Democratic Dilemmas of Data
By Grace Hunley, TReNDS Intern
Photo published in babelfish.
In its 2014 keystone report on mobilizing the data revolution for sustainable development, “A World that Counts”, the UNSG’s Independent Expert Advisory Group laid out a vision for a future in which closing gaps in data access and use would empower and inform citizens to democratically influence policymaking and hold governments accountable. Calls for open data, data sharing, data accessibility, and data literacy often coincide with this positive outlook on data as a democratizing force.
Over subsequent years, however, we have seen that the democratic potential of data is up against a competing force of data concentration. Centralized bodies – specifically big tech companies and governments – have been able to collect, commodify, monetize, and utilize for surveillance purposes ever-increasing amounts of data, often leaving individuals without meaningful control over their personal information (Couldry and Meijas 2019).
With more than 90% of the world using Google Search and over 3 billion people using applications owned by Facebook, it is difficult to avoid using and contributing to their vast mines of data wealth. Through their algorithms, these organizations wield enormous power in deciding what information is consumed by users. Similarly, issues arise concerning data misuse and misinformation by centralized bodies and the public alike, all of which have the potential to contribute to democratic regression.
Data governance practices need to be implemented to realize the democratic value of data and protect people from vulnerabilities associated with usage of personal data and concentrated power of data, the misuse and misinformation of data, and the power dynamics that producing and using data reinforces.
Data as a Democratizing Force
Many data sharing and open data initiatives have been pursued with the express purpose of allowing data to be used as an equalizing force. Indeed, a number of major open data tools have actually been created by centralized bodies, such as The European Union Open Data Portal, The World Bank Open Data site, and The InterAmerican Development Bank Data, to name just a few. Relatedly, citizen-generated data (CGD) can expand participation and bolster public trust in data. One of the most notable examples of CGD is Open Street Maps, an open map of the world built through local, voluntary data collection to the benefit of local data users. Such data initiatives can bring about decidedly democratic outcomes. For instance, teenagers in Rio de Janeiro were enlisted to fly kites mounted with cameras, and the resulting aerial data were then successfully used to push for the removal of garbage piles and infrastructure improvements. Under the right circumstances, democratizing data can empower underserved organizations or groups to address issues that may have been otherwise overlooked (Wu et al., 2021, Sawicki and Craig 2007, Roy and Reece, 2010).
Data-Driven Democratic Forces Face Growing Headwinds
Yet we are also observing the unequal concentration in ownership of ever-increasing amounts of data by centralized bodies, such as Big Data companies and governments. This trend has serious, undemocratic implications, and it is worsening accountability and existing power imbalances. These developments are made possible by efforts to promote the ambiguities around data ownership which can justify the appropriation of data under the assumption that data is open for capitalization (Couldry and Meijas, 2019).
Furthermore, in the age of ‘fake news’, mistrust in data sources and policymakers is rampant. Even when dealing with credible data sources, examples of misinformation abound. The VAERS (Vaccine Adverse Effects Reporting System) from the US CDC provides an open, citizen-generated dataset that has become highly politicized during the COVID-19 pandemic. Statistics from VAERS often show up online without proper context or analysis as part of anti-vax propaganda. Though the dataset has been touted for providing low-cost crowd-sourced information on side effects of the vaccines to inform vaccine efficacy and performance, its widespread misuse would suggest simply having access to more information does not directly lead to better decisions or to achieving democratic ideals.
Significant attention has been given to the need for increased literacy in our digital age, including data literacy, data infrastructure literacy, and critical big data literacy. While certainly important, such calls for literacy place the responsibility on data users, rather than data providers, to make decisions about the validity and usability of data. Corporate and governmental data providers also need to be held responsible for the information they collect and disseminate. Greater accountability can be achieved in part through strengthened governance mechanisms.
Data Governance to Realize Data’s Democratic Potential
Currently, it could be argued that data is being over-collected yet under-protected. The data revolution requires responsible data stewardship and, as suggested by Kevin O’Neil, a mixture of legal, institutional and technical governance tools, for it to live up to its democratic potential. In pursuit of a rights-based data revolution, a number of innovative approaches have been described and tested. This list includes:
Indigenous Data Sovereignty/Indigenous Data Governance, which gives the right of data ownership to each Native nation or indigenous group.
Data trusts and data collaboratives give fiduciary duties to trustees that make decisions on data rights and data usage for individuals.
De-identification tools protect the privacy of data subjects.
Transparent collection practices encourage inclusive data products.
Web decentralization mitigates the risk of government surveillance, censorship, and hackers targeting centralized data repositories. It also breaks up power held by a few select companies.
Publishing meta-data is important for open data programs to allow users to judge the quality of data and how to responsibly use it.
Open data for Main Street uses libraries to teach data literacy programs, foster trust in open data within communities, and increase equity and inclusion in using data.
The democratic potential of data will only be balanced against the concentrated power of data through implementing an effective data governance regime that includes a mixture of solutions that foster inclusion, equality and accountability.