Introducing 'Good Data': Data, Ethics and Law for the 21st Century
The book is available in various formats for free download here: http://networkcultures.org/blog/publication/tod-29-good-data/
One of the Good Data chapters was co-authored by Prof Chih-hsing Ho, who was a guest of Prof Daly and CUHK Law thanks to the CUHK University Academic Exchange Fund (China).
Profs Daly and Ho participated in two launch events for the Good Data book in Hong Kong: a public outreach event on 5 March 2019 at the ACO Bookshop in the Foo Tak Building which also featured Prof Jack L Qiu from CUHK School of Journalism and Communication; and an academic-oriented event hosted by CUHK Law’s Centre for Financial Regulation and Economic Development (CFRED) and chaired by Prof David Donald on 7 March 2019.
This blogpost presents an overview of the Good Data book and some future directions for the editors’ own research and practice.
Good Data was born from our frustrations with ‘bad data’ dystopia and from a wish to see a more ethical and just digital society and economy. What might a ‘Good Data’ future look like? How might we achieve it? What would the theory and practice of Good Data be? These were questions that drove the Good Data book project in the first place and for which we find some answers among the 20 chapters contributed by over 50 authors from around the world.
What is Good Data?
We chose the term ‘Good Data’ to describe our work and outlook given its expansiveness as a concept, open to various interpretations, beyond the sometimes narrow discussions of ‘ethics’.
We are certainly not the first people to think about ethical, moral and political questions surrounding data, and we acknowledge in particular the work on Stefania Milan and the DATACTIVE team at the University of Amsterdam and Lina Dencik and the Data Justice Lab at Cardiff University. There has also been extensive discussion of datafication from a privacy and data protection perspective in legal literature and a research community on Fairness, Accountability and Transparency in Machine Learning.
Instead, we view discussions of Good Data as encompassing these currents but also as breaking down some of the disciplinary silos in which they may exist, and also building on this body of work to shape our consideration of the practical ways in which Good Data theories and practices can be implemented.
We view Good Data considerations as permeating the whole process of creating and using data:
When the decision is made to collect data in the first place
When the data is collected
When the data is processed/analysed
When the data is used
When the data is re-used
Another important issue is who precisely is involved at each of these stages in the process, in the sense that it may be permissible and appropriate for certain actors to be involved in collecting certain data but that data should not be collected by others. This is an issue which we observe as generally being absent from the consideration of (western) data protection laws but clearly emerges from recent work by First Nations scholars on Indigenous Data Sovereignty - such as the chapter by Lovett et al we are very pleased to include in the Good Data book, and also Kukutai and Taylor’s open access book Indigenous Data Sovereignty (ANU Press 2016).
Overall, we view that data’s ‘goodness’ is an explicitly political (economic) question, and is always related to the degree which it is created and used to increase the wellbeing of society and especially to increase the power of the most marginalized and disenfranchised. This is of the utmost importance given the ways in which various contemporary collections and uses of data act in the opposite way, further entrenching disenfranchisement and marginalisation, such as the use of algorithms which entrench racial and gender discrimination as exposed in e.g. Noble’s recent book Algorithms of Oppression (NYU Press 2018).
Good Data considerations should permeate digital technology development, implementation and use at all stages of the process, from development to use to re-use. In order to achieve this, Good Data requires implementation via different tools, including Lessig’s four modalities of regulation: laws, norms, code and markets. Accordingly, we view next steps for us and others concerned about bad data practices and wanting to see better digitised futures as being both further theory and practice to implement Good Data initiatives and build communities of change. In particular, practical, user-friendly Good Data implementations in devices, software, apps and practices - as well as better laws to reign in bad data practices - are key to achieving a more fair and just digital economy and society.
Our own future research will involve developing and elaborating this set of Good Data principles and an investigation of whether the EU’s General Data Protection Regulation can be considered a ‘Good Data law’.
We welcome feedback on the Good Data book and the Good Data principles!
Angela Daly, CUHK Faculty of Law