Humanitarian organizations are facing increased incentives to collect and share disaggregated data about their operations. At the same time, they must ensure this data is only used for humanitarian purposes and does not cause harm to affected populations. Despite recent progress with sector-wide commitments and frameworks for data responsibility, there is an area in need of more specific guidance: data sharing between humanitarian organizations and the donors who fund their work.
In this post, ICRC’s Head of Data Protection Vincent Cassard, Stuart Campo from the OCHA Centre for Humanitarian Data, and Jonas Belina of the Swiss Federal Department of Foreign Affairs outline the opportunities and challenges of this relationship and provide six guidelines to support more responsible data sharing practices between humanitarian organizations and donors, drawing on a new framework developed through the Humanitarian Data and Trust Initiative.
Humanitarian action has long been based on the collection of timely and reliable information on people in need. From household-level survey results to details about the delivery of assistance to certain demographic groups or locations, disaggregated data from humanitarian contexts can unlock valuable insights. The increasing availability of such data in recent years has strengthened the incentives for humanitarian organizations to collect and share it, namely for the improved coordination, accountability, transparency, and efficiency of their operations.
Donors play a key role in this dynamic both as direct users of data and as drivers of its increased collection and analysis. Many donors now request various forms of disaggregated data from humanitarian organizations, such as gender, age, or disability markers. For example, a report on a nutrition programme might include data disaggregated by the age cohorts of the children assisted.
Data sharing requests may serve multiple purposes, including to check if intended objectives were reached, to enrich the donor’s understanding of different crises, and to support the donor’s public communications and help make a case for funding different organizations’ operations. But these requests can also come with significant risk.
For instance, sharing a disaggregated dataset may raise the risk of re-identification of individuals or groups. The potential or perceived use of humanitarian data for non-humanitarian purposes could also put affected populations at risk of harm and undermine trust between humanitarians and the people they seek to serve. Anticipating and mitigating such risks is key to principled humanitarian action. Getting this right requires collective action across the system.
Gaps in existing guidance for a diverse group of stakeholders
Humanitarians have invested in the development of frameworks for the responsible management of data in recent years. This includes, for example, the Inter-Agency Standing Committee (IASC) Operational Guidance on Data Responsibility in Humanitarian Action and the ICRC Handbook on Data Protection in Humanitarian Action. While these documents provide common guidance for humanitarians working in a range of complex contexts, they do not specifically address data sharing with donors.
Similarly, most frameworks for partnerships between humanitarian organizations and donors do not address the topic. Few donor reporting templates and contractual documents contain explicit clauses on data sharing – and when they do, these clauses vary depending on the partner and operating environment in question.
Most donors and humanitarian organizations alike are subject to legal frameworks governing the protection of personal data. Each partner may have different obligations that need to be observed. A multitude of actors are also involved in the process, as data shared with donors is often collected by the implementing partners of humanitarian organizations or shared with third-party monitors on behalf of the donor. Data sharing is not limited to formal reporting either: disaggregated data in and about humanitarian contexts may be shared during informal meetings or contacts.
Overall, these gaps in guidance and the diversity of stakeholders involved make it challenging to ensure consistent and responsible data sharing.
Finding the right balance through a common approach
In 2020, the Humanitarian Data and Trust Initiative (HDTI), convened by the United Nations Office for the Coordination of Humanitarian Affairs (OCHA) Centre for Humanitarian Data, the Swiss Federal Department of Foreign Affairs and the ICRC, initiated a dialogue process on responsible data sharing between humanitarian organizations and donors. The dialogue process consisted of two meetings convened through Wilton Park in September 2020 and in September 2021, two pieces of independent research conducted by the Global Public Policy Institute and the University of Manchester, and a series of stakeholder consultations.
Throughout this process, stakeholders acknowledged the importance of coming together to identify common challenges and to devise joint strategies for addressing them. The outcome was a Principled Framework for Responsible Data Sharing Between Humanitarian Organizations and Donors, which reflects the perspectives of a diverse group of participants and interests while offering a common framework for action.
The framework, which is the first of its kind, offers six guidelines to inform data sharing between humanitarian organizations and donors. It is not a legal document, but rather is designed to provide humanitarians and donors with practical steps they can adapt to their own data sharing work. The goal is to foster increased collaboration and build best practices through the experience of applying these guidelines in different settings. Each of the six guidelines is described below.
Six guidelines for responsible data sharing
The first guideline enacts one of the fundamental principles of humanitarian action: humanity. Donors and humanitarian organizations should work to ensure that data sharing processes keep affected people at the center. In other words, have affected people been adequately informed about the sharing of their data? Do they have the possibility to exercise their rights to information, access, correction and erasure of their data?
The second guideline focuses on clear communication of the purposes of data sharing. Whenever data is requested or shared, the reasons for doing so should be clearly articulated. In turn, this helps ensure that data sharing is minimized and limited to what is specified, necessary and agreed.
The third guideline underscores the importance of establishing clear and common requirements for responsible data sharing. Formalizing these technical and procedural requirements at the outset of a partnership allows for consistent engagement and monitoring over time.
The fourth guideline fosters a common approach to addressing potential risks related to data sharing. Humanitarian organizations and donors should collaborate to identify potential risks and mitigation measures throughout the course of the data sharing activity.
The fifth guideline emphasizes the need to invest in training and procedures for responsible data management. As the goal of the framework is to forge a common path forward for implementation, humanitarian organizations and donors should work together to provide their staff with clear instructions on how the framework applies in different operational settings.
The sixth guideline supports learning and accountability initiatives to foster – as well as advance – all the above guidelines. Humanitarian organizations and donors should support inter-sectoral collaboration and research to advance knowledge in this area.
We hope partners across the system find these guidelines useful and look forward to working together to advance the implementation of more responsible data sharing as a path to more principled and accountable humanitarian action.
- Fiona Terry, Fabien Dany, “Harnessing the power of Artificial Intelligence to uncover patterns of violence”, May 25, 2023
- Wim Zwijnenburg, “Data-driven environmental decision-making and action in armed conflict”, June 1, 2021
- Tess Bridgeman, “The viability of data-reliant predictive systems in armed conflict detention”, April 8, 2019