COVID-19 and behaviour change. A literature review.

Behavioural and social interventions have become an essential component of efforts to mitigate the effects of outbreaks because many interventions rely heavily on community engagement, participation and ownership and on intersectoral coordination and collaboration for prevention, control and mitigation strategies to work. Central to this shift in approach is the commitment to integrated, technically sound strategies that include effective health communication in outbreak control objectives. Experts have come to realize that community understanding of diseases and their spread is complex, context-dependent and culturally mediated.” – Communication for Behavioural Impact (COMBI), A toolkit for behavioural and social communication in outbreak response, WHO

Most of the measures involved in the fight against COVID-19, involve behaviour change. This has brought SBCC and its related disciplines to the forefront of the COVID-19 response.

The six major behaviours

There are six major behaviours that the public in most countries have been asked to adopt during this pandemic. These behaviours are as follows:

·      Staying home

·      Maintaining a social distance of 2m from other people

·      Wearing a mask

·      Washing hands frequently and for 20 seconds each time

·      Disinfecting surfaces

·      Coughing or sneezing into the elbow

Some of these behaviours like hand-washing are building on existing habits, but other behaviours are totally new, like social distancing. In fact, behavioural scientists have previously encouraged people to leave their homes and get active, which is generally the opposite of the requirements of social distancing!

The challenges of behavioural change

To change behaviour and establish new behaviours, we have to understand what the barriers are to people carrying out these behaviours. The COM model from UCL states that in order for people to perform a behaviour they have the Capability, Opportunity and Motivation to perform it. This means that they have to understand, for instance, how to wash their hands correctly; have running water and soap, and understand the importance of the action. Looking at the COM model can help to see where the barriers lie for a particular behaviour in a particular group of people. The Behavioural Insights Team have recently released a tool, which can also be used to deduce the barriers affecting the performance of behaviour in a particular situation.

As an added layer of complexity, humans are believed to have three basic psychological needs and these three needs have been found to remarkably consistent across a variety of countries, economies and settings. These are the needs for autonomy, competence and relatedness. Autonomy is about being able to make choices about your own needs. Competence is about interacting with your environment and seeing that you’re able to make a difference and relatedness is to do with a connection to other people. All individuals, whether or not they classify themselves as introverts or extroverts and ambitious or not, experience these needs.

Part of what makes behavioural change in the current COVID-19 setting so challenging is that the changes required compromise these three basic psychological needs. Firstly, autonomy is threatened through the fact that these behavioural modifications are essentially being imposed from outside sources. Secondly, our competence is threatened because during this pandemic we are being asked to ‘do nothing’ and ‘stay home’, in the hope that nothing much will change and we won’t get sick. This goes against our psychological need to take action and see something accomplished. Finally, social distancing compromises our psychological need for relatedness.

Keeping all this in mind, our interventions need to be designed to increase people’s sense of autonomy, competence and relatedness, whilst also overcoming the barriers to behaviour change for these six key behaviours.

Communication principles

There are several fundamental considerations that need to be borne in mind when planning behavioural interventions:

Levels of communication

Understanding how communication occurs at different levels and the internal dynamics of each level helps to design better communication interventions. There are five main levels:

·      Intrapersonal: These are the beliefs, attitudes, thought processes and values that predict individual behaviour and decision making. The individual or ‘intrapersonal’ level is profoundly influenced by the sociocultural context and ultimately shapes how we view the world. Effective communication at this level resonates with people’s values, beliefs, priorities, and social, cultural and economic circumstances.

·      Interpersonal: Health information is exchanged frequently at the interpersonal level (between people) and is critical in taking a person from the ‘decision making’ phase to the ‘acting on the decision’ phase. The sources of information being heeded at this interpersonal level must be identified. If the source is considered trustworthy, understanding and credible, the receiver will take the advice more seriously and is more likely to carry out the required behaviour. It is therefore vital that we understand whom people trust to deliver their health messages and whom they are currently listening to.

·      Group: Understanding the role of communication in different groups, such as health-care teams, village committees and families in making decisions that affect health-care behaviours will affect how messages are designed, put across and ultimately received. Gender differences and dynamics within groups are also important to consider at this level, as members of a household, community or family will have different perspectives and perhaps different opportunities to communicate their views depending on the society in which they live. The communication of behavioural objectives must also take into account the different burdens, constraints and decision-making power of individual members of a group.

