Policy narratives and digital compliance: comparative governance during COVID-19 in Indonesia, Japan, and China

1 Introduction

Digital technologies moved rapidly from administrative support tools to central instruments of public governance during the COVID-19 pandemic. Governments deployed contact-tracing applications, mobility-monitoring systems, digital health certificates, and extensive data infrastructures while operating under political pressure and considerable scientific uncertainty (Kędzior, 2021; Liu and Graham, 2021; Kitchin, 2014). Yet their political consequences were far from uniform. In some settings, digital measures came to be accepted as legitimate extensions of emergency public health policy. In others, comparable technologies became associated with distrust, political contestation, privacy concerns, and questions about the proper reach of state authority (Newlands et al., 2020; Shen et al., 2023).

The divergence is not readily explained by technological design alone. Governments confronting the same emergency adopted broadly similar digital instruments, but legitimacy and public compliance moved along different paths. Existing scholarship identifies institutional capacity, administrative coordination, regulatory quality, and technological capability as important explanatory conditions (Brummer and Ueno, 2024; Ho, 2023; Kshetri, 2014; Shen et al., 2023). These conditions clearly shape what governments can implement. They tell us less, however, about why comparable resources and technologies can generate different degrees of public acceptance once implementation is underway.

This comparison brings the politics of meaning into view. Under conditions of uncertainty, governments have to explain why extraordinary measures are necessary, proportionate, and worthy of public confidence (Stone, 2013; Shanahan et al., 2018). Administrative performance is part of that process, but it is not the whole of it. The legitimacy of an intervention also rests on how officials interpret unfolding events, justify difficult choices, and respond when circumstances change.

The problem becomes especially pronounced in digitally mediated crises. Official announcements, media accounts, expert interpretations, and public responses circulate together, often before implementation has settled into a stable form (Berry et al., 2007). Governments are judged by both their actions and whether the explanations attached to them remain intelligible as conditions evolve. Technical capability may make a policy possible; it does not, on its own, secure sustained cooperation.

The Narrative Policy Framework (NPF) provides a useful starting point because it explains how policy narratives define public problems, evaluate political authority, and support governmental action (Jones and McBeth, 2010; Jones et al., 2014; Dunlop et al., 2021; Schlaufer et al., 2022). Its application to digital governance remains comparatively limited. Much of the existing literature concentrates on institutional arrangements, legal frameworks, and technological performance, while giving less attention to the interpretive work through which governments attempt to sustain trust, legitimacy, and compliance as implementation proceeds. The gap is particularly visible in comparative studies of Asian governance.

A related limitation concerns time. NPF research has often examined narrative content—characters, plots, causal stories, and policy messages—without giving equal attention to what happens after a narrative enters a changing policy environment. In digital governance, that omission matters. Official interpretations can be revised, challenged, amplified, and recirculated while implementation is still in motion. Narrative adaptation consequently becomes part of governance itself, because governments must preserve the credibility of their explanations while the circumstances being explained continue to shift.

The analytical issue, then, concerns whether a persuasive narrative constructed at the outset can remain politically credible during implementation. New evidence emerges, public expectations move, institutional constraints become clearer, and political pressures change. Governing institutions must respond to these shifts without allowing the interpretive thread of the policy to disappear.

This temporal problem motivates the concept of Narrative Velocity: the capacity of official policy narratives to evolve during implementation while retaining enough coherence to sustain legitimacy. The concept directs attention to the temporal work of interpreting crises, defending policy choices, and revising explanations as conditions change.

Narrative Velocity therefore treats policy narratives as evolving components of governance and extends NPF by making narrative change during implementation analytically visible without altering the framework’s established foundations.

The empirical analysis compares Indonesia, Japan, and China. All three adopted broadly comparable forms of digital governance during the pandemic, but they did so within markedly different political, administrative, and communicative environments. That combination of policy similarity and institutional diversity provides leverage for examining why similar technologies generated divergent trajectories of legitimacy and observable compliance.

Indonesia represents a decentralized democratic system in which rapid digital expansion exposed persistent coordination problems across government institutions. Cybersecurity concerns, fragmented administrative authority, and uneven implementation capacity often challenged the government’s ability to maintain a coherent public account, contributing to declining confidence (Kurniasih et al., 2024; BBC Indonesia, 2024). Japan presents a different trajectory. Despite advanced technological infrastructure and comparatively strong administrative institutions, COCOA did not produce the sustained public engagement initially expected (Brummer and Ueno, 2024; Nakamoto et al., 2020). The case shows why technological sophistication and administrative competence do not necessarily translate into durable legitimacy or sustained behavioural compliance.

China followed a markedly different path. Its Health Code system operated within a highly centralized political environment in which state institutions exercised extensive control over digital infrastructure and public communication (Liang, 2020; Zhang, 2022). Under these conditions, centralized communication, administrative coordination, and policy implementation were closely connected within the broader architecture of crisis governance.

Considered collectively, the cases allow the narrative process to be examined without reducing the explanation to regime type or administrative capacity. Political institutions are treated as the settings in which policy narratives acquire credibility, encounter contestation, and acquire practical consequences.

