Digital Government for Sustainable Development: An Integrative Systematic Literature Review of Citizen Readiness, Governance Dynamics, and Public Value Co-Creation

Digital government has become a central governance instrument for advancing sustainable development. Across national and local contexts, governments increasingly deploy digital platforms, integrated service portals, mobile applications, data infrastructures, artificial intelligence, and participatory technologies to improve service delivery, administrative coordination, transparency, and citizen engagement. This transformation reflects a shift from earlier e-government models, which mainly emphasized online service provision and efficiency, toward digital government as a socio-technical transformation of public institutions, service ecosystems, and state-citizen relations [1, 2]. In this broader understanding, digital government is not only a modernization project. It is a governance project through which public institutions seek to produce inclusive, accountable, responsive, and sustainable public value [3, 4].

The relationship between digital government and sustainable development has become increasingly explicit in global policy discourse. The United Nations E-Government Survey positions digital transformation as a pathway for accelerating sustainable development through inclusive services, stronger local e-government, and digitally enabled institutions [5]. The OECD Digital Government Policy Framework similarly argues that mature digital government requires governments to become digital by design, data-driven, open by default, proactive, user-driven, and organized as platforms [6, 7]. The World Bank’s GovTech framework also emphasizes core government systems, online public service delivery, digital citizen engagement, and institutional enablers [8]. These frameworks indicate that digital government should be assessed not only by technological adoption, platform maturity, or service availability, but also by inclusion, equity, accountability, trust, resilience, and citizen empowerment.

However, the sustainability promise of digital government is neither automatic nor evenly distributed. Digital services may reduce transaction costs, expand access, improve responsiveness, and support data-informed decisions. Yet they may also reproduce exclusion when citizens lack stable connectivity, appropriate devices, digital literacy, trust, or the capacity to use digital systems meaningfully. Digital divide research shows that inequality is no longer limited to physical internet access; it also involves material access, skills, usage patterns, and unequal outcomes [9]. Citizens may be formally connected to digital services but practically excluded from their benefits. A sustainable digital government agenda must therefore move beyond infrastructure provision and service digitization toward citizen readiness.

In this article, citizen readiness refers to the multidimensional capacity that enables citizens not only to access and use digital public services, but also to engage in their design, implementation, evaluation, and improvement. It includes access, literacy, perceived usefulness, trust, privacy and security perceptions, motivation, empowerment, and institutional support. This distinction matters because digital government research has often treated citizens as users, adopters, consumers, or beneficiaries rather than as co-creators of public value [10, 11]. A citizen may use an online service because it is mandatory or convenient, yet still have no influence over how the service is designed or improved.

A second limitation concerns governance dynamics. In this article, governance dynamics refers to the adaptive and interactive processes through which public actors activate stakeholders, frame collective problems, mobilize institutional resources, coordinate service ecosystems, and interpret, coordinate, and potentially incorporate citizen input into service improvement and institutional learning. Existing digital government scholarship identifies leadership, regulatory clarity, inter-agency coordination, data governance, organizational learning, frontline capacity, stakeholder collaboration, and feedback mechanisms as relevant institutional conditions for digital transformation [1, 2]. Yet these conditions are often discussed separately, with limited attention to how they may operate as interconnected institutional processes shaping whether and how citizens’ digital readiness is recognized, coordinated, and associated with substantive public value co-creation.

Public value co-creation provides the theoretical bridge between digital government and sustainable development. Public service research argues that value is not generated solely through government provision, but through interactions among public organizations, citizens, communities, and other stakeholders [12, 13]. In digital government, co-creation occurs when citizens participate in identifying service needs, shaping service design, supporting implementation, evaluating performance, and pushing governments toward continuous improvement. Digital technologies can support such interaction, but they can also depersonalize services, bypass citizen interaction, or create new forms of exclusion [14]. The key issue is therefore whether governance systems enable meaningful citizen influence.

Existing scholarship remains fragmented. One stream focuses on adoption, acceptance, satisfaction, continuance intention, usability, and service quality. A second examines digital divide, literacy, accessibility, cybersecurity, privacy, and trust. A third addresses smart governance, platform governance, institutional capacity, data governance, and public sector transformation. A fourth explores citizen participation, e-participation, co-production, and co-creation. A fifth links digital government to sustainability and the Sustainable Development Goals. Existing reviews have advanced understanding of public value in e-government, public value creation in digital government, and the role of digital technologies in co-production [3, 4, 14]. Yet an integrative account of how citizen capacities, institutional processes, participatory arrangements, and sustainable public value may relate to one another remains underdeveloped. In particular, the literature rarely brings these dimensions together within a common analytical structure while preserving their conceptual distinctions and avoiding the assumption of a predetermined causal sequence.

This article addresses that gap through an integrative systematic literature review of digital government scholarship. It asks: How are sustainable development outcomes conceptualized and linked to digital government? What dimensions of citizen readiness are emphasized? How are governance dynamics conceptualized in relation to citizen participation and public value co-creation? How can citizen readiness, governance dynamics, and public value co-creation be integrated into a framework for sustainable digital government? The article makes four contributions. It reframes digital government as a sustainability-oriented public value project, conceptualizes citizen readiness as a multidimensional capacity for potential agency, positions governance dynamics as a proposed meso-level institutional process lens, distinguishes substantive public value co-creation from limited or tokenistic forms of digital participation, and develops the DR–GD–PVCC–SPV (digital readiness of citizens, governance dynamics, public value co-creation, and sustainable public value) framework as a theory-building synthesis of theoretically plausible relationships among citizen capacity, institutional processes, participatory value formation, and sustainable public value. The framework is intended to organize and guide subsequent empirical examination rather than to establish causal, mediating, or moderating relationships from the review itself.