·      Organisational: Understanding how organisations communicate with both internal and external audiences is important, as different sectors and agencies have to work together. A clear social mobilisation strategy can help to identify the roles and responsibilities of different organisations and how to mobilise their staff to support the behavioural goals. Clear, consistent messaging is also vital for allaying fear in times such as these so understanding this organisational level is key.

·      Social: It is important to consider the different ways in which health information is communicated through various channels. This will include use of media (mass and social) and programme interventions such as health education and health promotion.

The COMBI model from WHO addresses all these levels together to respond to the different communication needs of different groups for a specific behavioural result. Having a planning framework can help to ensure that activities are coordinated, which is particularly beneficial at a time like this when actions can seem somewhat chaotic. A systematic approach allows a better understanding of an event, which in turns helps the circulation of accurate, relevant, appropriate information. It also ensures that relationships with communities and agencies are cultivated and maintained by planning opportunities for communication, encouraging dialogue and listening.

The main goal of the COMBI plan is to achieve specified behavioural results through a structured communication strategy. Rapid research is prioritised to understand the current scenario in a particular community or sector and to understand what information is being communicated to them (whether that be accurate or inaccurate). When trying to understand this we need to ask about both first order and second-order beliefs, i.e. we need to ask ‘what do you know about COVID-19 and social distancing?’ and ‘what do you believe people in your community know or think about COVID-19 and social distancing?’ Messaging is put out only once messages are confidently established and have been tested through formative analysis. This ensures that all work done is effective and efficient; avoids wasting time and other valuable resources; and tackles the specific questions, myths and issues in a particular community or group.

Training

Whilst there is a need to respond very quickly to emergencies such as these, training of communicators must be done well. If people are unsure of the message they cannot deliver it correctly and accurately. Un-educated and un-prepared communicators are far more damaging than no communicators at all.

Key psychology concepts

There are several key psychology concepts that come into play in this scenario. An understanding of them is necessary for us to effect behaviour change through messaging and policy design.

The questionable role of emotions

The COVID-19 crisis is not the first time that behavioural scientists have tried to change people’s behaviour for good. The problem in the COVID-19 situation is that behavioural scientists need to work out how to generate long-term behaviour change when compliance is no longer proving exciting.

Many of the methods and techniques employed in this particular scenario have relied on the generation of emotions such as hope, anger or fear to kick-start new behaviours. This works remarkably well in the short term. However, over the long term, something more is needed to sustain these behaviours. Evidence regarding effective means of generating long term behaviour change is harder to come by. However, there are several examples that we can look at. The most successful technique appears to be the use of defaults. Individuals defaulted into two-sided printing for instance, seem to stick with that option long-term. Another successful technique proven effective in the long term is salience. For instance, placing vegetarian food on top of a menu makes it more likely that customers will choose it. The common factor across both of these two techniques is that the decision-maker does not consciously notice the change. It seems that the less conscious the nudges are the less they are likely to wear off or even backfire, regardless of whether the decision-maker ‘agrees’ with the principles behind the nudge or not.

In a recent article by Christina Gravett and colleagues, nudges were classified into two types – ‘pure’ and ‘moral’. Under this framework, defaults and salience are termed ‘pure’ nudges because they are changes in the preexisting choice environment meant to blend into the environment and counteract inattention or laziness. They should require no emotional involvement. ‘Moral’ nudges, on the other hand, are those that are fun or trigger fear, shame or pride. Moral nudges reward you for ‘doing the right thing’ and are meant to be noticed. Social proof scenarios are the best examples of moral nudges. Social proof and other moral nudges are powerful but potentially only in the short term, with long-term behaviour change being extremely rare as a consequence of these. These moral nudges also run the risk of having the opposite effect to that which was intended. For example, individuals asked to donate repeatedly will ultimately opt-out of communication altogether. Therefore, these authors propose that to create lasting behaviour, choice environments must be designed that can convert the initial emotional high and focused attention to long-term habits and norms that will be followed regardless of the presence or absence of emotional triggers. We need ‘pure’ nudges rather than ‘moral’ ones.

Having said this, the authors of this study believe that nudges alone whether moral or pure, will not be sufficient to stimulate the required long term behaviour change. The gap between what people want to do right now (return to normal) and what they need to do is too large. The authors thus purport, that whilst behavioural change is ultimately possible, it will require both pure nudges and also traditional economic incentives and regulations. Ultimately the use of emotion is a great kick-starter but is no longer going to be effective in the long-run.

