There are lots of individuals, communities and civil society groups calling for more people to be included in the data landscape. Amidst accelerating change in uses of data that affect how we live, society is in new uncharted and often unregulated territory without much say about what the future looks like. A recent video by Connected by Data video explains it well: "Data and AI are changing our world. Let's decide how, together."
At the ODI, we started our participatory data programme based on understanding the urgency of making data systems more inclusive and participatory, and aim to be a part of this process. In our work, we have encountered participation emerging in lots of different ways, which is creating excitement and success. We’ve seen how participation can increase data literacy and trust in institutions as well as support the development of data infrastructure to solve social needs (like crowdsourced information about lived experiences of city accessibility). Participation is also increasingly a part of data governance frameworks such as AI impact assessments. But what is the status of participatory data now? How much participation is there in our data ecosystems, and what kind of impact is it making? In a diverse and fragmented landscape we have found that it’s quite hard to say, but generally speaking we agree with our partners at Aapti Institute, whose research concluded that participation remains marginal. However, there is still a great diversity of brilliant initiatives that highlight the successes of participation. In this piece we look at the state of play of participation, starting here in the UK. The world of participation is vast, with lots of exciting initiatives around the world which would be impossible to cover in one piece. The participatory data programme is international by nature, and we’d like to explore participation in other contexts in the future. Here we offer three initial and exploratory reflections about where ‘participatory data’ is today in the UK.
‘Participatory data’ is diverse, serves many functions, and comes from many different places
Looking across what we call ‘data layers’, we have found that there are many different ways that people participate. At the data level, people can be involved in the creation, maintenance, use and sharing of data. For example, there are lots of projects which generate data with a ‘citizen view’ for use in science and policy making, like this example tackling air pollution in Reading, there is a campaign about sewage polluting rivers in England and Wales, and a project to understand where in the city of London needed to be better lit. At the organisational level, people can be involved in the governance of organisations that collect, maintain and share data. For example, there continue to be efforts to develop data trusts here in the UK, such as one in Brixham, as well as plenty of involvement of people and patients in decisions around access to health data like the Liverpool Civic Data Cooperative. At the policy level, people can be involved in determining the ‘rules of the game’, in the UK examples include activism, like the campaign to overturn the use of algorithms in A-Level results as well as bringing people into the development of data policy like the national level citizen juries on data sharing in the Covid-19 pandemic. Across this ecosystem, there is a diversity across various different factors:
- Parts of the data ecosystem. Are people participating in the creation and design of data, in decisions about how data is used, in policy about data, a mixture of these or in some other way.
- Who initiates participation. Is the initiative top-down (a government consultation or a scientific institution soliciting volunteers), or bottom up (activists presenting evidence to create change or develop their own initiative), collaborative, or somewhere in between? Who needs to support the initiative for impact to be achieved?
- Different names. You may know about ‘citizen science’ - where people create data about air or water pollution, species diversity, to have their voices heard in scientific monitoring and policy making. Or of the consultations during the Covid-19 pandemic helping the government to understand the public’s attitude to sharing their data in the efforts to combat the emergency. Or you may be aware of crowdsourced mapping initiatives where people contribute data about things like wheelchair accessibility in the city. These very different examples, using different terminologies, across quite different spaces of expertise, and are all part of what we understand to be ‘participatory data’.
- How ‘embedded’ vs ‘experimental’ are the participatory initiatives. Are they demonstrating the power of citizen voices, or emerging technologies that have yet to demonstrate how they can work to incorporate participants on a wider basis - or are they currently functioning to bring people into data ecosystems on a wider and sustainable basis.
- Who benefits, and how? Are the benefits mainly for the organisation soliciting participation with the aim of creating legitimacy, better targeted products or services? Are the benefits on a community level and in that case, for whom? Do participants themselves benefit from the process and outcomes of participation?