Against this background, the study addresses the following question:

• How did policy narratives evolve during implementation, and how were these dynamics associated with variations in observable digital compliance across Indonesia, Japan, and China?
• The analysis is qualitative and comparative; it does not test a deterministic causal sequence. It examines the interaction of Narrative Coherence, Narrative Velocity, and Narrative Centralization with institutional settings and asks how these processes are associated with different configurations of legitimacy and observable compliance. Digital technologies provide the empirical setting. The analytical focus is the political work through which governments interpret crises, justify policy choices, and seek to sustain cooperation under uncertainty.
• The study makes three related contributions. First, it extends the Narrative Policy Framework by bringing the temporal dynamics of narrative evolution into the analysis through Narrative Velocity. The shift is from treating narrative structures primarily as established configurations toward examining how governing narratives are revised as implementation unfolds.
• Second, broadly similar digital governance technologies can produce markedly different political outcomes even when governments confront the same policy problem. Institutional capacity and technological sophistication remain important, but the evidence points to their interaction with the narratives through which policies are explained, defended, revised, and interpreted over time.
• Third, narrative management is treated as a substantive dimension of state capacity. Sustaining a credible public account throughout implementation extends beyond political communication. In digitally mediated governance, it becomes part of governing itself, particularly when institutions must respond to changing political and social conditions without losing the coherence on which public confidence depends.

2 Literature review

2.1 Data governance, privacy, and competing approaches

COVID-19 digital governance reopened longstanding questions about privacy, surveillance, and the legitimate boundaries of state authority. Contact-tracing applications, mobility monitoring, and large-scale data analytics moved from more specialized administrative settings into public health governance (Poell et al., 2019; Kędzior, 2021; Liu and Graham, 2021; Couldry and Mejias, 2019; Zuboff, 2019). Governments presented these technologies as emergency instruments, but their use also expanded access to personal data and revived concerns about privacy, cybersecurity, and the possible persistence of exceptional surveillance practices (Newlands et al., 2020; Bengio et al., 2020).

Three broad explanatory approaches are prominent in this literature. The first emphasizes institutional conditions: legal frameworks, bureaucratic capacity, regulatory oversight, and the distribution of administrative responsibility. Digital governance is more likely to function effectively where mandates are clear, agencies coordinate well, and data-protection safeguards are credible (Brummer and Ueno, 2024; Ho, 2023; Shen et al., 2023). These conditions matter for implementation, but they are less sufficient when governments with broadly comparable institutional resources encounter different levels of public acceptance.

A second approach places technological capability at the centre. Digital infrastructure, cybersecurity resilience, algorithmic capacity, and information-management systems can strengthen administrative coordination and emergency response (Kshetri, 2014; Liang, 2020; Liu and Graham, 2021). Technical capacity, however, does not guarantee political legitimacy. System failures and data breaches can rapidly undermine confidence, and similar technological resources may coexist with very different patterns of public acceptance (Williams and Woodward, 2015; Ho, 2023).

A third perspective shifts attention from capacity to the politics of meaning. It asks how governments construct persuasive interpretations of extraordinary interventions and present compliance as serving a broader collective interest (Stone, 2013; Schmidt, 2008). This perspective is closely related to the NPF, which examines how narratives define problems, assign responsibility, and shape evaluations of governmental action (Roe, 1992; Jones and McBeth, 2010; Shanahan et al., 2018). In digital governance, such interpretive processes influence how surveillance is understood, how privacy trade-offs are judged, and how individual rights are weighed against collective responsibility.

Narrative explanations remain less developed in the digital governance literature than institutional and technological accounts. The imbalance matters because digitally mediated communication permits official accounts to be challenged, revised, and redistributed while implementation is still underway. Variation among apparently similar initiatives therefore requires attention to both the capacity with which governments deploy technology and the ways they construct and revise the policy narratives surrounding it.

2.2 Narrative Policy Framework and digital governance

The NPF provides the principal analytical basis for this interpretive dimension. It explains how policy narratives define problems, justify policy choices, and seek to persuade (Jones and McBeth, 2010; Shanahan et al., 2018; Schlaufer et al., 2022; Schmidt, 2008). Narratives give political actors a way to render complex and uncertain circumstances intelligible.

Within NPF, policy narratives are organized through recurring configurations of settings, characters, plots, and policy morals that turn fragmented events into intelligible accounts (Jones et al., 2014). Their role extends beyond transmitting information. Narratives assign responsibility, identify legitimate courses of action, and establish normative grounds for intervention. Research across environmental governance, health policy, and social welfare likewise shows that narratives can shape policy conflict, institutional continuity, and policy change (McBeth and Lybecker, 2018; Schlaufer et al., 2022).

These functions become particularly consequential in crises. Formal authority and technical expertise may not be enough when interventions affect individual freedoms under conditions of uncertainty. Governments also need accounts of risk, responsibility, and collective purpose that make extraordinary measures intelligible and compliance defensible (Shanahan et al., 2018). Narratives therefore operate as political resources through which legitimacy can be established and sustained.

Digital governance makes the temporal dimension of this process difficult to ignore. Official messages can be contested, amplified, reinterpreted, and redistributed almost immediately through interconnected information environments. Narrative evolution consequently becomes more iterative and more visible while implementation is still taking place.

Recent NPF scholarship has generated important insights into narrative structure, coalition formation, belief systems, and policy learning (Dunlop et al., 2021; Capano et al., 2023; Birkland, 2006). Less attention has been given to what happens after official narratives enter rapidly changing digital environments. Governments may have to absorb new evidence, institutional constraints, competing interpretations, and sustained scrutiny simultaneously. Under such conditions, maintaining an intelligible policy account may be more consequential than preserving an unchanged one. Recent reviews also point to the need for further conceptual development in digitally mediated governance (Schlaufer et al., 2022; Kuenzler et al., 2025).

For comparative governance, the implication is clear: a narrative that works at policy launch may lose its force during implementation. The analytical problem is how governments revise policy narratives while retaining enough continuity to preserve institutional credibility (Carter et al., 2014).