The concept of shared adversity

We do not cope well when we are afraid. Our instinct when we are afraid is for survival. Therefore, shared adversity can only lead us to care for others and work better together if we are not also individually afraid and stressed. Lessening individual fear is thus vital for getting people to work together and work to help each other during this time.

The term ‘virus’

There have also been interesting insights into the word ‘virus’. It has been seen that many people believe the word ‘virus’ to denote something milder, that is easy to get over. People generally believe that they have recovered from many viruses in their lives, and thus when the virus is referred to as coronavirus it evokes less fear than when it is referred to as COVID-19. We have a mental model for what a virus is, and a high mortality rate as a consequence of a virus is not part of that mental model. We, therefore, need to consider when we use the term coronavirus versus when we use the term COVID-19, depending on our audience and goals.

Small probability events

This is also an area that has received much interest from behavioural scientists during this pandemic. A good example of this is with texting when driving. When you first learn to drive you would never dream of texting as you’ve heard about the probability of people having accidents being higher when texting. However, after a while some ‘necessity’ will dictate that you will text and drive, and if you survive unscathed from the experience, the risk of texting when driving is diminished in your mental model. You believe that your experience shows that the risk is much lower than you previously thought and thus you do it more and more. Each time you do it and don’t have an accident the probability of a negative outcome in your mental model drops further.

This then relates to coronavirus because we go out and go to the supermarket and don’t get sick. This tells us that the risk of going out is lower than we thought originally. The next time we go, the probability of getting sick drops further in our mental model. Ultimately, we end up becoming very casual in our approach to the virus as the probability has dropped so low in our mental model that we consider ourselves ‘above’ risk. It is therefore important that we find ways to keep the danger ‘real’ for people, without eliciting unnecessary fear responses, which would work against the principle of shared adversity as stated above.

Invisible enemy

The virus is an ‘invisible’ enemy is also difficult. We are expected to behave differently in our usual contexts with no ‘enemy’ to directly avoid. This makes it more difficult to change our behaviour. Showing images where the virus is seen under lights on door handles and taps or putting stickers of germs onto door handles, could be an effective way of tackling this and helping to remind people that the ‘enemy’ is there.

Learned helplessness

We have a tremendous amount of resilience, but to have that resilience we have to have an idea of what to expect and when. We can stand adverse events if we can understand the why and the when. The problem with our current scenario is that inconsistency and uncertainty in messaging and communication. Political leaders are naturally having to adapt and change plans as the data comes in and this uncertainty leads to a loss of autonomy (one of the key psychological needs) and in more extreme cases the experience of learned helplessness. Creating some consistency and a feeling of understanding in people is thus crucial for their cooperation.

Identifiable victim effect

In the fight against COVID-19, we have all been recruited to ‘save lives’ and ‘play our part’. We are all, essentially, working very hard and investing lots of time and effort to save these lives. The issue is that this is not consistent with how we have behaved previously. To save the life of a child from malaria costs about $3400, yet as populations, we have generally not been too interested in doing that. We have gone about our normal lives and paid little attention to this global issue. There are lots of ways to save lives by reducing obesity, but we don’t do these. Similarly, we have previously not worried so much about the deaths of elderly patients. We have seen these as inevitable and not been in a rush to invest money in saving them. However, if you look at this time period, we are behaving very differently. If you were to look at just this period, you would say that people care a huge amount about those who are vulnerable and are willing to invest a lot of energy and effort into saving the lives of these people. Yet, even six months ago the picture was very different. Why is it that we seem to care so much now? In psychology, there is a concept called the identifiable victim effect. When we can identify the victim, we are willing to do far more to help. It shows that sometimes our actions are designed to satisfy our emotion and not to make the optimal global solution. With coronavirus, we have seen so much news coverage and so many images of people very sick. We have heard of people close to us dying. We are therefore willing to put in more resources than ever before into this issue, but not necessarily in the most rational way. Keeping victims ‘identifiable’ thus helps to ensure compliance, but this must be done without evoking unnecessary fear.

Disintegration of trust

During times of panic, trust is scarce. Trust takes time to build, so in an emergency situation, such as this, using familiar people to implement policies and to spread messages is very important. It is also worth considering whether uniforms or identifiable clothing or branding can be used to strengthen trust and credibility. Another way to counter this loss of trust can be to find community role models, who demonstrate their compliance and adherence to the measures. Local role models have been found to be far more relevant than celebrities in this context.

