Mistrust, uncertainty and health risks

ABSTRACT

Aspects of the relationship between mistrust, uncertainty and risk responses are examined. Identity Process Theory and Social Representation Theory are used to explain risk responses. The operation of mistrust, defined here as an active state of uncertainty about whether a source or its assertions are to be believed, is examined. Mistrust can also be a cognitive or emotional trait of a person, associated with being habitually suspicious, doubtful, or sceptical, and, as such, can be a preferred strategy for coping with threats to identity. The relationship between mistrust and uncertainty in shaping responses to health risks is examined. An illustration using media reporting of MMR vaccine hesitancy during 2019 is described. This indicated: politicians and health experts sometimes mistrust the public’s reaction to risk guidance; uncertainty once established is resilient against remediation; and, targeting of mistrust and blame is purposive.

KEYWORDS:

• Mistrust
• uncertainty
• risk
• identity
• social representation
• MMR

Introduction

This paper explores the relationships between mistrust, uncertainty and risk responses. There is a long tradition of research on trust, uncertainty and risk emanating from many different areas of social science – economics, history, sociology, ethnography, political science, besides psychology. This paper draws on insights from across this spectrum, but does not set out to systematically review of all the literature. The purpose here is to highlight the significance of mistrust in risk responses from a social psychological perspective. It uses Identity Process Theory (Breakwell, 2015a) and Social Representations Theory (Moscovici, 1988) to explain some of the dynamics underlying the relationships between personal and societal uncertainty and between uncertainty and mistrust.

Reactions to risk and uncertainty: individual differences and identity processes

In practice, risk and uncertainty are intimately enmeshed but they are conceptually distinct. Risk refers to the possibility that harm will occur. Some social science debate has focussed on whether risk can ever be objectively assessed or whether it is inevitably a social construction (Burgess, 2015). Either way, two dimensions of the construct ‘risk’ are distinguishable: probability and effect. In terms of probability, risk refers to the likelihood of occurrence of some specific event or combination of events (the ‘hazard’) that could lead to harm. In terms of effect, risk refers to the degree of harm the hazard might cause. Estimates of the magnitude of probability and effect, whether objective or socially constructed, are distinct. This distinction is important because both may independently vary in the level of uncertainty attributed to them (Breakwell, 2014).

Level of uncertainty in the estimates of probability and effect of a hazard is one determiner of reactions to it. Faced with the possibility of harm, responses are partly shaped by the information available about its origin, form, and the opportunities for its avoidance, mitigation or removal. The degree of certainty attributed to this information by its source is crucial in predisposing responses, especially if the estimate changes over time (Gaspar et al., 2014). In fact, the behavioural, physiological, cognitive, emotional and motivational effects of uncertainty on individuals may be significant. Uncertainty about the probability and effects of a hazard can induce anxiety, act as a stressor, and alter decision-making and the use of information (de Berker, Rutledge, & Mathys, 2016).