The article therefore deliberately avoids treating digital government as a single administrative innovation. It approaches the field as a contested governance space in which technology, citizen capacity, institutional responsiveness, and public value are potentially interrelated. This perspective is particularly important because governments increasingly claim that digital transformation advances sustainability, yet empirical and conceptual evidence suggests that such claims should be examined in relation to the conditions under which citizens can participate meaningfully and institutions can respond to and learn from citizen input. Limited citizen readiness may constrain meaningful engagement with digital public services; weak or fragmented institutional processes may restrict how citizen input is recognized and incorporated; and participation may remain symbolic when citizens have little substantive influence over service design, implementation, or evaluation. These relationships are treated as theoretically plausible conditions for subsequent empirical examination rather than as predetermined causal effects.

The review is structured to support this integrative inquiry. The conceptual background clarifies the core constructs. The methodology explains the integrative review design, search strategy, appraisal, and coding procedures. The findings synthesize publication trends, thematic clusters, and missing links. The integrative framework then brings citizen readiness, governance dynamics, public value co-creation, and sustainable public value into a common theory-building structure and proposes theoretically plausible relationships among them for subsequent empirical examination. The discussion and implications identify theoretical, methodological, and practical consequences for digital government research and public sector reform.

The focus on sustainable public value is also important for local and developing-country contexts, where digital transformation is often pursued under conditions of uneven infrastructure, unequal citizen capabilities, limited administrative coordination, and contested public trust. In such settings, digital government should not be evaluated solely through national maturity rankings or platform expansion. Its evaluation should also consider the everyday capacity of citizens to access services, the ability of institutions to respond to feedback, and the extent to which digital reforms are associated with greater inclusion or, conversely, with the reproduction of existing forms of exclusion. These concerns strengthen the theoretical and practical case for examining citizen readiness, governance dynamics, public value co-creation, and sustainable public value within a common analytical framework.

This positioning also clarifies the role of the empirical corpus. The review does not treat the 297 records merely as a bibliographic inventory; rather, it uses the systematically appraised and coded corpus to develop an evidence-informed, theory-building synthesis of recurring conceptual relationships and gaps. The tables, figures, and supplementary materials serve as reporting and audit devices that document corpus construction, analytical weighting, thematic coding, and the evidence-to-concept progression underlying the proposed framework. The bibliometric visualizations provide descriptive context, whereas the article-level thematic synthesis constitutes the principal evidence base for identifying conceptual fragmentation and developing the proposed theory-building relationships.

3.1 Review design

This study employed an integrative systematic literature review to synthesize and theorize the relationship among digital government, citizen readiness, governance dynamics, public value co-creation, and sustainable development. The design was selected because the article aims not only to map digital government scholarship, but also to integrate fragmented conceptual streams and develop an integrative conceptual framework that proposes theoretically plausible relationships among the focal constructs for future empirical examination. Systematic reviews are appropriate for identifying, appraising, and synthesizing a body of knowledge through transparent and reproducible procedures [18, 19]. Because the purpose of this article is conceptual integration rather than effect-size estimation or empirical testing of causal or mediating relationships, it follows an integrative review logic, which is suitable for fields characterized by theoretical fragmentation, methodological heterogeneity, and emerging debates [20, 21]. Accordingly, the relationships incorporated into the proposed framework are treated as theory-building propositions derived from the synthesis of the reviewed literature, rather than as empirically established causal mechanisms (see Table 1).

Table 1. Review protocol

Note: The table summarizes the review design, database, publication period, document type, language, corpus size, and analytical stages used in the integrative systematic literature review.

3.2 Data source and search strategy

Scopus was selected as the primary database because it provides broad multidisciplinary coverage across public administration, information systems, governance, sustainability, public policy, and social sciences. Although the use of a single database may exclude studies indexed elsewhere, Scopus offers a consistent and high-quality corpus suitable for integrative synthesis. The dataset consisted of 297 Scopus-indexed records exported on April 25, 2026. The corpus covered publications from 2015 to 2025 and included 271 journal articles and 26 review articles. All records were written in English, and no duplicate DOI entries were identified. Full texts were consulted where titles, abstracts, keywords, and metadata provided insufficient information for relevance assessment or thematic coding.

The review was conceptually organized around four analytical domains: digital government, citizen-related dimensions, governance and co-creation, and sustainability, but these domains should be distinguished from the Boolean query actually used to retrieve records from Scopus. The final retrieval strategy consisted of two TITLE-ABS-KEY blocks connected by the Boolean operator AND. The first block captured core digital-government terminology, whereas the second combined sustainability-related terms with public value, digital inclusion, citizen participation, and co-creation terminology. The exact query used to construct the corpus was:

TITLE-ABS-KEY ("digital government" OR "e-government" OR "digital governance" OR "smart governance" OR "GovTech" OR "digital public service") AND TITLE-ABS-KEY ("sustainable development" OR sustainability OR "public value" OR "digital inclusion" OR "digital literacy" OR "digital divide" OR "citizen participation" OR "co-creation" OR "co-production" OR "e-participation").