Role of religion

For those who are religious, the rules of religion often trump the rules of state. It is therefore difficult to manage rules such as lockdown during religious festivals. In Israel, the Orthodox and Muslim communities had large festivals scheduled during the lockdown period. In Israel it was found that giving people alternative options to practice their religion during this time was deemed acceptable, if the messages came from their own religious leaders. This is worth bearing in mind as the pandemic continues.

General message design

The Behavioural Insights Team had the following advice for creating messages during COVID-19. This is fairly standard messaging information, but they have found that it has been particularly important during COVID-19. They still advocate the use of their EAST model in terms of keeping things Easy, Attractive, Social and Timely, but these additional points were picked out in their recent webinar:

·      Reduce cognitive load – keep messages simple and use plain language. The total amount of effort required for a person’s working memory is called cognitive load. At this time we are all being bombarded with information so it is more important than ever than messages are simple and easy to digest.

·      Put the key information or desired action at the top. Order of presentation of information affects recall. Therefore, even if you are sending an e-mail you need to make sure that most salient point is the first one mentioned or is even in the subject line. In their poster design the Behavioural Insights Team found that the most salient points needed to be at the top of the document in order to maximise recall and positive behavioural intent.

·      Shorten your message and then shorten it again. You need to keep the number of words to the absolute bare minimum to keep the focus on the main message. You do not want to risk distracting people with superfluous language. One way to do this is to provide links, that take you to the additional but non-essential information. Alternatively, you can use a bullet points or tables to break up the text.

·      Keep messages precise. Do not try and achieve too much with one communication!

·      Use images and graphics that reinforce your message or use them instead of messages. The image or graphic must be simple, intuitive and easy to understand. You do not want people to be trying to work out what they are seeing.

·      Encourage social commitment, emphasise the importance of altruism, help people to plan their time and provide the rationale behind requests. All of these techniques have been shown to be important components of messaging strategies in the UK and US during the pandemic.  

·      Test what works to get your point across and iterate accordingly. Results can actually be surprising in this context so testing with focus groups or digital focus groups online is vital! You need to ask three questions to find out if a communication works:

1) What was the main action the communication advised you to do?

2) To what extent do you intend to change your behaviour now you have seen the communication?

3) Do you find this information easy to understand?

Dan Ariely commented further on the communication strategies employed during his recent webinar. He discussed the need for behavioural scientists to not only say what people must not do but also what they must do. This works towards the human need of competence, where people feel the need to take action in order to achieve something. As mentioned above, those celebrating religious festivals during this period have adapted, provided they were given alternative ways of practising their beliefs by leaders of their religion. They needed to be given something to do, not just told what they could not do.

Adaptive messaging

It is important to appreciate that most of the top behavioural scientists acknowledge that the messaging needed is not only going to vary from country to country, and within countries, but also over the course of the pandemic trajectory. The more rapid testing we can do of our messaging to ensure that it is appropriate is thus vital, as is the ability to adapt our messages over time. We cannot establish a set of messages and leave them as they are, but instead constant research needs to be conducted, and further iterations created.

Principles of messaging for certain demographics

There are several general principles of stimuli messaging for certain demographics that can be applied in the context of COVID-19. These are as follows:

·      Externalities – This is where people are encouraged to think about others, not just themselves. This is particularly important during COVID-19 with younger populations who are unlikely to become very unwell themselves. e.g. if you are infected – even if you feel ok – you are likely to cause the infection of 45 other people within 2 weeks, and 2000 people within a month. Of these, 8 people are likely to die. Stay home.

·      Prosociality – This highlights that it is not just about you, but about those you love, so makes it specific to your particular circle of contacts. e.g. Now it’s your turn – protect yourself, your loved ones and your neighbours. Stay home.

·      Social norms – This emphasizes that social distancing is something that everyone else is doing. e.g. The vast majority of Italians of all ages believes that the threat of COVID-19 should be taken extremely seriously. Stay home.

·      Expert source – This relies on the premise that people tend to respect experts more than laypeople. e.g. The Italian Medical Association reminds Italians of all ages to stay home. Stay home.

