“How can I keep quiet?” Motivations to participate in vaccination communication on Facebook

1 Introduction

The omnipresence of social media creates opportunities that alter risk and crisis communication (hereafter, RCC). Institutional actors are obliged to inform the public to limit, contain, mitigate, and reduce public harm during emergencies (Seeger and Reynolds, 2008). Yet, as an open arena of communication, social media lower the threshold for participation, allowing laypeople to contribute to conversations around risks, making RCC increasingly multivocal (Austin, Liu, and Jin, 2012; Frandsen and Johansen, 2016; Strekalova, 2016). The hybrid ecology of social media allows content producers with varying goals and motivations to operate side by side, often deploying the same media towards different ends (Benkler, 2007; Jenkins, Ito, and boyd, 2016). Messages from state authorities and affected organizations might coincide and compete with the diverse voices of laypeople (Boin, ’t Hart, Stern, and Sundelius, 2005).

The alignment between non-specialist and expert framing can influence the effectiveness of institutional RCC. For example, if social media users criticize official messaging, propose alternative frames, or spread rumors or misinformation, this can obstruct or undermine RCC strategies. However, support and acclamation on social media can reinforce and facilitate RCC.

Institutional actors, their goals, obligations, and strategies have traditionally been at the core of RCC research. However, audiences move beyond being just a target or consumer of institutional RCC in the social media context since some laypeople also contribute to communication in crises. That sets up a “public-public” RCC and additionally requires an audience-oriented approach (Fraustino and Fisher Liu, 2017). Therefore, to better understand the multivocality of RCC on social media, there is a need to understand laypeople’s participation, and its motivations and characteristics.

The focus of this study is on public health, specifically vaccination. Looking for health-related information is one of the most widespread online activities (Fox, 2011; Silver, Cohn, and Cornibert, 2019). Attention to public health issues skyrocketed in 2020 due to the COVID-19 pandemic, spotlighting many topics including vaccination. The World Health Organization (WHO) and national public health authorities advocate for and promote vaccination. Yet numerous voices on social media are against vaccination. Vaccine hesitancy has been named one of the top threats to global health (WHO, 2018). This goes for all types of vaccinations, in regular times and during pandemics. Surveys show that in the fall of 2021, 45 % in Russia (Levada-Centre, 2021), 19 % in the United States (NPR, 2021), and 4 % in Sweden (NOVUS, 2021) refused to get a COVID-19 vaccine.

Many aspects of healthcare vary between countries. Therefore, scholars call for country-specific research on health communication (Zhang, He, and Sang, 2013). Sweden presents an interesting case since vaccination is voluntary, yet vaccination rates are among the highest in the world. Such results require effective RCC and trust in institutions (Tafuri et al., 2014). However, Sweden has had a recent history of public concerns about vaccine safety, even before COVID-19. These are related to the side effects of the Pandemrix vaccine against the influenza A (H1N1) virus, also known as the swine flu, in 2009 (Läkemedelsverket, 2019) and suspected side effects of the human papillomavirus vaccine in 2014 (EMA, 2015). Furthermore, growing vaccine hesitancy has resulted in regular local outbreaks of measles (Folkhälsomyndigheten, 2020).

In Sweden, official vaccination communication operates in a complex, multivocal RCC environment that includes numerous public and closed social media groups. In addition, intense debates are common in commentary threads on thematic Facebook posts by news media (including alternative media, e. g., Rooke, 2021) and authorities. Social media users perceive these discussion spaces to be an integral part of the story (Barnes, 2015), which can change the overall framing of vaccination campaigns.

Considering the growing importance of user participation in vaccination communication on social media, its potential impact on RCC efforts, and the limited availability of research on this phenomenon, this study aims to explore this issue. It draws on in-depth interviews with a highly specific sample of Swedish Facebook users who actively communicate about vaccination (both in favor and against), to provide a detailed description and analysis of user participation following Dahlgren’s (2011) model of online participation.

