Date of Award

Spring 5-15-2017

Author's School

Graduate School of Arts and Sciences

Author's Department


Degree Name

Doctor of Philosophy (PhD)

Degree Type



This dissertation studies strategic social influence in the presence of a self-control problem and egalitarianism.The first chapter is a response to the literature that the standard weak Pareto principle is incompatible with the bundle-reducing transfer principle. Sprumont (2012) proposes the consensual leximin ordering that satisfies i) a Paretian axiom that is weaker than the standard Pareto principle and ii) a property called Dominance Aversion that is a stronger axiom than the bundle-reducing transfer principle. I introduce and study the \emph{leximin Paretian ordering}, which refines the consensual leximin ordering by adding the Pareto principle to the concept of lexicographic egalitarianism. I characterize the leximin Paretian ordering to show that this ordering satisfies stronger notions of Dominance Aversion and of the Paretian axiom used in Sprumont (2012). I also provide an alternative characterization of the consensual Rawlsian ordering. I introduce several new axioms, including the Permutation Pareto Principle and Internal Dominance, and study their logical relationships in this paper.The second chapter provides alternative axiomatic characterizations of the extended egalitarian rules (Moreno-Ternero and Roemer, 2006) in a fixed-population setting of the canonical resource allocation model based on individual capabilities (output functions). The main axioms are \emph{disability monotonicity} (no reduction in the amount of resources allocated to an agent after she becomes more disabled) and \emph{agreement} (when there is a change in agents' capabilities or total resources, all agents who remain unchanged should be influenced in the same direction: all unchanged agents get more or all get less or all get the same amount as before).The third chapter asks the following question: When people or institutions donate to a public good, why do they often announce their plans beforehand? I address this question in the menu choice context in which individuals first set their menus as restrictions on future action choices and later choose their actions. I study an environment in which two kinds of agents, called "selfish" and "normative," choose a level of public good provision. I assume that normative agents have a preference for incorporating reciprocity, but also a temptation to be selfish, as in Gul and Pesendorfer's temptation model. I determine under what conditions there exist subgame-perfect Nash equilibria in which selfish players contribute to the public good, avoiding the tragedy of the commons. I characterize two families of strongly renegotiation-proof equilibrium, and show that the best one exhibits a behavior that I call "moral leadership."


English (en)

Chair and Committee

Jonathan Weinstein

Committee Members

Paulo Natenzon, John Nachbar, Brian Rogers, Daniel Gottlieb,


Permanent URL: https://doi.org/10.7936/K7H70D7D

Available for download on Saturday, May 15, 2117