Unraveling the Dark Side of Social Norms—Toward a Research Agenda on the Challenges of Social Norms in Health Communication

ABSTRACT

Social norms are a promising means in health crisis communication because they can guide collective action to reduce risk. However, recent research on the COVID-19 pandemic suggests that social norms may have not fully supported strategic goals and even contributed to phenomena that hindered risk reduction, calling into question the potential of social norms campaigns. This became most evident during the COVID-19 pandemic in the emergence of alternative norms of measure opposition, stigmatization of norm-deviant individuals, and the issue of free-riding. The article analyzes these phenomena from a social identity and communication perspective and outlines areas for further inquiry in health and crisis communication. The goal is to pave the way for a research agenda dedicated to the dark side of social norms to unlock the full potential of social norms in times of (health) crisis.

Social norms are a promising means of health crisis communication for several reasons. Collective threats require collective actions to mitigate risk (Diekmann, 2022), and social norms are a key mechanism for coordinating collective actions (Gelfand et al., 2021). Social norms indicate what others do (referred to as descriptive norms) and what others approve of (known as injunctive norms; Cialdini et al., 1990) and thus illustrate what is an effective and socially approved action. Public health crises, such as the COVID-19 pandemic, are times of uncertainty accompanied by a great need for orientation, including the need to identify functional and appropriate behaviors. In these times, social norms are particularly relevant and influential as they convey information about others’ behaviors and attitudes (Rimal & Storey, 2020). Lastly, social norms regulate compliance with prevention measures socially instead of by force through legal rules (Carbonara, 2017). Thus, the establishment of social norms is consistent with the principles of democracy and represents a legitimate long-term means of risk control in democratic societies. Given this strategic potential of social norms, health communication scholars and experts strongly recommended studying social norms in the context of the COVID-19 pandemic and employing norms-based communication to establish prevention behavior in societies (Gelfand et al., 2021; Krenn, 2021; Rimal & Storey, 2020).

However, research and observations during the COVID-19 pandemic suggest that social norms may have not fully supported strategic goals and even contributed to phenomena detrimental to risk reduction—calling into question the potential of social norms campaigns. This article highlights three norms-related phenomena that became salient during the pandemic and not only counteracted the effectiveness of prevention measures but also potentially undermined social cohesion, which is considered a fundamental societal resource for coping with crises (Jewett et al., 2021) and a lack of which can become a crisis itself (Allcott et al., 2020; Jungkunz, 2021). These phenomena are (1) the development of alternative norms in juxtaposition to the government-introduced prevention measures and population-wide prevention norms within a significant part of the population (Frei et al., 2021); (2) the stigmatization of norm-deviant people via prejudice, devaluation, and unfair treatment, for instance, for not wearing masks (Kwon, 2022); and (3) free-riding, where people benefitted from others’ compliance with measures while not contributing to public safety themselves, for instance, in the context of the COVID-19 vaccination campaign (Yong & Choy, 2021). These three phenomena reflect what I call the dark side of social norms as they underscore the risk of developing prevention-detrimental norms and the undesirable side effects of strong prevention norms. With its focus on the potential of social norms to promote health behaviors, scholarship thus far has widely disregarded the dark side of social norms. This article argues that health and crisis communication scholarship has a mission to understand the negative aspects of social norms to ensure that norms-based communication can unlock its full potential in times of crisis.

In this article, I will outline areas for further inquiry for health and crisis communication scholarship by analyzing the three phenomena of alternative norms, stigmatization, and free-riding that occurred during the COVID-19 pandemic. To this end, I will first provide relevant definitions of social norms as well as crisis communication. I will, second, discuss the three phenomena by referring to the social identity perspective (Tajfel & Turner, 1979; Turner et al., 1987) and the related communication perspective on social norms, as propagated in current social norms research (e.g., Geber & Hefner, 2019; Hogg & Reid, 2006; Lapinski & Rimal, 2005; Yanovitzky & Rimal, 2006). The social identity perspective encompasses social identity (Tajfel & Turner, 1979) and self-categorization theory (Turner et al., 1987) and bases its argument on the human need for a positive social identity. The communication perspective on social norms focuses on the meaning of communication in normative social influences (Hogg & Reid, 2006; Lapinski & Rimal, 2005; Yanovitzky & Rimal, 2006) and includes propositions on the role of media in norm formation (Geber & Hefner, 2019; Geber & Sedlander, 2022). In combination, both perspectives help understand factors and dynamics contributing to these phenomena. Third, I will derive broader recommendations for a research agenda in health and crisis communication on understanding the dark side of social norms in times of crisis.

