Data Governance Models

##California ##transparency ##accountability ##NewEngland ##Australia ##waterdata ##fisheriesdata ##LatinAmerica ##usability #risk #extractivism #datatrust #datacooperative

What are they?

Environmental data is an essential tool for monitoring and evaluating environmental issues. Navigating these often complex issues requires context-specific data practices to ensure accuracy, accessibility, responsible use, and overall collective benefit of the data. Environmental data stewards can use a data governance model to contextualize their data practices.

Data governance models are frameworks or sets of practices that outline how an organization will manage and use their environmental data. They encompass the relational, legal, and technical aspects of the data lifecycle, from collection to dissemination, in order “to minimize risks, ensure accountability, and optimize data assets” (The World Bank).

The goal of data governance models is to reduce harm, avoid extractive practices, and maximize benefits by defining the responsibilities and methods of managing data throughout its lifecycle. Some of the key principles that data governance models work to achieve are transparency, stewardship, and privacy and control.

Transparency is key to accessibility and building trust. Data transparency involves being open about the procedures in place for handling data. This provides teams with the information they need to properly collect and manage data, and external users with context they may need to understand and responsibly use the data.
Stewardship ensures accountability by defining and assigning responsibilities for how the data will be used, managed, and shared. Stewardship provides structure and accountability for data workflows and sharing (Aapti Institute).
Privacy and control are fundamental parts of any governance model to ensure all identifying private information is kept safe, from collection to dissemination.

These principles have helped develop many models, each with a different approach and focus, but that can be easily adjusted to the needs of an organization, and serve to enhance the management of data, from collection to dissemination. Below are some commonly accepted data governance models:

Data Trusts are legal frameworks that assign the role of data steward to uphold and maintain data rights. The data subjects—those whose data is being collected—define the role of data steward and establish their expectations, goals, and standards for the data. The steward is expected to make decisions regarding the data with the established rights as their guidelines to ensure accountability in decisions. In addition, trusts increase the control the data subjects have over their personal data and who has access to it. However, there is potential for power imbalances and unclear or poorly defined interests when the data subjects are not all on the same page. Overall, data trusts are still a fairly new form of governance for environmental data.

For more information on the application of data trusts, see:
- Australia's National Farmers Federation (ANFF) has developed the Australian Farm Data Code in order to provide farmers control over how their data is being used and shared.
- For information on adapting data trusts for climate action, read Enabling Data Sharing for Social Benefit Through Data Trusts: Data Trusts in Climate.
- For an example of how data trusts can be used to maintain data rights among groups of people in New England, read Building A Fishermen-First Data Ecosystem.
- To learn more about the feasibility of using data trusts for climate action read Co-designing data trusts for climate action by Vinay Narayan and Joe Massey.
Data Collaboratives or Cooperatives emphasize the importance of multi-stakeholder management and ownership to establish shared data collection, sharing, and processing practices. The goal of a data collaborative is to drive innovation and derive collective benefit for all. They also help foster trust between organizations and ensure the overall safety of the data by defining the data sharing process.

For more information on the application of data collaboratives and cooperatives, see these examples of working models:
- The California Data Collaborative (CaDC) is a coalition of water utilities working to “revolutionize the data infrastructure of California's water management… to automate the collection and analysis of metered water use from participating agencies. This information will allow for the creation of a more accurate dataset that details how much, when and where water was used by California residents.” (The Data Economy Lab)
- PescaData is a cooperative platform model that enables fishing communities in Latin America to collect more accurate data to ensure sustainable fisheries.
- GovLab’s Data Collaboratives Explorer documents many different types of collaborative and cooperative models in the environmental sector.

Note

When establishing a new data governance model, consider these questions:

What is your goal of establishing a data governance model?
Are there policies and procedures already in place to control user data access, limiting what data can be accessed depending on job roles and responsibilities?
Are there standards for data integration and information exchange?
- For more questions and guidance on how to select the right model, visit ITRC’s Environmental Data Management Best Practices and Aapti’s Data Stewardship Explorer.

Why does it matter?

Well-designed data governance models promote #accessibility, #trust, and #usability of environmental data, and can prevent creating data silos caused by uncoordinated methods of data management. Data governance models can support project teams in developing common #management practices that increase accessibility and usability, thus supporting understanding and engagement among communities from which the data was collected. Additionally, a well-designed data governance model includes transparent risk minimization processes; this can foster trust among different actors or users within the project, especially when users are kept well-informed about changes.

**Mentioned and additional resources:

For an introduction to the importance of data governance models and examples, including data trusts, guilds, and collaboratives, see Data Governance Models and the Environmental Context: Part 1.
To understand the differences between individualized control models and shared control models, see Equitable Data Governance Models for the Participatory Sciences.
To review the benefits and challenges of four models—data sharing pools, data cooperatives, public data trusts and personal data sovereignty—see Emerging models of data governance in the age of datafication.
To explore the considerations needed to apply a data governance model to an environmental project, see Data Governance Models and the Environmental Context: Part 3.
For a discussion on the best approaches to governing data for “greater social and environmental impact,” see the GovLab and The Democratic Society’s report, Governing the Environment-Related Data Space.
To read Aapti Institute’s findings from research with the Omidyar Network on understanding data stewardship, see Data Stewardship: A Taxonomy.
To explore how different data governance models were used across 6 different countries in the Americas, see Open Data and Climate Change: Experiences from the Americas.