Crucially, responses to uncertainty in risk information are determined by the way that individuals actively interpret that information. Individuals differ in the way they perceive the same message about the same estimate. So they may respond differently to what is ostensibly the same level of uncertainty about a hazard. Psychological research on the factors affecting responses to uncertainty came to the fore in the 1970s (Fischhoff & Kadvany, 2011) and subsequently has clustered around models of certain key issues. The first describes information-processing in conditions of uncertainty and showed judgements reveal consistent biases and that these can be predicted on the basis of certain heuristics that people employ (e.g. hindsight bias, optimistic bias) and led to Risk Framing and Prospect Theory (Kahnemann & Tversky, 1979). People vary in the extent to which they deploy these heuristics. The second cluster focuses on work on the perceived characteristics of hazards, known now as ‘the psychometric paradigm’ (Slovic, 2000) which essentially locates hazards in terms of two orthogonal dimensions: dread and controllability. Again people differ in their ratings of hazards on these dimensions. Indeed, psychometric methods rely on this. The third cluster specifically explored individual differences in risk reactions. Socio-demographic factors were considered (age, gender, ethnicity, nationality and disadvantage) and studies, which were often subject to methodological critiques, yielded complex evidence of interactions between these factors in predictions of risk reactions (Breakwell, 2014, pp. 70–83). In this cluster, personality was examined, particularly the ‘Big Five’ personality factors: impulsive sensation-seeking, neuroticism-anxiety, aggression-hostility, sociability-extraversion, and activity (Zuckerman, 1991). Impulsivity and neuroticism regularly emerge as related to differences in risk perception and risk-taking but any of the five can be significant under specific conditions (Barnett & Breakwell, 2001). In addition, some people are found habitually to have a greater tolerance of uncertainty and desire to control it than others (Shuper, Sorrentino, & Otsubo, 2004). This characteristic is labelled the ‘uncertainty orientation’ and is associated with the personality trait ‘openness to experience’ and with authoritarianism (Hodson & Sorrentino, 1999). Individual differences in abiding ‘cognitive styles’ have also been examined. For instance, a pessimistic cognitive style has been found to be associated with unrealistic pessimism about personal vulnerability to health risk (van der Velde, Hooykaas, & van der Joop, 1992).

This work has shown that individuals will respond differently when exposed to uncertainty. Building on this, current social psychological theory has begun to explore how variations in uncertainty and risk perceptions and reactions may be purposive, that is to say be motivated by some desired personal outcomes. It is in this context that identity processes have a strong bearing upon uncertainty reactions. Identity Process Theory (Breakwell, 2015a; Jaspal, Nerlich, & Cinnirella, 2014) proposes that individuals are motivated to maintain and protect four aspects identity: their self-esteem, self-efficacy, distinctiveness and continuity. They use coping strategies at the intra-psychic, interpersonal, and intergroup levels when these aspects of identity are threatened. Sometimes uncertainty or risk threaten these identity elements. If they do, a range of coping strategies will be deployed, differing according to social context and constraints. This approach to the analysis of reactions to uncertainty and risk assumes that the individual is an active agent trying to manage the ultimate impact that the hazard will have upon their identity – and upon their life more generally. The relationship between these four aspects of identity and responses to uncertainty have been examined. For instance, variations in levels of self-efficacy are associated with seeking to reduce uncertainty about their own health status by choosing to have genetic screening for disease (Hendy, Lyons, & Breakwell, 2006). Perceptions of uncertainties in environmental harm estimates are biased by the significance of the specific environment to the identity elements of self-esteem and distinctiveness (Bonaiuto, Breakwell, & Cano, 1996). Denial of anthropogenic climate change, which involves emphasising uncertainties in the scientific data, has also been shown to be linked to identity processes (Jaspal et al., 2014; Pechar, Bernauer, & Mayer, 2018).

Identity-threat coping strategies also often involve social engagement. Hogg (2007) postulates that feelings of ‘self-uncertainty’ motivate people to identify with groups in order to reduce their uncertainty and this can lead to participation in group and societal extremism. This idea seems to fit well with both Identity Process Theory and Tajfel’s (1978) Social Identity Theory. Each of these models emphasise the centrality of the connection between identity processes, uncertainty and social engagement.

While identity processes are important, reactions to uncertainty are rarely wholly individual or individualised. Responses can emerge at interpersonal, group, community, national, or international levels. Shared representations of the uncertainty are constructed across time and place. An individual’s reaction to uncertainty in contexts where the hazard is shared will be influenced by how others respond to it and particularly by how and what they communicate about it. However, it may be anticipated that the individual’s engagement with the shared representation will be mediated by the desire to defend their identity and, inevitably, there will be disparities in the amount of influence different individuals have over the representations of uncertainty which come to be accepted or contested. Mistrust has a major role in this and is considered later.