Social norms and crisis communication

Social norms are rules shared by members of a group or society that guide or constrain their behavior (Cialdini & Trost, 1998). According to the work of Cialdini et al., (1990), social norms can be differentiated into descriptive and injunctive norms, that is, the prevalence and the social approval of a behavior in a social group. Descriptive norms are influential because of people’s motivation to do the right thing, and injunctive norms motivate behavioral change because of the motivation for affiliation with others (Cialdini et al., 1990; Rimal & Lapinski, 2015). Social norms are group-specific per definition, meaning that they are defined in social groups and may vary between them. People are especially motivated to follow the norms of groups they belong to or aspire to belong to, also called “reference groups” (Merton & Rossi, 1968). Reference groups can vary according to the level they refer to—ranging form the direct social environment to the overall population (Patrick et al., 2012). There is theoretical and empirical evidence that closer groups—encompassing people with whom individuals identify and whom they trust—exert stronger normative influences than abstract collectives (Geber & Sedlander, 2022; Woolf et al., 2014). In a public crisis context, however, the overall population is assumed to be influential as well because of the need for collective action and thus widespread compliance with risk reduction measures in the public (Diekmann, 2022; Schneider & van der Linden, 2023).

Successful crisis management requires a well-adjusted combination of crisis measures and communication at all stages of a crisis, that is, from the risk of the crisis, to eruption, to recovery on into evaluation (Reynolds & Seeger, 2005). Besides providing information about risks and hazards with the intention of creating risk awareness and understanding, crisis communication employs persuasive strategies to motivate the public to engage in measures of risk reduction (Weaver et al., 2008). Because of their ability to coordinate actions, social norms campaigns that deliver messages about behaviors and attitudes of relevant others (i.e., social norms) are a promising means in (health) crisis communication (Berkowitz, 2004). However, insights of the COVID-19 pandemic suggest that the strategic use of social norms has limitations in this regard. This has become evident in the emergence of alternative norms of measure opposition, the stigmatization of norm-deviant people, and the issue of free-riding.

Alternative norms of measure opposition

In many countries, the majority of people engaged in the collective effort and complied with prevention measures during the COVID-19 pandemic, such as mask wearing, social distancing, and vaccination (Gelfand et al., 2021; Georgieva et al., 2021; Petherick et al., 2021). Thus, there were majority norms of prevention behavior in several societies (Collis et al., 2022). Yet it also became apparent that a significant number of people did not follow the prevention norms, among others, for political reasons (Hannawa & Stojanov, 2022; Reinemann et al., 2022). To analyze the phenomenon of politcally motivated measure opposition, I adopt the term “alternative” from research on alternative news media (Frischlich et al., 2023) and introduce the notion of alternative norms. Following the conceptualization of alternative media as media that position themselves as correctives of the government and the mainstream news media (Holt, 2018; Holt et al., 2019), alternative norms are defined as norms of measure rejection, critique, and protest, in contrast to mainstream prevention norms. Though these alternative norms were followed only by a specific and small group and have not motivated large parts of the population to adopt them, they implied a twofold threat to society: first, they have affected the effectiveness of the COVID-19 measures (Aschwanden, 2021) and, second, they have reinforced social and political polarization by highlighting divides within the society (Allcott et al., 2020; Jungkunz, 2021).

