Exploring stakeholder engagement in urban village renovation projects through a mixed-method approach to social network analysis: a case study of Tianjin

As China’s urbanization accelerates, urban village renovation has become a key initiative for high-quality urban development but faces challenges including complex stakeholder interests, unclear property rights, and weak governance. Mishandling diverse demands of government, developers, and residents can trigger severe social risks. Collaborative decision-making among interdependent stakeholders is crucial. Existing work has analyzed stakeholder roles and relations but often treats them statically, overlooking lifecycle dynamics where stakeholder composition and importance change by stage. This study, drawing on stakeholder theory and social network analysis (SNA), examines a Tianjin urban village renovation case across its lifecycle to identify stakeholders per stage, analyze their interactions, trace role evolution, and propose strategies for multi-stakeholder collaborative governance.

Urban village renovation, rooted in China’s urban–rural land duality, entails complex property rights and multi-stakeholder governance challenges. International experiences (UK, Singapore, Brazil) highlight shifts from government-led clearance to pluralistic, inclusive governance, institutional design, and public participation. Project lifecycle segmentation varies across studies, but commonly includes initiation, preparation, implementation, and completion. Stakeholder theory in this context focuses on government, enterprises, residents, and the roles of social organizations, planners, and media. Differences in stakeholder demands and the need for coordination are central, with studies employing stakeholder analysis, SNA, and hybrid frameworks to identify key actors, conflicts, and governance models. SNA effectively captures networked stakeholder relations, individual positions, and quantifies roles; it has been applied in BIM projects, urban renewal comparisons, and environmental program management. Research gaps: (1) limited attention to diverse stakeholder relations beyond core actors and to collaborative governance from a network perspective; (2) scarce staged, dynamic analyses across the project lifecycle. This study addresses these by applying SNA to a full lifecycle case to uncover network evolution, role shifts, and collaborative strategies.

Design: Mixed-method approach combining a single case study with Social Network Analysis (SNA). The single-case study enables longitudinal tracking of stakeholder behavior across stages; SNA quantitatively analyzes network structure, subgroup characteristics, and node centrality. Case: Urban village renovation in Beichen District, Tianjin, China. Project initiated 2005; demolition/relocation progressed from 2015; first batch relocated in 2021; >3,900 units completed and 1,307 under construction; expected move-in by early 2024. Development mode: government-led, village–enterprise collaboration, market-oriented operation. Lifecycle stages defined: Preliminary Preparation (PP), Demolition and Relocation Compensation (DC), Development/Construction Implementation (CI), and Post-Maintenance/Operation (MO). Stakeholder identification: Literature review plus focus group with six practitioners (district officials, developer and constructor project managers, design and consultant managers). Initial 11 stakeholders; added media and property management via fieldwork, yielding 13 stakeholders: Government (G), Construction Management Unit (CM), Land Consolidation and Acquisition unit (LCA), Construction unit (CU), Constructor (C), Consulting Service unit (CS), Supplier (S), Financial institution (F), Legal institution (L), Community Council (CC), Property Management (PM), Public (P), Media (M). Participation varies by stage per Table 2 in the article. Data collection: Four survey waves aligned to stages: Jun–Jul 2014; Sep–Oct 2016; Jun–Jul 2018; Sep–Oct 2021. Valid responses: 142, 130, 136, 152; total 560, >94% response rate. Respondents represented the 13 stakeholder categories with urban village renovation experience. Relationship strength measurement: Based on rights/interests: contractual relationship=5; subordinate/mutual interests=3; communication opportunity=1; none=0. Constructed four stage-specific 13×13 adjacency matrices (Supplementary Tables S1–S4). Analysis tools and indicators: UCINET 6.0 for network metrics; NETDRAW for visualization. Overall network: density D=2m/[n(n−1)], average geodesic distance L. Group structure: core–periphery decomposition and small clique (CONCOR) analysis. Node roles: centrality measures—degree (communication capacity), betweenness (brokerage/control), closeness (independence/dependence). Centralities normalized for cross-stage comparison (Supplementary Tables S5–S7).