Terms appearing in the broader conceptual framework but not in this final Boolean query including “citizen readiness,” “digital skills,” “trust,” “privacy,” “security,” “governance capacity,” “collaborative governance,” “public service innovation,” “SDGs,” “inclusive public service,” “accountability,” “equity,” and “resilience” are therefore not reported as database retrieval terms. Instead, where relevant, these concepts informed the subsequent relevance assessment, data extraction, and deductive–inductive thematic coding of the retrieved literature. This distinction separates the reproducible database search from the broader conceptual vocabulary used to interpret the corpus.

The initial Scopus search yielded 1,896 records. Sequential eligibility and database filters relating to publication period, document type, publication stage, journal source, language, and access status were subsequently applied as detailed in Sections 3.3–3.4 and Supplementary Material (Appendix B), resulting in the final corpus of 297 records. Supplementary Material (Appendix A) reports the exact retrieval query and distinguishes it from the broader analytical concepts used during subsequent synthesis.

3.3 Inclusion and exclusion criteria

The inclusion criteria were designed to ensure relevance, transparency, and replicability (see Table 2). Records were included when they were indexed in Scopus, published between 2015 and 2025, written in English, classified as journal articles or review articles, published in final form in journal sources, and identified as open access in Scopus at the time of retrieval. Topical scope at the retrieval stage was defined by the two-block TITLE-ABS-KEY query reported in Section 3.2 and Supplementary Material (Appendix A2). No additional undocumented substantive-exclusion step was applied after the database filters had produced the 297-record corpus. Instead, variation in the degree of digital-government, sustainability, public-value, and focal-construct relevance was evaluated subsequently through the structured post-retrieval relevance appraisal described in Section 3.5.

The open-access criterion was applied to ensure that full texts could be consulted when titles, abstracts, keywords, and metadata were insufficient for relevance assessment, methodological-transparency appraisal, or thematic coding. Open-access status was therefore treated as an accessibility criterion rather than an indicator of methodological quality. Because this restriction may underrepresent relevant subscription-based literature and may disproportionately capture journals with extensive open-access publishing, it constitutes a potential source of selection bias and is explicitly recognized as a boundary of the review corpus.

The focal constructs citizen readiness, governance dynamics, public value co-creation, and sustainable public value outcomes were not treated as four cumulative inclusion requirements. Rather, they were used after retrieval to assess conceptual relevance and to classify the contribution of each record during appraisal and thematic coding. Accordingly, a study was not excluded solely because it did not address every focal construct.

Conference papers, book chapters, editorials, notes, non-English records, non-journal sources, records without open-access availability under the applied Scopus filter, and studies lacking substantive public-governance or sustainability relevance were excluded through the corpus-construction filters. Records differing in substantive public-governance, sustainability, public-value, or focal-construct relevance were not removed through an additional post-retrieval exclusion stage; these differences were captured through the relevance-classification procedure reported in Section 3.5.

Table 2. Inclusion and exclusion criteria

Note: The table distinguishes corpus-construction filters from post-retrieval analytical classification. Publication period, document type, publication stage, source type, language, and open-access status were applied as corpus-construction filters. The topical scope was defined by the retrieval query, whereas variation in substantive digital-government, sustainability, public-value, and focal-construct relevance was assessed after retrieval through the relevance-appraisal procedure described in Section 3.5. The post-retrieval appraisal assigned analytical weight to records and did not constitute an additional undocumented database-exclusion stage.

3.4 Screening process and PRISMA logic

The identification and corpus-construction process followed a PRISMA 2020-informed reporting logic (Figure 1).

tu_pian_1.png

Database filters were applied sequentially and are reported separately from subsequent conceptual relevance assessment. Restricting the publication period to 2015–2025 reduced the dataset from 1,896 to 1,315 records, excluding 581 records. Limiting the corpus to journal articles and review articles reduced the dataset to 692 records, excluding a further 623 records. Filters for final publication stage, journal source, and English language reduced the dataset to 630 records, excluding 62 additional records. Finally, the open-access filter applied during corpus construction retained 297 records and excluded 333 records. These 297 records constituted the defined review corpus for descriptive mapping and subsequent appraisal.

Open-access status is reported as a corpus-construction criterion rather than a quality criterion. As explained in Section 3.3, the restriction facilitated full-text consultation where metadata and abstracts were insufficient for relevance assessment and coding, but it may introduce selection bias by underrepresenting subscription-based publications. The review does not claim that the 297 records represent the complete universe of digital-government scholarship.

Following corpus construction, the 297 records were assessed using titles, abstracts, keywords, metadata, and, where necessary, full texts to determine their degree of conceptual relevance. This stage did not constitute an additional database retrieval step or an additional substantive-exclusion stage. Rather, all 297 records remained within the defined review corpus and were analytically classified through the relevance appraisal described in Section 3.5. On the basis of that appraisal, 118 records were classified as highly relevant, 63 as relevant, 83 as moderately relevant, and 33 as supporting/background. The 264 highly relevant, relevant, and moderately relevant records contributed to the broad thematic synthesis, whereas the 33 supporting/background records were retained primarily for contextual and descriptive mapping. The 181 highly relevant and relevant records formed the core evidence base used to develop the integrative conceptual framework, whereas the 83 moderately relevant records informed the broader synthesis without being treated as core framework evidence. The relevance categories therefore represented analytical weighting within the corpus rather than eligibility decisions or a second stage of study exclusion.

Thus, Figure 1 distinguishes three analytically different stages: (1) identification and database filtering, (2) construction of the 297-record review corpus, and (3) post-retrieval relevance appraisal and analytical classification. No additional records were removed from the defined corpus during Stage 3; instead, records were assigned different analytical roles according to their relevance scores. Supplementary Material (Appendix B) reports the detailed corpus-construction and relevance-classification flow.