The following section presents the theoretical framework, followed by an explanation of the methods and data used in this study. Later, the main results are presented. The final section discusses the implications of these results for RCC in the public health context.

2 Theoretical framework

Participation refers to the act of taking part in an activity to address “issues of public concern” (APA, 2012). Late 20^th-century changes in media and communication technologies facilitated new online public arenas, including social media. These spaces gave rise to a distinct type of online participation (Oser, Hooghe, and Marien, 2013).

Social media, in general, have received significant research attention. Many studies approach social media use broadly (e. g., Park, Kee, and Valenzuela, 2009; Whiting and Williams, 2013), which encompasses the doubly articulated character of social media audiences (Livingstone, 2007): spectatorship and active forms of participation. In addition, some studies specifically look at the latter and analyze production and (re)distribution of user content (e. g., Leung, 2009; Schaedel and Clement, 2010). However, a limitation of previous studies is the lack of context-specificity (Daugherty, Eastin, and Bright, 2008).

The current study focuses on active forms of online participation in a specific context. It will provide a conceptual toolbox for an explorative study, contrasting frameworks with a preset list of categories (for instance, in the uses and gratifications approach) (Katz, Blumler, and Gurevitch, 1973; Rubin, 2010; Whiting and Williams, 2013). Considering these factors, the current study draws on Peter Dahlgren’s (2011) model of online participation. In contrast to other approaches, this model was explicitly developed for operationalizing participation in the digital sphere and is intended to be generally applicable, which allows researchers to apply it in a wide variety of cases and account for contextual factors.

In a more recent work, Dahlgren and Hill (2020) draw attention to conceptual differences between participation and engagement. They highlight a broader character of the latter and state that participation is a manifestation of engagement with the observable behavior (Dahlgren and Hill, 2020, p. 6). Although some operational parameters of media engagement and online participation overlap, the model from 2011 is tailor-made for analyzing online participation and, thus, provides a better fit for the current study.

The model breaks online participation down into five parameters, which can be analyzed separately or in combination: socially situated context (trajectories), form of communication (modalities), intentionality (motivations), norms and practices (sociality), and the manifestation of participation (visibility) (Dahlgren, 2011).

This study draws upon these parameters. Trajectories and modalities are limited by the setup. First, the focus is on online participation regarding a particular public health issue, vaccination. Second, the modalities are limited to Facebook communication. The originality of this study rests upon the research questions that address the latter three of Dahlgren’s parameters: motivations, sociality, and visibility.

Motivation generally refers to the drives that underpin human behavior (Hollyforde and Whiddett, 2002) and the internal processes that activate, guide, and maintain it (Baron, 1991). It focuses specifically on the individually reported motivations behind participation.

Context-specific studies on motivations for social media users’ participation in RCC have so far focused on natural disasters and political protests. For example, an interview-based study took the 2011 Queensland floods in Australia as a case study in online participation. It identified four main motivations: information monitoring, community, affirmation, and awareness (Brunce, Partidge, and Davis, 2012). Another example is a study that analyzed political protests. It found two motivations for social media participation: distrust in mainstream media coverage, and the need to provide alternative views (Bal and Baruh, 2015). The present study aims to expand this research stream into the public health context. Previous studies on vaccination communication have approached user participation by analyzing the content of posts and comments on Facebook (Faasse, Chatman, and Martin, 2016; Tustin et al., 2018) and other social media. Moreover, there have been calls for more attention to participants of vaccination communication and their intentionality (Ward, Peretti-Watel, and Verger, 2016).

The first research question is the following:

RQ1: What are the motivational factors for lay social media users to participate in risk and crisis communication in the public health context, specifically on vaccination?