The internet age, an age of uncertainty – the role of social representations

Galbraith (1977) popularised the idea of ‘the age of uncertainty’. He depicted industrial societies in transition (reflecting changes in the means of production, control of wealth, geopolitical distribution of power, and population mobility) from the certainties of nineteenth century philosophical and economic thought to the growing doubt of the twentieth century. These changes, he argued, nurtured a state of societal uncertainty manifest in doubt about, or rejection of, previously held aspirations, beliefs and values. This uncertainty was not about ‘tangible’ hazards (like disease or poverty), it was centred on meaning and motives. It can be argued that in the first quarter of the twenty-first century this form of societal uncertainty is still present (Taleb, 2007). Actually, Breakwell (2020) suggested that societal uncertainty is growing because of certain fundamental changes that have happened since 1990 when the World Wide Web was opened to the public, giving unprecedented channels of communication and platforms for information exchange, dissent and the undermining of established authorities. The reason this is a source of uncertainty is not simply that the rate of change has accelerated, it is primarily because individuals and society as a whole are less confident about assimilating and using this, often conflicting, vast data reservoir. The fear of ‘fake news’ and disinformation (deliberately fabricated) is widespread (Lazer et al., 2018). Fact-checking websites are multiplying as people seek for information integrity. The internet age has added a new dimension to the age of uncertainty: making distal risk more visible and proximate while simultaneously eroding the ostensible certainty that traditional authorities (such as scientists or health practitioners, Wellcome, 2018) offered. If seen historically, we may live in a more certain and predictable world but it just does not seem that way to a very large number of people.

Societal uncertainty is a complex dynamic system. Any particular uncertainty will be seen against the context of this societal uncertainty. So, for instance, uncertainties about climate change and those about disease control impact upon each other in diverse ways in the public narratives of societal risks. This system of societal uncertainty is itself the product of active, continual interpretation of events through the agency of individuals and many other social actors (for example, activists, media, governments, and professions). Societal uncertainty is a many-faceted entity reconstituted continually through social processes. Moscovici (1988), in the Theory of Social Representations, described some of these processes. He studied how people give meaning to novel social realities, things and situations that they have not seen before. Moscovici described how ‘anchoring’ and ‘objectification’ processes come into play when a community is facing something novel. Both processes contextualise the novel and by doing so make it part of an intelligible set of meanings: anchoring ascribes meaning to a new phenomenon by linking it to pre-extant understandings or explanations; objectification gives it substance by associating it with experiential exemplars.

New or changed uncertainties thus become manageable by being reconceptualised, normalised or identified within the pre-existing system of societal uncertainty – often by reference to existing social representations of comparable events. In relation to health risks, Barnett and Breakwell (2003) described how these ‘hazard sequences’ and ‘hazard templates’ develop. They examined the history of oral conceptive health scares showing the way that the public is notified about a health risk over time builds representations of the hazard. These hazard templates are part of the fabric of shared societal uncertainties. An example at the time of writing in 2020 is the way the uncertainties surrounding the emergence of the COVID-19 virus (2019-2020) were articulated by policy makers and the media relied on understandings, or at least awareness, of other infectious fatal diseases (notably the 1918 influenza pandemic, H1N1, MERS-CoV, SARS, Ebola, measles and tuberculosis). The threat of the new coronavirus was anchored and objectified by reference to past experience of health crises that shared some of its features. The object was to make the levels of uncertainty concerning the consequences of the new viral threat more intelligible to the general public by grounding the explanations in commonly available (if not comprehensive) understandings of what had happened before and so what might happen in the future. In this case, the social representational processes deliberately served to lead the public to expect uncertainty to continue. Solutions (treatments, cures, vaccinations, containment) would be sought but when they would be found was left uncertain. Acknowledging uncertainty has been found in some contexts to enhance trust in the risk communication. But the effect of persistently acknowledging uncertainty has not been determined.

