Democratic governance has been one of the key areas of work in international development over the last decades. Over this period, innumerable number of initiatives, programmes and projects have pursued the strengthening of governments’ transparency, accountability and citizen participation in decision making. Another key feature in this processes has been promoting, expanding and strengthening decentralisation in order to bring the power of decision making as close as possible to those that are affected by those decisions. The irruption of data for development has, in general terms, discontinued this trend, and the concept of the bid data divide helps us understanding this trend.
As discussed in the previous post of this series, the big data divide refers to the asymmetric relationship between those who collect, store and mine large quantities of data, and those whose data is being collected. This concept is useful when thinking about the governance of big data in that it highlights key dimensions about data: ownership, management, and who participates in defining what data should be used for.
National statistical offices, government agencies, have traditionally been, for the most, those collecting and processing large quantities of citizens data, but the arrival of big data has changed this picture drastically. In the current state of affairs, it is no longer government or their agencies playing this role. Instead, it is now private companies, with the required capabilities, assuming this role. The change has been so spectacular that when looking at this globally, we have witnessed a move from state data monopolies to a situation of global private quasi-monopolies. This move has several important implications for those whose data is being collected. On the one hand, their data is now collected and controlled by private organisations, often residing outside the territories where the data is being collected, which they cannot hold accountable. On the other hand, their data is being collected, controlled and used by private organisations with defined interest, profit making, which greatly differ from those of public institutions, providing public services and serving the population. As such, the big data divide contributes to maintaining and often enlarging existing power imbalances.
Diego Silva, from the Big Data for Development Network, defines governance as the ’form how society organises to make decisions’, which in the context of big data ecosystems refers to the form of ’the different interactions among relevant actors in charge of the multiple dimensions of information and data management’. The concept of big data divide helps us integrating the power imbalances in Silva’s definition, thus big data governance is not only about interactions among stakeholders but also about tackling the power imbalance among these stakeholders. This in turn helps us identifying a number of elements that should be taken into account in shaping big data governance, namely: ownership, access, and participation in the application of data.
Experiences from the Global South help us illustrating this elements in addition to providing inspiration about the ways forward.
Data ownership and governance can focus in ensuring that data ownership rights moves away from those collecting the data and rests as close as possible to the data subjects. In this regard, data sovereignty initiatives in Australia can help illustrating key features of this element. The initiative from the Maiam nayri Windara Indigenous Data Sovereignty Collective on indigenous data sovereignty seeks to develop a set of protocols for data sovereignty and governance that ensure the right of Indigenous peoples to govern the creation, collection, ownership and application of their data. This enables Indigenous peoples and their representative and governing bodies to accurately reflect their stories by ensuring what/who is counted, as well as identifying what works and why. It further enables them to ensure the data reflects their priorities, values and diversity, as well as grants them the right not to participate in data processes that do not respect those.
Access to data and governance can focus in ensuring that data access is not restricted by purchase and processing capacity. In this regard, open data initiatives provide free access to existing data sets so these can be used by communities. While open data per se does not guarantee that citizens can make use of it, due to processing capacity deficits, it does allow for other actors such us civil society to access the data and processed on their behalf and with their participations. There are plenty of examples of open data initiatives, as documented by Jacinta Rivera in our blog with the experience of atuservicio.uy in Uruguay.
Data application can focus in ensuring that the interests and understanding of those subject to the decisions made on the basis of big data are integrated in developing the focus of data processing. In this regard, initiatives like Yawuru data sovereignty show positive results in its application to understanding and promoting community wellbeing. The key to this results can be traced back to ensuring community data ownership and participation in defining how wellbeing of the Yawuru people should be conceptualised.
Taking into account these elements in the development of big data governance forms is likely to balance the power imbalances, bring automated decision processes closer to the people’s priority and needs and open opportunities for harnessing the potential impact of big data in development.