Leveraging AI to Understand Public Policy Dynamics

This research project explores how machine learning can be leveraged to analyze whether and how public policies evolve over time. By integrating political science research with advanced data science methodologies, the project investigates policy dynamics in critical areas of the green transition.

About the research project

Addressing pressing societal challenges—such as natural resource management, land use conflicts, and climate change—requires policy adaptation and transformation. However, research in both economics and social sciences consistently indicates that policies tend to remain stable, and the mechanisms driving policy change remain elusive.

Traditional studies of policy change have primarily relied on qualitative case studies or survey-based statistical analyses. These approaches often struggle to detect incremental changes or may overemphasize large-scale shifts, limiting their ability to capture the complexity of policy evolution. Given that public policy processes are non-linear, iterative, and deeply influenced by historical decisions, studying them requires longitudinal designs, extensive datasets, and analytical methods that account for serial correlations.

This project addresses these challenges by applying machine-learning techniques to analyze policy dynamics over time. By combining expertise in political science with data science methodologies, the research will examine key policy areas relevant to the green transition, beginning with an in-depth study of Swedish mineral policy developments over the past 30 years. How can machine larning help us study policy change over time?

Funding

The project is funded through a postdoctoral research grant from SUN Natural Resources for Sustainable Transitions at Luleå University of Technology.

Participators

• Ahmed Alregal
• Simon Matti
• Felicia Robertson
• Annica Sandström