Responses to COVID-19 particularly illustrate the significance of the societal pervasiveness of uncertainty. The uncertainties posed by such a global pandemic encompass all aspects of human existence – including not just physical and psychological well-being but also economic and political security. It is barely possible to conceptualise the range and complexities of the social representation processes that would be needed to objectify and anchor these uncertainties. Such processes are inevitably iterative and often apparently chaotic. They are also open to contestation – conflicting representations will emerge (e.g. conspiracy theories that COVID-19 was deliberately manufactured). Competition between alternative representations, often targeting different aspects of a complex hazard against a background of societal uncertainty, will be unlikely to result in a coherent, common understanding of the risk.

The dynamic system of societal uncertainty as a whole has not been subjected to empirical research. Data analytics making use of many simultaneous data streams in real time might begin to offer at least a cartoon image of it in its entirety in the future. But a precursor to that has to be greater conceptual clarity about the principal components that should feature in such a model. One vital component has to be communication. Anchoring and objectification processes rely on various types of communication to establish shared (and contested) representations of uncertainty. In regard to health risks, a series of studies illustrated how social representations are negotiated (Joffe, 1996 – AIDS; Joffe & Haarhoff, 2002 – Ebola; Joffe & Lee, 2004 – Avian flu). The changes that have occurred in the form and control of communication channels over the last thirty years mean that the construction of social representations of risk and uncertainty is now less dominated by a limited range of influencers. Information control is less restricted to elites. Opportunities online to input material to social representations have proliferated and become more accessible to more people.

The social representations emerging take many forms. One form, which has been a target for social science analysis (e.g. COMPACT, 2020; Sunstein & Vermeule, 2009), manifests in ‘conspiracy theories’. These come in many shapes and sizes (some being more social media ‘conspiracy rumours’ than theories). As theories, they tend to represent the origin of an event as the result of a secret plot by exceptionally powerful and cunning conspirators to achieve a malevolent end. Interestingly, belief in conspiracy theories (so-called ‘conspiracy ideation’) is correlated with certain psychological characteristics, one being paranoia. But uncertainty or hazards which threaten aspects of individual identity (such as continuity or self-esteem) can also arouse support for conspiracy theories. Douglas, Sutton, and Cichocka (2017) note that belief in conspiracy theories appears to be driven by motives that can be characterised as epistemic (understanding one’s environment), existential (being safe and in control of one’s environment), and social (maintaining a positive image of the self and the social group). The same can be said of support for any social representation of risk and uncertainty. In fact, Breakwell (2015b) suggested that identity processes will result in individual differences in the awareness, understanding, acceptance, assimilation and salience of any social representation.

The implications of relatively universal access to awareness of events, change and uncertainty through multiple channels of communication are manifold. It offers the opportunity for great commonality of understanding and action. Equally, it offers the potential for great disparities and conflict in interpretation. There is little evidence, currently, that the communication opportunities now available build harmony or common purpose (though some argue they do serve to shrink individuality and autonomy, Byung-Chul Han, 2017). However, they do often serve to heighten the perception of threat and mistrust (Jodelet, Vala, & Drozda-Senkowska, 2020).

Personal uncertainty and mistrust

Recipients of any information face the task of deciding whether it is believable. Until a decision is made the recipient can be said to be in a state of personal uncertainty. This state of personal uncertainty which often occurs in response to descriptions of risk is an important stage in coming to the eventual acceptance or rejection of uncertainty estimates.

