This article is written by Kulani Abendroth-Dias, behavioural scientist and PhD candidate at the Graduate Institute of International and Development Studies Geneva

A few days ago, I was participating in a virtual meeting on the use of behavioural insights to combat the Covid-19 pandemic. The conversation was taking place between behavioural scientists, public policy officials, and citizens with no prior background in behavioural science.

After a round of introductions and a discussion of the salience of behavioural science for combating the Covid-19 pandemic through the management of targeted behaviours such as handwashing, physical distancing, and wearing masks, the moderator asked, “So what can you say about how behavioural science can be used to win the war against the coronavirus?” The conversation quickly devolved into a discussion of the validity of various nudges that have been tested, from messaging on masks to the design of written communications.

This reminded me of my first day as a behavioural scientist at an international institution, when my boss said to me during our first team meeting, “OK, give us the behavioural science spiel.” I responded with a quick presentation on the different biases and limitations that have been found to guide human behaviour, from the halo effect to loss version, to cognitive capacity and bandwidth tax, and a quick selection of associated nudges and BI interventions across developed and developing contexts. At the end of my presentation, my boss looked at my team and said, “Great job! Let’s talk about these nudges. How can we implement them in our context?”'

It was evident that the first part of my presentation, on the biases that guide all human behaviour had been lost against the setting of the flashier nudges. My team’s understanding of my job was simple: design nudges for social problems within our innovation unit. Present, test, tweak, implement, repeat. This is in part due to the exciting innovation quick-in business and workshop models that have been proposed by burgeoning behavioural science practices worldwide. They often gloss over the years of interdisciplinary theoretical and applied research, psychological and economic principles, and rigour in analysis that have gone into the design of our first, most famous nudges.

The questions I kept getting were about what kinds of nudges have been used and how we can adapt these to target better waste disposal/financial management/illness vaccination policies. The question should be, what types of cognitive and environmental biases hinder proper waste disposal behaviour within this specific sector of our population or this geographical location.

In the same way, when it comes to the coronavirus, we need to avoid solely basing our discussions on the types of nudges that can be adapted to increase physical distancing and handwashing behaviours. Our starting point should be the cognitive and environmental biases that may hinder physical distancing and handwashing behaviours. Nudges are important, but they’re not the full story for policy practice.

Viral behaviours

The Covid-19 pandemic is a fascinating case study of the importance of understanding the biases that guide human behaviour in the development of policy. Here we have a publicly salient crisis where managing certain very specific behaviours is crucial to actually save lives. Physical distancing, handwashing, wearing specific types of masks, avoiding hoarding, and sharing misinformation are key factors that can contribute to the spread or containment of the virus.

How well we manage to accomplish these behaviours can have ramifications on the strain on our healthcare workers and systems, how soon we can return to economic productivity, our financial markets, managing our cybersecurity, and at the end of the day, how many lives we can save. People who don't physically distance can spread the virus, prolong containment, and further derail economic activity. For the first time in a long time, the tissue that connects all our systems across silos is becoming salient. Shifts in specific behaviours have repercussions across sectors. Changes in specific policies have implications across borders.

Behavioural science is not a discipline to be housed solely within your innovation department. It’s a set of principles and tools to be mainstreamed across your organisation

Given the nature of the pandemic which demands behavioural interventions at the macro level and across sectors, there have been behaviourally-informed interventions, mostly in developed countries, to bolster empathy, trust, science communication, and healthy mindsets and target the spread of conspiracy theories, discrimination, zero-sum thinking, hoarding of essentials, and gender-based violence. Given the evolving nature of the virus and the trajectory and ramifications of its spread globally, there has been a need to ensure that research happens at a fast pace, is quickly updated and adapted, and measured often. The power of social referents i.e. people with a higher capacity for socialisation from whom others derive their behavioural cues, is being utilised to influence more handwashing behaviour.

Framing effects in the content of messaging to encourage physical distancing are being tested across audiences and geographical locations. These are exciting developments, which if measured long-term across samples, can have implications for communication and policy design post-Covid-19. However, a core lesson from the pandemic is already abundantly clear: we don’t have an effective default playbook to face a global health crisis, let alone potential other crises such as the escalating effects of climate change. This needs to change.

What does mainstreaming Behavioural Insights across organisational functions look like?

