Yu Jianxing: Practices and Frontiers of Digital Public Governance

Digital public governance refers to public governance conducted in an era marked by digital technology. It encompasses "governance by technology," which utilizes digital technology as a tool for managing public affairs; "governance of technology," which addresses digital technology itself and its derivative problems, risks, and issues; and the governance of the tension between technical logic and the logic of public values, seeking institutional reconciliation and the reconstruction of order. What are the current directions and trends in the development of digital public governance? What limits are contracting the boundaries of its extrapolation? How should we shape the future of digital public governance? In the context of the New Era's deep advancement of the Digital China ^[1] initiative, answering these questions represents both a forward-looking, normative reflection on digital public governance theory and a practical outlook on the future of digital development for China and the world.

Directions and Trends of Digital Public Governance

"Direction" answers where governance might lead amidst the technological wave, while "trends" reveal the momentum and inertia already manifest behind these trajectories. The reason we must synthesize the directions and trends of digital public governance is not to prescribe a fixed path, but to capture the clues of governance evolution amidst uncertainty, thereby understanding how digital public governance continuously breaks through old boundaries and generates new orders.

1. Knowledge Base: From Empirical Induction to Data-Driven

Traditional decision-making models rely primarily on two types of knowledge sources: empirical judgments accumulated by practitioners through long-term practice, and theoretical inductions formed by experts and scholars based on historical data and existing research. While experience and theory could constitute a feasible basis for decision-making and support policy generation and adjustment within the relatively stable environment and limited information scale of industrial society, this logic of decision-making also possesses clear limitations, such as information lags, narrow data sources, and a lack of foresight regarding complex situations.

With the widespread application of digital technologies like big data and artificial intelligence, the knowledge base of public governance is undergoing a profound transformation. More and more information regarding societal operations can be collected, processed, and analyzed in real-time; data has become an indispensable foundational component of public governance. This epistemological approach, which takes data as its starting point, is driving public governance toward a new knowledge paradigm—a data-driven logic of cognition and decision-making. The State Council’s "Guiding Opinions on Strengthening the Construction of a Digital Government" ^[2] proposes to "fully leverage the role of data as a foundational resource and an innovation engine, and to improve the scientific level of government decision-making and the efficiency of management and services." This indicates that data-driven approaches have become a major direction for current governance transformation and are gradually ascending into institutional arrangements.

In the process of governance operations, this shift in the knowledge base can be summarized as the "data-modeling-simulation" cognitive chain. On one hand, the scope and frequency of data collection are constantly expanding; many facets of social life are being digitally recorded, providing rich and real-time knowledge material for governance. On the other hand, algorithmic modeling is increasingly embedded in institutional operations. Through structural analysis of massive data, models not only perform problem identification but also serve as important cognitive tools for risk assessment and trend prediction. Looked at more deeply, simulation and emulation technologies are beginning to be used in policy processes, enabling the government to run simulations of different schemes in a virtual environment, thereby accumulating knowledge and experience before a policy is implemented. Consequently, the generation of governance knowledge will exhibit new characteristics: moving from reliance on limited experience and static induction toward a dynamic iteration more reliant on data; and moving from a focus on ex-post summation toward a greater emphasis on ex-ante foresight.

2. Methodology: From Linear Reasoning to Prediction and Experimentation

In existing public governance research, policy analysis and formulation rely heavily on linear reasoning and empirical judgment. This research path mostly rests on two major assumptions: first, that social processes are relatively stable and causal relationships can be grasped through ex-post induction; second, that the subjects of governance are somewhat homogeneous, and that a "greatest common denominator" or "mean value" can represent the overall law. However, the real world is often complex and volatile, and these assumptions face severe challenges: policy outcomes are usually accompanied by feedback effects and path dependence, meaning causal relationships are not simply linear; the differentiation among governance subjects is constantly expanding, and average values often mask the true situation of certain groups.

The digital era has brought tools and methods to break through these limitations. Big data analysis can identify patterns and trends within massive amounts of information, making up for the deficiencies of traditional statistical methods in scale and real-time capability. Machine learning algorithms optimize prediction and classification abilities through continuous training, enabling governance to respond dynamically to complex environments. Emulation and simulation technologies allow policies to test multiple schemes in virtual environments beforehand, without having to rely entirely on single-point trials.

