Convergence in Policy Recommendation? Analyzing Academic Research in Transportation Using Computational Social Science

• In Event: Student Research: Politics of Public Policy

Abstract

Academic journals serve as critical venues for communicating policy-relevant research findings. Within this context, policy recommendations from research artivles play a pivotal role, serving as a bridge between theoretical insights and practical applications. Scientists should serve as "honest brokers," expanding policy options without explicitly advocating for specific choices. However, the interplay between empirical findings ("is" statements) and normative recommendations ("ought" statements) highlights the dual necessity of both for effective policy-making. Together, they ensure that research not only informs but also guides policy formulation and decision-making processes. However, crafting effective recommendations involves balancing generalizability and contextual specificity, a challenge compounded by disciplinary norms and methodological constraints.

This study addresses a largely underexplored phenomenon: the convergence of policy recommendations across academic research, particularly in the context of electric vehicle (EV) policy in transportation. It highlights how reliance on shared theoretical frameworks, cross-jurisdictional learning, and the increasing use of large language models (LLMs) may inadvertently reinforce similarity in policy prescriptions, potentially fostering broader policy convergence. Employing computational social science methods and leveraging LLM capabilities, the study develops a novel measure of policy recommendation convergence, assesses its prevalence in EV research, and investigates the factors driving this trend. The findings aim to deepen understanding of how policy recommendations evolve and the implications for evidence-based policy-making and the integration of computational tools in academic scholarship.

Authors

• Zhizhao Li, The Hong Kong University of Science and Technology
  Presenting Author

• Xuan ZHANG
  Non-Presenting Co-Author