These messages were tested by Harvard Business School against the standard ‘Stay home’ message. This was tested in Italy at the height of their pandemic and it was found that little benefit was found with any of the messages when all demographic groups were examined together. The reason for this is felt to be that people in Italy were seeing the social proof of the dangers of COVID-19 and thus already adhering to these measures. However, when the groups were looked at in a more granular fashion more interesting finding came to light. Firstly, young people were more likely to disclose the fact that they had symptoms and stay home if messages containing externalities and social norms were included. Similarly, men were also more likely to stay home and disclose their symptoms if externalities and social norm messaging were used. As these two groups were seen to be the least likely to respond to any kind of messaging, these findings are quite significant. They also fit with our understanding that many men are more willing to take action to protect their loved ones than to take action to protect themselves. This could be an interesting concept to test in the African context.

SMS communication

SMS communication has been used effectively in many countries across the world during the COVID-19 pandemic. NHSx, the branch of the NHS aiming to make healthcare digitally friendly, worked on a series of messages which were sent if a person tested positive and was asked to self-quarantine for 7 days. These messages applied various behavioural science principles. For instance, they provided a rationale to explain why people should act in specific ways as this has been shown to increase compliance. They encouraged social commitment from individuals to build on the need for connection during isolation. They helped people to plan and chunk their time due to the research showing that maintaining a structure and schedule is good for managing anxiety during a crisis and finally, they emphasised the altruistic benefits of the actions of individuals, particularly towards the end of isolation periods when people may be feeling better and thus want to socialise again. Also, as the research at that stage suggested that day 6 of the disease marked a turning point in the condition – either towards recovery or towards the need for more care – a message was sent to check on symptoms on that day. Interestingly, NHSx reported receiving multiple replies to their messages, and as many of these replies contained practical questions that were then responded to by staff members, they advise using SMS communications to open dialogue with patients rather than as a one-way communication strategy.

The Behavioural Insights Team praised the medium of SMS for contacting individuals, as they feel that being forced by the medium of SMS to keep messages simple and short made them much more effective. As a recall of information is also shown to be a challenge during periods of stress, such as this one, keeping information concise and relevant is of vital importance.

‘We’ vs ‘you’ in messaging

We need to use ‘we’ as much as possible in our communications to create a feeling of unity. This has been found to be much more effective than ‘you’.

Physical distancing vs social distancing

It is generally considered best to refer to physical distancing as opposed to social distancing, as we want people to remain physically distant but connected socially. This has now been testing in various countries and found to be more popular. However, it is difficult to change once the term ‘social distancing’ has been adopted into local vernacular.

How to present numbers in messaging

Fear is only a short-term motivator at best, and ultimately removes the sense of agency from people and leads them to feel helpless. We, therefore, need to be careful when presenting numbers to the general population. Cumulative totals are fear-inducing, but daily numbers are less so. Cumulative totals can (by definition) only increase, whereas daily numbers can start to improve or help to put things more into perspective, as people can visualise small numbers but not bigger ones. Another option is, for instance, to present the numbers of people dying from coronavirus in a week against the number of people dying in one week normally or in that same week last year. This again can create perspective. It is therefore important that we choose how to present numbers based on the effect we are trying to create, and the audience we are speaking to.

Make value more apparent to increase cooperation

If we are able to make the ‘value’ of a course of action more visible to people they are more likely to cooperate with it. For instance, asking them to reflect on the good that can be gained by testing, makes them more likely to test. Therefore making messaging transparent can be very helpful. For instance, Dan Ariely suggests we make our goals transparent by saying ‘once the infection rate reaches x% then we can reopen schools’ or ‘once the children have been tested for COVID-19 they can come back to school’. This incentivises by giving a clear reason for taking action. Giving a target for people to work towards makes them more likely to comply.

The identifiable victim phenomenon in action

In messaging it is worthwhile asking people to list the people in their community or the people whom they know that are over 60. By reflecting on these specific individuals, and explaining that behaviour changes are necessary to save those ‘identifiable’ people, people are more likely to comply and follow the guidance. Giving people a potentially identifiable victim galvanises support for the message.

Using existing contacts

Harvard Assistant Professor Jon M. Jachimowicz recommends contacting databases of existing contacts in the following ways:

·      Via text message, to ask people to respond with a pledge to engage in social distancing measures. Pledges have been found to be useful in some contexts and asking people to respond tends to make them engage better with the content. However, there is a risk of being perceived as ‘spam’.

