Behavioural Sciences During the Lockdown

We’ve heard and read a lot in recent months about behavioural sciences, and how they are being used to inform the response to the Covid-19 lockdown. What’s less clear is how behavioural sciences are incorporated into public communication and service operations, and on what basis? Does behavioural science stand up to scrutiny? Is there enough light being shed on the assumptions that inform behavioural science approaches? Is there enough discussion about the processes and concepts that are associated with behavioural science? Is behavioural science really up to the job of helping people to stay safe and respond pro-socially to the pandemic?

According to Matteo M Galizzi, an Assistant Professor of Behavioural Science at the Department of Social Policy at the London School of Economics, behavioural science is a “cross-disciplinary, open-minded science of understanding how people behave.” According to Galizzi, the advantage of the behavioural science approach is that it “cross-fertilises and brings closer together insights and methods from a variety of fields and disciplines, from experimental and behavioural economics to social and cognitive psychology, from judgement and decision-making, to marketing and consumer behaviour, from health and biology to neuroscience, from philosophy to happiness and wellbeing research.” Behavioural science clearly has ambitions to explain a lot about the human world.

When combined with a systematic investigative approach, such as the use of randomised controlled experiments in the lab and the field, behavioural sciences are said to enable a greater understanding of how people behave and respond to specific stimuli in different social situations. There are two broad ways that this insight is gained. First, people and the patterns of behaviour that they demonstrate in daily life are considered in relation to the way that we are said to be information processing agents, each faced with different stimuli in differing social situations. Secondly, people’s behaviours are observed and the relationships and social connections that they enact are modelled. What behavioural sciences tries to understand is on what basis, and in what way are we able to establish and maintain those relationships.

Behavioural sciences has been turbo-charged in recent years with the widespread uses of digital data tracking and large-scale information modelling techniques. The aim of data modelling is to map and observe how our behaviours are linked, how they correspond with our cognitive preferences, and how they fit with the social setting in which people are circulating. This has given rise to the datafication of our lives, in which the pervasive tech giants are warehousing enormous quantities of information, analysing it with artificial intelligence, and predicting our lifestyle, political and consumer choices (Fenton et al., 2020). Cognitive biases are viewed negatively, as a hangover from our evolutionary adaptation, and unsuited to life in the postmodern age of instant global communications.

Underpinning behavioural science, then, is the expectation that we make rational judgements and calculations, as if we are a psycho-technical processing unit, which lead us into certain predictable patterns of social action. These patterns of action are said to be based on our cognitive capacity for individual and social functionality. Changing these patterns of behaviour is said to simply require a ‘nudge’ in order to assist whole populations make either pro-social or anti-social adjustments (Thaler & Sunstein, 2009). To put it simply, if we change the inputs that people are given, it is assumed by behaviourists that we get different outputs in the form of behaviours of different kinds.

This model of human nature, moreover, is also characterised by a focus on animalistic survival and reproduction mechanisms. Behaviourism is informed by the suggestion that, as with all other animals, humans are subject to stimulus and response processes. These deep-rooted response functions, it is said, are hard wired into our DNA, and have evolved over millennia. We can’t easily shake them off, as they are part of the human cognitive systems. These basic behaviours are part of our inherent wiring and serve a bio-evolutionary function. They were formed and established through long periods of evolutionary pre-history when, as a primate species, we lived closer to nature, preceding the formation of our group associations and civilisations.

Behaviourism, therefore, assumes that we are driven by a social logic which, in an ideal world and all things being equal, is subservient to species replication and sustainability. In this model of human nature we are programmed by the need to replicate as a species, and this therefore prioritises autonomous self-interested social functions first. This is similar to the way that Richard Dawkins argued that memes, i.e. units of social information, are analogous to genetic distribution in the population. The behaviourist approach assumes that we spread ideas and information, and therefore behaviours, in dependable and uniformly coherently ways. As a consequence, our behaviours can be modelled as an expression of the quality of the information that we have been socially programmed with. If we have faulty information, then we will likely witness resulting faulty behaviours.

The complementary view of behaviourism, in addition to information processing models, focusses on relationships and the ability that people have secure of bonds within their social frameworks. As with the memetic model of social behaviour, there is a focus on the processes of replication and reproduction, not only biologically, but socially and culturally. If the first model is about information processing, the second model, or flipside of the coin, is about relationship forming and social network management. This component of behavioural sciences seeks to understand the psychology of relationships, kinship patters and social networks, as they are manifested in relationships, social interaction, communication networks, collective associations and “relational strategies or dynamics between organisms or cognitive entities in a social system.”