At the same time, new tools and methods are not merely isolated or mechanical additions. They act simultaneously across the entire governance process, forming a multidimensional shift: in the temporal dimension, governance gradually moves away from "ex-post summation" toward "ex-ante foresight" based on real-time data and predictive models; in the logical dimension, policy generation no longer relies on static causal chains but is corrected through dynamic iteration; in the dimension of subjects, policies no longer refer only to the "majority" but respond to differences between groups and regions through finer-grained data analysis; and in the spatial dimension, policy experimentation sheds the constraints of single-point trials, using virtual simulation and digital twins to conduct parallel dress rehearsals and comparisons across multiple scenarios. The holistic reshaping of methodology by digital tools makes public governance more forward-looking, flexible, sensitive to differences, and spatially extended.

3. Subject Relations: From Single-Center to Multi-Dimensional and Multi-Plural

The relationship between subjects is a classic topic in public governance discussions. Government-centrism is a common feature of traditional public management theory and practice. With social development and the diversification of public needs, single-center structures have gradually revealed their governance limitations. Since the late 20th century, the academic understanding of subject relations has gradually shifted toward pluralistic interaction, and the focus of governance has moved from control toward synergy.

In the digital era, the evolution of subject relations has upgraded from "plural" to "multi-dimensional." Not only has the number of subjects participating in governance surged and their types expanded, but their modes of existence and logics of agency have also become heterogeneous. "Subjects" in the governance system are no longer limited to human individuals and organizations but extend to technology itself, forming complex interactions across physical, social, and digital spaces, constituting multi-dimensional and multi-plural subject relations.

The emergence of such subject relations is the result of mutual reconstruction between digital technology and governance structures, manifesting in practice as trends toward synergization, platformization, and networking. For example, when facing public issues like climate change or platform economy regulation, a single subject finds it difficult to cope alone; the effectiveness of governance depends on the complementarity of capabilities and resources among different subjects. Governance methods have moved from departmental fragmentation toward cross-boundary collaboration, and this synergy relies on platform-based technical and institutional conditions. In this process, the functional boundaries of digital platforms are being reshaped, transforming from information channels into intermediaries and operating systems for interaction among multiple subjects, connecting government, enterprises, society, citizens, and technological subjects. Amidst the evolution of public governance subject relations toward a multi-dimensional direction, state authority has not dissolved; rather, it is re-embedded into the network through institutional design, data regulation, and algorithmic governance, becoming a key force for maintaining overall coordination and public values. This shift indicates that in the digital era, public governance is moving from centralization to distribution, and from a logic of control to a logic of synergy.

4. Governance Ecosystem: From Network Logic to Ecological Logic

Traditional bureaucratic systems maintain order through hierarchies and rules, emphasizing advantages in resource mobilization and unified action. The subsequent rise of governance network theory attempted to explain coordination logic under multi-subject interaction, but this assumption has limited applicability in highly dynamic and uncertain situations. In response, academia has introduced the "ecosystem" metaphor and interpretive framework, emphasizing that public governance does not operate in isolation but is embedded within a whole constituted by the interaction of multiple subjects, institutional arrangements, and the techno-social environment.

The digital era has further amplified the limitations of traditional governance while giving rise to new conditions for response. The scale and flow of data are increasing exponentially, meaning governance processes can no longer rely on single-center collection and processing but must form distributed response networks. The widespread application of digital platforms and algorithms has reshaped resource and power structures, with government, market, and social subjects nesting and contending on platforms in new ways. Governance scenarios are constantly expanding, with virtual and physical spaces intertwining, making governance objects and boundaries more fluid. The resolution of complex problems requires policies to iterate continuously and adjust in real-time, rather than being one-off formulations and executions. Compared to traditional governance logic, "ecological logic" demonstrates stronger explanatory power and practical adaptability due to its emphasis on system self-adaptation, continuous learning, and trust mechanisms.

