Sangam, an Indian restaurant in Philadelphia, offers an all-you-can-eat lunch buf-
fet for $5. Customers pay $5 at the door, and no matter how many times they refill their plates, there is no additional charge. One day, as a goodwill gesture, the owner of the restaurant tells 20 randomly selected guests that their lunch is on the house. The remaining guests pay the usual price. If all diners are rational, will there be any difference in the average quantity of food consumed by people in these two groups?
Having eaten their first helping, diners in each group confront the following question: “Should I go back for another helping?” For rational diners, if the benefit of doing so exceeds the cost, the answer is yes; otherwise it is no. Note that at the moment of decision about a second helping, the $5 charge for the lunch is a sunk cost. Those who paid it have no way to recover it. Thus, for both groups, the (extra) cost of another helping is exactly zero. And since the people who received the free lunch were chosen at random, there is no reason to suppose that their appetites or incomes are different from those of other diners. The benefit of another helping thus should be the same, on average, for people in both groups. And since their respective costs and benefits of an additional helping are the same, the two groups should eat the same number of helpings, on average.
Psychologists and economists have experimental evidence, however, that people in such groups do not eat similar amounts.
In particular, those for whom the luncheon charge is not waived tend to eat substantially more than those for whom the charge is waived. People in the former group seem somehow determined to “get their money’s worth.” Their implicit goal is apparently to minimize the average cost per bite of the food they eat. Yet minimizing average cost is not a particularly sensible objective. It brings to mind the man who drove his car on the highway at night, even though he had nowhere to go, because he wanted to boost his average fuel economy. The irony is that diners who are determined to get their money’s worth usually end up eating too much, as evidenced later by their regrets about having gone back for their last helpings.
The fact that the cost-benefit criterion failed the test of prediction in this example does nothing to invalidate its advice about what people should do. If you are letting sunk costs influence your decisions, you can do better by changing your behavior. In addition to paying attention to costs and benefits that should be ignored, people often use incorrect measures of the relevant costs and benefits. This error often occurs when we must choose the extent to which an activity should be pursued (as opposed to choosing whether to pursue it at all). We can apply the Cost-Benefit Principle in such situations by repeatedly asking the question “Should I increase the level at which I am currently pursuing the activity?”
In attempting to answer this question, the focus should always be on the benefit and cost of an additional unit of activity. To emphasize this focus, economists refer to the cost of an additional unit of activity as the marginal cost of the activity. Similarly, the benefit of an additional unit of the activity is the marginal benefit of the activity.
When the problem is to discover the proper level at which to pursue an activity, the cost-benefit rule is to keep increasing the level as long as the marginal benefit of the activity exceeds its marginal cost. As the following example illustrates, however, people often fail to apply this rule correctly.
fet for $5. Customers pay $5 at the door, and no matter how many times they refill their plates, there is no additional charge. One day, as a goodwill gesture, the owner of the restaurant tells 20 randomly selected guests that their lunch is on the house. The remaining guests pay the usual price. If all diners are rational, will there be any difference in the average quantity of food consumed by people in these two groups?
Having eaten their first helping, diners in each group confront the following question: “Should I go back for another helping?” For rational diners, if the benefit of doing so exceeds the cost, the answer is yes; otherwise it is no. Note that at the moment of decision about a second helping, the $5 charge for the lunch is a sunk cost. Those who paid it have no way to recover it. Thus, for both groups, the (extra) cost of another helping is exactly zero. And since the people who received the free lunch were chosen at random, there is no reason to suppose that their appetites or incomes are different from those of other diners. The benefit of another helping thus should be the same, on average, for people in both groups. And since their respective costs and benefits of an additional helping are the same, the two groups should eat the same number of helpings, on average.
Psychologists and economists have experimental evidence, however, that people in such groups do not eat similar amounts.
In particular, those for whom the luncheon charge is not waived tend to eat substantially more than those for whom the charge is waived. People in the former group seem somehow determined to “get their money’s worth.” Their implicit goal is apparently to minimize the average cost per bite of the food they eat. Yet minimizing average cost is not a particularly sensible objective. It brings to mind the man who drove his car on the highway at night, even though he had nowhere to go, because he wanted to boost his average fuel economy. The irony is that diners who are determined to get their money’s worth usually end up eating too much, as evidenced later by their regrets about having gone back for their last helpings.
The fact that the cost-benefit criterion failed the test of prediction in this example does nothing to invalidate its advice about what people should do. If you are letting sunk costs influence your decisions, you can do better by changing your behavior. In addition to paying attention to costs and benefits that should be ignored, people often use incorrect measures of the relevant costs and benefits. This error often occurs when we must choose the extent to which an activity should be pursued (as opposed to choosing whether to pursue it at all). We can apply the Cost-Benefit Principle in such situations by repeatedly asking the question “Should I increase the level at which I am currently pursuing the activity?”
In attempting to answer this question, the focus should always be on the benefit and cost of an additional unit of activity. To emphasize this focus, economists refer to the cost of an additional unit of activity as the marginal cost of the activity. Similarly, the benefit of an additional unit of the activity is the marginal benefit of the activity.
When the problem is to discover the proper level at which to pursue an activity, the cost-benefit rule is to keep increasing the level as long as the marginal benefit of the activity exceeds its marginal cost. As the following example illustrates, however, people often fail to apply this rule correctly.
No comments:
Post a Comment