Mechanism Design and Strategic Complexity

Abstract

This thesis concerns the problem of strategic complexity in economics. Every economic theory that models human behavior has limits, which includes limiting what a human can do. Often, economic theory supposes that humans are strategically simple, making at most a handful of decisions. This thesis investigates enriched economic models where humans are strategically complex, being allowed to take a myriad of interrelated strategic decisions. The central goal is building a framework for understanding when strategic complexity is empirically relevant, can reverse the economic predictions, and undermine policy commendations. The thesis has two parts, the first being empirical (Chapters 1 and 2) and the second being theoretical (Chapters 3-5). In Chapter 1, I study groundwater-usage in the American Midwest using a dataset that accounts for 3% of US irrigated agriculture. I develop new econometric tests to show that standard, strategically simple models fail to explain real-world behavior, thereby motivating the need for more strategically complex models. Chapter 2 investigates the interaction between public debate in the media and national-level asylum practices in Europe, where I show that governments faced a much more strategically complex environment than what has been previously suggested in the literature. In Chapter 3, I develop a mechanism design framework for overcoming strategic complexity. This framework allows me to characterize all social goals that can be implemented in a way that is robust to strategic complexity. In Chapter 4, I focus on a class of somewhat infamous theorems in economics— commonly referred to as “folk theorems”—which suggest that strategic complexity in the form of dynamic strategies can render economic models uninterpretable. I propose a novel methodology for resolving folk theorems, which enables researchers to make sharp predictions notwithstanding strategic complexity. Finally, in Chapter 5, I develop new mathematical tools for identifying economic predictions in a way that is robust to strategic complexity. Taken together, this thesis pioneers a new framework for studying economic models, resolving the problem of strategic complexity, and designing more robust policies for real-world economic systems.