Kahneman presents another puzzle, asking the reader to set compensation for a victim of a violent crime—in this case, a man who lost his arm after being shot during a robbery. The puzzle asks, should the store in which the man was shot (either his regular store, or a different store he rarely went to) make a difference in the compensation amount? When people evaluate these two scenarios at the same time (called joint evaluation), most people believe that the compensation should be the same in both situations, because location should not be a factor.
When the two situations are considered together, people are able to understand that location should not be pertinent to their calculation of victim compensation. In this instance, they are able to mobilize System 2 in order to evaluate what information is and is not relevant to the task at hand.
When people are given one option or the other (rather than both together), they assign a much higher compensation to the version in which the store is not his regular store, because they see it as more poignant. Seeing only one option is how people normally experience life: alternatives that might change our minds or affect our decisions are absent.
In comparison with the example above, there is a clear instance of WYSIATI here. The information that people have been provided is the only thing they consider—a bias inherent in System 1. This demonstrates how difficult it is to mobilize System 2 in a case like this, because often we don’t know what alternatives might change our mind.
Another economic experiment focused on these kinds of reversals. In this experiment, participants are told that they have a choice between two bets, played on a roulette wheel with 16 sectors. Bet A: 11/36 to win $160, 25/36 to lose $15. Bet B: 35/36 to win $40, 1/36 to lose $10. People usually prefer bet B. But, if they imagine that they own each bet and must determine the lowest price at which they would sell it, the selling price is higher for A than for B. This experiment was surprising to economists, but they accepted that individual choice can depend on the context—a clear violation of the idea that people are rational and make the same choice in every scenario.
Again, reversals constitute a particularly interesting problem of the human mind. When we evaluate certain odds, we rely on our intuition—a feature of System 1. But when our intuition appears inconsistent, we have a difficult time understanding what our basic underlying beliefs are.
The fact that evaluation depends on context and categories is proven in this example: John is 6. He is 5 feet tall. Jim is 16. He is 5 feet 1 inches tall. Individually, everyone will agree that John is tall, and Jim is not. But the question “Is John as tall as Jim?” yields a different answer, because it prompts a direct comparison.
System 1 understands features of a given category, and direct comparisons among a category are easy to do. But the next example demonstrates how comparisons across categories—even though we have intuitions about them—often yield inconsistencies.
Kahneman next asks readers to imagine that they have been asked to contribute money to help set up pollution-free breeding locations for dolphins. People can come up with an amount by translating their attitudes towards dolphins onto the scale of their normal contributions (particularly referencing contributions to environmental issues).
This example builds upon Kahneman’s earlier discussion of substitution. We make our evaluation easier by noting how much we care about dolphins, perhaps in comparison to other sea creatures, and by matching that affinity to an appropriate dollar amount.
On another occasion, readers might be approached to support medical check-ups for farmworkers, who have a higher rate of skin cancer. People’s contributions reflect how urgent they feel the issue is, particularly in comparison with other medical concerns. In single evaluation, the dolphins generally attract larger contributions. But taken together, the issues are represented differently, because people feel that humans deserve more aid than animals.
Here is where the example becomes more difficult: we have intuitions both about how much we like dolphins and about how much we care about farmers, but when we have to compare the two directly our intuitions drastically shift. These questions demonstrate how much we need to simplify questions in order to answer them, whether it is by comparing dolphins to other sea animals or farmers’ risk of cancer to other medical issues.
Reversals can be found in the justice system as well. In mock juries, people assessed pairs of cases. In the first case, a child suffered moderate burns when his pajamas caught fire; the firm that produced the pajamas had not made them fire resistant. In the second case, the unscrupulous dealings of a bank cost another bank $10 million. In single evaluation, the bank was awarded higher damages than the child. In joint evaluation, sympathy toward the individual victim prevailed and the jurors awarded higher damages to the child.
The inconsistencies in the justice system are troubling because we do not have inherent anchors for different crimes. Any jury might be affected by a case that came before or after it, because their frame of reference for damages in various crimes is relatively narrow. In this example, the damages for the bank becomes an anchor, and so the child is given more money than he might have otherwise received. Again, this shows how moral intuitions can vary with context.
In another study, psychologists compared different administrative punishments that can be imposed by government agencies like OSHA and the EPA. The punishments given by an individual agency seem sensible, but between agencies appear incoherent. A fine for a serious violation of regulations concerning worker safety is capped at $7,000, while a violation of the Wild Bird Conservation Act can result in a fine up to $25,000. One can see the absurdity only when the two cases are viewed together.
This study shows how far-reaching the inconsistencies are. Even between entities that are controlled by the same governing body (which would presumably have a greater body of knowledge than a jury), inconsistencies abound in evaluating violations against humans versus violations against wildlife. Once again Kahneman demonstrates that biases are not unique to individuals or even one system but are far-reaching in society.