The Evolving Trade-offs between Data Collection, Disclosure and Privacy: An Analysis of COVID-19 Dashboards and Implications for Data Governance
This article is written by Veronica Li, a first-year PhD student in the Department of Science, Technology, Engineering and Public Policy (STEaPP) at University College London. She is interested in exploring effective data governance and opportunities for citizen participation in sustainable smart cities. This is a follow-up to the article Increasing resilience via the use of personal data: Lessons from COVID-19 dashboards on data governance for the public good in the Data & Policy Journal, which she wrote with her MPhil supervisor Professor Masaru Yarime at the Division of Public Policy at The Hong Kong University of Science and Technology.
How can COVID-19 case data be used for the public good?
As the COVID-19 pandemic evolves, many governments continue to use public data platforms to keep citizens updated about the COVID-19 situation in their local communities. Data platforms of note are dashboards like the ones by Johns Hopkins University and the World Health Organization, which act as a snapshot of the spread of COVID-19 across the world and within countries at a specific time period. This has in some cases led to data privacy protections being temporarily rolled back so that key data about COVID-19 patients could be acquired and shared. If done incorrectly, this approach could lead to rising concerns about data privacy violations and erode trust in governments.
Hence, I submitted an article to Data & Policy last year exploring the lessons that could be learned from COVID-19 dashboards regarding the salient management of personal data for the public good. Through a comparative analysis of COVID-19 dashboards around the world as well as an in-depth case study of Hong Kong, I came to the following conclusions:
1. COVID-19 case data can be treated as a public good with free-riders.
Although 60% of survey respondents would demand sensitive data (e.g., locations of residence or quarantine), only 40% would be willing to share such data about themselves.
2. Contextual factors such as culture and population density could affect how COVID-19 case data is presented.
Disclosing a patient’s building of residence poses a much lower risk of re-identification in a city where buildings could house hundreds of people, compared to in the suburbs where houses are only occupied by one household each. Cultural beliefs about data privacy could also affect the types of data that governments decide to disclose, as the breakdown of individual cases was more prevalent in cities in Eastern countries (e.g., Hong Kong, Singapore, Tokyo) than those in Western countries (e.g., London, New York).
3. Salient data standards should be co-created with citizens and other relevant stakeholders.
In Hong Kong, the high level of personal data disclosure was a result of continuous interactions where citizens demanded data, the Centre for Health Protection supplied it, and the local media served as an intermediary. Although this process is not perfect — minority groups are often underrepresented in Hong Kong’s local media — this collaborative approach allows for data standards to be agreed upon and for stakeholders to trust in the data that is collected and shared. Other jurisdictions should consider using similar approaches to determine contextually appropriate trade-offs between data disclosure and privacy.
What has changed since the article was published (November 2021)?
The Omicron variant of COVID-19 emerged, overtaking Delta as the dominant variant and leading to a global surge in cases. Even Hong Kong, which previously declared zero local cases over the course of several months, suddenly became an epicentre of the pandemic with daily cases surpassing the thousands. Hong Kong’s dashboard reflects these dramatic changes in several ways, from the plastering of red dots across the city map representing the prevalence of COVID-19 in the past 14 days to the replacement of individual case data with the data of affected buildings in different districts. Similar changes were made to the Tokyo Metropolitan Government dashboard, which no longer has an embedded table of individual cases, and the Seoul City dashboard, which has replaced individual case data with aggregate statistics. These governments likely decided on these changes in case reporting because the sheer volume of daily cases has affected their capacity to collect granular data — this appears to be the case for Tokyo, which is missing information on their latest patients’ residence, occupations, and pathways of infection because the data is still being collected. Another reason for this change could be that cities such as Shanghai and Singapore are experiencing far more asymptomatic cases due to a combination of factors such as virus mutation, high vaccination rates, and early detection. This could lead governments to lower the priority of collecting granular data on individual cases.
What are the policy implications of these changes?
The recent changes in the design of COVID-19 dashboards reflect the need for adaptable and resilient data systems in the face of crises, such that governments can maintain their capacity to collect, process and visualize important data regardless of the scale of the problem. Governments will constantly need to maintain a careful balance between meeting stakeholder demands for data and preserving privacy protections as they navigate an increasingly complex and rapidly changing world, all while being constrained by resources such as manpower and time. This balancing act will become increasingly difficult, but governments and citizens could continuously collaborate in the data governance process to make this possible.
