Open Data in Developing Countries: Different models, new approaches
Lauran Potter · July 15, 2014
Today, we’re sharing the first wave of results from our “Exploring the Emerging Impacts of Open Data in Developing Countries” (ODDC) project. This project marks the first major study of the use and impact of open data in countries across the developing world. Over a hundred researchers from the global South have been involved in developing 17 qualitative case studies with findings that span 13 countries, from Indonesia to Brazil.
These studies describe a wide range of open data efforts, including top-down initiated projects, led by governments and donors; bottom-up efforts, led by technology communities or civil society organisations; and sector-specific initiatives focussed on very specific datasets. In addition to producing rigorous research, the project is also working to build open data capacity, supporting researchers with the local knowledge needed to drive the open data agenda in developing countries. Over recent months, ODDC partners have run local presentations and workshops to share findings, including sessions with government ministers, NGOs, academics and local activists.
Alongside research posters and case studies from the ODDC partners, we’re sharing a synthesis report to highlight fifteen emerging insights and common themes. When it comes to the question of “emerging impacts,” our report suggests considering two different models for thinking about the impact of open data in developing countries. We’ve termed these the “domino effect” and the “ripple effect.”
In a domino effect, there are a number of different pieces that need to be lined up before open data can trigger outcomes and impacts. These may be pieces relating to data supply, such as data quality and good meta-data; pieces related to intermediary activity, such as the interest, awareness and skills of the media or NGOs related to the open data movement; or pieces related to the wider political context, such as the receptiveness of the public or of decision-makers to data-driven analysis and arguments. Impacts based on such a model rely on a chain-reaction from datasets to outcome. While it’s a powerful model, many of the domino pieces are underdeveloped or absent in developing countries, making it challenging to secure the desired impacts.
In a ripple effect, the outcomes of open data may be indirect—starting from an open data intervention but spreading out into new areas. Our case studies show that the process of thinking about open data may encourage governments to change their systems for collecting data, which may affect how they use data internally and may, in turn, have consequent effects on policy and planning. We note that open data initiatives are creating new spaces for government and civil society to work together, building trust or enabling more targeted NGO actions. Even unsuccessful attempts to access or use specific datasets from government as part of an open data project may still build interest in having better data, and so lead to civil society generating new data that can be used in advocacy or in the delivery of public services. It may be hard to attribute these impacts entirely to the use of a specific open dataset; rather, they emerge around the introduction of open data ideas and practices in general. In such a ripple effect, interventions may start small, but, with the appropriate support, can scale outwards.
Of course, these two concepts are not mutually exclusive—the most successful projects typically have elements of each. Ripples spreading out from open data interventions may lead to more sustainable but subtle change, whereas a successful domino effect is likely to lead to more visible and high-profile stories of impact.
Other emerging insights from the study include:
- There is often a significant mismatch of supply of and demand for open data in developing countries. All too often, sensitive datasets that could greatly enhance transparency are not being made available for political or other concerns. Meanwhile, economically valuable datasets are not being collected due to a lack of capacity in government statistical agencies. Until there are better connections between open data demand and supply in developing countries, many potential impacts are likely to be missed. Studies from Kenya, India, and the Philippines demonstrated some of these challenges.
- Some of the greatest impacts of open data are found at local government level. Many of our case studies demonstrate that an outsider mindset of seeing open data as primarily a central government issue, implemented through national data portals and policies, should be challenged. Local government data, city data, and data from the judicial and legislative branches are all important, and for many citizens, the data of most relevance to their day-to-day lives exists at a local level. The Opening the Cities project provides compelling examples of this from four major cities across Latin America.
- Digital Divides create Data Divides…but mobile technologies are not the silver bullet. Unsurprisingly, many cases highlight how a lack of access to technology, connectivity and digital skills is a barrier to both publication and use of data in many countries. These challenges need to be solved before the full power of open data can be harnessed across the developing world. One case from Kenya suggested that users preferred to access open data tools on desktop computers, while work in Uganda has noted the importance of offline methods, community radio and trusted local intermediaries in making data accessible to all.
This is just a snapshot of some of the early findings emerging from the ODDC project— our researchers have delivered a treasure trove of information, and you can access all of the cases and supporting material on the Open Data Research network website. In the coming months, we’ll be sifting and analysing these findings in more detail, and plan to release a series of thematic papers starting in Quarter 4, 2014. You can receive these by signing up for our newsletter.
You can also discuss findings of the study in the Open Data Research Network LinkedIn group, where we are interested in hearing if there is a particular theme you’d like us to explore in one of these papers.