Reorienting "data for development"
"Data for development" enjoys high currency in global policy circles. It has come to stand for the idea that developing countries need to acquire the capacity to produce "quality, accessible, timely and reliable disaggregated data" and use the promise of big data to overcome data gaps in national statistics. However, despite spectacular strides in technology, a data revolution adequate to tackle real life problems at scale is still only a hope for a number of reasons.
First, available big data is not always big enough (covering the volumes needed for accurate modelling around "wicked" development problems) nor is it sufficiently representative (covering the most vulnerable populations whose futures are in question). Also, the complexity of development requires knowledge that is contextual, requiring conventional theory-building that uses causation-based models, rather than the merely correlation-based ones most often employed in big data techniques. Small data and local verifiability therefore remain hugely relevant, but tend to be ignored in the frequent fixation with techno-solutionism, something that UN Global Pulse, a UN initiative which aims to harness big data safely and responsibly as a public good, also underlines.
In addition, many "partnerships" in big data seem to have sidestepped concerns around privacy and data ethics, recognized as non-negotiable by the UN Development Group, revealing a bandwagon mind-set in the pursuit of techno-innovations.
Finally, the largest data sets on human behaviour today are controlled by digital behemoths in closed, proprietary systems. They hold the key not only to the digital exhausts of users—the infinitely reusable raw material for data analytics—but also to the digital intelligence produced through these closed systems. Development in the digital age is thus predicated on how digital corporations will re-carve markets in developing countries based on their data power. For instance, Bayer’s imminent acquisition of Monsanto is driven by the former’s need for the granular insight into seed and soil data that the latter controls. This intelligence is what will ensure Bayer’s dominance over agriculture input markets in pesticides and fertilizers, but not necessarily produce inclusive and sustainable development.
Today, "data for development" discussions, based on myths and misunderstandings about big data, are not pointing us in the right direction. Rather, they seem to ignore the flagrant disregard for the first principles of the right to development which underlies the current extractive global data regime. The debate therefore needs to be reoriented to answer the following questions: Data for whom? Controlled by whom? And to what end?
Right to development, right to data
The UN Declaration on the Right to Development acknowledges that "the right to development belongs to everyone, individually and collectively, with no discrimination and with their participation." In this view, all individuals and peoples have the right to self-determine their development trajectories, and assert full sovereignty over natural wealth and resources. It is from this vantage point that data innovations for development must proceed.
It is increasingly being seen as vital to guarantee individual rights in the digital age through data governance frameworks. From data protection guarantees and inclusive data taxonomies and systems, to the rights to participate in the design of data systems, discrimination-free algorithms and appeal against outcomes of automated decision-making processes, key rights related to data are currently hot topics of discussion.
However, from the perspective of the right to development, data practices need to uphold the right to self-determination not only of individuals but also of collectivities. This means that the right of developing countries to determine their path to development includes control over data and digital intelligence. "Data for development" must therefore encompass the idea of the public value of data and the role of data infrastructure as a public good, respecting citizens’ digital rights individually and collectively.
Data as a public good
Developed countries like Singapore and Canada are already creating an enabling public infrastructure for data and digital intelligence in key sectors like urban mobility and energy monitoring. Capturing sectoral data and building digital intelligence systems for core socio-economic domains, such as agriculture, transport, trade and commerce and so on, can go a long way in supporting public policy decisions in developing countries.
Some ways to do this include:
- Digitizing and converging data sets captured through national statistical machinery and building digital intelligence solutions on top of such data sets, with due attention to privacy, transparency and accountability safeguards.
- Mandating that private platforms operating in key sectors disclose critical data they collect to state agencies, with safeguards for protecting user and citizen privacy. Platform companies collect copious amounts of citizen data on a daily basis, using the intelligence obtained from such data to successfully thwart competition. To support essential public services, like city transport or health care, companies must be obligated to open up this public data for use in the public interest. For example, the municipality of Curitiba in Brazil has enacted legislation that requires platform companies such as Uber to share trip-related information with municipal authorities, except routing information, which may compromise user anonymity.
- Building public digital infrastructure that can, over time, contribute to the aggregation of data and the creation of digital intelligence solutions. For example, a state-run online agriculture market—such as India’s e-NAM—can, with wider adoption by producers and traders, become the data ecosystem for effective, intelligence-based, policy implementation directed at the farm sector.
A community’s right to its data
The idea of a collective right to data may also be seen as a community’s right to its data. "Communities" here refers to ethnic, cultural, indigenous or religious groups, usually at the margins of the national mainstream. As the Maori Data Sovereignty Network in New Zealand has asserted, any indigenous community has the right to govern the collection, ownership and application of its own data that derives from its inherent right over its natural resources.
According to the UN Declaration on the Rights of Indigenous People, every community has a right to develop, maintain and protect their cultural institutions, control their traditional knowledge, and prevent the destruction or forced assimilation of their culture. Extending this principle to data frameworks would mean that the application of data-supported solutions or creation of digital intelligence systems cannot violate the cultural integrity of a community.
Scholars have pointed out that big data methodologies have now made it possible to profile minority communities, even if they do not compromise individual privacy. Personal data protection frameworks like the European Union’s General Data Protection Regulation do not offer any redress in such a situation. This means we need to go back to the drawing board to re-imagine privacy from a group or collective right-against-discrimination perspective.
Understanding the value of the data commons
In order to harness the gains of the Fourth Industrial Revolution for equitable and sustainable development, public policy frameworks need to understand the value of the data commons. Today, data is naturalized as commodity, given its irreplaceable value for profits. It is time to explore new institutional frameworks to harness the public value of data and promote a community right to data.
Such frameworks must be embedded within the architecture of human rights, with due attention to the right to development of collectivities and peoples. The rule of law is non-negotiable in determining definitions and delimiting boundaries for the collection, processing and use of data. To appropriate data for development as a collective resource and a public good, institutional mechanisms of governance need to be revamped so that federated data publics can self-determine their data rules.
This article was originally published in the UNRISD blog.