3.5 Quality appraisal and relevance assessment

The appraisal combined conceptual relevance and methodological transparency and was used to differentiate the analytical contribution of records to the review rather than as a formal risk-of-bias assessment. Each record was assessed using five criteria: relevance to digital government, relevance to sustainable development, clarity of conceptual contribution, methodological transparency, and relevance to the focal constructs of citizen readiness, governance dynamics, public value co-creation, and sustainable public value outcomes (Table 3).

Each criterion was scored from 0 to 3, yielding a total possible score of 0–15 for each record. Total scores were classified using pre-specified thresholds: 12–15 = Highly Relevant, 9–11 = Relevant, 6–8 = Moderately Relevant, and 0–5 = Supporting/Background. These thresholds were applied consistently across the 297-record corpus, and the article-level scores are reported in Supplementary Material (Appendix D).

Across the review process, one primary reviewer conducted the screening, relevance appraisal, data extraction, and coding. Independent duplicate screening, appraisal, extraction, or coding was not undertaken. To strengthen procedural consistency, the co-authors reviewed the appraisal and coding rules and the resulting synthesis, while ambiguous cases were discussed among the authors before final classification. The appraisal was also supported by the predefined scoring rubric, repeated examination of ambiguous records, and an article-level audit trail. No formal inter-rater or inter-coder reliability coefficient was calculated; accordingly, the resulting relevance bands are treated as a structured analytical classification rather than as a statistically validated measure of reviewer agreement.

Table 3. Relevance and methodological-transparency appraisal rubric

Note: Each record was assessed using five criteria scored from 0 to 3, yielding a maximum total score of 15. Pre-specified thresholds classified records as highly relevant (12–15), relevant (9–11), moderately relevant (6–8), or supporting/background (0–5). The rubric was used to assess analytical relevance and methodological transparency rather than as a formal risk-of-bias instrument. Article-level scores and classifications are reported in Supplementary Material (Appendix D) to provide an auditable link between the appraisal procedure and the analytical use of the reviewed records.

Application of the scoring criteria resulted in 118 highly relevant records, 63 relevant records, 83 moderately relevant records, and 33 supporting/background records. Thus, 264 records were used for broad thematic synthesis, while 181 highly relevant and relevant records formed the core evidence base for integrative conceptual framework development. The 83 moderately relevant records contributed to the broader thematic synthesis but were not treated as core framework evidence, whereas the 33 supporting/background records were used primarily for contextual and descriptive mapping. These relevance bands therefore indicate analytical weighting within the review rather than methodological validity or study exclusion.

3.6 Data extraction and coding

Data extraction was conducted through a structured matrix covering author, year, title, journal, country or context cue, method, theoretical lens, digital government focus, citizen-related construct, governance construct, co-creation or participation construct, sustainability outcome, key findings, and research gap. The structured extraction matrix was applied consistently across the review corpus to create an article-level audit trail linking bibliographic information, substantive content, and subsequent thematic classification. The complete extraction structure and article-level evidence are reported in the Supplementary Material.

Coding combined deductive and inductive procedures. Deductive codes were derived from the review questions and conceptual background, including digital government, citizen readiness, governance dynamics, public value co-creation, and sustainable public value outcomes. These deductive categories provided the initial coding structure but were not treated as mutually exclusive eligibility criteria. Rather, they were used to identify the extent to which individual studies contributed evidence to one or more focal dimensions of the review.

Inductive codes were added when recurring concepts emerged from the corpus, such as trust readiness, feedback synthesis, symbolic participation, platform governance, public value co-destruction, and equity-sensitive evaluation. Emergent codes were retained when they represented recurring conceptual patterns that were not adequately captured by the initial deductive categories. Where an article contained conceptually ambiguous or overlapping material, coding decisions were reviewed against the operational definitions in the codebook and the article-level extraction record before final classification.

Consistent with the appraisal procedure described in Section 3.5, data extraction and initial coding were undertaken by the primary reviewer using the predefined extraction matrix and coding framework. The co-authors reviewed the coding logic and synthesis, and ambiguous classifications were discussed before the final thematic categories were established. No formal inter-coder reliability coefficient was calculated; therefore, the coding results are interpreted as a structured and auditable thematic synthesis rather than as a statistically validated measure of coder agreement.

Because thematic coding was non-mutually exclusive, a single record could contribute to more than one thematic cluster. Consequently, thematic frequencies should not be summed to represent the number of unique studies in the corpus. Instead, the counts indicate the number of records contributing evidence to each thematic domain. This coding rule underlies the reported thematic frequencies for citizen readiness, governance dynamics, public value co-creation, and sustainable public value outcomes and allows the thematic counts to be traced to the article-level coding matrix.

3.7 Data analysis and thematic synthesis

Data analysis proceeded in four stages. First, descriptive mapping summarized publication trends, document types, source distribution, and relevance bands. Second, conceptual mapping examined how the literature defined and operationalized digital government, sustainability, citizen readiness, governance dynamics, and co-creation. Third, thematic synthesis identified recurring patterns and conceptual tensions across the corpus. Fourth, integrative framework development synthesized recurring evidence patterns, conceptual gaps, and theoretically plausible relationships among the focal constructs within the proposed DR–GD–PVCC–SPV framework. Because the review did not empirically test causal, mediating, or moderating relationships, the relationships incorporated into the framework are treated as theory-building propositions for subsequent empirical examination rather than as empirically established mechanisms.