Many factors determine whether a recipient believes information and reaches personal certainty. The psychological literature introduced earlier indicates that it matters whether the information accords with recipients’ self-interest, attitudinal and value preconceptions, past experience or personal knowledge, and their emotional state when they receive the information. Evidence is less likely to be believed that runs counter to your political beliefs, challenges elements of your identity, or contradicts evidence you have, or is incongruent with your mood. Another factor is how the information is structured and it’s medium of communication. Information specificity predisposes belief, especially if accompanied by independent corroborating evidence. Furthermore, other things being equal, information from a trusted source is more likely to be believed than the same information from a less trusted source. Trust in a source offering risk assessments has been shown to depend on: perceptions of their competence, integrity (evidenced in commitment, predictability, consistency and caring), and whether they share the salient values of the audience (Terwel, Harinck, Ellemers, & Daamen, 2009), all of which can be affected by stereotypes of the groups to which they belong. Trust in a source may be changed by engaging in risk communication (because of what is communicated, what context it is communicated in, when it is communicated, to whom it is communicated and with what purpose). Moreover, risk communication impact may be changed by levels of ambient public trust, relevant institutional trust, and specific trust in relation to the particular hazard at that time (Breakwell, 2014, pp. 158–159). The interaction of these factors in determining belief in information is complex. For instance, Longman, Turner, King, and McCaffery (2012) showed that how quantitative health risk estimates are specified can change the perceived credibility of their source. Point estimates rather than ranges were associated with greater credibility of the source and increased perceptions of risk. Such studies emphasise that trust in a source is not fixed and may react to new evidence and situational variables. Personal uncertainty, once triggered, can also be associated with reduced trust in a source. Vaccari and Chadwick (2020) examined ‘deepfakes’ (videos synthesised with fake content that closely resemble real videos) and online disinformation effects. They found that people are more likely to feel uncertain than to be misled by deepfakes, but this resulting uncertainty, is associated with subsequent reduced trust in news on social media.

Trust is a topic that has been examined from every social science perspective and from analytic levels at the systemic (Sztompka, 1999) through to the neurophysiological (Kosfeld, Heinrichs, Zak, Fischbacher, & Fehr, 2005). While sociologists focus on the position and role of trust in social systems (e.g. Giddens, 1984) and psychologists have been concerned with the development of the capacity to trust with its implications for personality and well-being, social psychologists have examined the role of trust in social influence, group membership and identity processes (Tanis & Postmes, 2005). The concept of trust, as a form of social capital and factor in decision-making, is also central to many models in economics. With the advent of the ‘post-truth’ world (McIntyre, 2018) the significance of trust may have changed. There is a shift to encompass a situation where ‘objective facts’ or ‘evidence’ are less influential in shaping public opinion than appeals to emotion and personal belief. Of course, it can be argued that ‘truth’ has always (and long before the label ‘post truth’) been viewed through the lens of emotions and beliefs. Nevertheless, in a time even less concerned with verified facts, trust, or its absence, becomes even more important because it can act as a filter which limits the effect of appeals to emotion and belief. It can also bolster the impact of those appeals, heightening their emotional resonances and their belief-relevance.

From a social psychological perspective, trusting can be viewed as a process of evaluation (one not necessarily based on facts, reason or logic) which involves three possible outcomes: trust, distrust or mistrust. Treated as a process, judging trustworthiness involves cognitive and emotional activity that may entail in quick succession (and in any order) mistrust, distrust and trust but will not necessarily produce a single stable outcome. Empirically, trust, mistrust or distrust of something can appear to co-exist (Huvila, 2017) though this may be explained by their transience and instability in certain contexts. Trust is usually defined as a belief in the reliability, strength or veracity of a person or thing. Distrust is the opposite. While mistrust is sometimes treated as synonymous with distrust or with misplaced trust, more often it is used to refer to uncertainty about whether trust should be offered or not. Mistrust, in this sense, is about being actively suspicious about whether to trust a person or thing but without having actually decided. This definition of mistrust is used in this paper. Mistrust is a state of active uncertainty which may be fleeting or prolonged. Mistrust is not just about cognitions, it is also often embedded emotional responses (for instance, fear, anger, love). Mistrust is of particular interest when considering responses to estimates of risk that are themselves uncertain. Mistrust makes it difficult for risk communicators to know how to target their messages or predict responses to them.

