Standing United or Falling Divided? High Stakes Bargaining in a TV Game Show

摘要

Bargaining is ubiquitous in our professional and private lives. Not surprisingly, bargaining has received considerable research attention. Because real-world settings generally entail a lack of control, most empirical insights derive from laboratory experiments. The external validity of laboratory findings is, however, still an open question (Levitt and List 2007; Camerer 2015; Baltussen, van den Assem, and van Dolder 2015). Two of the grounds for concern are the frequent use of student subjects and the small or hypothetical stakes. In the present study, we use data from the British TV show Divided. This game show combines high stakes and a diverse subject pool within a controlled setting. We find that individual behavior and outcomes are strongly influenced by equity concerns: those who contributed more to the jackpot claim larger shares, are less likely to make concessions, and take home larger amounts. Threatening to play hardball is ineffective: although contestants who announce that they will not back down do well relative to others, they do not secure larger absolute amounts and they harm others. There is no evidence of a first-mover advantage and little evidence that demographic characteristics matter.