While people who followed alternative norms and opposed the recommended prevention measures had different political and ideological backgrounds, ranging from right-wing to left-wing ideologies and from esoteric to conspiracy beliefs, they shared a unifying commonality—namely, their strong identity as critics of the measures and the elites in politics, traditional media, and science (Frei et al., 2021; Pantenburg et al., 2021). This was also reflected in their high distrust in politics, the media, and the science, as well as their populist attitudes (Reinemann et al., 2022). These populist attitudes, reflecting a perceived antagonism between the people and the elite (Mudde, 2004), may have driven their social categorization into “us,” the ingroup, and “them,” the outgroup (Frei et al., 2021). From a social identity perspective, this suggests that people who followed alternative norms can be considered a social group defining its own identity relative to the elite groups (Turner et al., 1987). Thus, their noncompliance with and disapproval of the measures can be interpreted both as a rejection of the measures and as a means of making their self-categorization and group identity as opponents visible in public, for instance, by not wearing masks (Grunawalt, 2021; Klinenberg & Sherman, 2021). Qualitative interviews have further revealed that these people considered it their duty to identify the social ills generated by the crisis measures, wake up people, and protest against the measures (Nachtwey et al., 2020).

Considering the key role that communication plays in forming norms (e.g., Hogg & Reid, 2006; Lapinski & Rimal, 2005; Yanovitzky & Rimal, 2006), alternative news media may have been important in communicating and distributing alternative norms. Research shows that alternative news media turned out to be relevant information sources among people who opposed the recommended prevention measures (Reinemann et al., 2022) and shared “overly critical, even anti-systemic messages, opposing the view of the mainstream news media and the political establishment” (Boberg et al., 2020, p. 1). It is thus likely that exposure to this content played a crucial role in alternative norm formation. In addition, research revealed that opponents to measures turned to mainstream news media as well to stay informed about the COVID-19 pandemic (Reinemann et al., 2022).

Directions for future research on alternative norms

The findings from the COVID-19 pandemic highlight the need to study the effects of alternative and mainstream news media coverage on perceived alternative norms and measure opposition. Noting that there are additional communicative sources for norm formation, such as social media and interpersonal communication (Geber & Hefner, 2019), I focus here on news media, because they have been found to be the most relevant source of information during the COVID-19 pandemic (Friemel et al., 2020; Reinemann et al., 2022) and are the most important channels in crisis communication (Reynolds & Seeger, 2005).

The tradition of media effects research (Valkenburg et al., 2016) provides a variety of theoretical explanations on how news media contribute to the formation of normative perceptions. Taken together, it highlights direct effects of exposure to news coverage (e.g., cultivation theory, Gerbner, 1969; social cognitive theory of mass communication; Bandura, 2001) and indirect effects based on news users’ presumptions of media influences on others (i.e., presumed influence, Gunther & Storey, 2003/persuasive press inference, Gunther, 1998). Direct effects refer to the idea that normative cues in media coverage (such as “60% of the Swiss population is already vaccinated”) form perceptions of social norms, for instance, through cultivating ideas about social reality (Gerbner, 1969) or social learning processes (Bandura, 2001). Indirect effects are based on media users’ presumptions that others, specifically, their behaviors and attitudes, are affected by media coverage (Gunther & Storey, 2003). Because perceptions about others' behaviors and attitudes refer to social norms (i.e., perceived descriptive and injunctive norms; Cialdini et al., 1990), presumed influence processes have already been discussed as a factor in norm formation (Geber & Hefner, 2019; Geber & Sedlander, 2022) and empirically shown to be meaningful in this regard (Gunther et al., 2006; Hong & Kim, 2019). Thus, future research should consider both direct and indirect effects on alternative and mainstream news media coverage on perceived alternative norms in times of crisis.

In addition to direct and indirect effects, findings from the COVID-19 pandemic suggest that moderators of the relationship between media exposure and perceived norms must be considered to fully understand the emergence and establishment of (alternative) norms. Among others, hostile media perceptions—that is, the tendency to perceive that media coverage is unfairly biased against the own side and in favor of the outgroup’s point of view (Vallone et al., 1985)—or populist attitudes might moderate media effects: For instance, exposure to mainstream media coverage might have different effects on persons that see their views represented in media coverage than on people who hold hostile media perceptions. These elaborations on media effects and normative influences are in line with the differential susceptible to media effects model (Valkenburg & Peter, 2013) and the theory of normative social influence (Rimal & Real, 2005), which consider media effects and normative influences not to be the same for the entire audience, but to depend on individual characteristics.