Trust and distrust are contingent upon circumstance to some degree and either may be temporary resting places in the trusting process. Mistrust is less a resting place and more a state of unrest, characterised by suspicion, doubt and ambivalence, and, sometimes, anxiety. It is typified by being uncertain what to believe. Mistrust can be a situationally specific response to a particular piece of information or person. But it can also be a chronic state. The individual may be habitually suspicious, doubtful, ambivalent, lacking in conviction, anxious and sceptical. At this level, mistrust is a cognitive and emotional trait of the person, having stability over time. Erikson (1963,) in his theory of psychosocial development, argued that experiencing mistrust very early in life shapes the personality. Mistrust can also be a preferred and regularly deployed coping strategy for dealing with threats to identity, acting as the basis for effective denial of evidence that would otherwise be damaging to self-esteem or distinctiveness (Breakwell, 2015a).

Mistrust is not just an individual trait. Just as some societies are based on trust, others are based on mistrust. Ethnographers have described societies dominated by cultures of mistrust. In early studies, Banfield (1958) characterised them as ‘backward’ plagued by solitude, anomie and pitiless mutual predation. More recently, Carey (2017) studying contemporary European communities (e.g. the Ukraine) where mistrust is widespread has argued that it does not need to be corrosive and can have utility. Simmel (1950, pp. 318–19) describes trust as ‘a hypothesis regarding future behaviour, a hypothesis certain enough to serve as a basis for practical conduct’. Carey suggests the same is true of mistrust – ‘it is an alternative hypothesis and one that gives rise to social forms of its own’ impacting on practices of communication, co-operation and politics (p. 3). Cultures adapted to pervasive mistrust will deal with novel uncertainty differently to those founded on assumptions of trust. Obviously, this begs the question: if we live in an age of societal uncertainty, are we already a culture of pervasive mistrust?

Mistrust, uncertainty and a health risk

Examples of how mistrust affects social representational processes and behaviour are common in relation to health risks. Calman (2002) pointed out that ‘hypothetical’ risks (those which are possible, biologically plausible, but with no conclusive evidence of their aetiology, their level or their effect) are the most difficult to communicate. Such hypothetical risks are usually a focus for mistrust and challenges to health policies and practice. The absence of medical certainty, which is not uncommon when a new health hazard emerges (such as AIDS/HIV, BSE, or coronaviruses), allows room for social representation processes to construct explanations of what is happening and elaborate predictions of what will happen in the future. The representation activity is not necessarily stimulated by the prior existence of mistrust. However, prior mistrust will shape individual and group participation in the communication that determines what representations emerge (Eicher & Bangerter, 2015).

An illustration of the relationship between uncertainty and mistrust comes from vaccine hesitancy (MacDonald, 2015). The World Health Organisation listed it as one of the top ten global health threats of 2019. Vaccine hesitancy encompasses refusal to vaccinate, delaying vaccination, accepting vaccines but remaining uncertain about their safety, or using vaccines only selectively. Such reactions are documented since the invention of vaccines despite being challenged by overwhelming scientific consensus about the safety and efficacy of vaccines. Vaccine hesitancy is uncontrovertibly linked to disease outbreaks and an increased fatalities from vaccine-preventable diseases. Yet the ‘anti-vax’ movement now continues to grow; promulgating elaborate conspiracy theories to explain how vaccination harms and using social media platforms to mobilise support (where celebrity endorsements are used to sell the message) (Jolley, Douglas, & Tripp, 2014).

The anti-vax phenomenon has stimulated much social science analysis over many years (Hornsey, Harris, & Fielding, 2018). This is particularly true of anti-vaccination activism against the MMR (measles, mumps, rubella) vaccine that was introduced in 1988 (Deer, 2020). In 1998, the Lancet published an article by Wakefield claiming the MMR vaccine was linked to children developing autism. There is now evidence that there is no link between MMR vaccination and autism (https://vk.ovg.ox.ac.uk/vk/mmr-vaccine) and, in 2010, the UK General Medical Council struck Wakefield off the medical register. It is not known how far the uncertainty aroused by Wakefield explains the decline (from 92% in 1996–61% in 2003) in the vaccination uptake by those eligible in the UK. But by 2008 measles was declared endemic again in England. In 2018, across Europe there were 82,500 measles cases, 15 times as many as in 2016. Measles, which can blind, cause brain damage and, even, kill, is the main risk from MMR vaccination refusal.