A lot of participation is ‘experimental’ rather than systematically embedded
Participation has a long history in the UK, and has now long been common across all different types of environmental governance, and has played a role in planning and architecture for several decades. Some even argue that environmental monitoring would not be where it is today without the work of millions of volunteers. However, many of the examples of participation in the data ecosystem that we come across are emerging and experimental, unsurprisingly corresponding to a world where the ways we use data are changing. As our digital world becomes more sophisticated, we are able to better build interfaces for people to participate across all different areas. Technology can now better support participation and enable better accessibility, while the proliferation of the smartphone means more people can now participate online (2020 research suggested that 84% of UK adults had a smartphone for private use). It’s not only technologies that are changing, though. Data also appears to be experiencing the ‘deliberative wave’ across policy-making, where participation is being employed to tackle complex policy problems like governance decisions during the Covid-19 pandemic, the climate crisis or infrastructure investment decisions. There is a lot of hope riding on these types of approaches, breaking down the divide between the technocrats and citizens in an increasingly technologically driven world, but these approaches are often in the experimental phase. With these major changes, we need to ‘keep up’ to understand all the prospects of participatory data in developing new forms of inclusion amongst new forms of data and its changing importance in society. Just as regulation must keep up with the pace of change of technology, the ways we think about involving people in the data ecosystem must adapt as well.
There is a lot to think about in making participation impactful
We know that participation isn’t inherently a good thing. Across the world of participation Arnstein's Ladder of Participation is commonly used to assess interventions, demonstrating the spectrum from tokenistic efforts to those that empower people. There is real concern about such approaches being co-opted for ‘participation washing’, which recognises the extractive logic behind many of these tokenistic efforts. Researchers at the Ada Lovelace Institute developed their own framework explaining the spectrum of participatory approaches to data based on Arnstein's Ladder. This Framework highlights that any organisation involving people in decision-making should consider how far people are truly able to impact data ecosystems. For those interested in participation from across different sectors and industries, there is a broad awareness of additional concerns about the fairness, effectiveness and impact of participatory data approaches. The degree to which people are involved in participatory practices is important, and there are additional lens to view participation through, including:
- Representation. Who is able to participate, and do they represent the people who are at risk from exclusion, discrimination, or who might benefit most from having a say? As above, with the world of data and opportunities for participation rapidly changing along with technological change, there is a question of how to ensure that digital transformation does not leave people behind.
- Motivations, incentives and remuneration. How are people compensated for their contributions when they participate in data practices? Why do people participate and how can incentives be used to make fairer and more representative data systems? What is the motivation for an organisation or government to implement these approaches?
- Sustainability of participation. Will people continue to participate over time, or are we creating ‘one-off’ experiences?
- Experiences of participation. What do people learn, do they experience a sense of empowerment themselves?
Conclusion
In so many diverse circumstances in the UK, participation appears to be working to generate excitement about improving data systems and bring people closer to decision-makers using data and data-enabled systems. But just because there are many examples, does not mean the data systems themselves have yet achieved a satisfactory level of inclusion, and the level of participation and inclusion in data systems is generally quite low here in the UK. To make data more participatory, we’ve seen calls from different parts of the data ecosystem, for:
- Political parties to include participation in their manifestos for the upcoming 2024 UK general election, such as to develop structures that include people in government use of data (like a ‘data republic’, as it was termed in this blog).
- Data intermediaries between institutions and citizens that help make citizen involvement fit the institutional needs and vice versa.
- Participation by design, where the developers of technology take responsibility for embedding participatory processes into the development, design, and deployment of new products and systems
We are interested in bringing together the different voices of this world and learning about how to join up knowledge, efforts, and experience and be a part of moving data ecosystems to a more participatory ‘culture’ all across the board. If you’d like to work with us on participatory data, or have any thoughts about this work, please get in touch at [email protected].
A major thanks to the experts who we spoke to about the participatory data landscape in the UK for this blog and provided their valuable reflections - in particular Reema Patel and Michelle Mackie, Research Directors at Ipsos, Muki Haklay, Professor of Geography at UCL, and Gefion Thuermer, Research Fellow at King College London and now Head of Research at the ODI.