From Trees to Forests: Reflections on Open Data Initiatives in Asia
I was fortunate to be invited to Civic Data Lab’s (CDL) celebration of open data for public good. The two-day conference brought together nearly 200 data scientists, activists, advocates, scholars, and enthusiasts at the India International Center in New Delhi on 6–7 November 2023.
I first became aware of CDL’s work when I met Gaurav Gadwahni, CDL’s founder and CEO, in Buenos Aires at the International Open Data Conference some five years ago. Gaurav was a very impressive, young, brilliant, and skilled socio-technologist, passionate about finding ways to unlock the power of data to achieve lasting impact. Since then, I have closely followed the work of CDL and have been glad to see how it has grown, partnering with organisations across India and globally.
I was asked to join a panel of awesome people: Gurman Bhatia, a data journalist turned data-preneur with a mission to make data beautiful; Shrayana Bhattacharya, economist and author who advocates, among other things, for the use of pleasure as a mental model to guide research conduct and use; and Siddharth Hande of Kabadiwalla Connect, whose work on inclusive waste value chains fascinated me a lot. Instead of looking at examples of how (open) data initiatives can grow and deliver impact, I reflected on the work I have done or studied in the last five to ten years and what it means to those who aim to design data initiatives that can change society for the better. I focused on four points that I think are critical in this respect:
1. There is no singular road map through the forest.
This is just a fancy way of saying, let’s not be method- or approach-focused. When we started the Open Data Lab Jakarta some eight years ago, we developed and tested what we called the “Responsive Open Data Model”. We argued (and we still argue now) that disclosure of a few data sets that users need is much better than disclosure of a plethora of datasets that are not relevant. Building capacity and use cases around those relevant datasets will lead to more impactful results, especially when products from the process influence policy and practice within government.
Almost at the same time as this was happening, PetaJakarta was changing information availability and disaster response in Jakarta (now going global as PetaBencana). Their innovative solution followed a different logic than what we did, but it significantly impacted Jakarta’s disaster response and management. It harnessed the power of data from social media to locate and respond to flooding emergencies and partnered with the city government to deliver emergency interventions.
2. In data ecosystems, stakeholder categories and relationships are fluid and defy rigid definitions.
I was part of a group of scholars who started looking at how open data can help deliver impact in developing countries. One of the results of our scholarly work is a book edited by Stefan Verlhurst and Andrew Young of the GovLab — Open Data in Developing Economies. Looking at the different cases across Asia, Africa, and Latin America, the editors created the diagram below to show the journey of data from publication to impact.
But diagrams such as this are a static representation of an evolving reality. Governments need not necessarily be the source of data, and non-government organisations need not be the ones demanding and using them. One of the things that Jin Sato of the University of Tokyo presented at a climate conference in Bangkok a few months ago was this image of a woman in Northern Thailand — measuring the water level of the Mekong region and aggregating these along with other data collected across the river bank, to inform decision-making processes at the individual or household level.
The woman here is a data producer, data intermediary, and data user at the same time. The relationships established were different from the ones contemplated in the diagram above. She used old and customary tools to do their work. She has not used any sophisticated gadgets or apps that will help them predict the ebbs and flows of the river. Indeed, we can argue that there is a better way to do this, but do we discount the power of this community-based innovation just because it does not conform with our reasoning and expectations?
3) Data accessibility and understanding may be more critical than our fascination with formats and licenses.
Many years back, I have been involved in the production of three editions of the Open Data Barometer, the predecessor of the current Global Data Barometer. In those three editions (from 2013 to 2017), we found that while a lot of data is available online, only very few can be considered strictly as open data, using the standard definition.
However, global and local challenges are increasing where data can play a crucial role, and data remains not openly accessible in several countries. As a matter of fact, in some areas, open data initiatives have died, and open data portals have become data graveyards. I argued elsewhere that designers of data initiatives should not start with the data, but with the problems people face. When access to data is restricted, some actors work with whatever they have access to or sign non-disclosure agreements to use the data and influence policy and practice.
Don’t get me wrong; there is value in advocating for openness of public-interest data. I believe we don’t need to go through that process of bargaining and negotiation to get access to the data that, in the first place, should have been made public. But we will have to make do with whatever we have momentarily, while our advocacies have not gotten us the kind of result that we desired.
4) Finally, our advocacy for data openness and use may cause greater opacity and use restrictions.
The world has significantly changed since the open data movement gained international traction. We have seen the rise of strong men who do not care about transparency and accountability. We have also seen the unfortunate restrictions on civic space and even the death of activists and journalists.
In the most recent edition of the Global Data Barometer, most of the use cases are on financial transparency and accountability, but only half of these have resulted in impact. The report mentioned that persistent problems of impunity and weak accountability mechanisms cause the lack of impact. Indeed, Claudia Juech of Bloomberg Philanthropies is right when she said, in one of the plenaries of the CDL conference, that sometimes we put so much premium on data and technology, forgetting power, accountability mechanisms and incentives on why data or tech solutions don’t work.
In Southeast Asia, where I do most of my work, there are more incentives for governments to suppress the disclosure of data that can incriminate them, especially data that can be used to reveal corrupt practices. Despite strong advocacy from civil society organisations, data availability and access have not improved and sometimes have deteriorated. In some contexts, there is an observed decline of not just access to data, but also of spaces for contestation and discourse.
But not all things are unwell. Despite the challenging political environment, several civil society organisations in India, some of which have put up booths and exhibits during the two-day event, have shown how partnership and collaboration between advocates and data scientists have allowed better use of resources, more responsive services to realise protection of children, or address the climate crisis. I have also seen how these initiatives have empowered individual activists, the organisations they worked for, and the people they have committed to serve.
I’m looking forward to more stories of change, more pathways to impact, and better access to data leading to social change and the promotion of public good.