Analytical transparency was supported through the exact retrieval query, explicit eligibility criteria, a predefined relevance-scoring rubric, structured article-level extraction, a documented coding framework, and supplementary audit materials. The main limitation is that the corpus is limited to Scopus-indexed, English-language, open-access journal articles and review articles meeting the Scopus access filter applied during corpus construction. The open-access restriction facilitated full-text consultation where required for appraisal and coding but may underrepresent subscription-based literature and influence the distribution of journals represented in the corpus. Therefore, the findings should be interpreted as a synthesis of this defined corpus rather than the entire universe of digital government scholarship.

The exact two-block Scopus retrieval query is reported in Section 3.2 and Supplementary Material (Appendix A2). Broader concepts such as citizen readiness, trust, governance capacity, accountability, equity, resilience, and related constructs that did not appear in the final Boolean retrieval query were used only during post-retrieval relevance appraisal, data extraction, and deductive–inductive thematic coding. This distinction separates database retrieval terms from the analytical vocabulary used to interpret the corpus.

Descriptive and bibliometric mapping were used to characterize publication patterns, source distributions, conceptual structures, and keyword relationships, whereas thematic synthesis provided the principal evidence base for framework development. Bibliometric visualizations were therefore treated as descriptive and structural evidence rather than as evidence of causal relationships. The software, data fields, term-cleaning procedures, and figure-specific analytical settings used for the bibliometric outputs are reported in Section 3.8. Because the visualizations were generated through different analytical procedures, methodological parameters are reported only where applicable rather than assumed to be identical across all figures. Supplementary descriptive and bibliometric outputs are reported in Supplementary Material (Figures I1–I8).

The analytical audit trail is documented across the review protocol, eligibility criteria, PRISMA-informed corpus-construction flow, relevance rubric, article-level appraisal scores, codebook, and extraction matrix. These materials allow the progression from corpus construction to coding, thematic synthesis, and framework development to be traced through an article-level evidence trail. Analytical weighting followed the relevance classification described in Section 3.5. The 264 highly relevant, relevant, and moderately relevant records contributed to the broad thematic synthesis, whereas the 181 highly relevant and relevant records constituted the core evidence base for integrative framework development. The 33 supporting/background records were retained primarily for contextual and descriptive mapping.

Analysis operated at two complementary levels. At the metadata level, titles, abstracts, keywords, document types, years, journals, and context cues were used for descriptive mapping and preliminary classification. At the conceptual level, abstracts and full texts were examined to identify how each article contributed to the focal constructs. As described in Section 3.6, the predefined deductive categories were supplemented by recurring inductive concepts, including trust readiness, feedback synthesis, tokenistic participation, platform governance, public value co-destruction, and equity-sensitive evaluation.

Tables, figures, and supplementary materials were used as complementary reporting and audit devices rather than as independent sources of methodological triangulation. The main interpretive claims were derived from the article-level thematic synthesis, while the supplementary materials preserve the information required to inspect corpus composition, appraisal decisions, coding categories, and evidence extraction.

Thematic synthesis followed an evidence-to-concept logic. Each thematic claim was supported by recurring patterns in the corpus rather than by isolated examples. Citizen readiness was coded when records addressed access, literacy, skills, trust, motivation, empowerment, or participatory capacity. Governance dynamics was coded when records addressed coordination, leadership, data governance, platform governance, stakeholder mobilization, or feedback use. Public value co-creation was coded when records addressed co-production, co-creation, participation, digital commons, complaint systems, or user-centered service improvement. Sustainable public value was coded when records addressed inclusion, accountability, equity, trust, resilience, legitimacy, SDGs, or service continuity. Because coding was non-mutually exclusive, these thematic frequencies represent records contributing evidence to each domain and should not be summed as unique studies.

The synthesis was designed to identify both conceptual contributions and persistent absences. A record could contribute strongly to digital inclusion while remaining weak on governance mechanisms, or it could discuss sustainability without explaining co-creation. These asymmetries were treated as evidence of conceptual fragmentation rather than as coding errors. Accordingly, the proposed integrative framework was developed from recurring patterns of conceptual convergence as well as systematically identified disconnections among citizen readiness, governance dynamics, co-creation, and sustainable public value.

Throughout the analysis, descriptive frequency was distinguished from theoretical contribution. High thematic frequency was not interpreted as evidence of a causal relationship or conceptual integration; framework development instead relied on the substantive patterns, gaps, and connections identified across the core evidence base.

3.8 Bibliometric mapping procedures

Bibliometric and descriptive analyses were employed as complementary mapping procedures to characterize the development, intellectual structure, conceptual organization, and collaboration patterns of the reviewed literature. These analyses were conducted separately from the thematic synthesis described in Section 3.7. Whereas article-level thematic coding provided the primary evidence for conceptual interpretation and framework development, bibliometric mapping was used to describe publication patterns, source distributions, international collaboration, keyword relationships, temporal evolution, and thematic positioning within the defined review corpus.

The bibliometric analyses were conducted using VOSviewer version 1.6.20 and Biblioshiny version 5.0. Biblioshiny version 5.0 was used for the author-keyword co-occurrence network, descriptive bibliometric analysis, and the thematic map. VOSviewer version 1.6.20 was used for the temporal-overlay visualization, the author-keyword density visualization reported in Supplementary Material (Figure I8), and the country co-authorship network reported in Supplementary Material (Figure I4). Author keywords were used as the source field for keyword-based analyses.