Stigmatization

Stigmatization occurred during the COVID-19 pandemic in the context of public behavior, such as social distancing and mask wearing. People were attacked and insulted or publicly shamed for not wearing a mask (Klinenberg & Sherman, 2021; Kwon, 2022) or for not keeping a distance (Travaglino & Moon, 2021). Notably, research also highlights that conversely—depending on the policies and norms during this highly dynamic time—people were stigmatized for engaging in prevention behavior, for instance, for wearing a mask when the majority had abandoned mask wearing (Kwon, 2022). From a social identity perspective (Tajfel & Turner, 1979; Tajfel et al., 1971), stigmatization might occur when the individual’s social identity, defined by group membership, is threatened by their awareness of the presence of an outgroup, especially in times of uncertainty, when there is a great need for a normative orientation. Stigmatization is then the process by which people recognize others as belonging to the outgroup because of their deviance from the ingroup’s norm and jeopardizing the group’s collective effort; they consequently subject them to prejudice, devaluation, and unfair treatment (Link & Phelan, 2001). Specifically, it is the injunctive aspect of the ingroup norm, the “what ought to be done” part (and not so much the descriptive aspect), that is likely to enforce stigmatization as its disregard is accompanied by social sanctions (Cialdini et al., 1990). Indeed, scholars note that COVID-19 measures were constructed as “moral imperatives” and “civil duties” (Capurro et al., 2022; Prosser et al., 2020).

Notably, the media seemed to play a vital role in the sanctioning of noncompliance with these moral norms. A study of Canadian newspaper coverage of COVID-19 found that those who did not comply with the measures were construed as a threat to public health and the moral order (Capurro et al., 2022). Other studies report about online videos showcasing citizens going toe‐to‐toe in public places because of someone’s refusal to wear a mask (Klinenberg & Sherman, 2021) or about social media campaigns using hashtags, such as “covidiots,” to ostracize individuals failing to abide by the new norms (Travaglino & Moon, 2021). While social enforcement is arguably part of the potential of social norms to influence favorable behavior (as introduced above), stigmatization lies far beyond effective and appropriate means of social sanctioning. Indeed, research shows that stigmatization is a painful experience that pushes people toward identifying with others who also feel excluded and thereby undermines social cohesion and supports polarization in the society (Täuber, 2018).

Directions for future research on stigmatization

While there is already some research on ethnic-related stigmatization and stigmatization of health-care workers during the COVID-19 pandemic (Schubert et al., 2021; Yuan et al., 2022), there is limited scientific knowledge of stigmatization related to violation of risk mitigation norms. Acknowledging the public dimension of stigmatization, the communication perspective on social norms (e.g., Hogg & Reid, 2006; Lapinski & Rimal, 2005) might be particularly instructive in this regard. Following its propositions on the role of media in norm formation (Geber & Hefner, 2019; Geber & Sedlander, 2022), future research should monitor the normative discussion of (non-)compliance with crisis measures and depictions of opponents to those measures in the media and examine the effects of media messages on the willingness to engage in a respectful and integrative negotiation of norms. Such a monitoring could include the identification of explicit norms messages, specifically injunctive norms messages (“No excuse—we all need to wear masks”), and/or identify hate speech (i.e., negative stereotyping, expressions of violence) and offensive language (i.e., insults, slurs, degrading metaphors) toward specific groups (Paasch-Colberg et al., 2022). While computational classifiers for detecting offensive language and hate speech have already been developed, applied, and tested (Malik et al., 2022), computational approaches to detect normative language are still in their infancy but promising (Olson & Forbus, 2021). Such computational approaches would enable the examination of large amounts of content across a variety of news media outlets and social media platforms. This would allow, for instance, tracing how normative language, hate speech, and offensive language in news media affect hostility in user comments. In the combination with survey data, the consideration of moderators, such as group identity our outgroup beliefs, is promising for understanding differential communicative (Valkenburg & Peter, 2013) and normative influences (Rimal & Real, 2005).

