[Spotlight] Open considerations for effective Data Partnerships and Good Governance
A Re-cap of ODC’s Implementation Working Group meeting on February 28, 2023
To kick off a brand new ODC IWG year, we asked the Open Data Charter Implementation Working Group members what the most pressing topics were for their communities. The resounding response was around data partnerships and data governance within the context of open data. We strive to quickly address community needs and so, for 2023’s opening presentation last month, we were joined by Andrea Barenque, founder and CEO of Datamorfosis; and Christian Medina, Global programs manager at Open North, to discuss the interconnected nature of data partnerships and good data governance.
Defining Data Governance
As the data agenda has evolved worldwide, so too has the definition of “data governance” along with its various approaches. Andrea reminds us that definitions of governance are much more logical when applied to data. It determines that data is “fit for purpose” by defining the rules that ensures data is collected, used, and shared in the “right way”. For Datamorfosis, effective data governance should answer four key questions:
- What is your vision/strategy?
- What is your data infrastructure?
- What are the roles and responsibilities?
- And what are the processes/tools/standards needed?
In this vein, data governance is no longer just about prepping data for publication. Instead, it sets the basis for everyone to have a minimum threshold for built-in processes that are more sustainable and can adapt easily to the ever-changing data landscape. In fact, a lack of sustainable and adaptable processes is why Andrea believes many data projects fail.
Christian’s take was slightly different. At Open North, they recognize the challenge in reaching a consensus on what data governance means. To face this challenge, their approach has been to build a taxonomy of different discussions around data governance — particularly in the City of Montreal, Quebec, where Open North is based in Canada. They then monitor and evaluate frameworks to determine whether their assumptions are correct. The underlying assumption that they choose to begin with is an understanding that data in and of itself, has no power. Rather, as data is integrated into data governance frameworks and used to shape decisions, that is where it gains said power. With this in mind, when structuring data governance decision-making, it is important to consider:
- Which type of value is pursued by which stakeholder throughout the data life cycle?
- Which stakeholder is enabled or unable to influence which decision?
- How can data governance explicitly center the structurally underrepresented groups that inherently have less power and influence?
Both approaches provide a robust list of considerations for establishing comprehensive data governance frameworks.
Data governance driving Data Partnerships
Open North is among Canada’s leading not for profits, focusing on data governance, open and shared data, open smart cities, public consultation, and open government. Their mission centers on research and solution design, capacity building, and collaboration; while their vision is to enable the responsible and effective use of data and technology in service to transparent, accountable, and inclusive communities.
To achieve these objectives, it has been necessary that Open North goes beyond open data to focus more broadly on value in relation to transparency, economic benefit, and inclusion. Recognizing that the value of openness can often conflict with other values, like privacy protection of vulnerable populations, Open North leverages partnerships as the vehicle for value realization. They define data partnerships as any initiative where two or more organizations align around a common goal and parties engage in the sharing of data to fulfill a given value proposition. In so doing, partnerships become a way to build trust and collaboration, and can help organizations recognize that no data governance framework is the same, as each brings different stakeholders to the table. These frameworks are key to understanding how data is being used. Moreover, thinking about data and society from the perspective of desired civic value means asking:
- What impacts do you want to generate?
- What principles guide your strategy?
- And how can an expanded data governance framework help imbed transparency, inclusivity, and accountability into decision-making?
Evaluating responses to the key questions associated with defining data governance as well as the desired civic value questions, ensures that both our internal objectives and the needs of our partners are reflected in the resulting data governance framework.
Principles in practice
Both Andrea and Christian provided excellent examples of the application of these principles. Datamorfosis specializes in consultancy dedicated to providing services that build data-ready teams, for example supporting the Government of Colombia in establishing their 2021 National Strategy. Their initiative, which was implemented by PIT policy lab, was a textbook case of how to integrate effective data governance with a data infrastructure governance model and implementation roadmap. The Colombian model included seven distinct elements that formed part of their plan:
- Data quality
- Searchability, accessibility, interoperability and reusability
- Data security and protection
- Privacy by design and by default
- Public trust and ethical data management
- Standardization and interoperability
- And strategic sectorization
The model had three levels from strategic, to tactical, to operational to illustrate how each element interacts with the other, while also ensuring alignment with priorities and strategic direction.
At each level, there are different roles, each of which have both unique and inter-related responsibilities:
Together, these elements identify the data that Colombia needed imminently, to solve its most pressing problems. Datamorfosis played an integral role in helping the government connect the dots between policy, process, technology, and people to promote democratic data governance.
Open North’s example demonstrates their emphasis on civic value on a far more local scale. Citing the health research company PULSAR at the University of Laval in Quebec, Open North partnered with them to develop a data governance framework that respects the needs of health researchers as well as the needs of the community; i.e., those people being researched, and the policy makers that may use the research. This is an example where values come into conflict with personal use and sensitive data. It begs the question of how we consider this conflict and still advance the research in a manner that produces positive impact for society through better use of health resources, and better-informed policies with real-time data.
Many questions arose in the process, from both the researcher and the public. For example, the researcher asks about legal and regulatory frameworks, data needs, role and responsibilities; while the public may ask about data privacy and confidentiality, and consent. Effective data governance encompasses answers for both parties by considering every factor shaping decisions about the data.
In the end, both Andrea’s and Christian’s case studies highlight the idea that data governance compliments open data, particularly when we think of it in terms of the data spectrum (see image below). Discussion among IWG members further spurred the idea that we can no longer view data as existing along a binary, where it is either top secret or radically exposed. Rather, we can evolve our understanding to reintegrate “open” into the data spectrum, meaning data ranges from open to shared to closed — and everything in between. Where your data lands along this spectrum is defined by good data governance, based on its size, audience, composition, among other issues.
Armed with good governance, data providers can more easily open data or share it, or close back up if needed, in a highly trusted way, for the most relevant people. Ultimately, mutually beneficial partnerships are enabled in this environment. This is why open data and data governance are not mutually exclusive because good governance enables everything else, just like good open data is an enabler to the broader data system.
Watch the session here. If you would like to join our next Implementation Working Group meeting in April, please don’t hesitate to get in touch: info@opendatacharter.org.
