Consider your friend who has a final paper due in a relatively important class. The paper will be worth about half of his grade in the class. It is Friday at 7:00 P.M., and he must complete the paper by Monday morning at 9:00 A.M. He believes that the paper will take about six hours of hard work to complete. However, he is in the middle of a role-playing video game and is really itching to play the game for at least another 20 minutes before diving into his paper. Upon considering his options, he believes the 20 minutes will not set him back too far because it is such a small amount of time. Further, the 20 minutes would not allow him to make very much progress at all on his paper relative to the amount of progress he might be able to make on the game. Thus, he decides to play. About 20 minutes later, he faces the same decision. But 20 minutes still seems like such a short period compared to the amount of time he has left to finish the paper. So, he continues to play. At about 3:00 A.M., after playing for eight hours, he decides to go to bed and work on the paper the next day. He regrets the amount of time he spent playing video games, and realizes that he could have made significant progress on the paper and could perhaps have completed it by then. He laments, “How could I have played video games for eight hours instead of working on my paper?”
Consider, alternatively, a company headed by two people: a CEO and a president of operations. The CEO has ultimate responsibility for the company, and the president directly oversees the eight team leaders who make up the senior management. Each of the team leaders has some autonomy and can decide on the particular projects that they will undertake. The CEO is worried that although the company is profitable, profits are smaller than she would like. Wondering the root cause of the problem, she calls on the president to consider the problem. The president meets one by one with each team leader, asking for detailed accounts of their decision-making process. He finds that each has engaged in relatively safe projects. As well, each has turned down some risky projects. For example, a typical risky project would return $20 million with about 50 percent probability and lose $10 million with 50 percent probability. Considering the overall budget of this individual team is only around $20 million, the president considers the decision to scuttle such risky projects sound. He compiles his report detailing the similar decision processes and abandoned opportunities from each of the eight teams to the CEO. The CEO, upon reading about the potential for eight projects, each returning about $20 million with 50 percent probability, and losing $10 million with 50 percent probability, calculates the expected value to be $40 million, with less than a 15 percent chance of losing any money. She immediately fires the president of the company.
Much like the framing effects discussed in previous chapters, how decisions are grouped together can have a tremendous impact on which outcomes look most attractive. Although few in the position of needing to work on a substantial class project in the next 48 hours would choose to play video games for several hours straight before beginning the project, many might make several individual decisions to play for 20 minutes at a time. Further, whereas a single investment choice might sound prohibitively risky, a group of many choices each bearing the same risk might sound very attractive. Bracketing or choice bracketing refers to how choices are grouped together. Often choices are grouped naturally by their placement in time. For example, one generally decides on what to have for breakfast before deciding on what to have for lunch or dinner. Nonetheless, decisions regarding the size and content of breakfast can severely affect what may practically be eaten for lunch. Once you have had an especially large lunch, you might feel uncomfortable having a particularly heavy dinner in addition.
Decision bracketing is closely related to hedonic editing or framing. However, hedonic editing and framing deal only with how people evaluate events (valuing them more or less), whereas choice bracketing deals directly with how the decisions themselves are made (which tradeoffs or variables are considered when making the choice). Choice bracketing has implications both for risky and riskless choice. However, discussing risky choice requires some review of rational choice under uncertainty, hence the placement of this chapter in the Information and Uncertainty section of the course. We review rational models of multiple decisions as well as the most widely accepted rational model of decision under risk: expected utility theory. This model is covered in greater detail in Chapter 9, including its axiomatic foundation (i.e., why we consider it rational).