My oldest son went to James Madison University. When he was a junior, he traveled to Antwerp, Belgium for a wonderful study abroad program provided by JMU’s College of Business. My wife and I were fortunate enough to visit him for a week. This was in Fall, 2015. The day we were leaving out of Brussels coincided with scary terrorist attacks occurring the night before. This was the time when attacks occurred in Paris, killing 100s of people. We learned of it on the way to the Brussels airport. Of course, Belgium is not far from Paris. We were very concerned about leaving our son in Belgium.
In our son’s study abroad program, about half the families recalled their college students and had them return to the U.S. immediately following the attacks. We faced the same decision. What would you do? As you can imagine, my wife and my first response was to place our son on the first plane out of Europe! But after further reflection, we came to a different, Bayesian-inspired conclusion.
My thinking was as follows:
The prior probability of a terrorist attack was fixed. Meaning, we knew that Europe had a certain chance of an attack. Relatively low but possible. We went into our son’s study abroad program with this known attack probability.
After the attack, security in the area increased SUBSTANTIALLY. This meant the local authorities were out in force, making their presence known. The local authorities were reinforced by the French Armed Forces. The message to people was that it was safe and the police were there to make sure this sort of terrorist attack would not be tolerated. The increased security was meant to both calm the local populace and provide a deterrent to would-be terrorists.
This all had the effect of reducing the ability of terrorists to execute a terrorist attack following the Paris attack.
As a result, I believe our son was significantly safer following the Paris attacks than prior to the attacks. So if we were ok with him going to Europe before, given the probability of a terrorist attack, we should be EVEN MORE ok with him being in Europe following the terrorist attack. By the way, this is the way Bayesian inference works.
I did have conversations with people I knew in the U.S. State Department and checked publically available statistics. My Bayesian reasoning was sound. There was evidence that increased security following terrorist attacks did decrease the likelihood of an attack occurring in the near future.
Now, explaining this to my wife was another challenge. She certainly understood my reasoning, but Mama Bear was not having it. All she knew was that people were dying near her cub. It was the availability of scary information that had a great impact. This is also known as availability bias. People tend to overweigh the most available information (a deadly terrorist attack near our loved one) above otherwise less available “rational” information like the prior probability of an attack combined with the increased security following such a terrorist attack.
Ultimately, we presented our son with our thinking and let him make the call. He decided to stay and even traveled on his own after the program ended. He said he felt safe the entire time. Mama Bear was not happy but she came around.