2.3 Extending the Narrative Policy Framework in digital governance

Existing NPF scholarship has generated substantial knowledge about narrative structure, persuasion, coalition formation, and policy learning, but it has paid less attention to how official narratives evolve after adoption. The omission becomes especially significant during crises, when governments face new information, shifting expectations, and competing interpretations at the same time.

The study uses Narrative Velocity to capture that temporal process. It refers to the capacity to adjust official policy narratives as conditions change while retaining sufficient interpretive coherence to sustain legitimacy (Rogers, 2003). The focus is the pattern and consequences of narrative adaptation: successive explanations matter insofar as they remain connected enough for the policy to retain an intelligible rationale.

Narrative Velocity complements the established NPF. Narrative content still matters because it defines problems, assigns responsibility, and justifies intervention; the temporal extension asks how those structures evolve when governments respond to epidemiological evidence, implementation difficulties, judicial scrutiny, media reporting, and public debate. The concern is continuity of meaning across successive adjustments.

The concept also occupies a distinct analytical space from adjacent approaches. Agenda-setting concerns political attention, framing concerns interpretation, and policy diffusion concerns movement across jurisdictions. Narrative Velocity focuses on the evolution of official narratives during implementation and on whether that evolution preserves coherence under changing conditions.

The comparative puzzle makes this extension useful. Similar digital technologies can produce different trajectories because institutional arrangements and technological capacity do not fully determine how governments sustain public confidence once policies begin to change. A temporal perspective helps distinguish adaptation that preserves interpretive continuity from adjustment that fragments credibility.

Narrative Coherence, Narrative Centralization, and Narrative Velocity are treated as distinct but interconnected dimensions. Their interaction structures the comparison of how policy narratives evolve during implementation and how those dynamics bear on legitimacy and observable compliance across Indonesia, Japan, and China.

Figure 1 sets out the analytical logic used in the comparison. Institutional context shapes the communication environment in which narratives are produced and contested. Narrative Centralization, Narrative Velocity, and Narrative Contestation operate within that environment, and their interaction bears on Narrative Coherence. Adaptation can respond to contestation, but its consequences depend on whether coherence is retained.

These processes may produce narrative stabilization or erosion. Stabilization occurs when changing conditions and competing interpretations are absorbed without substantial loss of continuity; erosion occurs when adaptation becomes fragmented or contradictory. These conditions are examined in relation to institutional trust and, alongside enforcement conditions, to observable digital compliance. The framework is interpretive rather than deterministic: it identifies processes through which outcomes may differ across the cases.

Table 1 summarizes the principal analytical constructs, their conceptual definitions, and the empirical indicators used to guide the comparison.

Table 1. Conceptual operationalization of the analytical framework.

Narrative stabilization is treated as an emergent analytical condition rather than an independent explanatory variable. It refers to the extent to which narrative adaptation and contestation are absorbed without substantial loss of interpretive continuity.

3 Methodology

3.1 Research design and comparative logic

The study uses a qualitative comparative design to examine how governments justified, communicated, and sustained digital governance during the COVID-19 pandemic. Particular attention is given to policy narratives as they changed during implementation and to their relationship with institutional trust and observable compliance across Indonesia, Japan, and China.

The comparison follows a Most-Different Systems Design (MDSD). Indonesia, Japan, and China differ in political institutions, administrative traditions, governance arrangements, and state–society relations, yet they confronted a broadly comparable policy problem: managing COVID-19 through digitally mediated governance. The contrast provides a basis for examining why similar instruments produced different trajectories.

All three cases employed comparable digital measures, including contact-tracing applications, mobility monitoring, and large-scale data collection. Their outcomes nonetheless differed in public acceptance, institutional trust, and observable compliance. Accordingly, the analysis examines how policy objectives were framed, how interventions were justified, and how official narratives changed as implementation became more difficult.

Institutional conditions are treated as the context within which these narrative processes operate. Privacy protection, enforcement intensity, media openness, technological capacity, and governance arrangements shape how narratives are produced, contested, and interpreted. Tables 2, 3 summarizes these differences.

Table 2. Comparative institutional and policy context across the three cases.

Table 3. Documentary search strategy.

The comparative design is not intended to establish universally generalizable causal relationships. Its purpose is narrower: to identify recurring narrative mechanisms that help account for variation across contrasting political systems while preserving the institutional distinctiveness of each case.

3.2 Documentary corpus construction

The corpus was assembled purposively to capture documents that illuminate how governments justified digital interventions, communicated policy objectives, and adapted official narratives during implementation. It was not designed as an exhaustive archive.

3.2.1 Document search and selection process

Documents were collected between January and March 2026 from four complementary source categories: official government publications, mainstream media reports, civil society reports or publications, and academic literature retrieved through Google Scholar. The combination allowed official policy narratives to be read alongside external interpretations and critical assessments, reducing reliance on a single institutional perspective.

Documents were included when they were publicly accessible, addressed digital governance or digital compliance during the COVID-19 period, contained substantive policy narratives relevant to the analytical framework, and came from one of the predefined source categories. Duplicates, records focused primarily on epidemiological statistics, materials without relevant narrative content, and documents outside the study scope were excluded.

After duplicate removal, titles and abstracts were screened for relevance, followed by full-text eligibility assessment. The search produced 128 records; 18 duplicates were removed, leaving 110 for screening. Application of the eligibility criteria produced the final corpus of 45 documents.