This subdivision of the behavioural sciences is informed by social psychology, social network analysis, dynamic network analysis, agent-based model and microsimulation. It is tied with the use and development of information tracking and communications engagement using associative models. The great insight of companies like Facebook, for example, is not simply the amount of information that they have access to, which is considerable, but their ability to cross-reference this information with different or alternative subsets of information. This ability cross-reference data offers a potentially lucrative ability to predictive changes in social attitudes and social behaviours. Think Trump and Brexit and the controversies around the use of big data to align political campaigns through social media and network modelling.

According to Klaus Schneewind there is much debate and disagreement in behavioural and social sciences about the motivation that people are informed by when making choices and acting in particular ways. The question, according to Schneewind, is are we driven by goals or are we driven by values? If we base our assumptions about people’s motivation on one principle, for example goals, then we may arrive at a very different set of outcomes and views than we would see if we assume that people are driven by values. The process by which we normalise and govern the socialisation process, if it is goal driven, will be very different from that which is needed if our socialisation process is values driven. Likewise, the theories that apply for understanding human behaviour in terms of goals, is very different from what is needed to understand human behaviour in terms of values. As Schneedwind argues,

“Although goals and values are undoubtedly objects of socialisation theories and research, this controversy remains unresolved. However, even if the originators of socialization theories deliberately restrict themselves to descriptive and explanatory statements, such theories can nevertheless contribute to a great extent to the clarification of socialisation goals, and to a better understanding of their antecedents and consequences.” K.A. Schneewind, in International Encyclopaedia of the Social & Behavioural Sciences, 2001.

In other words, while theories of socialisation my help us to anticipate practical information that can be deployed in real-world situations, we will be left with a degree of doubt and incongruity that what we might consider, on an individual basis, might not be applicable on a universal basis. What is resolvable for individuals might not be applicable to social groups. There is always a need to find the exception that proves the rule.

What behavioural science attempts to do, then, is to look for cases and examples of human behaviour that can be replicable because they are stable (L.J. Harris, in Consciousness and Cognition, 2007). Human behaviour is expected to demonstrate transferable characteristics that operate across different domains and disciplines. The same or similar sets of cognitive processes are expected to be in operation regardless of whether we are thinking about “economics, political science, psychology, and sociology” and so on. While each will have their own specialised “training, expertise, reward structures, associations, networks, and norms,” they will share a view that human nature shares a set of common cognitive processes and biases (F.J. Levine, in International Encyclopaedia of the Social & Behavioural Sciences, 2001). The task of the behavioural science researcher is to collect and identify the variables, and to seek out the entry points and places where different ideas and behaviours become relevant, and as Kelso suggests, to “prune away” those areas that are irrelevant while trying to understand those elements that are essential.

As Kelso goes on to describes “in the social and behavioural sciences the key collective variables characterizing the system’s state space are seldom known a priori and have to be identified.” It is assumed that any self-organising system has a specific set of processes and dynamics for managing and mediating change, and the role of the behavioural scientist is to identify what those regulating processes might be. As Kelso goes on to suggest, this involves thinking about qualitative changes as well as quantitative changes, fuelled by the need to understand and draw distinctions between “one pattern and another.”  The behavioural scientist’s task, therefore, is to “identify the key pattern variables that define behavioural states.” (J.A.S. Kelso, in International Encyclopaedia of the Social & Behavioural Sciences, 2001).

This takes behavioural sciences into a wider domain of social enquiry, and starts to encompass “anthropology, sociology, and psychology,” while also broadening the scope of application of the discipline to inform predictive behavioural actions, such as social media engagement, consumer purchases pricing, consumer goods recommendations, and many other spaces of human engagement that have a need to focus and direct social responses. The promise is that behavioural science is a mechanism for “predicting the future,” based on an “understanding how people have behaved in the past.” According to this model, it is likely that we might get a sense of “how people will behave in the future.” The focus of attention following this, is then on the process of converting people’s needs and social predispositions into purchases, votes or actions. Wash your hands more often, wear a mask, and so on.

Systems designers, developers of user interfaces, public policy developers, educators, public service providers, to name only a few, are intensely interested in learning about social behaviour and how they might develop effective systems for engaging with people. However, rather than participating in protracted modification processes and service analysis, supported by prolonged and longitudinal engagement with clients and service users, behavioural sciences are increasingly adopted for their alleged potential to short-cut and speed-up the development process. With a better understanding of people, people’s biases, their limitations and expectation, the development of platforms for social engagement are being introduced based on narrower and more limited testing cycles as they are aligned with faster, iterative development practices.

