Data Governance in Public Health Crisis: Trade-off Mechanisms, Temporal Changes, and Regional Variances of China’s Health Code During COVID-19

Abstract

During the COVID-19 pandemic, China launched the Health Code, the first contact tracing application (CTA) around the world, marking the government's shift from traditional governance to data governance in dealing with public health crises. The Health Code practice demonstrated the complexity of data governance. First, in terms of technology design, China's Health Code demonstrated a unique governmental trade-off between the governance efficiency of curbing the pandemic and data privacy. China strengthened surveillance while significantly weakening personal data protection, which is in sharp contrast to some other countries developing their own CTAs. Second, in the process of technology implementation, the operational status of Health Code data governance showed significant temporal differences, from efficient operation in the early stage to disorder in the later stage, and finally to decommissioning. Third, the operations of Health Codes in different regions of China also varied, which could be rooted in the capability of local governments in political authority and resources. The complexity of the Health Code practice provides a valuable window for us to understand the efficiency-privacy trade-offs, cyclical changes, local adaptation, and national capacity boundaries in data governance.

Given the above phenomenon, this thesis aims to answer the following questions: 1) What mechanism shaped China's data governance model that strengthened governance efficiency in controlling the pandemic and weakened privacy protection? Why was the model adopted by China significantly different from that adopted by other countries? 2) Why did the operational status of Health Code data governance show cyclical changes? 3) Were there regional differences in the Health Code implementation in different regions of China, and what were the causes and impacts of these variances?

Existing literature has not yet provided a comprehensive and dynamic framework to explain these issues. To this end, we propose a three-dimensional framework of data governance to examine data governance from the three dimensions of information and communication technology, data protection law and government policy. These dimensions may interact differently with each other in different countries or regions and at various time points, thus jointly shaping multiple models of data governance.

This study uses a multi-case comparative analysis method to study the rich differences within data governance. In the cross-national comparison, we selected the Chinese National Health Code, the Shanghai Health Code, and the Shenzhen Health Code as Chinese cases, and supplemented them with Singapore's CTAs TraceTogether and SafeEntry and Germany's CTA Corona-Warn-App for comparison; in the temporal comparison, we focused on three key stages in the data governance of Health Code, namely the Wuhan lockdown, the Shanghai lockdown and the Beijing protest, to discover the mechanisms shaping the operational status of data governance; in the regional comparison, we investigated the operation of the Health Code systems in City A, City B and County C.

Through the analysis of privacy policies of CTAs in China, Germany and Singapore, we found that three legal-technical dynamics shaped three typical models of data governance during the technology design phase. Specifically, under a "Technology challenges law" dynamic, a centralized and intrusive model that features weak data protection but intense surveillance for high governance efficiency emerged in China; shaped by a "Law constrains technology" dynamic, a decentralized and unintrusive model that trades governance efficiency for individual privacy appeared in Germany; and Singapore struck a balance between law and technology, thus opting for a semi-centralized and semi-intrusive model balancing the surveillance and data protection.

By analyzing news reports on the three stages of Health Code implementation, the Wuhan lockdown, the Shanghai lockdown and the Beijing protest, we concluded that the interplay of government policy and digital technology shaped the operational status of data governance. During the three years of the COVID-19 pandemic, the Chinese central government maintained campaign-style governance to efficiently mobilize society to adopt the Health Code and fight against the virus. Technology initially facilitated pandemic control effectively, but its effectiveness waned due to changes in the external environment. Thus, Health Code data governance shifted from the initial high-efficiency status to a disordered status. The disorder ignited citizens' protests against China's Zero-COVID goal, which finally prompted the Chinese government to end its campaign-style governance for the Zero-COVID goal and terminate the use of the Health Code.

We investigated the regional operation of the Health Code in City A, City B and County C by interviewing grassroots government staff and conducting participant observation with the role of visitors and residents in the three regions. Our analysis reveals that City A, which has sufficient horizontal political authority and abundant financial resources, presented a data pool model of data governance. City B, with limited horizontal political authority but abundant financial resources, adopted a data flow model in Health Code implementation. County C, which has limited horizontal political authority and limited financial resources, opted for a manual control model to fight against the pandemic. These findings imply that local governments need to leverage their political authority and resources to execute the policy goal assigned by the central government. The variances in the capabilities of local governments in these two dimensions can shape their technology implementation. These models had their own advantages in achieving high efficiency of pandemic control, but they were all affected by the environmental changes and gradually became ineffective.

The findings of this thesis reveal the complexity and multidimensionality of data governance, emphasizing the dynamic interaction of law, technology and policy, as well as the impact of state power, temporal evolution and local differences on governance models. The three-dimensional framework of data governance we proposed fills the gap in existing literature and provides a new perspective for understanding the trade-off between efficiency and privacy, technical limitations and differences in policy implementation. In practice, our study inspires the design and management of future data governance tools, emphasizing the need to strengthen legal protection of citizens' data privacy rights while pursuing efficiency and building a resilient mechanism to adapt to environmental changes.

Date of Award	7 May 2026
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Fen LIN (Supervisor)