Searching ceased when additional documents repeatedly confirmed established narrative patterns without generating substantively new analytical insights. The stopping point followed thematic sufficiency rather than numerical exhaustiveness. Supplementary Appendix A provides the complete document inventory.

Government documents were retrieved from official websites, including Ministries of Health, communication and digital agencies, Cabinet Offices, and the National Health Commission of China. Mainstream media sources consisted of nationally recognised news organisations, while academic publications were located through Google Scholar using predefined search strings. All searches were completed between January and March 2026.

The resulting records were screened sequentially through identification, relevance assessment, eligibility review, and final inclusion. Figure 2 presents the search and selection process leading to the analytical corpus.

The source categories serve different evidentiary purposes. Government publications document official policy narratives; media and civil society sources expose competing interpretations and public controversies; academic studies provide contextual and analytical evidence. The objective is convergence around recurring narrative mechanisms rather than statistical representativeness.

3.2.2 Temporal alignment of the three country cases

Comparative analysis requires temporal equivalence rather than identical chronological events. Indonesia, Japan, and China introduced digital governance measures under different institutional and political conditions, but all confronted the same public health emergency and later entered a period of post-pandemic adjustment. Comparable phases of policy development are therefore aligned without attempting to match individual events or calendar dates.

The comparison is organized around five common phases: policy introduction and initial adoption; implementation and enforcement; major controversies or technical challenges; public response and compliance trajectory; and transition to post-pandemic arrangements. These phases provide a shared temporal reference without assuming that the countries experienced identical events at identical moments.

3.3 Analytical framework

The conceptual framework developed in Section 2 provides the interpretive basis for the comparison. Policy narratives are treated as changing components of governance rather than static texts, particularly because digital technologies, regulations, and public communication were repeatedly adjusted during the crisis.

The analysis begins with the Narrative Policy Framework (NPF) (Jones and McBeth, 2010; Shanahan et al., 2018; Weaver, 2015), which provides a common structure for identifying the principal elements of policy narratives. Narrative settings, characters, plots, and policy morals are examined to determine how governments defined policy problems, attributed responsibility, and justified digital interventions across the three cases.

Four dimensions—narrative adaptation, narrative responsiveness, narrative continuity, and institutional incorporation—are used to operationalize Narrative Velocity. They are treated as overlapping processes, not sequential stages, with attention to how governments responded to implementation problems, institutional constraints, and changing political conditions.

The analysis also distinguishes these dimensions from related constructs. Narrative contestation concerns competing interpretations; Narrative Coherence concerns consistency across institutions and over time; Narrative Centralization concerns coordination in official communication; and Narrative Velocity concerns the capacity to adapt while retaining interpretive continuity. These distinctions organize the cross-case reading of trust and compliance.

The framework functions as an interpretive guide rather than a set of directly measurable variables. Applying the same lens across the three cases permits systematic comparison while leaving room for institutional differences.

3.4 Coding strategy and analytical procedure

The documentary corpus was analysed through an iterative qualitative coding process informed by the framework developed in Section 2. The procedure combined a common conceptual vocabulary with enough flexibility for patterns to emerge from the documentary evidence.

Open coding began with each document being read for narrative settings, characters, plots, policy morals, and references to institutional trust, legitimacy, and compliance. This first reading remained deliberately inductive so that documentary patterns could inform the subsequent comparison rather than merely reproduce prior theoretical expectations.

A second pass applied the dimensions associated with Narrative Velocity—adaptation, responsiveness, continuity, and institutional incorporation—alongside contestation, coherence, centralization, trust, and compliance. Coding focused on the relationships among these processes across implementation stages, particularly whether adaptation absorbed new challenges while preserving continuity or instead intensified contestation and weakened coherence (Table 4).

Table 4. Comparative timeline of digital governance implementation and policy response across Indonesia, Japan, and China.

The timeline distinguishes the principal comparative period from subsequent developments. The common analytical period is 2020–2023, while post-pandemic developments are included only where they provide contextual information about the subsequent trajectory of digital governance. The 2024 Indonesian PDNS ransomware incident is therefore not treated as evidence of compliance with PeduliLindungi during the pandemic. The 2020–2023 period constitutes the principal evidentiary basis for assessing narrative development and digital compliance. Later developments are used only as contextual evidence about the subsequent trajectory of digital governance and do not alter the comparative assessment of pandemic-period compliance.

Passages were then assessed against the operational criteria and decision rules in Table 5 and Supplementary Appendix B. Codes were assigned only when the substantive evidence met the relevant criterion; passages containing distinct processes could receive multiple codes. Ambiguous cases were revisited in their documentary context before the coding decision was retained.

Table 5. Analytical coding framework and operational decision rules.

Coding decisions were based on substantive documentary evidence rather than the mere presence of keywords. Where a passage contained more than one analytically distinct process, multiple codes could be assigned. Narrative stabilization was treated as an emergent condition derived from the interaction among adaptation, contestation, and coherence rather than as an independent explanatory variable.

Table 4 records the operational coding rules; the full codebook and representative coded extracts are supplied in Supplementary Appendices B,C.

3.4.1 Coding verification and multilingual source handling

The first author conducted the coding using the operational codebook. Relevant passages were assessed against the applicable definitions and decision rules, with multiple codes permitted where a passage contained distinct analytical processes.

Coding decisions and emerging interpretations were periodically reviewed with the co-author. The purpose was conceptual consistency rather than statistical intercoder reliability. Where passages remained ambiguous, they were returned to their documentary context and discussed until a consistent interpretation was reached.