According to Stephanie Habif “effective behaviour design is not a linear process,” because “it requires rapid experimentation.” Behavioural scientists are useful because they “know how to experiment, and many companies use behavioural science to make the world a better place.” What isn’t accounted for in this speeded-up process, however, are the temporary and momentary habits that form around novel technologies that have no embedded role in our social structures or psyches.

Subsequently, the knowledge and information offered by behavioural sciences is used to encourage greater levels of engagement and participation in public and social campaigns. What works well, for example, are calls-to-action that are ground in trusted sources of information, and which are supported by endorsements from known and trusted figures who resonate with people in the targeted networks. Social media is a vast and active living laboratory of behavioural engagement. Social media platforms are designed principally to elicit and record data, and it does this in real time, tracking and enabling developers to make adjustment in on the fly.

If a message isn’t effective with one part of the population, at a particular time, then it can be adjusted and modified accordingly. Resulting changes in behaviour are demonstrable in the way that people respond in real life as tracked in the data that is gathered and shared. During the lockdown, for example, one of the key indicators of behaviour change was the reduction in people travelling, either on public transport, or via individual car journeys. The measure of success of the governments messaging was demonstrable in the number of car journeys, the time of those journeys, and the distance travelled. Popular compliance with the lockdown was noted and anticipated by the number of people out and about in their cars.

Message effectiveness was also measured by the number of people who shared or engaged with messages. The patterns of sharing, liking and searching can be used to track the ‘spreadability’ of messages that have resonance within specific market segments (Jenkins et al., 2013). If we are passing messages on, we are likely to be observing the expected behaviours that are associated with those messages. Clapping in the street for the NHS was a strong indication of how many people had taken on board the importance of isolating and staying away from in-person social events.

Behavioural science, then, is the promise of a shortcut to action. It does this by making messaging distinctive, easy to digest and inviting. Participation in the process of sharing and reinforcing the message becomes as important as the process of transmission itself. Supporters of the message are able to shape the campaign themselves by inviting their supporters and network cohabitees to share the goals of the project, thereby further enhancing the efficaciousness of the communication process. The argument is that “with each successful campaign, the supporter is more likely to become confident in her or his ability to perform harder tasks to help the organisation meet its goals and the chances of them becoming active supporters is increased.”

In the behaviourist model people are motivated by, either intrinsic or extrinsic factors. People volunteer to undertake certain actions or responses on the basis that they are driven by different types of reward factors. These can be maternalistic, or they can be non-materialistic. According to behavioural models “some people are motivated by extrinsic motivators such as fame and social status while others are motivated by intrinsic motivations such as altruism, friendships, and duty.” The challenge is to find out which people are aligned with which motivational framework, and to match social messaging in such a way that they appeal to both in due proportion to the social groups that are most at risk.

Behavioural change is made easier, moreover, if the communication model also leverages established social information and norms. For example,social media campaigns that tap on the effects of social norms, which are further heightened by social media features that broadcast supporters’ actions to their social networks, are typically more effective than campaigns that do not.” Those information campaigns that diversify and reinforce messages across platforms and media domains have a greater chance of being effective, particularly if they resonate with what it is that people care about and which they view as their most agreeable human characteristic.

Michelle Chen identifies how

“Campaigns that inform supporters of how similar people compare to them (i.e. descriptive social norms) and how well their participation is perceived by people they care about (i.e. injunctive social norms) fosters supporters’ internal drive to conform to the community’s shared norms. This is also why visual representation of goals such as number of people in their network interested in a virtual event or donated to a cause provide powerful social information to supporters, which can motivate them to engage in the activity” (Michelle Chen).

The strongest forms of social communication, according to behavioural models, are those that bring together social norms which override temporary anti-social actions, in order to achieve a greater good that has wider and more immediate needs: such as protecting the NHS or saving lives. This is a process that brings together and aligns the social needs and tacit expectations of people across a range of different social strata and milieu. It brings into alignment a set of interests for the common good that quickly becomes a pro-social norm, rather than an anti-social behaviour founded in freedom of expression and extreme forms of social liberalism.

Maximising buy-in, then, is achieved by ensuring that the key message resonates with what is already valued in a mutually cohesive society that shares common goals, aims and beliefs. If these goals and aims are highly individualistic, however, or based on disruptive self-interest, then they will fail to gain wider traction and are easily disregarded. The Cummings affair is a text-book example of how to disrupt social unity with a high-profile example of individualistic action.