The next parameter of online participation is sociality. As Dahlgren (2011) highlights, sociality serves to support (or discourage) participation (p. 97). It refers to a broad range of attributes of communicative interaction: spoken and unspoken ethics governing appropriate behavior, for example, in post sharing, commenting, and in disputes. The nature of social interactions is inextricably intertwined with communication technologies through which they take place (Hutchby, 2001). This is expressed by the concept of affordances (Gibson, 1979). These are a range of preset possibilities for how users may interact with others and the platform (Faraj and Azad, 2012). The specific affordances of online platforms shape the types of sociality that emerge within them. For instance, social media users can leave comments on a shared post, but only if allowed. At the same time, limits placed on user behavior by affordances will lead to users finding creative ways to circumvent them. For example, in the absence of a comments section, users may repost content to add their opinion and thus open a new discussion space. Given how malleable online platforms are to user initiative, researchers have called for more studies into how online sociality affects vaccination communication (Ward et al., 2016). This study proposes that affordances enable and constrain participation in ways that affect users’ motivations (Jung and Sundar, 2020). The second research question aims to explore this connection:

RQ2: What characterizes the sociality of participation on Facebook, and how does it relate to motivations?

The last parameter of online participation is visibility. Social media is a complex digital space offering various ways for communication and interaction. User-defined settings allow individuals to manage the degree of visibility of their profiles, posts, pages, and groups to other users. Facebook has private, semi-public, and public degrees of visibility, in order of ascending accessibility. The private mode is the most restricted form of access, where any content posted will only be available to users with a direct connection to the poster (friends or followers). Semi-public visibility occurs whenever content can be accessed by people outside the user’s connections as long as they are part of the same group or page. Public visibility allows content to be visible for everyone on a social media platform. The degree of visibility is an adjustable setting and may be used as a communication strategy that targets different audiences (Enke and Borchers, 2019). Previous studies reveal a relationship between such self-disclosure choices and users’ motivations (Bazarova and Choi, 2014; Luo and Hancock, 2020) and call for more context-specific research. Therefore, the third question addresses this under-researched issue:

RQ3: What characterizes the visibility of participation on Facebook, and how does it relate to motivations?

3 Methodological considerations

The scientific method chosen for this study is personal interviews, a suitable approach for exploratory research (Swedberg, 2020). It provides a medium through which study participants can describe their perceptions and ideas, offering a deep, contextualized understanding of the experience that cannot be observed directly (for instance, motivations) (Brinkmann and Kvale, 2015; James and Busher, 2012).

The recruitment of research subjects was strategic. First, the study was limited to Facebook, as it is the dominant social media platform for discussions in Sweden, especially those relating to public risks and crises (Eriksson and Olsson, 2016). Second, subjects had to meet the following selection criteria: (1) participation in communication about vaccination on Facebook (moderating theme-specific groups/pages, sharing posts, or leaving comments), (2) clearly expressed attitudes towards vaccination (favorable or critical), and (3) a high degree of communicative activity on Facebook. These criteria were imposed to create a sample providing research data with high information power (Malterud, Siersma, and Guassora, 2016), which yields analytic value and achieves data richness by magnifying the experiences, knowledge, or other properties of a few individuals from the highly specific sample.

The recruitment process started with setting up a tailor-made Facebook account to secure the anonymity of all participants. The second step was to identify through a general search all relevant public and semi-public Swedish Facebook groups (meaning that they are searchable and visible on the platform for non-group members, leaving aside secret and private groups). Some of them welcome users with different views (e. g., Vaccindiskussioner [Vaccine discussions]), while others target users with particular positions or experiences (e. g., Vaccinkriget (V-kriget) [The vaccine war], Vaccinationens baksida [The downside of vaccination]). The most active users from all relevant groups were contacted with an interview invitation with a link to a full study description on a university web page. In order to anticipate potential mistrust, those contacted were informed that the study focuses on social media use rather than on vaccination per se, and that neither Swedish public health authorities nor Facebook had funded the research. All participating informants expressed informed consent to be interviewed and audio-recorded.

The interview grid consisted of several sections and did not contain questions that asked for sensitive personal data as defined by the General Data Protection Regulation (GDPR). The first section focused on overall experiences with communicating about vaccination on Facebook (groups/pages, activity over time, types of participation). Later, based on the mentioned examples of communicative activities, interviewees were asked to elaborate on what motivates their participation. The questions were asked separately about activities in private, semi-public, and public sub-arenas on Facebook. At the end, interviewees were offered an opportunity to talk about their experiences of communicative interactions during their online participation and how these affected their motivations.