The facts (vaccines are effective and safe, while the diseases are increasing) have been heavily publicised, so why don’t more people vaccinate their children? There is no evidence that they are uncertain about what message they are being given by the medical authorities. However, they also receive other messages: the conspiracy theories of the anti-vaxxers which discount public health pronouncements as driven by the ineptitude or self-interest of powerful elites. Faced with flatly contradictory messages, uncertainty is justifiable and, for some people, mistrust of the message sources will rise. How far both uncertainty and mistrust take shape for particular individuals and groups will be determined at least in part by the factors described earlier that typically explain variation in responses to hazards. Chief amongst them will be the need to protect threatened aspects of identity and the social representation processes which are at work.

To illustrate how this may work in practice, the following thematic analysis of media reports is presented. In September 2019, there was an upswing of public debate about MMR vaccination, co-terminus with the growing incidence of measles and increased anti-vax social media activity. In order to explore the alternative public narratives evolving during this time, media coverage was analysed. A media monitoring company (Kantar Media) commissioned by the NHS in England does a comprehensive daily search of all news media to capture health-related coverage. Their data for the period 2nd September to 8th November 2019 were used here. Ninety five articles concerning MMR from national newspapers were reported. The topic and message of each report were analysed to be presented here. Table 1 summarises chronologically the MMR-related themes and issues in the reports over the period. No attempt is made here to quantify the number of different newspapers that reported each theme or the overall space given to it. The purpose is merely to capture themes emerging for illustrative purposes. The thematic analysis was conducted by a single analyst and its reliability was not assessed through inter-rater corroboration. No attempt was made in this study to examine other MMR messages available at the time through other media. Consequently, the voices of the anti-vaxxers, predominantly available through social media, are not represented.

Table 1. Media reports MMR 2019.

Event or argument reported
Early September	Anti-vax momentum identified on social media
Mid-Late September	UK Prime Minister urges increase in uptake of vaccination
	New data are released that uptake has dropped
	Health minister moots fines/mandatory jabs/ school bans for decliners
	Vaccination decliners are accused of putting others at risk
	Public health calamity is predicted
Early October	Homeopath blamed for false information given to parents
	Calls for action against social media companies allowing decliners access
	Criticisms of NHS information provision
	PM rules out compulsory jabs
	Calls for other ways to make ‘jabs’ available (not through GPs)
	Tory austerity blamed for withdrawal of immunisation co-ordinators
Mid-October	MMR rejection claimed to be a middle class revolt
	Anti-vax myths said to be promulgated at school gates between mothers
	World Health Organisation declares UK no longer free of measles
	Concerns raised for loss of herd immunity regarding measles
	UNICEF explains how measles kills large numbers
	Head of Public Health England (PHE) blames complacency, lack of awareness and mythology for declining
	GPs say PHE should inform people more effectively
	Health Minister says parents have the moral duty to vaccinate
Late October – Early
November	Claims that NHS is inconsistent in provision of jabs
	NHS targets 500,000 young adults for jabs
	Homeopaths attacked for their bogus cures and scare stories
	Claims that measles makes the body forget to fight infections
	Attacks on social media against anti-vax
	MMR vaccine banned from export amid global shortages