Supplementary Appendix C links representative extracts to the source documents in Supplementary Appendix A and records the corresponding analytical codes. Supplementary Appendix B contains the operational definitions, inclusion and exclusion criteria, and decision rules.

Japanese and Chinese materials were consulted in official English versions where available. When no official English version existed, relevant passages were translated by the researchers, with machine-translation tools such as Google Translate or DeepL used as an aid where necessary. The resulting translations were checked against the original-language text and, where available, relevant parallel English-language sources. Indonesian materials were translated by the authors where necessary. Ambiguities were revisited against the original-language document and relevant policy or scholarly sources.

Document availability differed across the three cases, so the study did not seek numerical equivalence in source distribution. Selection remained purposive and substantively driven. Where a comparable source type was unavailable, interpretations relied on converging evidence from other categories rather than treating source absence as evidence that a narrative was absent.

3.5 Operationalization of key analytical constructs

Digital compliance, institutional trust, policy legitimacy, and public consent are related but analytically distinct. A mandatory digital system may generate observable adherence without demonstrating trust, legitimacy, or voluntary consent.

Digital compliance is therefore treated as a multidimensional outcome. The analysis considers implementation, enforcement, sustained use, and documented reception across the pandemic period, drawing on converging evidence rather than a single indicator.

Administrative use may result from legal obligation; institutional trust concerns confidence in the institutions responsible for implementation; policy legitimacy concerns the perceived appropriateness of governmental action; and public consent concerns voluntary acceptance beyond formal requirements. Keeping these distinctions separate prevents observable compliance from being treated as a proxy for legitimacy.

The analysis focuses on observable compliance while treating legitimacy primarily as an interpretive and institutional condition through which compliance may become sustained or fragile. Table 6 specifies the definitions and indicators used in the comparison.

Table 6. Operationalization of compliance and related analytical constructs.

Source: Developed by the authors.

The classifications of fragmented compliance (Indonesia), passive compliance (Japan), and high compliance (China) are interpretive assessments based on converging documentary evidence. They are not direct measures of adoption rates, institutional trust, or voluntary consent, nor should they be reduced to enforcement intensity alone.

The same definitions were applied across the three cases, while interpretation remained attentive to differences in institutional arrangements and implementation conditions.

3.6 Researcher reflexivity

Because documentary analysis depends on interpretation, the researchers treated the analytical framework as a guide rather than a fixed template. Reading moved between the corpus and the emerging interpretation so that empirical patterns could refine the analysis without abandoning a common conceptual vocabulary (Finlay, 2002).

Disagreement among sources was treated as evidence rather than noise. Government documents, media reports, civil society materials, and academic studies often described the same developments differently; those differences reveal how policy narratives were contested and reshaped within particular institutional settings.

The common coding procedures and operational definitions provided a basis for cross-case comparison. At the same time, interpretations were revisited when evidence challenged an emerging pattern. The codebook and representative extracts in Appendices B and C document this process.

3.7 Trustworthiness, validity, and study limitations

Credibility rests on comparative analysis, source triangulation, and consistent application of the analytical framework. Drawing on government, media, civil society, and academic sources allowed narrative patterns to be examined from different institutional positions (Lincoln and Guba, 1985).

Supplementary Appendix A records the corpus, Supplementary Appendix B the operational codebook, and Supplementary Appendix C representative coded extracts. Together they provide an audit trail from document selection through coding and comparative interpretation.

Triangulation reduced reliance on any single documentary perspective. Shared coding criteria supported cross-case consistency, while reflexive review allowed emerging interpretations to be reconsidered against the available evidence (Tracy, 2010).

The study has two principal limitations. First, it relies on documentary evidence and therefore does not directly measure individual attitudes or behaviour through interviews or surveys. The findings concern how policy narratives were constructed, communicated, and contested in public sources. Second, a purposive corpus cannot capture every narrative that circulated during the pandemic.

These limits define the scope of the study. It offers context-sensitive explanations of narrative development across three contrasting governance systems rather than universally generalizable causal claims. Its value lies in identifying recurring mechanisms that can be examined in future comparative research.

4 Findings

Indonesia, Japan, and China incorporated broadly comparable digital technologies into their COVID-19 responses, yet implementation produced different patterns of public acceptance, institutional trust, enforcement, and sustained digital compliance. The evidence makes it possible to examine what changed once similar instruments entered distinct political and narrative environments.

The divergence cannot be attributed to technology or institutional design alone. The documentary record also points to the ways governments justified intervention, responded to emerging problems, and revised official narratives as implementation unfolded.

Table 7 shifts from outcome evidence to the narrative conditions associated with those outcomes, examining narrative coherence, narrative velocity, and narrative centralization in relation to institutional trust and governance arrangements.

Table 7. Comparative narrative dynamics and compliance outcomes.

Narrative Velocity categories are qualitative comparative classifications derived from the converging pattern of narrative adaptation, responsiveness, continuity, and institutional incorporation across implementation stages. “Controlled–high” denotes high narrative velocity under strong institutional coordination; “controlled” is a contextual qualifier reflecting narrative centralization, not a separate dimension of narrative velocity.

4.1 Indonesia: narrative fragmentation and trust erosion

Indonesia introduced PeduliLindungi as a central component of its pandemic response. The platform was presented as a means of strengthening contact tracing while supporting the gradual reopening of economic and social activity. Official communication linked public health protection, safe mobility, and democratic governance, producing a relatively coherent narrative during the initial phase of implementation (Ministry of Health of the Republic of Indonesia, 2021; Ministry of State Secretariat of the Republic of Indonesia, 2021; Ministry of Communication and Informatics of the Republic of Indonesia, 2021a; Kurniasih et al., 2024; Ministry of State Secretariat of the Republic of Indonesia, 2023). At the same time, PeduliLindungi recorded more than 90 million downloads and became a requirement for entry to many public facilities during the peak of the pandemic (Table 8; Ministry of Health of the Republic of Indonesia, 2020).