Free-riding

Free-riding potentially occurs in the context of collective action, where people must cooperate to achieve the best possible effect (Diekmann, 2022; Yong & Choy, 2021). Getting vaccinated is an exemplary cooperative behavior, which became apparent in the discussion of the concept of herd immunity and the related idea that there was a way back to normality if enough people cooperated and got vaccinated (Aschwanden, 2021). However, the problem of free-riding has also been discussed regarding social distancing (Cato et al., 2020) and mask wearing in the COVID-19 context (Bir & Widmar, 2021). Free-riders are people who benefit from others’ cooperation while avoiding individual costs and not contributing to the public good, in this case, health protection, themselves (Yong & Choy, 2021). From a social identity perspective, free-riding might occur if people define their identities first and foremost as individuals and less as members of a social group (Turner et al., 1987), implying that group identity and thus the social pressure that the group can potentially exert are low among free-riders (Agranov et al., 2021; Neville et al., 2021).

Social norms likely provide a basis for a free-riding calculus. Unlike the case of stigmatization, the descriptive norm aspect specifically seems relevant for free-riding because it indicates the prevalence of a behavior and thus provides crucial information for efficient decision-making (Cialdini et al., 1990). While there is some empirical evidence that increasing descriptive information and perceptions of others’ vaccination willingness increased vaccination intention (Moehring et al., 2023), studies on the COVID-19 vaccination campaign (Agranov et al., 2021) and the influenza vaccine (Galizzi et al., 2022; Lau et al., 2019) have also found that particularly high levels of perceived descriptive norms may counteract this effect, suggesting that a “free-riding effect may kick in at high levels of the coverage rate” (Galizzi et al., 2022, p. 641). Notably, this research shows that the possible underlying free-riding effect depends on individual variables, such as political attitudes, and may also be dependent on cultural moderators, such as collectivism/individualism (Geber et al., 2023; see also Betsch et al., 2017).

Directions for future research on free-riding

Free-riding has often been discussed as an underlying mechanism and explanation for the limited or counter-theoretical effects of social norms on COVID-19 prevention behavior, and it has a long tradition in research on vaccinations in general (Betsch et al., 2013; Korn et al., 2020). Beyond the health context, free-riding is likely to occur in the context of other crises and risk reduction measures that require collective action. These situations can be understood as public goods dilemma, where individuals need to choose between cooperating to maintain the public good (i.e., community health protection, mitigating risks of climate and energy crisis) at some personal cost (i.e., getting vaccinated, committing to a low-carbon lifestyle) versus free-riding if there are already enough cooperating (high descriptive norm in the population; Yong & Choy, 2021). Yet there is a lack of research explicitly examining the free-riding calculus in the context of high social norms and low intention to adopt prevention or risk reduction behavior.

Future studies dedicated to free-riding might include perceived descriptive norms, perceived personal costs (e.g., perceived threat of vaccination’s side effects), as well as perceived personal and community benefits (e.g., perceived protection for oneself and the community by vaccination) as predictors of intention to engage in the prevention behavior (e.g., intention to get vaccinated) or even the free-riding rationale itself (e.g., by items such as “If everyone is vaccinated, I don’t need to be vaccinated too”). Besides the linkage of such survey data with content data on descriptive norms, experimental designs allow testing messages that are promising in reducing free-riding and improving compliance with crisis measures (e.g., messages that emphasize the prosocial aspect of the prevention or risk reduction behavior; Böhm & Betsch, 2022). Again, the effectiveness of certain messages and the weight of personal cost and social benefits in the decision to free-ride are likely to depend on moderators, such as group identity or collectivism/individualism.

Conclusion

To realize the full potential of social norms campaigns in times of (health) crisis, scholarship must devote more attention to the challenges social norms may pose. This article analyzed three phenomena—the emergence of alternative norms of measure opposition, stigmatization of norm-deviant people, and the issue of free-riding—that reflect the dark side of social norms and provide concrete areas for further inquiry for health and crisis communication scholarship. While these phenomena have been analyzed in the context of the COVID-19 pandemic, they are likely to occur also in other crisis contexts that require collective action to mitigate risks. This applies to future health crises, such as the next epidemic or pandemic (“Disease X”; Simpson et al., 2020), but also to other current crises such as the climate and energy crises. Similar to the COVID-19 crisis, the climate and energy crises require collective action to mitigate their risks, and social norms have been already discussed as a powerful lever “for engendering large-scale behavior change” in these domains (Schneider & van der Linden, 2023, p. 346). However, there is a lack of examination of the promise and limitations of social norms in risk reduction and implications for health and risk communication. The present analysis yielded four broader recommendations for an agenda in health and crisis communication that considers the dark side of social norms.

