Language of Bargaining

Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics Volume 1: Long Papers, pages 13161–13185 July 9-14, 2023 ©2023 Association for Computational Linguistics Language of Bargaining Mourad Heddaya University of Chicago mourad@uchicago.edu Solomon Dworkin University of Chicago solomon.dworkin@gmail.com Chenhao Tan University of Chicago chenhao@uchicago.edu Rob Voigt Northwestern University robvoigt@northwestern.edu Alexander Zentefis Yale University alexander.zentefis@yale.edu Abstract Leveraging an established exercise in negoti- ation education, we build a novel dataset for studying how the use of language shapes bilat- eral bargaining. Our dataset extends existing work in two ways: 1) we recruit participants via behavioral labs instead of crowdsourcing platforms and allow participants to negotiate through audio, enabling more naturalistic in- teractions; 2) we add a control setting where participants negotiate only through alternating, written numeric offers. Despite the two con- trasting forms of communication, we find that the average agreed prices of the two treatments are identical. But when subjects can talk, fewer offers are exchanged, negotiations finish faster, the likelihood of reaching agreement rises, and the variance of prices at which subjects agree drops substantially. We further propose a tax- onomy of speech acts in negotiation and enrich the dataset with annotated speech acts. We set up prediction tasks to predict negotiation success and find that being reactive to the argu- ments of the other party is advantageous over driving the negotiation. 1 Introduction Bilateral bargaining, in the sense of a goal-oriented negotiation between two parties, is a fundamental human social behavior that takes shape in many areas of social experience. Driven by a desire to better understand this form of interaction, a rich body of work in economics and psychology has evolved to study bargaining (Rubin and Brown, 1975; Bazerman et al., 2000; Roth, 2020). How- ever, this work has seldom paid careful attention to the use of language and its fine-grained impacts on bargaining conversations; indeed, many studies operationalize bargaining as simply the back-and- forth exchange of numerical values. Meanwhile, there is growing interest in bargaining in NLP ori- ented towards the goal of building dialogue systems capable of engaging in effective negotiation (Zhan et al., 2022; Fu et al., 2023). In this work, we aim to bridge these two lines of work and develop a com- putational understanding of how language shapes bilateral bargaining. To do so, building on a widely used exer- cise involving the bargaining over the price of a house used in negotiation education, we develop a controlled experimental environment to collect a dataset of bargaining conversations.1 The treatment in our experiment is the manner in which subjects communicate: either through alternating, written, numeric offers (the alternating offers or AO condi- tion) or unstructured, verbal communication (the natural language or NL condition). Furthermore, to encourage naturalistic interactions, we recruit participants via behavioral labs and allow partici- pants to negotiate in a conversational setting using audio on Zoom instead of crowdingsourcing text conversations as prior work has done (Asher et al., 2016; Lewis et al., 2017; He et al., 2018). In total, …

Content truncated to 500 words