fbpx Science of Behavior Change | Harvard Kennedy School
Todd Rogers Photo

Todd Rogers

Professor of Public Policy

Over the last 30 years, psychologists and economists have joined forces to study how people process information and actually make decisions, rather than how they would make decisions if they were fully rational and selfish.  This research program (dubbed behavioral economics or behavioral science) has provided an understanding of how people’s decisions deviate from “optimal” choices as well as the consequences of such deviations. This course is devoted to understanding the nature, causes, implications and applications of these limitations.  This course focuses on how these judgment, decision-making and behavior tendencies can inform the design and development of welfare-enhancing interventions.

The Science of Behavior Change (MLD 304) has one central objective: to improve students’ abilities to design policies and interventions that improve societal well-being.  It accomplishes this by focusing on how to leverage insights about human decision making to develop interventions (“nudges”).  This will be accomplished by building on the toolbox that standard economics provides for influencing behavior (namely, incentives and information) with the insights from behavioral science. 

There are three additional, though secondary, goals for this class.  First, it will help you better understand the science of how humans make judgments and decisions.  We will review research on human thinking from social psychology, cognitive psychology, political science, organizational behavior, decision science, and economics.  In the process you will also learn how randomized experiments work and why they are critical for making inferences about causal relationships.

Second, this course aims to improve the quality of your own judgments and decisions.  People are poor intuitive statisticians, meaning that when they “just think” about situations for which some data or casual observations exist, they tend to make serious inferential errors, in turn leading to systematically biased decisions. We will study some errors that are particularly important for real world problems and look for easy-to-implement solutions. 

Third, this course aims to increase your familiarity with randomized experiments so you can be a smarter consumer of claims that interventions cause certain outcomes.  The class will be suffused with randomized experiments and we will repeatedly discuss how confident one can be that intervention X causes outcome Y.   

Applications of the material covered in this course include policy design, healthcare, energy, politics, education, finance, negotiation, risk management, diversity, human resource management, and organization of teams, among others.