First, the social identity perspective (Tajfel & Turner, 1979; Turner et al., 1987) and the communication perspective on social norms (e.g., Geber & Hefner, 2019; Hogg & Reid, 2006; Lapinski & Rimal, 2005; Yanovitzky & Rimal, 2006) were instructive in the analysis of these phenomena and should serve as the theoretical basis for a future research agenda on social norms and the role of communication in this regard. A communication-driven perspective is more relevant than ever given the increasing digitalization of the media landscape, encompassing the emergence of new (alternative) news outlets as well as communication spaces on social media platforms or instant messaging services. These new communication spaces provide a fertile ground for identity-based communication, such as the negotiation of alternative norms as well as hate speech and offensive language toward outgroups.

Second, and related to this, future social norms research must disentangle communicative (Valkenburg & Peter, 2013) and normative influences (Rimal & Real, 2005) in terms of their processes and conditions. This includes the consideration of mediators that transmit communicative and normative influences, such as media users’ beliefs about influences on others, and moderators, such as social identity, that account for differences in the relationships between media, social norms, and behavior. This requires sophisticated methodological designs, such as “linkage studies” that combine content and survey data (Vreese et al., 2017) and longitudinal designs that allow for examining causal influences. Specifically, the linkage of (computational generated) content data on norms messages and survey data on normative perceptions, beliefs, and individual behaviors in a longitudinal design seem most appropriate to examine causality and conditions of the relationships between media, social norms, and behavior. The expected findings are instructive for future crisis communication as they indicate which target group can be addressed by which means (i.e., channels and messages) to improve compliance with risk reduction measures and to prevent stigmatization, free-riding, and, more general, further polarization.

Third, in this context, it is important to consider that communication (Ball-Rokeach, 2008) and social norms (Edberg & Krieger, 2020; Gelfand, 2012) are culture-dependent: The role of media in shaping normative perceptions might be contingent on media system-related factors, and the strength of social norms in guiding behaviors depends on cultural values (e.g., Geber & Ho, 2023; Geber et al., 2023), such as collectivism and individualism (Triandis, 2018). Future research needs to take into account the cultural context and move beyond the current focus in social norms research on Western countries (Shulman et al., 2017).

Fourth, the attributes of the behavior itself must be considered and differentiated (Diekmann, 2022; Rimal et al., 2011). In the emergence of norm-related phenomena discussed here, the question of whether the behavior serves a collective goal (beyond or instead of an individual goal)—which is likely the case in a crisis context (Diekmann, 2022; Schneider & van der Linden, 2023) but not necessarily true for other health prevention behaviors and the differentiation between public and private behaviors proved crucial (Rimal & Lapinski, 2021; Rimal et al., 2011). Systematic reviews that synthesize the state of research on social norms across a variety of behaviors and crisis contexts can provide an evidence-based understanding of the meaning of behavioral attributes and might result in a comprehensive classification of behaviors along their main attributes. This would be instructive for health and crisis communication in order to anticipate the effectiveness of social norms campaigns and the risk of side-effects (e.g., stigmatization or free-riding) across risk issues.

Together, the areas for further inquiry and recommendations provide a starting point for a research agenda dedicated to understanding the dark side of social norms in health and crisis communication. Such a research agenda is imperative to create a foundation for future (health) crisis communication that aims at realizing the full potential of crisis measures and preventing further polarization in society (Schneider & van der Linden, 2023). Its findings will be more relevant than ever given the likelihood of the next pandemic (Simpson et al., 2020) and the fact that, also beyond health crises, humanity is currently facing severe crises, such as the climate and energy crises (Homer-Dixon et al., 2022).

Acknowledgements

I thank Dr. Dorothée Hefner and Dr. Thomas Zerback for their valuable feedback on an earlier version of this manuscript. I also thank the editors of this special issue—Dr. Maria Knight Lapinski and Dr. Rajiv Rimal—and the two anonymous reviewers for the constructive review process and helpful comments on the manuscript.

Disclosure statement

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

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.