Poster #12 - Beyond Local Solutions: The Operationalization of 'Wicked Problems' Using Citizen Request Big Data

• In Event: Friday Poster Reception

Abstract

The concept of 'wicked problems', introduced by Rittel and Webber (1973), describes complex issues resistant to clear definitions or solutions. Despite its widespread use, operationalizing this concept remains challenging, hindering both theoretical advancement and practical application in addressing complex societal issues. Recent digital governance initiatives have enhanced public service delivery and citizen-government interaction globally. These platforms offer multi-million level data with unique opportunities to study persistent issues that challenge local governments, often necessitating escalation to higher administrative levels and or inter-local collaborations. This phenomenon provides a valuable lens through which to examine and potentially operationalize the concept of wicked problems.

This study aims to provide a more precise characterization and operationalization of 'wicked problems' by examining the characteristics of complex issues that local governments struggle to resolve. Using 2 million citizens request in China between 2012 and 2023. We analyze citizen complaints escalated from local service platforms (12345 platforms) to a national-level platform ("Message Board for Leaders") in China, focusing on the patterns and features that make these problems particularly challenging for single local governments to address. By using an innovative supervised Latent Dirichlet Allocation (S-LDA) process with LLMs, we are able to build and verify a systematic typology of wicked problems in local governments and propose an operationalization strategy accordingly.

This work contributes to the ongoing scholarly debate on the nature and management of wicked problems, particularly in the context of digital governance.Theoretically, we offer a reproducible and scalable method for identifying wicked problems, enabling future research to assess their distribution, evolution, and policy consequences. Methodologically, we demonstrate the utility of combining LDA and LLMs in structuring unstructured citizen feedback into meaningful categories of administrative burden and complexity. Practically, our findings inform strategies for improving public service delivery by identifying cases where local governments require support from higher-level institutions or interjurisdictional collaboration. This work has implications for institutional design, especially for designing escalation mechanisms, task allocation, and early detection systems in digital service delivery. By proposing a measurable and interpretable framework for identifying wicked problems, our study helps bridge the gap between theoretical discourse and empirical diagnosis. It opens new avenues for using computational tools to enhance responsiveness and problem-solving in local governance.

Authors

• Corey Kewei XU, Hong Kong University of Science and Technology（Guangzhou)
  Non-Presenting Co-Author

• Ruoyun Wang, The Hong Kong University of Science and Technology(Guangzhou)
  Presenting Author