Table 8. Comparative evidence supporting digital compliance assessments across Indonesia, Japan, and China.

Compliance assessments are qualitative comparative interpretations derived from converging evidence across the implementation period, adoption patterns, documented actual use, enforcement conditions, and public attitudes. Evidence is triangulated using government documents, mainstream media reports, civil society publications, and academic literature included in the 45-document analytical corpus. The classifications of fragmented, passive, and high compliance should not be interpreted as direct measures of institutional trust, policy legitimacy, or voluntary public consent. Rather, they represent comparative assessments of observed behavioural compliance within each country’s institutional and political context.

The narrative became harder to sustain as implementation shifted public attention from epidemiological objectives toward cybersecurity, personal data protection, and governmental accountability (SAFEnet, 2020; BSSN, 2020; CNBC Indonesia, 2022; Kompas.com, 2022a, 2022b; Ministry of Communication and Informatics of the Republic of Indonesia, 2021b). Official communication continued to stress disease control, while media and civil society accounts increasingly questioned privacy safeguards and institutional capacity (Ministry of Communication and Informatics of the Republic of Indonesia, 2021c; ELSAM, 2021; The Jakarta Post, 2022; Wahyuni et al., 2024). The subsequent full implementation of Indonesia’s Personal Data Protection Law further institutionalized data-protection requirements (Tempo, 2022; Kompas, 2024). Public awareness of personal-data protection was also substantial, although gaps remained in how individuals protected their data (Katadata Insight Center, 2021). The technology changed incrementally; the political meaning attached to it changed more quickly.

The 2024 PDNS ransomware incident is treated separately as a post-pandemic development. It subsequently reinforced concerns about institutional reliability and digital governance capacity, but it is not used as evidence of pandemic-period compliance or trust. The incident does not alter the assessment of PeduliLindungi during the principal 2020–2023 implementation period (BBC Indonesia, 2024).

Government agencies responded to emerging controversies, but the documentary record provides less evidence that these responses consolidated into a common interpretive framework. Successive explanations often addressed immediate pressures, leaving competing accounts in circulation across government institutions, media, and civil society. The result was a less stable narrative environment and greater difficulty in sustaining institutional trust.

PeduliLindungi use was widespread during the peak of the pandemic because QR-code verification was required in transportation, workplaces, commercial centres, and public facilities. As restrictions were relaxed, routine use declined and privacy and data-security concerns became more prominent (Table 8). Continued adherence consequently appears to have depended increasingly on administrative requirements rather than durable acceptance of the policy rationale.

Indonesia therefore represents fragmented compliance. Rapid deployment and administrative capacity generated substantial initial participation, but neither secured durable adherence once enforcement weakened. The documentary evidence links that fragility to the deterioration of narrative coherence and institutional trust.

4.2 Japan: coherent narratives and limited public resonance

Japan adopted a substantially different model. The government introduced COCOA as a voluntary instrument intended to support collective public health while preserving individual autonomy (Cabinet Secretariat, Government of Japan, 2021; Ministry of Health, Labour and Welfare, 2021; Naikaku Kanbō, 2020; Ministry of Health, Labour and Welfare, 2020a). Official communication consistently presented participation as a form of civic responsibility grounded in privacy protection, mutual trust, and administrative restraint (Nakamoto et al., 2020; Brummer and Ueno, 2024; Ishii, 2020; Asia Centre, 2024; Personal Information Protection Commission, 2020; Yomiuri Shimbun, 2021). Within the documentary corpus, this core narrative remained comparatively stable, with limited evidence of sustained political contestation directed at the underlying policy rationale.

Implementation posed the main difficulty. Software malfunctions, delayed exposure notifications, and other operational problems increasingly shaped perceptions of COCOA. Government assessments and reporting by The Japan Times (2021) and Yomiuri Shimbun (2023) reinforced the salience of these shortcomings (NHK, 2023; Omae et al., 2021). The evidence points less to rejection of the policy rationale than to declining confidence in the instrument’s practical value (Ministry of Health, Labour and Welfare, 2020b).

Narrative coherence and behavioural engagement consequently moved along different tracks. Official messaging remained comparatively stable, but its persuasive force increasingly depended on whether the application worked as expected. Technical unreliability weakened the basis for sustained participation without producing the broader narrative fragmentation observed in Indonesia.

Because participation was voluntary, continued use depended on citizens’ assessments of COCOA’s reliability and usefulness (Asahi Shimbun, 2023; Ōiso et al., 2021; Shimpo and Japan Civil Liberties Union, 2022). Downloads therefore cannot be treated as evidence of sustained engagement. Initial uptake was substantial, but routine participation became more limited as technical problems accumulated. The resulting pattern is best characterized as passive compliance.

Japan presents a different pathway. The evidence points to implementation performance rather than a broad legitimacy challenge directed at government intervention. Narrative coherence remained relatively strong, but it could not by itself compensate for a digital instrument whose practical performance fell short of expectations.