·      Via e-mail, calling on people (most likely staff) to talk to five of their friends that may not be as bought into social distancing. In this way communicating with your contacts should reach more people and means that the second tier of people are being contacted by friends who they are more likely to be able to influence.

·      Via a social media campaign or competition that exemplifies ‘good’ normative behaviours.

Using role models

Using relevant and appropriate role models in messaging is vital for message acceptance. This has been shown repeatedly in behavioural science studies. In this particular scenario, encouraging people to make public commitments on social media or wash their hands on social media (in a similar vein to the #IceBucketChallenge) could engage people. Even if these messages are light-hearted if they come from relevant role models they are likely to be taken more seriously.

Creating a policy

The failings of behavioural public policy

Behavioural public policy was no more ready for a pandemic than any other field. As Varun Gauri states, “We had no behavioural playbook on nudges, defaults and other strategies for improving social distancing, mask use, remote learning, home-based work and social transfers”. This is a primary, if common, failing and it demonstrates an unwillingness generally across multiple sectors, to look back historically and to prepare accordingly.

In this COVID-19 scenario, our ability to determine sound public policy was found to be grossly inadequate. The behavioural public relies on assumptions that biases, heuristics and mental models determine behaviour. However, in this crisis scenario, these models are no longer sufficient to describe or predict human behaviour. People are currently responding to the behaviour of other individuals and organisations, and each acting under the influence of each other in this novel context of multiple biases, heuristics and mental models. This chaos is standard in a crisis and is even described as far back as the 1918 flu pandemic, but the behavioural public policy did not look back historically and prepare for this. Understanding the interactions between the various risk perceptions, stereotypes, disposition to cooperate, mental models of disease and trust in government and science is, unfortunately, way beyond the scope of the current behavioural policy set-up.

The other major failing of behavioural public policy at this time is the unwillingness of behavioural scientists, in the main, to trust their intuition. Behavioural public policy relies on randomized control trials (RCTs). Without an RCT most behavioural scientists are inclined to hesitate. The search for evidence is, of course, vital, but during times of crisis, we sometimes need to be able to make recommendations faster based more on experienced intuition than proven research findings. This involves employing a dynamic ‘think-act-check-change’ iterative approach to behavioural science that many behavioural scientists are uncomfortable with. In a crisis such as this, however, if behavioural scientists fail to trust their experienced intuition nothing is ventured, and, as the saying goes, nothing is gained.

The Behaviour Insights Team use a tool called Predictor, which allows for the rapid construction of testing scenarios for communication material. This avoids the need for RCTs and can test scenarios rapidly in a digital format with large numbers of participants. Over time iterations of the process can be run to refine the ultimate end product. However, the software allows a rapid starting point to be found for behavioural messaging and policy design. This software was found to be useful in the context of the UK and USA but is much more difficult to use in lower-income countries.

Policy setting

We need to be aware when setting policies that if policies are unrealistic they will be deemed not to apply for ‘people like us’ and be ignored. For example, asking people to wash their hands regularly when there is no running water is simply impossible, but would lead to the rejection of all other policies as well. All policies need to be tailor made for their particular scenario and presented and delivered to the relevant populations accordingly. This is in line with Dan Ariely’s statement mentioned earlier that states that people need advice as to what to do (as opposed to just what not to do) but builds on this, in that we need to be advising people ‘how’ to do what they need to do, not just telling them ‘what’ they need to do.

Community involvement

During the COVID-19 response, the importance of community acceptance and involvement has been greater than ever before. For example, with the ‘shielding’ approach put forwards by the London School of Health and Tropical Medicine, the importance of the implementation of the measures being community-led is emphasised. They acknowledge the need for “appropriate communication of accurate and consistent information, as well as proactive community engagement and participation in the design and local implementation of the shielding approach”. They recognise that if measures such as these are implemented by the authorities, as opposed to being community-led, they are likely to be perceived as coercive or oppressive.

Key to the involvement of communities is also the use of individuals known to the community in policy implementation. If, for example, healthcare workers have been working in a particular area for some time, albeit for a different purpose, these are the ideal people to implement the policy in this area e.g. testing. These people have established trust in the area and during a time of fear and panic, trust and credibility are highly prized. If healthcare workers have not been based in the area, recruiting local people to implement a particular policy or task is recommended for these reasons.