The messages most often reported in the press are those of politicians and health specialists. Their key messages are that vaccination refusal is a product of ignorance, fake news, mythology, conspiracy theories and complacency. Blaming is high on the agenda in these messages. There are a variety of candidates for blame: parents for failing in their moral duty, social media channels, Tory austerity, NHS disorganisation and staff shortages, homeopaths, and, of course, the ‘anti-vax myth-makers’. The blame-target chosen varies across sources. Mistrust of ‘the general public’ by political and health authorities may explain some of these messages. They are uncertain about the motivations and future behaviour of the some parts of the public. They certainly cannot rely on them to be voluntarily compliant. This doubt and suspicion – not knowing quite what to believe and expect of the public – is at the heart of mistrust. This mistrust may explain the use of threats (e.g. of legislation/school exclusion) and of blaming (e.g. chastisement for irresponsibility). It could be that the press coverage has somehow lost the nuances of some of these messages. However, the complex blame-game and challenge to individual freedoms remains the context for evolving shared social representations of vaccine refusal. The blame arguments are explicitly designed to arouse public mistrust and rejection of those targeted. But, once aroused, this mistrust may be generalised to the very authorities that initiate it.

This MMR illustration highlights the way mistrust affects risk communication and management. The extent to which ‘experts’ mistrust the public’s ability to comprehend risk estimates and behave rationally in response to them has received considerable attention previously and has been shown to result in poor policy decisions (BSE Inquiry, 2000). This sort of mistrust skews the type of risk communication that is used away from focussing on the actual reasons for public rejection of advice or guidance (which may be rational or practical at a personal level).

Uncertainty once present is resilient against remediation. Doubt about the safety of MMR was not immediately challenged in 1998. It had several years to be elaborated before it was formally rejected by the medical establishment (Deer, 2020). Such doubt, once promulgated through social representation processes, became part of the social perception of MMR. The doubt sits like a shadow over MMR. Doubt is part of the ongoing social representation of MMR. It can be called forth when events make it relevant again (as they did in 2019 with the WHO report on prevalence of measles). Since these sorts of social representations now have an incalculable half-life on social media they are particularly virulent and accessible. In addition, it may be difficult for someone who belongs to the anti-vax movement to abandon their beliefs because being a member of the movement may be important to their identity (Jolley et al., 2014). Group dynamics perpetuate mistrust.

Uncertainty about the risk itself and mistrust of the actors responding to it interact in different ways at various stages in the MMR story. Mistrust of new groups is advocated along the way. At the time of the media coverage analysed, homeopaths are highlighted. As scientific evidence for the certainty of MMR vaccine safety grew, the burden of mistrust shifted from vaccine manufacturers, to quack doctors, to social media, to alternative medicines, and so on. Changes in social representations rely on conflicting explanations of the risk and of the vaccine hesitancy being propounded and popularised. Explaining who should be mistrusted is part of the process of justifying the need for a revision in the social representation. Changing the categories of people who are to be mistrusted can also trigger threat to identity in an audience. For instance, people who have used them may find the blaming of homeopaths challenges their own self-esteem or self-efficacy. To protect their own identity, they may deny the credibility of blaming homeopaths and, as a consequence, mistrust those who blame homeopaths. Individuals will construct their interpretation of blame messages to protect themselves and this affects who they mistrust.

Conclusion

The MMR story illustrates the manifestations of mistrust and uncertainty in relation to risk reactions. In an age of societal uncertainty, conceptual clarity in modelling the relationship between uncertainty, mistrust, and the cognitive, emotional and behavioural responses to risk is needed. The level of uncertainty in the estimates of probability and effect of harm is important in determining reactions to a hazard but reactions are also determined by identity processes and social representation processes. Uncertainties become manageable by being reconceptualised, normalised or identified within the pre-existing system of societal uncertainty – often by reference to existing social representations of comparable events. However, this does not result inevitably in consensus about the risk. In the context of contested uncertainties, mistrust (either of a source by the audience or vice versa) will play a significant part in determining the risk response. Mistrust, characterised by active scepticism, doubt and suspicion without resolving into a stable estimate of the honesty or accuracy of the risk estimate, motivates continued ambivalence and unpredictability in relation to its target. In practice, establishing the factors influencing mistrust should be valuable in allowing contested uncertainties to be reconciled.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes on contributor

Glynis M. Breakwell developed Identity Process Theory and conducts research on the psychology of risk, mistrust, and social representation processes.