Four recommendations for building trust in multistakeholder governance of the data economy
The 2021 World Development Report, Data for Better Lives, argues that a multistakeholder approach to data governance is best suited to govern complex data ecosystems in ways that are transparent and inclusive and that reflect the interests of all stakeholders. This approach is an essential component of the ‘trust framework’ that strengthens the social contract around data use.
Participants in an online panel discussion in September discussed multistakeholder governance models. The panel discussion brought together speakers from the World Bank, Wellcome Trust, the Wildlife Conservation Society, the GovLab and the Global Partnership for Sustainable Development Data to(available to watch here) explore how effective multistakeholder governance creates transparency and accountability and empowers people and communities.
Here’s four recommendations from the discussion for building trust in data governance:
1. Multistakeholder governance processes must be designed to handle a broad range of views to enable meaningful participation. Participation can be messy, but it must lead to effective decisions and outcomes. However, consensus among all stakeholders is not always possible due to diverging interests. Building consensus and buy-in on the process of designing governance models can ensure that people who disagree with the final decisions will nonetheless trust the approach adopted to reach those decisions. Basic principles, safeguards, and ground rules can contribute to achieving buy-in on decision making processes and ensuring that these are seen as fair despite potential disagreement on the final outcomes.
For example, challenging existing health data governance practices, the Wellcome Trust developed a new “learning data governance” model. This model adds two steps to the traditional data lifecycle aimed at building trust in the process and increasing representation of different views in a sustainable way. According to this approach, the outcomes from the use of data by third parties (whether positive, negative, or neither) are reported back to those who granted access to the data. Citizens’ panels or forums are then able to scrutinize and learn from previous decisions. This introduces a loop of accountability into data governance and keeps scrutiny of outcomes separate from data access decisions which increases trust for individuals and data subjects. The approach also helps improve the quality of decision-making by using new information on outcomes from past projects and the dynamic views gleaned through the participatory process.
2. Voices of stakeholders with less power and fewer resources must be effectively included to ensure that power imbalances do not affect participation. Governance processes and partnerships include negotiations among people and communities with differing interests, resources, and power. Recognizing these dimensions is a first step to including those who may have an equal stake in data governance but less power and fewer resources. To achieve better inclusion, stakeholders with less power and resources must be given the opportunity to meaningfully contribute in defining expected outcomes that reflect their specific needs and aspirations. Enabling this will require targeted outreach efforts and support.
Citizens assemblies, like those organized by the GovLab in New York during the COVID-19 pandemic, can involve communities in discussions around how their data should be used, taking into account people’s needs and aspirations. In New York City, partnering with libraries was pivotal to ensure that assemblies were representative of local communities and perceived as trustworthy by people at risk of being marginalized.
3. Rules must be clearly set and made explicit from the beginning, so they can be contested and even re-negotiated if necessary. Rules do not need to be set in stone—they can be adaptive and even tested over time before being confirmed or discarded. However, even the testing period should be clearly defined, so that stakeholders’ expectations of the rule-making process are aligned.
4. Innovate beyond outdated multistakeholder governance models that originated in previous eras. Stakeholders today have access to new tools and more data to establish innovative governance models. More experimentation is needed to push the boundaries of what multi-stakeholder governance means and broaden the range of participants beyond those who have traditionally had seats at the table, including those who cannot be physically present at negotiations.
The 100 Questions Initiative of the GovLab challenges the way in which data governance is often approached by switching the focus from data to the key questions that data should address.
There is no one-size-fits all approach to building trust in multistakeholder governance models. Setting up successful governance arrangements requires understanding the specific context, interests, and power relations of stakeholders involved. Processes must be guided by needs at the ground level. When done well, these processes can lead to governance models with improved long-term sustainability that are equipped to empower all relevant stakeholders.
The discussion highlighted a range of approaches. The Wellcome Trust, for example, uses deliberative research and citizens’ juries to explore people’s needs and views around third party use of health data and identify the most relevant governance models to make those available. The GovLab, on the other hand, works in all its projects to establish specific social licenses for reusing data which take into account the expectations and values of stakeholders and indicate what can and cannot be done with the data to steer responsible reuse.
Quick wins are difficult to achieve. Setting up multistakeholder governance solutions requires time and energy. Nonetheless, effective multistakeholder governance can change people’s mindsets concerning the value of data and increase data literacy, creating more meaningful engagement and participation in data decision making.
The panel drew on the expertise of the following experts and practitioners. You can watch the discussion here.
- Adele Barzelay, Counsel, Data and Digital Development (LEGOP), World Bank
- Dr. Natalie Banner, Understanding Patient Data Lead, Wellcome Trust
- Mariana Varese, Director of Amazon Landscapes & Citizen Science for the Amazon Project at the Wildlife Conservation Society (WCS), Amazonia
- Stefaan Verhulst, Co-founder and Chief Research and Development Officer, The GovLab
- Victor Ohuruogu, Senior Africa Regional Manager, Global Partnership for Sustainable Development Data