Those time periods don’t link themselves to creating a vivid scenario about terrorism, at least not to the degree of the flights themselves, so you don’t bring risk events to mind and your subjective probability seems low.

Using probabilities in decision-making

In a two-person political race, like many of US presidential elections, we may not know the true probability for candidate A or for candidate B, but we do know that only one of them will be elected and we know that no one else will be elected. But, many other sorts of events aren’t well defined, sometimes it’s because of imprecision in their stated probabilities and often a frustrating example is weather forecast.

What exactly is meant by a 50% chance of rain tomorrow? Does that mean a 50% chance for rain somewhere and sometime in your local area a 50% chance at every location ended every time in your local area. Or, most importantly, should you or should you not buy a ticket for tomorrow’s baseball game of the outdoor stadium?

The single probability of 50% chance of rain could describe any of a very large set of potential events, only some of which might dampen your experience of the baseball game. But, for now let’s consider decision situations in which the probabilities of different potential outcomes can be known or can be estimated with precision. How then should you use those probabilities in your decisions?

Probability weights

Behavioral economists refer to the way in which objective probabilities are transformed in the subjective influences on a decision as probability weights. This term may seem like jargon, but it actually maps on well to our intuitions. When we say that some factor weighs heavily on her decision, we mean that an objective difference that factor has outsize influence compared to other factors.

Probability weighting just means the same thing: probabilities that are overweighted are treated as more important than they should be and probabilities that are under weighted are treated as less important than they should be. A rational decision maker should assign probability weights to different events in a manner that is both consistent and complete.

Just as 50% is twice as likely as 25% objectively, consistency requires a 50% seen twice as likely as 25% subjectively so that twice as much weight is given 50% in a decision. Completeness requires that the total probabilities for all possible events add up to 100% not more and not less. These may seem so obvious that they aren’t even worth discussing.

The availability bias

Let me turn the question around and ask it to you in a different way how would you form a probability judgment for something like term life insurance during a flight to Thailand? One thing you cannot do is rely on your personal experience. I’ll say that you’ve never died while on a flight to Thailand, instead you have to estimate the probability of some rare event like that by constructing scenarios in your mind about what might happen.

If you bring to mind a vivid scenario about an event it will seem more real and thus more probable. This phenomenon has been given several names but it is typically called the availability bias. In the flight insurance example earlier thinking about each flight independently forces people think not only about what might happen on that flight, but also about possible differences between the two flights.

Bringing past events to mind they might think that flights to the United States are more dangerous than flights away from the United States and those vivid scenarios serve to increase the subjective probability of a terrorist event, but when considering insurance for your entire trip you’ll think about spending time in your hotel meeting colleagues and friends touring local sites and eating at restaurants.

Those time periods don’t link themselves to creating a vivid scenario about terrorism, at least not to the degree of the flights themselves, so you don’t bring risk events to mind and your subjective probability seems low.