Three Essays on Strategic Behavior and Policy.

Abstract

This dissertation contains three chapters that study the strategic behavior of economic agents in environments shaped by legal and regulatory policy. Two chapters examine strategic behavior as it pertains to the market outcomes of firms competing as oligopolists; a third studies the negotiation outcomes of a firm and consumer bargaining in the shadow of the law. Each chapter extends or applies fundamental models of strategic behavior to develop testable implications. It then exploits the historical change induced by a policy intervention to empirically test the model's mechanisms, conditions, and predictions. Finally, it interprets the meaning of empirical estimates for policy toward antitrust and other types of law.

The first chapter analyzes the sustainability and consequencs of collusion in a high-technology industry with learning-by-doing. It shows how the dynamic cost savings that occur through learning, combined with multiproduct competition, can reverse the standard predictions of static supergame models of collusion. It illustrates this by building an oligopolistic model of learning-by-doing embedded within a supergame. It estimates the price change attributable to collusion by using firm-level data before, during, and after explicit collusion in the electronic memory chip market.

The second chapter assesses the predictions of the Priest-Klein model of pre-trial settlement bargaining. It applies the model to an auto insurance negotiation setting to show how parties' uncertainty in a legal standard drives the likelihood of claim settlement. It then exploits state- and time-level variation in states' adoption of tort liability for insurance bad faith law to find that tort liability significantly increased the likelihood of a claim to settle with litigation in the short-term, but significantly reduced this likelihood in the long-term.

The third chapter uses data from a price fixing cartel in the Korean petrochemical industry to test empirical screens of collusion. It exploits the information disclosed by the Korean antitrust regulator to formulate a hypothesis of the effect of collusion on price-cost asymmetry. It uses price data before and after the breakdown of the cartel to find evidence consistent with a change from high to low price-cost asymmetry before and after the cartel's collapse.