The questions that are left handing, and which we need to ask about these ideas are urgent and important. For example, to what extent can behavioural research be directly manipulated by the experimenter? What happens if the experimenter starts by basing their working models of human behaviour that is faulty, distrustful or based or misguided assumptions and reasoning? What happens when groups that are unable to engage with the modelling process as equals are targeted by the researcher? Who gets to decide and rank the order of importance for pro-social outcomes and expressions of modified behaviour? What happens if these changes are not brought about through consent, but have to be imposed through active police interventions and control? What happens when behaviourists don’t account for their own biases and levels of socialisation? What is the model of observation and evidence gathering that is used to validate approved forms of behaviour change?

Aside from the many operational questions that can be raised about the behaviourist approach, there are some fundamental questions that need to be asked about the deterministic theories that inform behaviourism at a deeper level. Is human nature quantifiable in terms of general laws? Do these laws apply to all organisms? What is the difference between organisms that are responding to stimuli, and those that are capable of forms of conscious adaptation and change? The reductive appeal of behaviourism dissolves human agency and ethics into a limited field of action and reaction, in which people are equated with animals, units of information, or patients that need to be cared for.

Indeed, modern society has displaced the defining trope of master and servant which was used to define human relations for millennia. It is only now, in the modern age that humanistic individualism has been the dominant model for social engagement. How does behaviourism account for this? Emerging in the Eighteenth and the Nineteenth and centuries, the model of individualism based on inalienable human rights did not exist. Most people where social subjects bound by their relationship to their social masters. If there are no general and universal laws of human nature, only those that have currency in each era, how can we apply behaviourist solutions that minimise harm and maximise empowerment and growth?

How we choose to use behavioural sciences is critically important, because not all models of human nature can or should be affixed in perpetuity, or applied to all. Human experience is contingent and developmental, and therefore can’t be affixed to one set of guiding criteria or assumptions. As Karen Glanz and Donald Bishop point out

“Effective health promotion and public health depend on marshalling the most appropriate theory and practice strategies for a given situation.  The choice of a suitable theory should begin with identifying the problem, goal, and units of practice, not with selecting a theoretical framework because it is intriguing, familiar, or in vogue.”

Behaviourism, therefore, is not the only model of human interaction that can – or should – be used to determine public policy. As Marie-Louise Von Franz, made clear about the insights of Carl Jung,  

“Jung to some extent took the opposite approach to that of the behaviourists, that is, he did not observe people from the outside, did not ask how we behave, how we greet one another, how we mate, how we take care of our young. Instead, he studied what we feel and what we fantasise while we are doing those things. For Jung, archetypes are not only elementary ideas, but just as much elementary feelings, elementary fantasies, elementary visions” (Franz, 2001, p. 6).

The attention and emphasis that is placed on behaviourism is likely to prove to be misplaced, particularly as evidence is gathered of the failure of the communications model and the limitations of the practice of behavioural sciences during the pandemic becomes clear. A one-size-fits-all, mass-media approach to communications and behaviour change might have contributed to excess deaths that could otherwise have been avoided had other forms of civic and social engagement been tried and prioritised. It is likely that the normative symbolic framework used during the pandemic has excluded whole sections and subsections of the population. These are communities that mainstream media practitioners and producers have little awareness of, their symbolic needs and their sense of social identity. The communications models that have been deployed using these behavioural approaches mean that some communities are very easy to overlook or disregard. It is not enough to say that communications during the pandemic has been guided by science, we need to dig deep into the assumptions that are being made, and we need to ask those that have made them to explain and test their validity. As James Hollis points out, this is as much a psychic as a physiological crisis, and

“Until we can find that which links us to that which transcends us, in whatever arena we may find it, we will be torn apart by the opposites; until then, our conflicts have brought us only suffering without meaning” (Hollis, 2020, p. 116).

Fenton, N., Freedman, D., Schlosberg, J., & Dencik, L. (2020). The Media Manifesto. Polity Press.

Franz, M.-L. V. (2001). Psyche and Matter. Shambhala Publications.

Hollis, J. (2020). Living Between Worlds – Finding Personal Resilience in Changing Times. Sounds True.

Jenkins, H., Ford, S., & Green, J. (2013). Spreadable Media. New York University Press.

Thaler, R. H., & Sunstein, C. R. (2009). Nudge – Improving Decisions About Health, Wealth and Happiness. Penguin.


Liked it? Take a second to support Decentered Media on Patreon!