The transition from a network logic to an ecological logic does not negate hierarchy or networks but places both within a larger ecosystem to be understood: the resilience and efficacy of governance depend more on the quality of multi-subject interaction and whether trust and learning mechanisms can be institutionalized, contextualized, and continuously operated to support the continuous generation of the "public" nature of governance. From a trend perspective, the governance ecosystem is gradually evolving from localized, problem-oriented collaborative practices into an institutionalized systemic logic. Future digital public governance may manifest more as an open system across organizations, fields, and levels, emphasizing self-adaptive adjustment and resilience. This means the governance ecosystem is not merely a passive response to complexity but has become an active framework for reshaping the governance order.

The Limits of Digital Public Governance

Currently, digital technology seems to offer unprecedented possibilities for public governance, but possibility does not equate to infinity. Digital public governance has its inherent limits; these are neither simple technical defects nor solitary institutional obstacles, but structural boundaries emerging from the interaction between technology and society.

1. The Limits of Technology Itself

Public governance concerns the public interest, and its objects are complex, volatile social systems and agents with human agency. When digital technology is introduced into the field of public governance, potentially existing technical defects will touch upon citizens' basic rights, social equity, and even national security; they may even evolve into unbearable social risks and crises of trust.

First, the limit of authenticity constituted by data distortion. Translating complex social reality into processable data inevitably involves information loss and selection bias. Therefore, when any governance object is represented as data to enter governance workflows, measurement errors, selection biases, and loss of context are unavoidable.

Second, the stability challenges brought by algorithmic bias. Social inequalities inherent in training data can be learned and solidified by algorithms, thereby forming technical discrimination. At the same time, digital systems operate within highly non-stationary social environments; as time, context, and institutions change, models and algorithms often experience "distribution shift," causing their predictions and judgments to lose stability, thus affecting the fairness and reliability of governance decisions.

Third, the dilemma of controllability caused by the amplification of systemic risks. The scale, connectivity, and automation speed of technology mean a single error can instantly evolve into a systemic risk. For example, if a financial market circuit breaker fails due to an algorithmic error, it could trigger a chain reaction, resulting in enormous losses.

Fourth, the ambiguity of responsibility attribution weakens accountability. Cross-level "cloud-edge-end" architectures, closed-source models, and transnational supply chains make the decision-making paths and responsibility attribution in digital public governance complex and difficult to trace. Meanwhile, the "black box" nature of algorithms makes it difficult for the public to understand decision-making processes, let alone engage in effective appeals and corrections.

2. The Limits of Instrumental Rationality

The over-expression of instrumental rationality was a malady born in the industrial era, and it is even more pronounced in the digital era. When instrumental rational tendencies such as "efficiency first," "scale priority," and "optimal solution orientation" are applied to public governance, they may conflict seriously with values such as equity, plurality, and trade-offs.

First, the efficiency paradox. In algorithmic optimization and automated systems, efficiency is usually set as the primary goal. However, an increase in administrative efficiency does not necessarily lead to an enhancement in governance efficacy. Overemphasizing the logic of efficiency can erode the public foundation upon which governance exists. For instance, although algorithmic credit can speed up lending, it may solidify discrimination against certain groups due to model bias.

Second, scale heterogeneity. The development of digital technology relies on data and computing power; its inherent tendency is toward standardization and scale. But governance objects are highly heterogeneous, and conditions among groups and regions vary greatly. Standardized platforms often find it difficult to respond to this diversity, which can lead to new forms of digital exclusion. For example, a unified government service system may leave the elderly or those with insufficient digital skills feeling lost.

Third, the illusion of the "optimal solution." Technical systems rely on optimization models, assuming that a problem has a unique "best" answer. However, in the context of governance, public problems are mostly...

"Wicked problems" lack a single optimal solution; public governance faces multiple value tensions that require trade-offs and reconciliation. Evidently, the "optimality" of an algorithm cannot substitute for the consultation, compromise, and institutional balance required for governance.

3. Limits of Institutional Adaptation Institutions are composed of three foundational elements—regulative, normative, and cultural-cognitive. Respectively, these elements require institutions to remain predictable in time, to define what is acceptable in terms of values, and to clarify accountability in terms of cognition. However, digital technology brings structural tensions, forming a triple limit on institutional adaptation.

First is the rate mismatch in the regulative dimension. Institutions provide stable expectations for public governance over time through written rules, processes, and procedures; conversely, digital technology is characterized by high-frequency iteration and continuous deployment. A natural mismatch in speed exists between the two.

