What Makes an Effective Negotiator? Measuring Sophistication, Behavior, and Learning in Bargains

Abstract

From children's arguments over toys to executives' dialogues during business deals, negotiation is a key factor in many aspects of economic life. Bargaining is extensively studied in the economics, psychology, and computer science literatures, and these academic findings may shape the way consumers and business leaders think about bargaining strategy. However, much of this work is either theoretical or based on experiments, as there are few rich data sources on bargaining in the field. In this thesis, using a recently-released dataset of millions of bargaining outcomes on eBay, I employ competitive skill models to empirically study the factors that make negotiators effective. Using these models, I corroborate the relationship between patience and bargaining power posited in theory, and find empirical measures that distinguish between latent (agent-specific) and local (bargain-specific) patience. Furthermore, I analyze the behavioral effects of round numbers and the 50-50 focal point, and find that i) more sophisticated sellers use marked-up, rounded listings, which garner more attention to their products, and ii) more sophisticated buyers use the 50-50 split, which improves the chances their offers are accepted. Finally, I use a dynamic model of skill to measure changes in bidder sophistication. This model suggests that bidders who bargain better over time tend to become more patient and take better advantage of behavioral factors. Together, this work demonstrates the power of empirically measuring skill with competitive models, which can provide more comprehensive insight into what makes negotiators effective.