The use of role model narratives in campaign and policy design is also important. A campaign in rural India was successful at increasing handwashing by creating a story around a ‘supermom’ or SuperAmma – a role model mother raising children who wash their hands.

Building in competition

In the US, Dan Ariely designed an index to create competition between city mayors on certain crucial behaviours and measures required by citizens. This was not popular with the mayors, to begin with, but has been effective in generating change.

Privacy

‘Privacy’ is a big buzz word in society currently, and whilst the right to privacy is obviously vital to protect, there potentially needs to be a more nuanced understanding of the term. Sharing of information during a global crisis is vital if progress is to be made. Data shows that many people are unwilling to test or report symptoms due to ‘privacy’ concerns. People need to be reassured that a) their data will only be used for their own benefit and research purposes, for which it will be anonymised, that b) their willingness to cooperate is beneficial for the ‘greater good’ and that c) the data captured through testing is in fact far less than the data taken by a standard website or mobile phone provider. Policy design however needs to be sympathetic to these privacy concerns and all data handled must be treated confidentially and securely. Making mistakes with people’s data at this time would be catastrophic for national trust.

Nudges

As discussed earlier, ‘pure’ nudges are far more likely to be effective in the long term with COVID-19 than ‘moral’ nudges. An example of a pure nudge is the use of footprints on the floor of supermarkets to remind people to keep apart or putting bottles of hand sanitizer around public places. The potential usefulness of these sort of measures should not be overlooked but then needs to be applied optimally. For example, advice is that footprints are better than other shapes for creating distance in queues, as they create a feeling of shared humanity and the use of the colour blue has been found through research to be good for denoting hygiene. It is recommended that colours such as red are avoided as these can create a sense of danger and panic. Krukow Behavioural Design has created a kit on these nudges and has recruited artists to decorate the footprints for specific locations.

A ‘moral’ nudge that may be effective at this time, could involve using the term ‘proudly South African’. For instance, in Italy they did several campaigns using ‘Join all proud Italians and stay home’ and found this messaging to be effective. This could especially be good and unifying, at a time when mistrust of strangers is particularly high.

Game theory uses

Creating ‘games’ or ‘fun activities’ out of some of the behaviours required could be beneficial. For example, creating tracks of footprints that have been decorated by the children at a school to the sink after using the bathroom could encourage cooperation. Similarly, putting toys into the middle of blocks of soap has been shown to encourage children to wash their hands more.

Future implications

Efficiency vs resilience

Pre-COVID-19 we designed our systems and models around efficiency. The only system that is not truly designed around efficiency in most countries is the military. The military does not look to meet the needs right now but instead plan and prepare for the future. They are adaptable and have to be ready for a variety of eventualities. COVID-19 has shown us that we need greater resilience in our systems and greater adaptability. An efficient hospital, for instance, has beds that are full i.e. limited extra capacity, but in reality, we need to offset this need for efficiency with some degree of resilience by having beds that can be made available in a crisis such as this.

Connectivity

This experience has shown that whilst we may all function as separate nations, when it comes to healthcare and protecting our future against scenarios like this we are ultimately all so interconnected. All countries need to show cooperation and a willingness to assist each other in order to survive in a world where life is so fragile and so at the mercy of nature. If 10% of the world decides not to comply and cooperate in reducing the spread of COVID-19 the effect will be much greater than just the 10% they represent. A small number of people in a small number of countries can really cause huge disruption if we do not work together. The fact that the world is so connected, has sadly not led to greater cooperation but in many instances greater distrust. Healthcare needs to cross these barriers and data sharing needs to be prioritised.

Dangers of excess fear

Ultimately we are going to need to live with coronavirus for some time, even if life returns to some kind of ‘normal’. Generating too much fear will make this transition difficult particularly for those who are vulnerable. We need to be aware of this risk and understanding of those affected. We need to try and allay excess fear with our messaging and encourage safe, small steps back towards normality.

Trust issues

We are likely to be more afraid of people we don’t know in the future. This is to do with the phenomenon of small probability events, as the people, we see every day don’t make us sick so we assume they are safe and others are risky. We need to be aware of this and discuss this in our messaging, to avoid separation and mistrust in society.

Closure

It is vitally important that once the pandemic is under control, this information is communicated well. People need closure and so common in these scenarios we fail to provide it. Instead, we move onto the next headline or set of headlines and fail to acknowledge the ending of the crisis. This may be a way away in the future at this stage, but we need to plan to do this well and efficiently once the situation is suitably controlled and this kind of messaging is deemed appropriate. This can even be done on a smaller scale in individual communities where the worst of the pandemic has passed, for instance. This builds trust for future events.

