This course aims at providing a hands-on introduction to evidence-based regulation and public policy. The course will be heavily discussion- and participation-based, and will be organized in two modules: (i) a short introduction to fundamentals of behavioral insights and to nudging and experimental tools; (ii) applications to the digital domain, including hypernudging, algorithmic bias and fairness, dark patterns, personalization, recommendation systems, privacy. At the end of this course students will have earned an understanding of what behavioral sciences and behavioral insights are, how they can, should and should not be applied to public policy, and how they are related to digital media and tools. Students will learn theoretical and practical tools during the first part of the course and they will then be asked to put such instruments to use in imagining digital nudges. Project work will target a specific issue chosen from the macro-areas discussed in the weeks preceeding students' presentations: students will provide a hypothesis of policy recommendation based on behavioral insights and will provide an experimental design intended to validate the hypothesis put forth in the project.

Students will learn about the application of behavioral insights drawn from an array of disciplines to the design, implementation and enforcement of public policies. By the end of the course, students will have acquired innovative tools for public policy design to enrich their abilities to better understand, analyze and involve themselves in policy design.

• In the first module we will review behavioral notions and insights relevant for public policy design. In particular we will review the notion of bounded rationality and its relevance for public policy, i.e. how policy makers can use boundedly rational (biased) judgment and decision making to nudge citizens towards better choices, and how to cognitively empower citizens to overcome bias. • In the second module we will look at the main instruments for evidence-based public policy: experiments and policy tools based on behavioral insights and useful to promote behavior change through evidence-driven public policy. Students will learn how to design and run experiments, in order to create evidence validating and justifying the use of behavioral insights in public policy design. • In the third module, students will delve into a specific problem in digital behavioral public policy and build a project around it. Macro-areas from which topics will be chosen include some or all of: digital choice architecture, online defaults and friction, dark patterns and sludge, personalized nudging, recommender systems, AI assistants and copilots, algorithmic bias and fairness, digital public services, consumer protection, privacy and data disclosure, financial decision-making, administrative burden, risk communication, and AI-supported regulatory or judicial decision-making. Students are encouraged to propose relevant topics not included in the above list. By the end of the second module, students will have selected a topic among the ones discussed in class. The topic will be developed into a group project, which may take the form of a digital nudge prototype, a dark-pattern or sludge audit, an A/B testing design, an algorithmic fairness analysis, or a policy/legal assessment of a digital behavioral intervention. The project will be presented to the rest of the class during the third module of the course, in a format inspired by applied AI/behavioral-science project presentations.

On classic nudging, the textbook is Michael Hallsworth and Elspeth Kirkman, Behavioral Insights, MIT Press 2020, pp. 248, of which we will read parts during the first part of the course. Instructors also will provide a list of readings/lecture notes on digital nudging

Enquiry and Discussion Based course

Class Participation, In class and online presentations and discussions, Project work presentation, and Final Project.

talk to instructors.