The coronavirus pandemic illustrates the danger of working in silos. The OECD Secretary General Angel Gurria has called for “multidisciplinary expertise” in “Joint actions [to] win the war” against the coronavirus. Recommendations from AI, healthcare, economic, and educational experts (to name a few) have been the same: we need more knowledge sharing and open-sourced materials to develop faster, more efficient solutions.

Bill Gates famously committed to funding seven coronavirus vaccine candidates to ultimately pick the best two, although “billions of dollars spent on manufacturing would be abandoned.” The idea is from economics 101: simultaneously testing and building manufacturing (R&D)  capacity can more efficiently bolster product development. Developments in AI are being leveraged for more efficient contact tracing and medical imaging in the healthcare sector. In the same way, developments in behavioural science research are being used to understand how emphasising source accuracy, benefits to the recipient, and social consensus and approval (others in your group are all doing it, don’t you want to?), as well as changes to the content of a message (emotive versus factual, order of information) are influencing behaviour.

These types of findings with regard to the design and content of messages are relevant to any department in any organisation that sends out emails/contracts/paperwork that require employee responses. Much of my time as a behavioural science practitioner has been spent explaining to folks that often revamping their messaging boards or templates (as boring as that may sound to some) can bring about significant shifts in enrollment and behaviour. In a typical day, individuals often have to sift through informational overload to solve dozens of problems. This depletes their cognitive capacity from one task to the next.

Put simply, nobody wants to expend time going through all the options available to them when signing up for a healthcare program or phone contract. Puzzlingly, not all phone and healthcare providers have this in mind in the design and marketing of their products. Simply making the options more salient to a consumer without them having to sift through all the alternatives has been found to be effective in bolstering enrollment, savings, and increasing competition in the healthcare and phone contract markets. Nobody wants to go through paperwork. So why do we make everyone do it in our workplace?

BI is not as glamorous as you might think

Many public officials have probably never thought of behavioural science in the context of boring old paperwork. However, sifting through paperwork has led to the design of some of the most interesting behavioural interventions to date, from nudges in the design of content to choice architecture in setting the default to remove a certain amount of paperwork altogether.

All organisational departments have paperwork that behavioural scientists can streamline for efficiency and better enrollment (where required). Behavioural science is not a discipline to be housed solely within your innovation department. It’s a set of principles and tools to be mainstreamed across your organisation to bolster the efficiency of low investment/high impact interventions. Insights from behavioural research tell us that we have limited mental bandwidth, which can adversely influence our hiring practices. We tend to remember the first and final candidates during recruitment more so than those interviewed in between. We tend to imagine those with specific facial features in leadership positions. How do these biases influence our recruitment strategies and gender balance across our company? An awareness of our biases (sometimes in something as simple as when to schedule a meeting or interview) can have long-term implications across our organisations.

Behavioural science teaches us to listen to the data and not our intuitions. It also illustrates the importance of segmentation: one-size-fits-all policies don’t work

Instead of behavioural scientists from the innovation department “honing in” on your processes (a very human perception at the heart of many officials in innovating their processes), behavioural scientists from your own department should take ownership of streamlining your processes and paperwork. A post-Covid-19 “new normal” or default should use learnings from behavioural science to streamline scheduling and paperwork that would cut it down to the essentials: what content is required, how it should be timed and designed to maximise the desired outcome behaviour, and how it should be digitised, protected, and communicated in the face of a crisis. Applying these principles to paperwork across departments, from tax compliance to recruitment in HR can help us become much more agile in the face of the next global crisis.

If you think there won’t be another global crisis of this magnitude, remember that most of us didn’t see the Covid-19 pandemic coming either. This is known as the availability bias — we look to accessible examples to predict the future and thereby often fail to adequately calculate the possibility or risk of an event. This is precisely why we need a new default.

Data, data, data: One-size-fits-all policies don’t work

Contextualisation and segmentation are at the heart of behavioural science practice. This is why behavioural principles need to come in at the beginning of policy design, instead of at the intervention implementation.

As Daniel Kahneman, one of the principal architects behind behavioural science as we know it today, noted, “Policy is ultimately about people, what they want and what is best for them. Every policy question involves assumptions about human nature, in particular about the choices that people may make and the consequences of their choices for themselves and for society.” As policy officials, we may think that we know which policies would best bring about the most effective change. A look at the data often tells us otherwise.

