For data commons to become a (partial) solution to the issues caused by data monopolies, they need to be politicised. It is important to ask whose data is made common and, subsequently, whose city we will end up living in, write Gijs van Maanen and Anna Artyushina in the second of their two-part blog.
Infrastructure for a data commons
The pooling and sharing of data are the two widely discussed aspects of data commons. The effects of physical and digital infrastructure that are needed to produce, store, curate, clean, and share the data on their open and common character, can be easily overlooked. While, according to theorists, data can be shared continuously, without ever losing its value, the infrastructure that supports its existence may affect its value and pose significant restrictions to data reuse.
A telling historical example here is the attempt of the 17th century French government to compile a database of all French estates. The data produced by these lists and tables made it significantly easier for the French authorities to rank and tax those estates, and it functioned as an infrastructure that reified the relationship between national and local governments. From the perspective of these estates, however, the usage and imposition of this infrastructure was seen as an attempt to equate things that were thought to be incomparable. That was a period of time, when the French estates existed as independent institutions, all having their own rules, cultures, and privileges; imposing on them some kind of statistical relations, through the production of data, was largely perceived as an infringement upon the fundamental freedoms of their owners.
A more contemporary example illustrating the role of the infrastructure needed to produce and share data is a Dutch data sharing initiative, the Amsterdam Data Exchange (AMdEX). AMdEX explores ways to “harness the enormous potential of data for the good of society,” and conceptualises data commons as the “tool to achieve responsible data use.” AMdEX develops the data storage facilities, individual or collective “data pods,” and offers a set of rules for ‘safe’ and ‘fair’ data sharing. Apart from providing the infrastructure for data storage and sharing, AMdEX acts as the data controller responsible for the participants’ compliance with these rules. An open question, however, is how such an infrastructural focus on those who own and those who want data, will be of benefit to the citizens of Amsterdam who directly or indirectly contributed to the production of the data. Or, to put it differently, who decides which forms of data reuse are socially beneficial? AMdEX, those engaged in data sharing, or the inhabitants of Amsterdam?
Whose data?
Different data commons are conceptualised and implemented to serve the interests of different individuals, collectives, and communities. These different delineations of data commons illustrate the political nature of the commons, where balancing different interests in data is an immensely hard task. This also has to do with the fact that data is inherently relational—it says something about how individuals are related to others, and by doing so dispels the idea that individuals are always the best qualified to make data governance decisions. Moreover, data can be easily manipulated to build connections between individuals and groups that were previously thought to be outside of the data’s “scope.”
These last questions point to the often-overlooked challenge of the commons that hard choices need to be made regarding who is part of the arrangement, and whose interests it will prioritise. There are several ways in which such choices can be made. Several strands of data commons and governance literature suggest starting with the communities that have interests in certain data. Recent discussions on community rights in data in India are in line with this thinking, and are at the forefront of exploring legal possibilities to grant communities rights to their data. Along similar lines the proponents of Indigenous data sovereignty argue for the right to collect, own and process data by Indigenous communities. Not merely economic, but also decolonial arguments motivate such projects. Furthermore, there are scholars who argue that, to make the data economy less extractive, data should be framed as a public asset. Therefore, data commons should, at least potentially, include all members of the public.
In both academic and policy discourses, the notion of data commons seems to oversimplify the unimaginable plurality and multiplicity of interests, values, spheres, and modes of existence encountered in the real world; moreover, the notion of data commons seems to deflect our attention from the fact that data and technologies are capable of making and unmaking (data) worlds. While commons might be a useful means of self-organisation for one community, they can be a Leviathan for others, especially as data reuse often leads to wider social changes in the form of policies, economic implications, and new products. To make them equitable, data commons need to be politicised; questions of who governs what and for what purposes need to be addressed in any data sharing initiative. People and collectives who create data commons and seek to benefit from them need to be mindful of the exclusionary nature of such initiatives and avoid transgressions into other spheres of human life.
Conclusion
Based on these brief reflections on data commons as a form of collective data governance, we propose the following purposefully loosely formulated imperatives. Data governance, and thus data commons, are an inherently political practice that requires acknowledging and incorporating:
- Reflections, discussions, decisions and justifications about who is part of the practice, and who benefit (and who do not);
- Reflections and discussions on how the concerned initiative is situated and related to the city in which it is found;
- Reflections and discussions on whether data are always the good to be produced and preserved (and not e.g. justice, sovereignty, ecology, or the neighbourhood);
- Reflections and discussions on the roles infrastructures play when producing, preserving, sharing, and using data (and their relationship to the city);
- Reflections and discussions on the relationship between the researcher(s) or policy-maker(s), and those involved in the actual governance themselves (who am I / are we to suggest these and these regulatory measures);
- Reflections and discussions on the types of rules to be employed, and by whom.
Read part one of Whose data commons? Whose city?
Acknowledgments
This blog post is one of the results of a workshop organised by Gijs van Maanen, Nadya Purtova, Michiel de Lange, and Jörg Pohle, and was part of a seminar and workshop series on data commons. The workshop took place on October 6 2022 in Berlin at the Alexander von Humboldt Institute for Internet and Society (HIIG). Gijs van Maanen is part of Nadya Purtova’s INFO -LEG project and received support from an ERC grant (716971). Michiel de Lange’s research is supported by the Utrecht University focus area “Governing the Digital Society” (GDS). We thank the participants of this workshop for their presentations, our discussions, and Tasniem Anwar, Emeline Banzuzi, Jiska Engelbert, Tommaso Fia, Berna Keskindemir, Alexander Mörelius-Wulff, Michiel de Lange, Martijn de Waal, Alina Wernick, and Jörg Pohle for the feedback given to this blog post, and the HIIG for their hospitality.
Image by @rawpixel.com on freepik
The views and opinions expressed in this post are those of the author(s) and not necessarily those of the Bennett Institute for Public Policy.