Interesting suggestions

·      Increasing the opening hours for shops to allow for a greater spread of clients over time.

·      Using the infrastructure of street vendors etc to supply food to those most in need should lockdown conditions need to be imposed again. This stops people congregating in one area for ‘food parcels’. It also builds a better relationship between street vendors and the authorities, as their truly essential role in society, is seen to be recognised.

·      Putting toys into the middle of bars of soap has been shown to increase handwashing rates enormously in children.

·      Placing stickers of cartoon-like germs on door handles has caused an increase in handwashing.

Great resources

The COVID-19 Communication Network has so many resources here for SBCC work:

https://covid19communicationnetwork.org/

Regularly updated document with guidance on messaging:

https://covid19communicationnetwork.org/featured-resource/guidance-document/

Online course from Johns Hopkins on the COVID-19 basics:

https://coronavirus.jhu.edu/covid-19-basics/understanding-covid-19

Behavioural Insights Team have the Barrier Identification Tool.

Predictiv is great online behaviour testing tool.

Literature review by Skye Grove and Elizabeth Harrison

References

2012. Communication For Behavioural Impact (COMBI), A Toolkit For Behavioural And Social Communication In Outbreak Response. 1st ed. [ebook] The World Health Organisation. Available at: <https://www.who.int/ihr/publications/combi_toolkit_outbreaks/en/> [Accessed 5 June 2020].

Ariely, D., 2020. COVID-19. [video] Available at: <https://youtu.be/dIWT71sHM1Q> [Accessed 6 June 2020].

BIT Webinar: Crafting effective communications during the COVID-19 crisis. 2020. [video] Directed by H. Harper. UK: The Behavioural Insights Team.

Bucher, A., 2020. What The Global Coronavirus Pandemic Can Teach Designers About Designing For Behavior Change.

Carlsson, F., Gravert, C., Johansson-Stenman, O. and Kurz, V., 2020. Nudging as an Environmental Policy Instrument. Review of Environmental Economics and Policy,.

Datta, S., 2020. Responding To COVID-19 In The Developing World. [online] Behavioral Scientist. Available at: <https://behavioralscientist.org/responding-to-covid-19-in-the-developing-world/> [Accessed 6 June 2020].

Favas, C., 2020. Guidance For The Prevention Of COVID-19 Infections Among High-Risk Individuals In Urban Settings. [online] Lshtm.ac.uk. Available at: <https://www.lshtm.ac.uk/media/35726> [Accessed 6 June 2020].

Gauri, V., 2020. Behavioral Public Policy Faces A Crisis. [online] Behavioral Scientist. Available at: <https://behavioralscientist.org/behavioral-public-policy-faces-a-crisis/> [Accessed 6 June 2020].

Global Behavioral Economics Network, 2020. Dan Ariely And Ernst Fehr On COVID-19. [video] Available at: <https://youtu.be/74Z6lCW4ZMA> [Accessed 6 June 2020].

Gravert, C., 2020. Why Triggering Emotions Won't Lead To Lasting Behavior Change. [online] Behavioral Scientist. Available at: <https://behavioralscientist.org/why-triggering-emotions-wont-lead-to-lasting-behavior-change/> [Accessed 6 June 2020].

Harris, T., 2020. From Inform to Persuade: How Can Tech Step Up for Humanity?. [Blog] Medium by Center for Humane Technology, Available at: <https://medium.com/center-for-humane-technology/from-inform-to-persuade-how-can-tech-step-up-for-humanity-3ac21de4c53b> [Accessed 30 May 2020].

Krukow.net. 2020. COVID-19 Nudges. [online] Available at: <https://www.krukow.net/covid-19nudges> [Accessed 1 June 2020].

Mad*Pow and Harvard Business School, 2020. Detailing Interventions To Encourage Social Distancing. [image] Available at: <https://youtu.be/IPMqBExjryg> [Accessed 3 June 2020].

Miller, K., 2020. Let’s Aim for Physical Rather Than Social Distancing. [Blog] Nautilus, Available at: <https://nautil.us/blog/lets-aim-for-physical-rather-than-social-distancing> [Accessed 1 June 2020].