Community Buy-in and Participatory Design in Data Governance

This is the eighth post for the Community Data Hubs Documentation series. This series will document the thought and conversation trajectories within the process of creating the building blocks of our Community Data Hubs model and OEDP’s broader data stewardship work. The first of these blogs will document the progress of the Community Data Hubs Advisory Group, which is working alongside OEDP to tackle conceptual questions related to the model, including social and technical infrastructures, stewardship, and community data.

This post documents the sixth meeting of the Community Data Hubs Advisory Group on July 26, 2023. For this meeting, we asked members of the Advisory Group to read sections of an Open Data Institute (ODI) report documenting a pilot “data trust” program run in collaboration with the Greater London Authority (GLA) and the Royal Borough of Greenwich (RBG). The pilot sought to assess the feasibility of creating data trusts in relation to two specific data projects: one involving mobility data meant to inform parking policies in the borough and another involving energy use data meant to improve energy efficiency in its public housing. Interestingly, the report recommends “exhaust[ing] the options of other data-sharing arrangements” before pursuing data trusts, with the authors finding data trusts unworkable in this case in part because of limited community buy-in. Advisory group members spent much of their time identifying why this might have been the case and discussing what we might learn from this example as we build our model for Community Data Hubs.

Our conversation centered on three major questions:

1. What kind of design practices can cultivate trust among stakeholders within data stewardship infrastructures?
2. Given the great diversity of data projects, how can we build data governance procedures that fit the parameters of a specific project?
3. How do we define “community” and the “data” we hope to accumulate?

What kind of design practices can cultivate trust between various stakeholders within data stewardship infrastructures?

The ODI report was candid that project leaders encountered some challenges cultivating trust in the community, indicating that one should not “assume that if people know the benefits of data-sharing they will share and accept those benefits”. As group members considered what data governance procedures might cultivate trust in this and other projects, a few best practices emerged, especially around participatory design. All parties, whether they will be creating or consuming the data, should be involved in the process of articulating project goals and procedures. Group members noted that many projects — including many environmental data projects — start with the needs of eventual data consumers, be they city planners or environmental regulators. This leaves those involved in these projects with the task of pitching a pre-articulated data project to community members who may feel they are being asked to join as resources rather than co-stewards.

There are also broader social and political variables that shape the possibilities for trusting relationships among the various parties involved in a data project. In this case, the project was affiliated with the city and municipal government, begging the question: What positive or negative experiences might shape how community members respond to a data design effort that is headed by those bodies? How do we resolve legitimate fears or misunderstandings between government and community members? One group member indicated, for instance, that such questions might be especially fraught in the case of the energy-use data. Multiple variables may shape the encounter, such as whether people living in public housing feel a sense of obligation to the local government that could be wrongly mobilized as a recruitment tool. Given the low-income status of the public housing residents, the group wondered how to negotiate the steep power imbalances at play between city officials and these prospective project participants. Navigating these types of hierarchies and accounting for existing tensions within a given social fabric are essential to building trust.

Knowing that the work of building trust takes time, one group member noted that a broader culture change may be necessary in government and beyond to make these kinds of projects feasible. Currently, government agencies tend to marginalize data stewardship as supplemental, as one part of the work of say, city planners or administrators, who have various other professional responsibilities. Indeed, ODI listed low capacity as one of the limiting factors in getting a data trust up and running. Instead, this group member noted, we need to advocate for dedicated positions whose work is primarily data stewardship. Given the complexity of the social and technical processes involved, and the remarkable power data can wield, this is a key step to modernizing state infrastructure.

Given the great diversity of data projects, how can we build data governance procedures that fit the specific parameters of a given project?

Echoing themes from past advisory group meetings (see XYZ), group members reiterated that there is a wide spectrum of data projects and that where a particular project lands on that spectrum shapes the relevant questions around cultivating trust. On the one hand, some projects are largely inwardly looking in their orientation, hoping to contribute to a process of sense-making or self-discovery within a community. Others are outward facing, seeking to document a phenomenon that is well-known to community members for an outside, perhaps skeptical, audience, such as policymakers. Yet other projects incorporate components of both.

One group member noted that many environmental data projects are outward facing. In such cases, trust issues tend to flow in the opposite direction as those encountered by GLA and RBG. Rather than community members responding skeptically to an existing government initiative, it is often regulatory agencies who respond with doubt to data accumulated by community members seeking to document environmental harm. The processes for cultivating trust in such cases may look quite different, though there are definite overlaps. Sometimes technical fixes can help: as when agencies like the EPA allocate funding to study the problem themselves or pay for a community to access EPA-approved instruments. In other cases, building trust may require work more akin to political advocacy/organizing, as community members and aligned parties lobby relevant policymakers or regulators.

How do we define “community” and the “data” we hope to accumulate?

Group members asserted the importance of asking big framing questions as part of designing ethical data stewardship plans. Not only challenging in their scope, such questions surface the multitude of complex political forces at play in data governance. For instance: who is the community with whom a given data project interacts? In this example, the project followed the municipal boundaries of Greenwich. Group members indicated though that proximity or co-residence in the same municipality do not necessarily constitute community, in any meaningful sense. Are there genuine social bonds that connect the people of Greenwich? Which people specifically? Who is left out of those lines of connection?

Adding an additional layer of complexity, group members indicated that participatory design should fundamentally shape what constitutes “data” in the context of this research program. In other words, democratic control of data necessitates a process of negotiating the diverse inquiries, philosophies, even ontologies, of those involved. For instance, one group member elaborated an example of a community science initiative in New Zealand that sought to test the water quality of streams in one corner of the country. Māori participants in the design process, though, disagreed about how “steam health” was being defined by their partners steeped in the environmental sciences. Participatory design means we must revisit basic questions about the framing devices, cultural paradigms, disciplinary silos, and ideological commitments that shape our inquiries.

Insights and questions to revisit during the CDH Co-Design Process:

1. Data stewards should try to incorporate all parties involved in a given data project in an early and iterative design process — be they contributors of data, those shepherding a database, or the eventual intended users of that data. We can build trust by pairing transparency with real democratic processes that allow all stakeholders to participate. This is an ideal which may encounter various constraints in practice, whether that be a distrustful community member or a skeptical regulatory institution.
2. Our CDH model must recognize that to open data governance to democratic control is to allow data projects to shift course, potentially dramatically. Participatory design is not supplemental or additive but transformational.
3. Data projects need to take on the philosophical and paradigmatic scaffolding upon which their inquiry is built. This means unearthing the unexamined complexities just below the surface in some of our most commonly used terms. A word like “community” can paper over a mountain of political complexities regarding the direction and intensity of social bonds between different people in a given place or between places. A descriptive phrase like “stream health” may smuggle in culturally or disciplinary-specific understandings of water wellness that are out of step with the ontological tenets of Indigenous stewards. CDHs need to examine the constituent terms and concepts they mobilize in building data governance.