Before network construction, author keywords were cleaned using a thesaurus to reduce fragmentation caused by spelling variants, capitalization differences, abbreviations, singular–plural forms, and semantically equivalent expressions. Equivalent terms were harmonized before analysis so that conceptually identical terms were represented consistently across the network visualizations. The cleaned author-keyword dataset provided the common input for the co-occurrence, temporal-overlay, density, and thematic analyses, while each visualization was generated using the software and figure-specific procedure reported below.

The author-keyword co-occurrence network was generated in Biblioshiny version 5.0 from the cleaned author-keyword data. The visualization was used to examine patterns of keyword association and conceptual clustering within the corpus. Node size reflects the relative frequency or prominence of keywords within the network, links represent co-occurrence relationships, spatial proximity indicates relative conceptual association, and node colors distinguish groups of more closely connected keywords. Because this network was generated in Biblioshiny rather than VOSviewer, VOSviewer-specific normalization and counting parameters are not attributed to this figure.

The density visualization reported in Supplementary Material (Figure I8) was generated separately in VOSviewer from the cleaned author-keyword data and was used to identify areas of relatively greater keyword concentration within the corpus.

Temporal variation was examined using the VOSviewer overlay visualization. For this analysis, author-keyword co-occurrence was analyzed using full counting, association-strength normalization, and a minimum occurrence threshold of three. Average publication year was used as the temporal attribute assigned to each keyword. Accordingly, node coloration represents the average publication year of the documents in which a keyword occurred, allowing relatively earlier and more recent topics to be distinguished.

International collaboration was examined through a country co-authorship network generated in VOSviewer and reported in Supplementary Material (Figure I4). The analysis used full counting to represent co-authorship relationships among countries represented in the reviewed corpus. Countries were treated as the unit of analysis, and network links represent co-authorship connections between countries. The visualization was used descriptively to identify patterns of international research collaboration and was not incorporated into the empirical support for the theoretical relationships of the DR–GD–PVCC–SPV framework.

The thematic map was generated using Biblioshiny version 5.0 from the cleaned author-keyword dataset. The map positions keyword clusters according to their relative centrality and density. Centrality represents the degree to which a theme is connected to other themes in the conceptual structure of the field, whereas density represents the degree of internal development or cohesion of a theme. Consistent with the centrality–density structure, themes were interpreted across four quadrants: Motor Themes, Niche Themes, Emerging or Declining Themes, and Basic Themes. The thematic map was used descriptively to assess thematic positioning and relative structural development within the corpus rather than as direct empirical evidence for the relationships proposed in the integrative framework.

Annual publication frequencies and other descriptive outputs, including source and country distributions, were calculated directly from the cleaned bibliographic metadata without network normalization. Other descriptive outputs, including source and country distributions, were similarly derived from the cleaned bibliographic dataset. Supplementary Material (Appendix I) provides additional descriptive and bibliometric visualizations of relevance classification, source distribution, geographical distribution, country co-authorship, conceptual structure, and keyword density. Figures I1–I8 in the Supplementary Material are treated as supplementary descriptive or structural outputs and not as independent evidence supporting the proposed framework.

For methodological transparency, the software, source field, term-cleaning procedure, and visualization-specific analytical settings are distinguished according to the visualization to which they apply. The author-keyword co-occurrence network and the thematic map were generated in Biblioshiny version 5.0 using cleaned author-keyword data. The temporal overlay was generated in VOSviewer version 1.6.20 using author keywords, full counting, association-strength normalization, a minimum occurrence threshold of three, thesaurus-based cleaning, and average publication year as the temporal attribute. Figure I4 in the Supplementary Material was generated in VOSviewer using countries as the unit of analysis and full counting; Figure I8 in the Supplementary Material was generated from the cleaned author-keyword data as a supplementary density visualization.

5.1 Rationale of the integrative framework

Building on the thematic findings, this section develops the proposed DR–GD–PVCC–SPV framework as an integrative, theory-building synthesis of digital readiness of citizens, governance dynamics, public value co-creation, and sustainable public value outcomes (Figure 6). Rather than specifying an empirically established causal sequence, the framework organizes theoretically plausible relationships among these constructs that emerge from recurring patterns and conceptual disconnections across the reviewed literature.

tu_pian_6.png

Digital readiness refers specifically to citizens’ capacity to access, understand, trust, use, evaluate, and shape digital public services. Within the framework, readiness is therefore conceptualized not merely as a technological or adoption-related condition, but as a multidimensional citizen capacity that may support agency and participation in public value formation.

The framework responds to the central fragmentation identified through the thematic synthesis. The literature has developed substantial knowledge on digital government, inclusion, governance capacity, participation, and sustainability, but these themes are often examined separately. Many studies explain adoption, institutional capacity, or sustainability outcomes without specifying how they are connected.

The DR–GD–PVCC–SPV framework addresses this fragmentation by proposing a set of theoretically plausible relationships among the four domains. Citizen readiness is positioned as a potential micro-level capacity for agency; governance dynamics is proposed as a meso-level institutional process through which citizen capacities may be activated, coordinated, and incorporated into public-service processes; public value co-creation represents the domain in which citizens may participate across co-design, co-implementation, and co-evaluation; and sustainable public value represents the broader outcome domain associated with inclusion, accountability, legitimacy, resilience, and institutional learning. These relationships are derived from the integrative synthesis and should be treated as propositions for subsequent empirical examination rather than as demonstrated mediating, moderating, or causal effects.