4.3 China: centralized narrative adaptation and compliance stabilization

China operated through a markedly different institutional configuration. Rather than relying on a stand-alone contact-tracing application, the Health Code system was embedded within widely used digital platforms, including WeChat and Alipay (China Daily, 2020; CGTN, 2024). Health verification consequently became part of routine administrative practice. Official communication presented digital governance as an instrument for protecting collective security, maintaining social stability, and supporting epidemic control (Zhang, 2022; National Health Commission of China, 2020; People’s Daily, 2021; State Council of the People’s Republic of China, 2020; Ministry of Industry and Information Technology of China, 2020). Digital monitoring was framed from the outset as part of coordinated state administration.

The interpretive framework remained comparatively stable. Government agencies and state-affiliated media repeatedly linked digital governance to public safety, collective responsibility, and epidemic control (Liang, 2020; Yu, 2022). Policies changed as epidemiological conditions changed, but those adjustments were generally incorporated into an established narrative rather than accompanied by substantial divergence among official accounts (Tianjin Municipal Cyberspace Administration, 2020; Xinhua News Agency, 2021; Ministry of Industry and Information Technology of the People’s Republic of China, 2020).

The contrast with Indonesia and Japan is consequential. In China, adaptation largely reinforced continuity. Administrative changes could be presented as extensions of an established governance strategy, while centralized communication limited visible divergence among official narratives (Ministry of Industry and Information Technology of the People’s Republic of China, 2020).

Health Code systems became embedded in transportation, workplaces, residential communities, and access to public services (Yu, 2022). Participation was therefore closely tied to everyday administrative processes. Within the documentary evidence, centralized communication, institutional integration, and mandatory implementation coincided with consistently high observable compliance.

An important qualification is necessary. High observable compliance should not be interpreted as direct evidence of voluntary public trust, democratic legitimacy, or unrestricted public consent. These concepts remain analytically distinct from compliance as defined in this study.

The Chinese case points to the role of institutional coordination in implementation (Xinhua News Agency, 2021). Sustained compliance appears to have emerged from the combination of centralized narrative management, integrated administrative structures, and mandatory implementation. The available evidence does not establish whether this compliance reflected broader voluntary acceptance, so the case is used analytically rather than normatively.

4.4 Comparative narrative analysis

Across the three cases, broadly comparable digital technologies produced different governance trajectories once implementation moved beyond the initial emergency phase. The divergence is best understood through the interaction of policy narratives, institutional coordination, and governance arrangements as implementation unfolded.

Three configurations of Narrative Velocity emerge across the cases. Indonesia is classified as Moderate because narrative adaptation and responsiveness were evident as privacy, cybersecurity, and accountability concerns became more prominent, but these responses did not consistently consolidate into a common interpretive framework. Japan is classified as Low–Moderate because its core narrative remained comparatively stable, with limited adaptation despite implementation problems. China is classified as Controlled–High because narrative adaptation remained substantial as policy conditions changed, while centralized institutional coordination maintained continuity across official narratives.

The comparative pattern is relational rather than attributable to a single narrative dimension. Narrative Coherence mattered most when governments could sustain confidence in the policies their narratives justified; adaptation supported continuity when successive explanations reinforced one another; and centralization strengthened implementation stability under particular institutional conditions. These dimensions therefore bear on compliance through different institutional pathways and remain analytically distinct from trust, legitimacy, and consent.

The analysis extends the NPF in two related respects. Policy narratives operate as processes that continue through implementation, and their significance depends partly on the capacity to change without losing coherence. Narrative Coherence, Narrative Velocity, and Narrative Centralization therefore interact with institutional context rather than functioning as fixed attributes.

Tables 7 and 9 integrate established NPF elements with the additional dimensions developed in this study, providing a process-oriented account of how policy narratives evolve during implementation and bear on contrasting governance outcomes across political systems.

Table 9. Comparative narrative policy framework (NPF) elements across Indonesia, Japan, and China.

The table summarizes the principal Narrative Policy Framework (NPF) elements identified through qualitative document analysis across Indonesia, Japan, and China. Narrative coherence refers to the internal consistency of policy narratives over time; Narrative Velocity refers to the capacity and pattern of official narrative adaptation during policy implementation; and Narrative Centralization refers to the extent to which narrative construction is coordinated by central governmental actors. Representative documentary evidence supporting these comparative findings is presented in Table 10, while complete document metadata are provided in Supplementary Appendix A.

Table 10. Representative documentary evidence supporting narrative identification.

Representative excerpts illustrate how the principal Narrative Policy Framework (NPF) elements were identified and interpreted during qualitative coding. Complete document metadata are provided in Supplementary Appendix A.

5 Discussion

5.1 Research question and the central interpretive mechanism

Across the three cases, broadly similar digital governance measures produced different patterns of legitimacy and observable compliance. Administrative capacity, institutional design, technological sophistication, and regime type each matter. The evidence nevertheless draws attention to what happened to official policy narratives once implementation encountered operational difficulties, institutional constraints, and changing public expectations.

Narratives did more than communicate decisions. They mediated the political meaning of governmental action during implementation. Legal authority, administrative resources, technological infrastructure, and enforcement remained necessary, but their political consequences depended partly on how government action was explained, defended, revised, and interpreted. Institutions supplied capacity; narratives shaped how that capacity was understood.

The MDSD strengthens this interpretation because the cases differ substantially in political institutions and administrative traditions while confronting a broadly comparable public health problem. Institutional differences remain important, but they provide the settings in which the narrative mechanism can be observed. Across those settings, changes in governance outcomes repeatedly coincided with changes in the coherence, coordination, and evolution of official narratives. The recurring association does not make institutional context irrelevant; it indicates that the interpretive mechanism is not confined to one political system.

