To learn more about Open North’s approach to data governance, be sure to check out the first blog post in this series as well as our online course.
In a previous blog post we began introducing Open North’s stance on data governance. In that post we explained our perspective on why data is not just about data: data is actually about all the social and economic relations that make up society. To ensure equity rather than marginalization or exploitation we need a proactive and broad understanding of data governance. We wrote that, “data governance is not just about the use of information; it is about rethinking our social contract in a time of massive digital upheaval.” As one of our brilliant colleagues once said, data governance is like driving: sure, you want to avoid breaking the laws and drive well, but even more importantly you want to get somewhere worth going.
Put differently, data governance isn’t just about internal legal compliance or operational effectiveness; more fundamentally, it is about the kind of world we want to live in together. Data governance is a pathway to reimagine and rebalance power imbalance between data collectors and the individuals or communities the data represents. Data governance offers a powerful lever to redesign the social contract to shift that agency and create a system in which all participants have a say in what civic value1 is generated and how.
To help publicize how Open North sees data governance as crucial to attaining this goal in our times of constant digital transformation, we developed an online learning module ‘What is Data Governance (D201).’ This blog reviews the central ideas in the learning module and introduces our principles-based and structurally iterative stance on data governance.
The module describes how data governance’s potential exceeds compliance or organizational effectiveness: data always extends beyond the internal workings of an organization into all its stakeholders’ lives. Even the most impersonal data, like sensors for air quality, are, in the end, both about and for people — as should be the civic value it offers. Data governance is needed to shift agency and ensure that the most marginalized have an equitable say in how it is used, who benefits from it and how. Data governance is, thus, also about equity — not just equality.
But how does one pursue equity with data governance? What does that mean in practical terms?
There is a rapidly growing trove of research into different models being developed to create better social and economic relations: on data trusts, collaboratives, cooperatives, commons, and many variations of data stewardship in between. These are all important innovations that are expanding our field of options, but working one on one we take a different stance.
Our work with big cities like Montreal or Toronto or small to mid-sized municipalities like Trois-Rivières or the Town of Bridgewater, Nova Scotia, has highlighted that existing models can inspire, but each situation is unique and needs to develop its data governance approach within its own context and stakeholders.
Principles-Driven
The characteristics a data governance framework must incorporate are principles that form the conditions required for pursuing equity. These are: inclusivity and participation; transparency; accountability; and responsibility. Each can be operationalized in different ways, but all are necessary.
- Inclusivity and Participation: Stakeholders need to be meaningfully involved in the design of the framework, and it must include options for ongoing participation. Inclusivity is achieved by considering issues of power, inequality, and accessibility among stakeholders and designing the necessary mitigations.
- Transparency: All stakeholders should understand the purpose of the data, how the data is being produced and used, and who is accountable. The multiple interests in the data and its usage must be clearly communicated.
- Accountability: There must be clear mechanisms for contestation, compliance, and evaluation at each step along the data’s lifecycle, including who holds the ultimate accountability for adherence to the data governance framework.
- Responsibility: A good data governance framework acknowledges its responsibility towards its immediate stakeholders and more broadly to the society, environment, and future we all inhabit. As such, it must shift agency between collectors, subjects, and beneficiaries to share benefits in a way that generates civic value for the common good beyond short-term profit considerations.
The observant will have noticed that these principles are predominately about ensuring agency for data subjects and that everyone can equitably participate in having a say about how data is used/re-used. Data governance requires equitable design in order to enable equitable outcomes.
There are many important mechanisms and tools in a comprehensive data governance framework to ensure good governance across the data lifecycle. Components like privacy impact assessments, data quality and standards, auditability, cybersecurity, or data destruction protocols are essential. However, before developing the means the ends must first be equitably determined — and for this, equitable democratic participation is necessary.
Consider again the example of air quality data. Clearly, ensuring the good management of the data is important so that it is safely and effectively created/collected, stored, accessed, used, shared, and destroyed when no longer needed. However, before even embarking on such a project fundamental questions need to be answered: who is getting to make decisions about whether creating the data is the right solution? Where is the data collected? What are the risks and for which communities? What are the mitigation strategies? Can the data be shared and in what form(s)? What kinds of value could be produced, and for whom? These are crucial prior questions for equity, and the four principles provide clear guidance on them.
Structurally Iterative
This naturally also means that the process of developing a data governance framework is structurally iterative. If the air quality project team, for example, considers expanding to conduct analysis together with traffic data, they may find they need to work to include newly impacted stakeholders and communities, or develop new mechanisms to govern analytics and sharing, or reconfigure existing tools for privacy and risk assessment. However, the four principles provide the ethical bedrock and yardstick for equity with each iterative improvement.
This process of principles-based reflection and iterative improvement is at the heart of OpenNorth’s core mission to ensure that if and when data and technologies are used they are done so in a way that mitigates risk and provides agency for marginalized or less powerful stakeholders. As we shift into a world of digital transformation in which ever more complex tools — like those grouped under the umbrella term ‘AI’2 — are becoming more common, the consensus is clear that deep public participation on issues of risk, benefit, and value is the only path to equity.
For this reason, iterative reflection is also built into the design of ‘What is Data Governance (D201).’ Each section introduces a fundamental characteristic of data governance and then invites reflection on it and its implications for the reader. Working through the module more than once will create improved layers of reflection on the critical components. OpenNorth will also be presenting an interactive, bottom-up participation approach to this principles-driven and structurally iterative stance on data governance at Mozfest in March 2023 — come join us and help collaboratively develop a framework!
The same approach structures Open North’s more applied data governance work, including health data through a project with Université Laval and an advanced Data Governance Self-Evaluation Tool. This tool provides an extensive process for the assessment and implementation of the necessary tools and mechanisms for organizations to have the data maturity to participate in a data governance framework. First developed for the Montréal en commun project, the tool provides a systematic process for the iterative operationalization of the four principles described above. This process guides the user through 14 key points on the data lifecycle, each with clear core and operational tactics for implementation, to ensure the development of a comprehensive yet flexible framework.
This in-depth tool will be published in English and French on the project’s website soon and will be the focus of the next blog. However, we are currently field-testing a simplified, 90-minute version with partners! Please don’t hesitate to reach out to csn@opennorth.ca if you are interested in a free assessment of your data governance strategy to:
- Help identify stronger alignment between existing measures
- Find opportunities for improvement, or
- Develop solutions to barriers
Footnotes
1 We talk extensively about value or civic value, by which we mean outcomes and impacts like wellbeing, trust, or civic engagement as well as financial benefit. Our paper ‘Creating Civic Value in Open Smart Communities’ engages with the importance of this point at length.
2 The term ‘artificial intelligence’ can be significantly misleading, so we use it in scare quotes where space prohibits a better explanation.