Second is the loss of boundaries in the normative dimension. If institutions draw boundaries too broadly, they may stifle innovation and cause a waste of resources; if drawn too narrowly, they may leave a vacuum of risk. In practice, "principle-based legislation" and "mid-process refinement" are often adopted for dynamic adjustment, but this results in institutional boundaries being in a perpetual state of "catching up," unable to synchronize with technological development. This brings risks such as the disordering of institutional values and boundary confusion.

Third is the ambiguity of responsibility in the cultural-cognitive dimension. Institutions are meant to define "who decides, who executes, and who is responsible" through a shared cognitive framework. However, under the cross-platform and cross-departmental architecture of digital public governance, this framework is weakened. Especially in scenarios involving the participation of diverse subjects, it is often difficult to clarify whether final responsibility should be borne by the developer, the platform, the user, or the regulatory department.

4. Limits of Governance Logic "Governance logic" is the cognitive framework of public governance, aimed at transforming complex social phenomena into actionable schemes through problem definition, causal induction, and value prioritization. The widespread application of digital technology may lead to complex problems that transcend traditional governance classifications and cognitive frameworks.

First is logical mismatch. At the international level, technical governance emphasizes synergy and sharing, while the logic of geopolitics prioritizes relative gains. At the domestic level, technical logic pursues universalization and scalability, whereas local governance must respond to regional differences and group diversity.

Second is the rupture of causal chains. The development of digital technology has broken through the boundaries of traditional linear causal models. For example, Artificial Intelligence systems possess learning capabilities; their behavior is not entirely preset externally but evolves continuously with environmental feedback. Once smart contracts on a blockchain are deployed, they execute automatically, leaving no room for mid-course correction. This makes it difficult for governors to trace causes or predict subsequent evolution even if they identify a problem.

Third is the difficulty of value prioritization. Digital issues are often nested with conflicting value objectives. For instance, platform recommendation systems pursue both user stickiness and commercial profit, while also concerning information diversity and the health of public opinion. Emphasizing the former exacerbates "information cocoons" ^[3], while emphasizing the latter may weaken the sustainability of the business model. The prevalence of such multiple value tensions means that traditional single-objective prioritizations (such as "efficiency first" or "stability above all") are no longer effective.

The aforementioned limits of digital public governance lead, in the final analysis, to the decline of human centrality. The "public" in public governance originates fundamentally in "people"; its core value orientation is to respond to human needs, protect human dignity, and realize human development. However, from the perspective of algorithms and models, individuals are described as calculable variables. While this improves the operability of governance, it weakens the response to human difference and subjectivity. Meanwhile, excessive reliance on technical systems may lead to the degradation of critical thinking and consultative capacity of human decision-makers and the public. The accumulation of these factors makes it possible for the governance process to slide from "humanity as the end" toward "technology as the end."

Placing People at the Center of Digital Public Governance

On the quadrant of real-world development, the evolution of technology is certainly inevitable and is driven by market forces in many fields, its path full of uncertainty. However, the legitimacy and directionality of institutional evolution can be more clearly shaped around "people." Facing the future, digital public governance must place people at the center of institutional design and technical application.

1. Acknowledging Human Diversity The objects of digital public governance are highly heterogeneous. Differences in gender, age, class, region, and cultural background cause public needs to present a complex and diverse pattern in form and intensity. If governance logic relies excessively on the majority or the mean to set standards, it will inevitably create new inequalities. Therefore, diversity must be incorporated into institutional design so that the policy execution process embodies "difference sensitivity." That is, the needs of different groups should be identified and responded to, entering the governance process through the rational allocation of public resources, thereby establishing a more resilient and inclusive social foundation for digital public governance.

2. Humanity as the End This needs to be translated into actionable arrangements at the institutional and policy levels. First, in setting policy goals, the improvement of conditions for human existence and development must be placed first, rather than using technical feasibility or economic efficiency as the sole criterion. For example, in the digital transformation of employment services, one should not only look at the overall job-matching rate but also pay attention to the accessibility and experiential differences for vulnerable groups. Second, in institutional design, value boundaries should be established to clarify which areas of "publicness" cannot be replaced by instrumental logic. For instance, data usage must undergo informed consent procedures, and social assistance must retain channels for human intervention. Third, in evaluation systems, indicators oriented toward "human outcomes" should be introduced, rather than using efficiency and coverage as the only performance standards. For example, the effectiveness evaluation of a public health platform should not only count vaccination or consultation rates but also examine the affordability, sense of gain ^[4], and long-term health improvements of the people.