Overall network evolution: Collaboration exhibited a U-shaped trend across stages; information transmission efficiency showed an inverted U-shape. Quantitatively (Table 3): - PP: n=7, density=0.857, average distance=1.143, overall coordination. - DC: n=9, density=0.472, average distance=1.583, core–periphery, weakest connectivity. - CI: n=9, density=0.500, average distance=1.556, core–periphery, modest recovery. - MO: n=6, density=0.800, average distance=1.200, overall coordination. Core–periphery by stage (Table 4): - PP core: Government (core degree 0.64), Consulting Services (0.50), Construction unit (0.43). Periphery includes CM (0.32), Public (0.16), Community Council (0.14), Media (0.10). - DC core: Construction unit (0.87), Public (0.28), LCA (0.28). Periphery includes Financial institution (0.21), Government (0.14), CC (0.07), Legal (0.07), CM (0.06). - CI core: Constructor (0.61), Construction unit (0.61), Consulting Services (0.29). Periphery includes Supplier (0.28), Financial (0.26), Government (0.13), CM (0.11), Media (0.01). - MO core: Property Management (0.89), Public (0.42). Periphery includes Construction unit (0.29), Community Council (0.24), Government (0.10), Media (0.06). Clique structure (Fig. 6; Table 5): - PP: Three cliques—C1 (G, CU, P) strongly linked to C2 (CM, CS) and C3 (CC, M); sparse ties between C2–C3. - DC: Four cliques with a new C1 comprising CM-type bureaus separated from others; lower inter-clique exchange than PP. - CI: Pronounced fragmentation; C1 (CU, Constructor) and C2 (CM, CS, S, F) tightly connected; C3 (Public, Media) weakly connected to others; government did not form an alliance. - MO: Three cliques; Media operates as an independent clique with fewer ties; strong intra- and interconnections between the other two cliques. Role evolution (Fig. 7): - Government: Highest degree and betweenness in PP (leader); declines in DC as CU assumes coordination; in CI, CU’s degree remains much higher and betweenness slightly higher than government (CU more pivotal). In MO, both G and CU decrease, shifting to supervision roles. - Public: Degree and betweenness follow an N-shaped pattern: lowest in DC (limited interactions, paired with media clique), highest in MO (central in operations and services). - Property management: Emerges as core actor in MO. - Media: Lowest closeness (highest independence) throughout; functions as an edge supervisor. Identified collaboration weaknesses: - Low public participation in PP stage (process and capacity deficits). - Weak connectivity and trust asymmetries in DC stage (information asymmetry, conflict over interests; supportive actors like F, L, M loosely connected). - Significant clique fragmentation in CI stage (construction-focused clusters isolated from public/media).

The staged SNA reveals when and why collaboration falters and which actors can remedy gaps. Strong coordination in PP relies on government-led integration, but limited early public involvement seeds later disputes; enhancing two-way communication and participatory mechanisms in PP can reduce DC-stage conflicts. The DC stage’s low density and long paths reflect trust and information asymmetries among government, construction unit, and residents; targeted mechanisms to share information and mediate interests can strengthen ties. In CI, clique fragmentation isolates public/media from construction-centric clusters, risking accountability and responsiveness; integrating marginal stakeholders via shared platforms and transparent workflows can bridge clusters. Role evolution underscores shifting leadership: government transitions from leader to coordinator; construction unit acts as resource controller during DC–CI; public becomes central in MO, necessitating service-oriented management via property firms. These insights inform governance designs that leverage core stakeholders’ resource-driving advantages while institutionalizing inclusive participation and information sharing across stages, thereby improving project performance and social acceptance.

This study combines a single-case analysis with SNA to open the black box of stakeholder relations across the full lifecycle of an urban village renovation project in Tianjin. It quantifies evolving network structures, identifies stage-specific core stakeholders, maps clique formation and fragmentation, and traces role shifts (government from leader to coordinator; construction unit as resource controller; public from passive recipient to active participant; media as independent supervisor). Main collaboration constraints include low early public participation (PP), weak connectivity (DC), and clique fragmentation (CI). Recommended strategies are to align development modes with resident needs, elevate early-stage public participation, leverage core stakeholders’ resource advantages, and build an online information-sharing platform to connect cliques. The theoretical contribution links stakeholder theory with collaborative governance through dynamic SNA; methodologically, it demonstrates a mixed qualitative–quantitative framework; practically, it guides stage-tailored governance. Future research directions include improving relationship measurement (adding interaction intensity and frequency) and comparative analyses across different development modes to generalize findings.

The study centers on a single urban village renovation case in Tianjin, which may constrain generalizability. Relationship strength was operationalized primarily via rights and interests (values 5/3/1/0), not capturing finer-grained interaction intensity, communication, or cooperation frequencies. The authors suggest future work to refine measurement dimensions and conduct comparative studies under different development modes (government-led, government–enterprise cooperative, village collective self-initiated) to examine variability in stakeholder relationships.