All interviews were conducted in Swedish, transcribed in NVivo 12, and anonymized. The transcripts were continuously inductively coded based on the mentioned issues. Later, these codes were clustered together into topics relating to each research question. The collection of interviews continued until the gathered data reached a point of saturation when no new codes emerged from the interviews (Fusch and Ness, 2015; Guest, Bunce, and Johnson, 2006). The data collection was finished when two consecutive interviews did not produce new codes.

Information about interview participants (hereafter, IPs) is presented in Table 1. The average length of the interviews was 41 minutes. They were conducted via Skype or FaceTime (6), telephone (2), face-to-face (2), and Facebook Messenger (1), dependent on the informant’s preferences. The interviews were held in 2018 (5), 2019 (4), and 2020 (2).

All quotes were translated to English by the author and checked for accuracy by another researcher fluent in both languages.

4 Results

5 Discussion and conclusions

The aim of the current study has been to expand the body of knowledge about the multivocality of risk and crisis communication by bringing attention to lay social media users’ participation and its parameters. The study offers a look beyond existing research on behavior and compliance to recommendations regarding vaccinations (Esaiasson, Sohlberg, Ghersetti, and Johansson, 2020; Krishna, 2018), motivations behind vaccination decisions (Dodd et al., 2021; Reiter, Pennell, and Katz, 2020), and the content of vaccination communication on social media (Broniatowski et al., 2018; Schmidt, Zollo, Scala, Betsch, and Quattrociocchi, 2018). Drawing on rich data from a highly specific and, in many aspects, unique sample of Swedish Facebook users, the current study identifies three main motivational factors for online participation: personal interest, information brokerage, and persuasion. Facebook sociality is shown to be characterized by asynchronous communication, loops, and widespread hostility. The results suggest that online participation varies in Facebook sub-arenas with different degrees of content visibility: “zones of peace” and “zones of fight”.

From a societal perspective, online participation may both be an asset or an impediment to strategic RCC by organizations and public authorities. Social media users can support and reinforce as well as criticize and resist institutional actors’ framing. Moreover, this participation is independent from the affected institutions. The study results show that personal interest and experience are essential springboards for laypeople to participate in vaccination communication. This is in line with previous research, which highlights the role of personal dispositions as important predictors of social media use (Orchard, Fullwood, Galbraith, and Morris, 2014).

Some of the interview participants support the institutional framing of the vaccination issue, while others oppose it. Drawing parallels with the conceptual language of corporate communications, these social media users become “faith” and “hate” holders (Johansen, Johansen, and Weckesser, 2016): favorable to or critical of vaccinations, correspondingly. While existing studies stress the participation of medical professionals and political leaders in vaccination communication (Biasio et al., 2016; Zhang, Chughtai, Heywood, and MacIntyre, 2019), they tend to overlook the potential of “faith” holders (social media users supporting vaccination), who become voluntary online ambassadors by supporting, promoting, and advocating for institutional RCC. The challenge for institutional RCC in vaccination communication and public health at large is how to effectively deal with both supporting communication and potential risks coming from critical social media users.

Social media users from both groups become pivotal grassroots intermediaries (Jenkins, 2006) through the (re)distribution of organizational and news media content (turning it into so-called user-distributed content; Villi, 2012) as well as user-generated content. Here, one should consider the two-step flow of communication. It implies that people often learn things through opinion leaders, who selectively share information with others (Katz and Lazarsfeld, 2017). In this case, information brokerage can facilitate or disrupt institutional RCC depending on the accuracy of shared information. The motivation of information brokerage, according to previous research, can be grounded in epistemaphilia, a pleasure not simply in knowing but in exchanging knowledge (Baym, 1998; Jenkins, 2006), in status, or in narcissistic personality traits (Andreassen, Pallesen, and Griffiths, 2017).

