The Normalcy Bias in Evacuation Behavior

The issue is no longer about understanding public behavior. The challenge now is modifying public behavior.

One of the primary safety tasks in sailing is reefing the mainsail – lowering it partway in a heavy wind to make the boat easier to manage.  There is a classic sailor’s saying about reefing the mainsail: “if you are asking if it’s time to do it, it’s already past that time”.

I learned this lesson the hard way – on a short sail close to the California coast from Oceanside to Dana Point. The winds were brisk and steady. Fortunately, my husband, who is a much more experienced sailor, noticed the weather was changing and discussed reefing.  Then he quoted that aphorism, reefed the sails and when the wind gusted up we were fine.  But, this is a mistake lots of experienced sailors make – not denying the wind will come up, but denying it will come up so quickly that they can’t manage it.

This is part of “normalcy bias”; a mental state of denial when facing a disaster or pending danger.  Normalcy bias leads people to underestimate and minimize the possibility of a disaster actually happening. Until it does – and then they frantically try to reef their sails in heaving seas.  

We see normalcy bias a lot in emergency management.  Most recently, some of those folks in Oklahoma City, who finally decided to react to the impending danger (the disastrous tornadoes) and – against the strong requests from state and local officials to shelter-in-place and stay off the roads –spontaneously evacuated and gridlocked the freeways, which put a whole lot of people in danger.

I was reading some of the preliminary reports (lessons learned are still being sorted out), and thought about the blog I wrote earlier this year about evacuation behavior and evacuation zones around nuclear power plants.  It was about a report in which the NRC suggested large evacuations during an incident put the closest public at increased risk because of traffic conditions and delays as the public furthest from the risk evacuated – whether they were asked to or not. In Oklahoma City, the public was specifically asked NOT to evacuate, tried anyway, and put a lot of people directly in the tornado’s path.

Normalcy bias in evacuation behavior: denying or overreacting to a pending disaster, then acting inappropriately, thereby putting one’s self and others in danger.

This is frustrating because it is such an intractable problem. We can’t convince the public to be prepared; we can’t convince the public to follow directions when they aren’t prepared.

In an interview with NPR, Rick Smith the warning-coordination meteorologist for the NWS in Norman, OK, talked about warnings, information, and human behavior.  Smith allowed as this was a complex issue and tackling it would take a group effort by “people who understand — or attempt to understand — human behavior and messaging and all kinds of things.”

Good news! Some people have already done that! There is a lot of social science research, going back to 1990 (Mileti and Sorenson: Communication of emergency public warnings) and 1992 (Lindell and Perry: Behavioral Foundations of Community Emergency Planning). You can find these and more here.

The issue is no longer about understanding public behavior.

The challenge now is modifying public behavior.

What we need is a good, sticky message that targets normalcy bias.  Any ideas out there?