When Does Platform-Based Governance Generate Collaboration? Policy Domains, Public Attitudes and Digital Collaboration

• In Event: AI in Public Sector Service Delivery

Abstract

Digital government platforms are widely viewed as institutional infrastructures that enhance cross-department collaboration by integrating information flows and standardizing administrative processes. Existing research in collaborative public management and digital governance often assumes that collaboration emerges as a relatively automatic outcome of platform adoption. This paper challenges that assumption by arguing that platform-mediated collaboration is conditional rather than inherent. 

Specifically, this study examines how collaboration is shaped by two key factors: policy domains and public attitudes expressed through citizen claims. Different policy areas entail varying administrative responsibilities, coordination costs, and risk profiles, while public demands influence how government agencies prioritize issues and communicate responses. Together, these factors determine whether, how, and to what extent collaboration is articulated through digital platforms. 

Empirically, the analysis draws on 799,642 records of citizen–government online interactions from a large local complaint platform in China. Each record includes the citizen's request, the responding agency, the official reply, and timestamps. Although internal coordination processes are not directly observable, these data capture an important observable dimension of collaboration: how coordination is signaled and communicated in official responses. 

Methodologically, the study adopts a hybrid approach combining manual coding, dictionary-based classification, and supervised machine learning. A sample of claims is manually coded to identify policy domains and issue complexity, while functional dictionaries classify the full dataset. Public attitudes are measured through textual features capturing confrontational, demanding, or cooperative orientations. Collaboration is operationalized through discursive markers in government responses, ranging from general references to coordination to more specific procedural descriptions. 

The findings show that collaborative signaling is uneven across contexts. Coordination cues are more prevalent in policy domains characterized by fragmented and interdependent responsibilities, such as infrastructure and urban management, while domains with clearer administrative ownership exhibit more siloed responses. In addition, confrontational citizen demands are associated with more standardized and less specific coordination language, whereas less confrontational claims tend to elicit more detailed and procedural responses. 

This study contributes to collaborative governance research by demonstrating that digital platforms do not automatically generate collaboration, but instead condition how it is expressed. By conceptualizing collaboration as a communicative and context-dependent practice, the paper highlights the importance of policy domain characteristics and citizen-state interactions. Methodologically, it offers a scalable approach to studying collaboration using large-scale administrative text data, providing practical insights for improving responsiveness and coordination in digital government systems.

Author

• Jiahuan Qian, Zhejiang University
  Presenting Author