Algorithmic Governance: Machine Learning and Spatial Analytics in Public Policy

Session Submission Type: Panel

Abstract

This panel is one of the guaranteed sessions for the Korean Association for Policy Studies (KAPS) based on the MOU between APPAM and KAPS.This panel examines the intersection of emerging technologies, spatial dynamics, and institutional performance in modern public administration. As governments transition from theoretical policy frameworks to data-driven realities, they face complex challenges in maintaining equity, integrity, and fiscal health. By synthesizing research on machine learning, geographic information systems (GIS), and institutional behavior, this session offers a multi-dimensional analysis of how digital and spatial characteristics shape public service delivery and economic outcomes.The discussion begins by addressing spatial and digital inequality, exploring how urban spatial characteristics influence public transit ridership and whether digital capacity tends to cluster within specific local governments in South Korea. These findings highlight the risk of "spatial digital divides" that can exacerbate regional disparities. This spatial lens is further applied to resource allocation and municipal finance, utilizing machine learning to uncover inequities in municipal borrowing costs and to optimize emergency medical services. These papers demonstrate that while advanced algorithms can improve efficiency, they also reveal deep-seated systemic biases that policymakers must address to ensure equitable resource distribution.The panel then shifts to the psychological and institutional consequences of performance regimes. Research on South Korean State-Owned Enterprises reveals how high-stakes accountability can lead to "organizational cheating" and institutional decoupling, particularly when mediated by political patronage. This internal tension is mirrored in the workforce, where interpretable machine learning models identify the non-linear drivers of civil servant turnover. Finally, the panel connects these administrative phenomena to macroeconomic stability, presenting cross-country evidence on how government net worth serves as a foundation for long-term economic growth.Collectively, these papers provide a comprehensive framework for understanding the "rhetoric to reality" gap in innovation. Attendees will gain critical insights into how governments can leverage AI and spatial data to enhance public value while mitigating the ethical risks and inequalities inherent in high-stakes, data-driven environments.

Policy Area

• Methods and Tools of Analysis

Chair

• Minchang LEE, Chosun University

Discussant

• Minyoung Ku, Florida State University

Organizer

• Dohyeong Kim, University of Texas, Dallas

Individual Presentations

• Rethinking Urban Loneliness Policy From a Place-Based Perspective - Presenting Author: Hemin Choi, Incheon National University

• Complement or Amplifier? Regional Ecosystems and the Returns to Digital Capability in Korea's AI Transition - Presenting Author: Hanbyul Choi, Jeonbuk National University; Non-Presenting Co-Author: Young-in Cha

• Who Pays the Price? Inequity in Municipal Borrowing Costs: Evidence from Machine Learning - Non-Presenting Co-Author: Yoon-Jung Choi, Florida International University; Presenting Author: Il Hwan Chung, Sungkyunkwan University; Non-Presenting Co-Author: Byeong Ho LEE

• Machine Learning and Public Resource Allocation: Evidence from Emergency Medical Services in South Korea - Presenting Author: Dohyo Jeong, University of Kentucky; Non-Presenting Co-Author: Minyoung Ku, Florida State University

• Policy Implications of Experiential Programs for Expanding the AI Talent Base - Presenting Author: Seung Hyun Lee, Korea Foundation for Science and Creativity