Observable compliance remains analytically distinct from democratic legitimacy, institutional trust, and voluntary consent, while institutional explanations remain essential. The evidence suggests that institutional capability and policy narratives work together during implementation, with narratives shaping how administrative capacity was interpreted in relation to continuing political legitimacy.

This mediating role is central to the theoretical contribution developed through Narrative Velocity.

5.2 Extending the narrative policy framework: narrative velocity

The findings bring a temporal dimension of the NPF into clearer view at the implementation stage. Policies continue amid changing institutional conditions, technical problems, media scrutiny, and shifting public expectations, so official narratives may be revised or contested after adoption.

Narrative adjustment is treated here as an implementation problem. As evidence, technical performance, institutional learning, media scrutiny, and public contestation alter the policy environment, the analytical question becomes how successive explanations remain intelligibly connected.

Implementation changes the environment in which governments justify policy. The significance of narrative adaptation lies in how institutional actors incorporate new evidence and emerging problems into an intelligible policy account.

The extension retains the established NPF architecture: settings, characters, plots, counter-narratives, and policy morals remain central to how policy meaning is constructed, while their temporal evolution during implementation becomes analytically visible.

Its boundary with adjacent concepts remains clear. Framing concerns presentation and interpretation (Entman, 1993); agenda-setting concerns political attention; policy diffusion concerns movement across jurisdictions; and policy feedback concerns the effects of implemented policies on later politics. Narrative Velocity addresses the evolution of official narratives during implementation and the implications for continuity and governance legitimacy (Tables 10).

The cases illustrate why this temporal perspective matters. Indonesia shows how adjustment can coexist with declining coherence; Japan shows how narrative stability can coexist with weak practical resonance; and China shows how adaptation can be incorporated into an established interpretive framework. The analytical focus therefore shifts from the initial construction of narratives to their maintenance and revision during implementation.

Narrative Velocity identifies temporal management of governing narratives as an implementation problem that has received less direct attention in NPF research. Where legitimacy continues to be negotiated after adoption, the capacity to sustain credible narratives over time becomes relevant to explaining divergent governance trajectories.

5.3 Implications for comparative governance

The findings also speak to comparative governance beyond the NPF. Institutional design, administrative capacity, regulatory quality, and regime characteristics remain important, but their effects become clearer when considered alongside the narratives through which governmental action is interpreted during implementation.

Institutional capacity enabled deployment, but deployment did not automatically produce enduring public acceptance or stable behavioural compliance. Across the cases, the more consequential issue was whether policy narratives remained credible as governments responded to political, technological, and social change. Institutional performance and narrative credibility are therefore better understood as interacting dimensions than as rival explanations.

The findings add a communicative dimension to state capacity in digital governance. Legal authority, organisational competence, bureaucratic coordination, and technological infrastructure remain foundational. Yet implementation increasingly occurs under continuous scrutiny and competing interpretations. Capacity therefore includes not only the ability to implement a policy, but also the ability to explain and revise it without dissolving its coherence.

The implication extends across different regime settings, although the mechanisms are not identical. Political systems differ in authority, enforcement, and constitutional arrangements, as well as in how official narratives are produced, coordinated, contested, and revised. These differences help explain why comparable resources can accompany different governance trajectories.

More broadly, governance outcomes reflect an interaction between institutional capability and narrative adaptation. Public authority depends on the capacity to act, but also on the ability to preserve credible interpretations as implementation proceeds under changing conditions.

5.4 From pandemic governance to future digital governance

The empirical analysis is rooted in COVID-19, but the problem it identifies extends to ordinary digital administration. Artificial intelligence, digital public infrastructure, algorithmic decision-making, and data-driven systems will continue to require governments to adapt policy while maintaining political legitimacy.

Digital policies rarely remain unchanged once implementation begins. Technological innovation, judicial review, scientific developments, media scrutiny, and public contestation can all prompt adjustment. Administrative adaptation is necessary, but governments must also explain why changes are required without undermining the coherence on which confidence depends. Adaptation without credible interpretation can erode legitimacy; continual revision without continuity can produce uncertainty.

Narrative Velocity is useful here because it directs attention to how governments manage that tension. Its value lies not in determining whether adaptation occurs, but in examining whether interpretive continuity is retained as implementation evolves. The same question may arise in artificial intelligence governance, cybersecurity, digital public infrastructure, climate adaptation, and other fields in which policy revision is likely to be continuous.

Future research can examine whether the relationship among narrative adaptation, legitimacy, and policy performance travels across democratic, hybrid, and authoritarian settings and across longer implementation periods. It can also explore how emerging technologies alter the interaction between institutional authority, public communication, and behavioural response.

The contribution extends beyond COVID-19 governance. Resilient digital governance depends on the interaction between institutional capability and narrative adaptation. Governments must be able to act, but they must also preserve credible interpretations as policies change. The challenge is not adaptation alone; it is adapting without losing the coherence needed to sustain legitimacy over time.

Generative AI statement

The author(s) declared that Generative AI was used in the creation of this manuscript. Generative AI was used solely to assist with language refinement and editorial support during manuscript preparation. The intellectual content of the study, including its conceptualization, methodology, analysis, interpretation, and conclusions, was developed and verified by the authors. The authors retain full responsibility for the accuracy, integrity, and scholarly contribution of the manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpos.2026.1910732/full#supplementary-material

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Footnotes

Edited by: Aleixandre Brian Duche-Pérez, National University of Saint Augustine, Peru

Reviewed by: Delipiter Lase, University of Nias, Indonesia

Niladri Maiti, Central Asian University, Uzbekistan