The true value of data lies in the ability to share it. It is when public sector entities can securely share data that the foundations for smart cities are laid down, and government service delivery improves exponentially.
Globally, a desire for better service delivery is driving the use of data analytics in addressing public problems and finding effective solutions. Yet, despite advances, many of the data resources that could help solve public problems and revolutionise government service delivery, remain closed. Definitional issues and a lack of clear frameworks hinder data sharing in practice. Risks associated with sharing sensitive information, reputational concerns, a lack of dedicated personnel and unclear incentives can all hamper the quest for meaningful data collaborations.
It was in direct response to these challenges that we at Dubai Data Establishment came up with our Data Sharing Toolkit — a framework that builds trust and strengthens effective data collaboration. This Toolkit is designed to further grow an ecosystem where data can be shared securely, effectively, and for maximum benefit.
In 2019, we partnered with Nesta to research effective ways of unlocking the potential of data sharing. As an innovation foundation that specialises in working with partners to bring bold ideas to life and address key challenges in society, Nesta was the ideal partner to create a blueprint for more effective data sharing.
We held interactive workshops with attendees from both the public and the private sector, and presented our work to external experts in Dubai and London to test the early versions of the content of this toolkit. We explored several data models and then boiled them down to their common essential components. Our research led us to several key insights that guided our Toolkit. First, defining a specific use case or problem to be solved is crucial to successful data sharing agreements. Second, data governance must be addressed across three layers: overall governance, data infrastructure, and access rights. Then, scale and lifespan have an impact on initiative design. Further, data sharing agreements must confront the ethical questions of privacy and bias. Finally, data sharing initiatives require the involvement of a variety of partners, not only those that hold useful data.
The Toolkit addresses these questions across two key parts. The first is a Decision Matrix. This gives you six key decision points to guide discussions about the shape a data sharing agreement must take. Here, we consider why share data, what data to share, whom to involve, what is the governance structure of the partnership and the appropriate infrastructure, and lastly how data will be accessed.
The “Why” is the starting point of any project. Data sharing arrangements either answer a problem or work towards an opportunity. They can discover new insights, as is the case with the Amsterdam Data Exchange, which wants to provide marketplace access to data for researchers, businesses, governments, and individuals. They can unlock innovation and identify new sources of value, such as the Open banking initiative from the UK CMA that allows authorised providers to access customer transactional information through APIs. Data collaborations can accelerate decision-making, as is the case with Google Waze, which provides government agencies with anonymised crowdsourced traffic data to inform policy or deploy traffic assistance. Or, they can increase forecasting capability, as was the case with Flowminder using de-identified data from mobile operators to follow population displacement during the 2010 Haiti earthquake response.
“What” data needs to be shared is also crucial. Our Toolkit asks stakeholders to hold frank discussions on which datasets would help solve the problem to hand, whether these datasets are closed or open, and what ethical considerations and potential risks arise from sharing these datasets.
The “Who” is also key to project success. Who are the project stakeholders? Are they international organisations such as the UN, World Bank, or IMF? Public sector entities at the national, regional, or local level? Or are they commercial entities ranging from multinationals to local businesses? Perhaps individuals are involved, or third-sector organisations such as universities and NGOs.
This section of the Toolkit also looks at the overall governance structure; whether it’s driven top-down by an organisation or board, or whether it’s a more polycentric initiative with multiple stakeholders sharing control. In addition to that, it also sets out the options for data storage infrastructure, which enable the data exchange across stakeholders, and helps you choose which form of data access best fits your initiative.
Then, the Toolkit moves onto its second part — Project Foundations. A key insight here is that data sharing initiatives only succeed if the people involved are committed and engaged. Open and clear lines of communication, and effective working relationships, are essential.
With the Toolkit’s Project Foundations section, we set the stage for creating these effective working relationships. This section includes a checklist where we look at senior buy-in, and how to incentivise parties to work together, and discuss equitable contributions from all stakeholders. Then, we consider the legal, technical, and funding elements around a data sharing project. The aim is to identify and eliminate hurdles to effective data sharing, with a holistic approach where people and organisational structures are as important as technical considerations.
We urge other entities reading this article to consider how the Toolkit can shape their thinking on creating data sharing initiatives. At Dubai Data Establishment, we are on hand to help develop such ideas. We can assist with use cases, governance, partnership models, and even technical data sharing infrastructure. We’d love to see the Data Sharing Toolkit used widely to spark value-generating data collaborations that make a positive impact to service delivery and quality of life.
To access the Data Sharing Toolkit, click here.