionalised as part of the mechanisms of the data cooperative itself (this is likely to involve people in how the data itself is curated, collected and used and decisions made by the cooperative) and,
  3. After: In enabling data stewards to understand how best to act in ways that use that data effectively, to help aid assessment of how the initiative is working and how it might be able to be improved; and to act as a sense check that the data is being (re)used in the interests of beneficiaries.

There are a range of mechanisms for involving people in the use of their data as set out in the Ada Lovelace Institute report on participatory data stewardship. These range from creating data institutions such as cooperatives, using deliberative mini publics to involve people in questions about how their data can be used and on what terms, as well as a range of approaches that ensure people know what is happening to their data and their rights when it comes to data use. 

Involving people is not a one size fits all approach, only a multitude of approaches can make a data cooperative truly civic.

The author of this piece is writing in a personal and independent capacity, drawing from research undertaken in prior roles.

Read more about our project working with the Liverpool Civic Data Cooperative on the project page.