The Deception of Specific Cases: Conjunction Fallacy
Chris is thirty-five. He studied social philosophy and has had an interest in developing countries since he was a teenager. After graduation, he worked for two years with the Red Cross in West Africa and then for three years in its Geneva headquarters, where he rose to head of the African aid department. He then completed an MBA, writing his thesis on corporate social responsibility. What is more likely? (a) Chris works for a major bank or (b) Chris works for a major bank, where he runs its Third World foundation. A or B?
Most people will opt for B. Unfortunately, it’s the wrong answer. Option B does not only say that Chris works for a major bank but also that an additional condition has been met. Employees who work specifically within a bank’s Third World foundation comprise a tiny subset of bankers. Therefore, option A is much more likely. The conjunction fallacy is at play when such a subset seems larger than the entire set—which by definition cannot be the case. Amos Tversky and Nobel laureate Daniel Kahneman have studied this extensively.
We are easy prey for the conjunction fallacy because we have an innate attraction to “harmonious” or “plausible” stories. The more convincingly, impressively, or vividly that Chris the aid worker is portrayed, the greater the risk of false reasoning. If I (Rolf Dobelli) had put it a different way, you would have recognized the extra details as overly specific, for example: “Chris is thirty-five. What is more likely? (a) Chris works for a bank or (b) Chris works for a bank in New York, where his office is on the twenty-fourth floor, overlooking Central Park.”
Here’s another example: What is more likely? (a) “Seattle airport is closed. Flights are canceled,” or (b) “Seattle airport is closed due to bad weather. Flights are canceled.” A or B? This time, you have it: A is more likely since B implies that an additional condition has been met, namely, bad weather. It could be that a bomb threat, accident, or strike closed the airport; however, when faced with a “plausible” story, we don’t stop to consider such things. Now that you are aware of this, try it out with friends. You will see that most pick B.
Even experts are not immune to the conjunction fallacy. In 1982, at an international conference for future research, experts—all of them academics—were divided into two groups. To group A, Daniel Kahneman presented the following forecast for 1983: “Oil consumption will decrease by 30 percent.” Group B heard: “A dramatic rise in oil prices will lead to a 30 percent reduction in oil consumption.” Both groups had to indicate how likely they considered the scenarios. The result was clear: Group B felt much more strongly about its forecast than group A did.
Kahneman believes that two types of thinking exist: The first kind is intuitive, automatic, and direct. The second is conscious, rational, slow, laborious, and logical. Unfortunately, intuitive thinking draws conclusions long before the conscious mind does. For example, I experienced this after the 9/11 attacks on the World Trade Center. I wanted to take out travel insurance and came across a firm that offered special “terrorism cover.” Although other policies protected against all possible incidents (including terrorism), I automatically fell for the offer. The high point of the whole farce was that I was willing to pay even more for this enticing yet redundant add-on.
In conclusion: Forget about left brains and right brains: The difference between intuitive and conscious thinking is much more significant. With important decisions, remember that, at the intuitive level, we have a soft spot for plausible stories. Therefore, be on the lookout for convenient details and happy endings. Remember: If an additional condition has to be met, no matter how plausible it sounds, it will become less, not more, likely.
* Source: The Art of Thinking Clearly by Rolf Dobelli