At the same time, "humanity as the end" is not only a requirement for the governance side but should also run through the entire process of technological research and development (R&D). Humanities and social science education for the designers and developers of digital technology can be strengthened, integrating concepts such as fairness, responsibility, ethics, and inclusion into the early stages of R&D to achieve the integration of engineering and the humanities, preventing technology from decoupling from public values.

3. Cultivating the "Public Person" A "Public Person" refers to an individual capable of understanding public affairs, fulfilling public responsibilities, and participating in governance as an active subject within institutional and technological platforms. First, the "Public Person" needs to possess "publicness" awareness. In a digital governance environment, individuals are often shaped by algorithmic pushes and personalized recommendations; without public awareness, they easily fall into information cocoons or become part of the "silent majority." To this end, individuals should be guided through the education system, public media, and community organizations to transcend the limitations of private interest and recognize the intrinsic link between public affairs and individual well-being. Second, the "Public Person" needs to possess the capacity for governance participation. Digital platforms are not only carriers of services but also arenas for policy communication, public consultation, and collective decision-making. "Public Persons" should be able to express opinions, supervise policies, and participate in co-governance on these platforms. This requires institutions to provide accessible participation channels for citizens, such as open data platforms, digital hearing systems, and community co-governance applications, allowing individuals to form a stable subjective role in public governance. Third, the "Public Person" should also possess digital ethical literacy. Whether an individual can respect the rights of others and follow principles of fairness and responsibility in data sharing, information dissemination, and online interaction determines whether the public space can maintain basic trust and order.

4. Improving Digital Governance Capability The effective operation of digital public governance depends not only on institutional design and technical supply but also on whether governance subjects possess matching governance capabilities. This capability is different from general technical skills and is not limited to common sense in public affairs; it is a composite capability spanning technology and governance. Its content includes at least three aspects: First, the ability to understand digital governance structures. Digital governance systems usually present a complex network across departments, levels, and platforms, where linear causal logic struggles to reveal interactive relationships. This requires governance subjects to understand the coupling mechanisms between institutional rules, technical platforms, and social behavior, grasping the holism of governance operations. Second, the ability to apply and supervise technology. Against the backdrop of Big Data and AI being widely embedded in governance, individuals and organizations must not only learn to use these tools but also be able to identify the risks and biases therein—such as understanding the basic logic of algorithmic decision-making and identifying potential biases in information pushes. Third, the ability for cross-border synergy and institutional innovation. Many problems in digital governance, such as data sharing, personal privacy, and network security, transcend the boundaries of a single department or profession. This requires governance participants to possess not only the qualities of negotiation and cooperation but also the creativity to propose improvement schemes within the institutional framework.

To place people at the center of digital public governance, preventing risks or maintaining order is certainly important, but more important is stimulating human creativity through institutional design and technical application. This requires a dual effort: on the one hand, institutions should lower the threshold for using public data, computing power, and technical tools so that more social subjects can enter the innovation process equally; on the other hand, cultural and educational systems should cultivate a spirit of exploration and cooperation, enabling individuals to respond creatively through digital tools when facing uncertainty. Through this process, human agency and technological potential can be combined, providing the impetus for generating new public orders and institutional forms for digital society.

In summary, if we define the "frontier" as the field that is most challenging, most in need of breakthrough, and which will determine the future landscape, then without a doubt, the frontier of contemporary public governance is shifting toward digital public governance. This process is full of possibilities but also faces many risks and challenges. How to reshape order between technology and values, find breakthroughs between boundaries and possibilities, and open paths between risk and creativity will be an endless frontier.

(Author: Yu Jianxing, Secretary of the CPC Committee of Zhejiang Gongshang University and Dean of the Institute of Social Governance at Zhejiang University) Source: Guangming Daily (January 9, 2026) Web Editor: Huihui