I remember working to design a policy intervention to reduce the incidence of motor accidents in a developing country. Qualitative interviews with police officials revealed that they had focused on interventions targeting drunk driving and speeding for the past few decades. They assumed these were the most prevalent problems, possibly because they were the most reported and talked about, in hospitals, the media, and at the dinner table. They also often occurred during holidays and festivals, making them a more salient public issue.

However, a thorough analysis of the statistics on accidents over the past ten years (thanks to excellent data collection!) showed that the poor street lighting at night had led to a significantly high number of crashes that had gone unchecked throughout the year. They made up a higher proportion of crashes than those that occurred due to drunk driving and speeding combined but had been overlooked. The data had been collected but because the officials were intuitively convinced that drunk driving and speeding were more pertinent problems, they had only selectively used their data in the design of their policies and interventions.

The Covid-19 pandemic is a case study in how quickly our organisations are able to adapt to conditions brought on by a global crisis

Behavioural science teaches us to listen to the data and not our intuitions. It also illustrates the importance of segmentation: one-size-fits-all policies don’t work. When trying to reduce the incidence of traffic-related deaths, we may need to consider policies and interventions that target driving under the influence, speeding, and poor lighting at night. We may also need to consider the nature of the injuries — counting fatalities is important, but what about those classified as grievous injuries that may be overlooked? Along the same vein, interventions designed to reduce the spread of Covid-19 related misinformation need to segment their audience, across age-groups and geographical locations at the least.

Are counter-narratives being delivered online and offline? Do counter-narratives work? Is the content of messages emphasising statistics and narratives that represent the segmented audience? A 60 year old factory worker probably won’t pay attention to a message that says, “70% of Spring Breakers chose to stay home to fight the spread of Covid-19 this year.” In the same way that interventions need to be adapted at the granular level, policies should listen to the data in taking into account audiences within their target populations, across age, socio-economic status, political affiliation, and gender among others. This data should be digitised, protected, purpose-limited, and open-source where possible to encourage knowledge-sharing and the agile design of policies.

If you don’t measure policy, you won’t see change

The Covid-19 pandemic is a case study in how quickly our organisations are able to adapt to conditions brought on by a global crisis. How quickly were your company’s tax compliance/recruitment systems and business model(s) able to shift to the “new normal?” Were there freezes within certain departments because the current processes could simply not be adapted? What does the push from certain sectors to go back to “normal” tell us about the effectiveness of our current systems?

Can we use learnings from the Covid-19 response to target “less salient” problems such as climate action?

We should be measuring and documenting the limitations of our current systems to innovate and build resilience for the future. If we don’t, we won’t see any change when the next crisis hits. This is a clear illustration of how we should also be measuring our policies to measure impact and change. Many of us refrain from measuring long-term behavioural change because it can be expensive, cumbersome, and seemingly counterproductive to our goals (what if our intervention didn’t produce any long term change?) This is exactly why we should be measuring our policies — if we don’t record instances of when our policies didn’t achieve change six months or one year post-implementation, it gives us more data and learnings to work with to design the next policy. If not, aren’t we just making the same mistakes over and over again?

What does a post-Covid-19 world look like to you?

The blue skies resulting from the reduced emissions of curtailed economic activity have instigated discussions on a post-Covid-19 world: the dangers of our current systems and what we want our future to look like. Covid-19 is a particularly salient behavioural problem that has instigated policy-makers across the globe to react quickly. The effects of climate change, on the other hand, are gradual and less salient (so far), but its effects no less pertinent.

People’s lives are at risk as a result of our current industrial activities and meat consumption practices. The effects of climate change could result in the next global crisis. Can we use learnings from the Covid-19 response to target “less salient” problems such as climate action? Can we develop salient examples from the effects of the Covid-19 pandemic on our current modalities to demonstrate the risks of climate change more vividly? Will we adapt our policies and working models to become more agile to threats posed by the risks of future health or climate-related national or global crises?

We can start by mainstreaming behavioural insights to obtain a better understanding of the biases that govern the behaviours of our people across organisational functions to build more human-centric products and policies. We can start by embedding behavioural scientists throughout our organisation to streamline our paperwork: small tweaks to our systems much more agile in the face of the next global challenge. — Kulani Abendroth-Dias

(Picture credit: Unsplash)