5.2 Framework components

The theoretical contribution of the proposed framework lies in three forms of conceptual integration. First, it repositions citizen readiness from a conventional adoption variable into a multidimensional capacity for potential citizen agency. Citizens are not conceptualized merely as users or beneficiaries, but as actors who may contribute to defining, implementing, evaluating, and improving public value. This extends readiness beyond access and technology use to include the capacities required for meaningful engagement with digital public services.

Second, it conceptualizes governance dynamics as a meso-level institutional process that may activate, coordinate, and channel citizen capacities toward participation and public value co-creation. Governance is therefore not treated as a static institutional background, but as a dynamic process of activation, framing, mobilization, coordination, and feedback synthesis. The framework does not assume that governance dynamics empirically mediate or moderate the relationship between readiness and co-creation; rather, it proposes this relationship for subsequent empirical examination. Third, it reframes sustainability not as an automatic consequence of digitalization, but as a sustainable public value outcome domain that may be associated with citizen capacity, adaptive institutional processes, and substantive public value co-creation. This interpretation shifts attention from whether digital technologies are adopted to whether digital government is associated with inclusion, accountability, legitimacy, resilience, learning, and other durable forms of public value.

The first component, digital readiness of citizens (DR), consists of access, skills, trust, motivation, and participatory capacity. The second component, governance dynamics (GD), consists of activation, framing, mobilization, coordination, and feedback synthesis. The third component, public value co-creation (PVCC), consists of co-design, co-implementation, and co-evaluation. These phases are treated as analytically distinct forms of citizen involvement rather than as a mandatory linear sequence. The fourth component, sustainable public value outcomes (SPV), is interpreted through social inclusion outcomes, democratic-public value outcomes, and system sustainability outcomes. Together, the four components bring insights from digital inclusion, digital government, adaptive governance, public service logic, and public value scholarship into a common theory-building framework [1, 3, 12, 13, 17]. Their proposed relationships are conceptual rather than empirically verified and provide the basis for the propositions developed in the following section.

5.3 Proposed framework relationships

The framework organizes a set of theoretically plausible relationships among citizen readiness, governance dynamics, public value co-creation, and sustainable public value rather than specifying a linear or empirically established causal sequence. Citizen readiness is positioned as a potential micro-level capacity for engagement and agency. However, the thematic synthesis suggests that citizen capacity alone does not necessarily translate into substantive influence over digital public services; institutional processes may shape whether and how such capacity is recognized, coordinated, and incorporated into service processes. Governance dynamics is therefore proposed as a meso-level institutional process involving activation, framing, mobilization, coordination, and feedback synthesis. In the proposed framework, these processes may enable, structure, or constrain the relationship between citizen readiness and substantive participation. This formulation does not assume that governance dynamics statistically mediate or moderate the relationship between readiness and co-creation; rather, mediation, moderation, and alternative configurational relationships remain propositions for subsequent empirical testing.

Public value co-creation is positioned as the participatory domain in which citizen capacities and institutional arrangements may contribute to public value formation. Co-design allows citizens to shape problems, priorities, and service features. Co-implementation involves citizens in use, dissemination, support, and adaptation. Co-evaluation enables citizens to assess performance, provide feedback, and influence iterative improvement. These three dimensions are treated as analytically distinct forms of co-creation rather than as a mandatory sequential pathway. For the purposes of this review, digital participation is treated as substantive public value co-creation when citizen input has an identifiable role in shaping how public services or public value are defined, implemented, evaluated, or improved [11, 14].

Sustainable public value is conceptualized as the proposed outcome domain potentially associated with the interaction of citizen capacities, institutional governance processes, and substantive co-creation. The framework distinguishes three outcome dimensions: social inclusion, democratic-public value, and system sustainability. Social inclusion outcomes include equity, accessibility, empowerment, and reduced exclusion. Democratic-public value outcomes include accountability, transparency, trust, legitimacy, responsiveness, and citizen influence. System sustainability outcomes include resilience, service continuity, adaptive capacity, institutional learning, and long-term improvement.

This outcome structure shifts the evaluative focus beyond platform availability, adoption, or administrative efficiency toward whether digital government is associated with inclusive, legitimate, accountable, responsive, and resilient forms of public value. The proposed pathways shown in Figure 6 should therefore be read as theory-building relationships synthesized from the literature rather than as empirically verified causal, mediating, or moderating effects.

5.4 Proposed propositions

The framework generates four propositions for future empirical research. Proposition 1 (P1) proposes that higher levels of digital readiness reflected in access, skills, trust, motivation, and participatory capacity are more likely to support substantive public value co-creation.

Proposition 2 (P2) proposes that governance dynamics may enable and structure the relationship between citizen readiness and public value co-creation through activation, framing, mobilization, coordination, and feedback synthesis. Whether governance dynamics operates empirically as a mediator, moderator, or alternative configurational condition remains an open question for subsequent testing.

Proposition 3 (P3) proposes that substantive public value co-creation may contribute to sustainable public value by enabling citizen involvement across co-design, co-implementation, and co-evaluation. The proposition does not assume an empirically established mediating effect; rather, it identifies a theoretically plausible pathway for future examination.

Proposition 4 (P4) proposes that sustainable public value outcomes are more likely to be associated with digital government initiatives that combine inclusive citizen readiness, adaptive governance dynamics, and substantive co-creation than with initiatives centered primarily on platform provision or adoption.

Taken together, the propositions frame sustainable digital government as a relational and process-oriented phenomenon while leaving the direction, strength, and causal form of the proposed relationships open to empirical examination.