Online participation by lay social media users is essential for what scholars call a “public-public” RCC (Fraustino and Fisher Liu, 2017). Information brokerage plays a key role when users produce and share their own content. It touches upon a cornerstone issue of the blurred concept of expertise (Nichols, 2017) in a world where the internet search “renders everyone an instant expert” (Bruni, 2015). The study results show that some social media users consider themselves experts even without formal training in the field and communicate their positions. Indeed, they can be knowledgeable, and their participation can facilitate the outreach of science. However, such activity can also spread misleading and inaccurate information and disrupt RCC. The importance of this issue grows since the credibility of citizen or amateur experts is rising (Baym and Burnett, 2009) at the same time as the significance of institutional expertise in the eyes of the public is declining (Casiday, 2010).

Information brokerage occurs both in zones of peace and in zones of fight. This distinction corresponds, to a certain extent, to Goffman’s (1959) backstage and frontstage framework. In the former case, the participation takes place in private or semi-public sub-arenas on Facebook. It primarily targets users’ networks of contacts and like-minded users to interact freely and avoid constant confrontation. In the latter case, participation happens in public sub-arenas on Facebook, where users communicate in front of others. Previous research showed that, in the case of vaccination, social media users are segregated in polarized groups (Schmidt et al., 2018; Yuan, Schuchard, and Crooks, 2019). The current study suggests that, although the segregation of social media users with different attitudes towards vaccination is prominent in zones of peace, the communication between users with opposing views increasingly occurs in public sub-arenas, zones of fight.

Participation in public sub-arenas provides social media users with opportunities for persuading a general audience. The access to publicly visible communication offered by social media strengthens the motivation of social media users. The present study shows that the primary target of lay RCC participants are the undecided audiences without strong a priori issue attitudes. A similar focus has been suggested even for institutional RCC on vaccination (Betsch, Korn, and Holtmann, 2015). Persuasion can occur directly through open discussion, but even the “presence” of alternative views can be a tool of persuasion. Some studies find that exposure to information contradicting a user’s initial position can affect their vaccination decisions (e. g., Jolley and Douglas, 2014; Kim, Han, and Seo, 2020; Neubaum and Krämer, 2017). However, other studies do not support this correlation (e. g., Dixon, 2021; Schulz and Hartung, 2021). To date, there is no conclusive evidence that the “presence” of alternative views can lead to opinion changes.

Persuasion can also occur through a distortion of the perceived opinion climate. The study results show that vaccine-critical social media users aim for an extensive presence of alternative views in public sub-arenas to challenge the one-sided mainstream vaccine discourse and normalize discussions about vaccination risks. The increased visibility (in zones of fight) of ideas that are considered deviant and confined to a small minority can, according to Hallin (1986), alter the broader discursive issue framing. In this case, these ideas will be perceived instead as a legitimate controversy, open to questioning in mainstream discourse. From there, they can even become the new general consensus.

The solid motivation for separate zones of peace from zones of fight revealed in the study also aligns with Goffman’s (1959) framework. Recent studies that revisit Goffman’s concepts in a digital environment stress the dangers of “backstage mimicry” in public spaces, when feelings of anonymity and invisibility can lead to an increase in online hate, creating “the virtual stages of hate” (Kilvington, 2021). The present study adds that widespread hostility can both strengthen and erode motivations for online participation. However, despite their positions on vaccination, social media users seek more meaningful, genuine dialogue, thus breaking the vicious cycle of hate.

The limited sample of the present study does not allow for a generalization of the results. However, it does provide a schematic road map for factors that motivate online participation in RCC. Now that the broad contours of these motivations have been identified, further study is needed to ascertain the spread of these factors among social media users in a precise, quantified manner. Further study could also address the conditions under which forms of sociality and levels of visibility on social media undermine or strengthen motivations. The present study is also limited to the issue of vaccination. It is possible to expand the analysis of online participation in RCC in other timely public health issues, such as antibiotic resistance or obesity, and validate the results cross-nationally.