Applying Behavioural Insights to Emergency Decision-Making

3 July 2018

Categoriesbehaviour change, masterclass

Tagsbehavioural insights, case study, masterclass

Behavioural insights can help address emergency decision-making to support NSW agencies and policy-makers reduce fatalities in the community. We look at the issue of driving through floodwater.

Applying BI to Emergency Decision Making SES

Driving through floodwater

Flood-related fatalities have been an ongoing problem in Australia since the early 1900s. Deaths during floods rank second only to heat waves in natural disaster fatalities. Approximately 159 people died from flooding in Australia during the last 15 years, with half (53%) due to driving through floodwater. NSW, together with QLD, represent 74% of flood fatalities. Rural and regional areas in NSW are especially at-risk. 

Driving through flood water is a pressing issue, and has been a persistent problem behaviour that has been tough to shift. 

The following reflects how we worked through this behavioural issue, and brainstormed problems, during a recent masterclass we ran with NSW State Emergency Service (NSW SES). 

Defining the issue

Who is driving through floodwater? The evidence tells us that both young people (aged under 29 years) and those aged over 60 years are overrepresented to drive through floodwaters. Additionally, drivers in 4WDs and SUVs are especially likely to drive through floodwater (48%). 

Where is this happening? The available evidence tells us that the biggest flood fatalities occur when people are en route to a destination where they have a time-sensitive commitment (such as getting to work on time) or when they are driving home. 

When does this happen? We know that 58% of fatalities happen close to home (within 20 kilometres) and a similar proportion of fatalities are locals (58%). 

What is the problem behaviour that lends itself to a BI intervention? There can be many answers to this question, and part of BIU’s role was to work with NSW SES masterclass participants to narrow down some options. Is it simply to stop people driving through floodwater? Should people turn around when faced with floodwater? Or should the public be encouraged to stay in place and find shelter? 

To better understand the options, we looked at the barriers and triggers that lead people to drive through floodwater. 

Identify behavioural barriers or triggers

There are behavioural biases acting as barriers to the desired behaviour of avoiding the flood water. Drivers who venture through flood water seem to overestimate their ability to survive and underestimate the risk, displaying an optimism bias. Research by Julia Becker and colleagues suggests that this is even more the case if drivers are very familiar with the environment. This same familiarity with the environment and a particular road could also mean that the status quo bias is at work. Deviating from the usual route taken regularly, such as their route home, would require more effort and planning than sticking with the tried and tested.

Drivers may also consider themselves atypical, using the representativeness heuristic. They might think: “only old ladies or inexperienced youngsters get stuck”. Yet research shows the biggest flood fatalities in Australia are men (up to 80%). One of the most devastating floods in NSW in recent times occurred at the Shoalhaven, just over half (53%) of the people who drove into flood water were men, 38% were women, while the rest were unknown (9%). 

Based on their expertise, NSW SES discussed how the emotional decision-making was an especially strong pull. It’s not just about driving to a time-pressed situation or the desire to get home, but rather why individuals might feel compelled to continue to reach their destination. If people have children waiting for them, or cattle that need tending, or some other priority they feel strongly depends on them, they will drive through floodwater to fulfil their commitment. 

In Figure 1, we modified the work of Irving Janis and Leon Mann to look at critical decisions faced in emergencies.


Figure 1: Decision-making in an Emergency

From this decision pathway, the following reflections by the driver could be possible:

  • The first question a driver may ask could be “Is there a serious risk if I do nothing?” To this the answer could be “Yes my kids won’t be fetched from school”.
  • This would lead to the next consideration of whether there is a serious risk if action is taken. “Yes I could get stuck”.
  • This would bring on the question on whether more information would help: “yes it would help to know my kids are safe and it would help to know if it’s safe to cross”.
  • To answer these questions would require time and the driver may decide that there is insufficient time to research and deliberate, and so make a bad decision, venturing into the floodwater. 

Each of these decision points offers potential for a behavioural insights solution.

Design potential solutions using EAST

In BIU we often apply the Easy, Attractive, Social and Timely (EAST) framework and design interventions that could enable the desired behaviour. These were some of the solutions that NSW SES generated using this framework: 

  • Easy:
    • remove frictions to the desired behaviour, or
    • create more suitable defaults like practising a new route home.
  • Attractive:
    • offer rewards, or
    • personalising messages and making messages more relevant or novel. For example a message to at-risk members of the NSW SES could be framed more altruistically “You can’t help anyone if you’re stuck”.
  • Social
    • Applying social norms to break the representativeness heuristic, or
    • focusing on the social networks of drivers to remove the trigger that was causing the unwanted behaviour. For example, schools to develop plans with parents in case of floods, such as looking after children beyond normal school hours even keeping them overnight if required. This would remove the need to urgently collect children allowing drivers to make a wiser decision.
  • Timely:
    • text message to prompt drivers when they would be most receptive: “Today may be the day you need your Plan B”.

Prioritise the intervention

Of course, in the real world we are unable to take every good idea forward. There can be real world constraints such as timing, resources, and communication channels. Prioritisation based on ease of implementation versus payoff can be a valuable exercise to shortlist what ideas actually become implemented as solutions. 

How can you use this in your work?

This approach to finding a relevant solution can be used in other contexts. The steps you can take are:

  • Name the problem behaviour: be as specific as possible
  • Identify barriers to the desired behaviour, or alternatively, the triggers for undesirable behaviour: what are the risks that people might consider in their decision-making during an emergency?
  • Address the behavioural biases at play: what are the cognitive processes influencing people’s actions?
  • Designing interventions using EAST: how can we make it easier for people to do the right thing? How can we make the desired behaviour more attractive? Can we tap into social networks or norms to improve outcomes? How can we make the intervention timelier?
  • Prioritising interventions: in this case, we considered the potential impact versus the ease of the implementation. Other considerations may vary for other problems. What factors does your organisation need to weigh up when considering an intervention or campaign to influence behaviour?

If you would like to see other applications of BI to other policy issues or decision-making processes, check out our past projects or other resources.