Figure 6 visualizes these propositions as theory-building relationships rather than empirically verified causal paths. It also introduces a proposed feedback pathway through which sustainable public value outcomes may be associated with future readiness, trust, legitimacy, learning, and participation, as well as a potential co-destruction pathway under weak or exclusionary governance conditions. These recursive relationships remain conceptual propositions requiring empirical investigation.

The framework also provides a diagnostic tool for empirical analysis. Researchers and practitioners may use the framework to examine where disconnections occur, for example, whether citizen capacity is limited, institutional processes fail to incorporate participation, citizen involvement remains tokenistic, or sustainability claims are weakly connected to demonstrable public-value outcomes. This diagnostic use complements, rather than substitutes for, empirical testing of the proposed relationships.

Table 8 summarizes the four theory-building propositions derived from the integrative synthesis. They are not hypotheses estimated from the reviewed corpus, but conceptual relationships that can be examined through quantitative, qualitative, mixed-methods, comparative, longitudinal, or configurational designs.

Table 8. Theory-building propositions derived from the integrative review

Note: P1–P4 summarize theory-building propositions derived from the integrative synthesis of the reviewed literature. They represent theoretically plausible relationships among the focal constructs and should not be interpreted as empirically established causal, mediating, moderating, conditioning, or configurational effects. The propositions are intended to guide subsequent empirical examination using quantitative, qualitative, mixed-methods, comparative, longitudinal, process-tracing, or configurational designs.

The review has theoretical, methodological, and practical implications for understanding and evaluating sustainable digital government. Theoretically, the proposed DR-GD-PVCC-SPV framework reframes sustainable digital government as a set of potentially interrelated citizen-, institutional-, participatory-, and public-value dimensions rather than as an automatic consequence of technological modernization. This shifts the analytical focus from technology-centered explanations of digital transformation to a public value-centered perspective that asks whether and under what conditions digital government is associated with socially and institutionally meaningful outcomes. Existing studies have contributed important insights into digital government, adoption, institutional transformation, and public management [1, 2]. However, the synthesis indicates that sustainability should not be inferred from platform availability, service digitization, or user uptake alone. A broader evaluative perspective should complement adoption and performance indicators with public-value dimensions, including inclusion, equity, accountability, trust, responsiveness, legitimacy, and resilience [3, 4].

Methodologically, the framework provides a basis for examining relationships among micro-level citizen readiness, meso-level governance dynamics, participatory processes of public value co-creation, and sustainable public value outcomes without prespecifying a single causal structure. Quantitative studies can test whether governance dynamics functions as a mediator, moderator, or other conditional factor in the relationship between citizen readiness and public value co-creation. They can also examine whether public value co-creation is associated with variation in sustainable public value outcomes.

Qualitative studies can trace how citizen feedback is interpreted, filtered, ignored, or incorporated into or excluded from institutional learning and service redesign. Longitudinal and mixed-methods designs are particularly valuable because they can capture both changes in relevant outcomes and the institutional processes through which digital participation may acquire, lose, or fail to acquire substantive influence over time. This is consistent with adaptive governance scholarship, which emphasizes learning, responsiveness, and institutional adjustment in complex public systems [17].

Practically, the synthesis suggests that governments should approach digital transformation as both a technological and a governance challenge. Public agencies need to pay attention to digital capabilities, accessible and inclusive service design, trust and cybersecurity, meaningful participation channels, and differentiated support for citizens who face digital exclusion [32], marginalized communities [33], accessibility [34], digital literacy [35], digital divide, cybersecurity, and trust.

Beyond providing digital access, public agencies should develop institutional arrangements through which citizen input can be systematically received, interpreted, responded to, and, where appropriate, incorporated into service improvement. Such arrangements may include structured feedback procedures, co-design forums, cross-agency response protocols, iterative service review, equity-sensitive evaluation, and indicators that capture trust, accountability, accessibility, and responsiveness.

Where participation is collected but has little connection to institutional response, digital engagement may remain symbolic, reproduce unequal voice, or contribute to conditions associated with public value co-destruction. The practical challenge is therefore not only to digitize services, but also to create institutional conditions under which citizen capacities and participation can meaningfully inform more inclusive, accountable, responsive, and resilient public services.

For governments pursuing digital transformation under sustainability mandates, an additional implication concerns evaluation. Digital performance indicators should be complemented by public-value indicators. User numbers, transaction speed, and platform availability can be assessed alongside measures of inclusion, trust, responsiveness, accountability, accessibility, citizen influence, resilience, and learning. This broader approach helps distinguish administrative or technological performance from the wider question of whether digital transformation is associated with sustainable public value [10].

This integrative systematic literature review challenges technology-centered assumptions that digitalization, service availability, or adoption alone are sufficient indicators of sustainable digital government. Across the defined corpus of 297 Scopus-indexed articles and reviews, the synthesis indicates that evaluating sustainable digital government requires attention not only to technological and administrative performance but also to citizen capacity, institutional responsiveness, meaningful participation, and sustainable public value.

The proposed DR-GD-PVCC-SPV framework integrates these previously fragmented domains by positioning digital readiness as a potential capacity for citizen agency, governance dynamics as a meso-level institutional process, public value co-creation as a participatory domain spanning co-design, co-implementation, and co-evaluation, and sustainable public value as the broader outcome domain. These relationships constitute theory-building propositions derived from the integrative synthesis rather than empirically verified causal, mediating, or moderating effects.

Future research should therefore test, refine, and where necessary reconfigure the proposed relationships across diverse institutional, territorial, sectoral, and social contexts, including conditions under which digital transformation may contribute to public value co-destruction rather than public value creation.