Introduction: stigma and avoidance in health crises

Although the WHO (2020) issued warnings about stigmatization as early as February 2020, health policy makers largely ignored practices of social exclusion in the COVID-19 pandemic (Dionne and Turkmen 2020; Inman et al. 2021). Reports of discrimination, scapegoating and violence against marginalized communities show that people directed blame and avoidance at immigrants, particularly those of ascribed Asian descent (Dionne and Turkmen 2020; Joshi and Swarnakar 2021). Social scientists argue that a pandemic ‘can intensify the social boundaries between identity groups, with people trying to distance themselves and the identity groups with which they identify from those perceived to be unhealthy through stereotyping already stigmatized groups’ (Dionne and Turkmen 2020, p. 214). In many countries, healthcare workers were stigmatized early in the pandemic based on fear of infection (Bagcchi 2020). Research from the United States and Canada found a similar pattern, noting that individuals in specific occupations, in particular healthcare workers, were ostracized during the COVID-19 pandemic (Taylor et al. 2020). The paper aims to investigate stigmatizing practices in Austria and Germany: Who is intentionally avoiding allegedly contagious groups of people and what are the reasons for avoidant practices? How do people make sense of their experiences of being excluded and negotiate the legitimacy of preventive practices in their social context?

The process of othering often manifests in practices of social stigmatization and exclusion (Joshi and Swarnakar 2021). As the history of pandemics has repeatedly shown, avoiding 'the sick' can quickly turn into eliminating them (Nelkin and Gilman 1988). Blaming social groups and their behaviour for the spread of contagious disease has a long history and can offer individuals a sense of control when they cannot yet contain their anxiety by relying on an accurate medical explanation (e.g., in the AIDS or syphilis pandemic, cf. Nelkin and Gilman 1988). According to Dionne and Turkmen (2020, p. 214), this situation led ‘to a redistribution of blame according to existing patterns of discrimination and othering’ and ‘the exacerbation of existing tensions and inequalities between and within groups’ in the COVID-19 pandemic (ibid., p. 222).

Contemporary analysis of pandemic discrimination often relies on the social theory of stigma. It has been criticised for its overemphasis on the individual and the cognitive construction of stigmatizing labels in micro-level interactions that are reconstructed from the point of view of a disengaged researcher (Link and Phelan 2001). To address methodological issues of research on pandemic stigma, the paper applies social practice theory to the study of pandemic health crises. Drawing on Pierre Bourdieu’s social theory of classification struggles, I conceptualise schemes of pandemic othering in terms of Negative Health Classifications. Classifications, which are embodied by individuals and are part of structures of domination, denote a conceptual link between the symbolic order (i.e., cultural meaning and values) and the social order (structure and social hierarchies) of societies (Völker and Trinkhaus 2014). The practice of classifying is a dynamic part of social struggles visible particularly in times of social crises even though it may remain relatively stable in periods of social hegemony. During pandemic crises, classification struggles require more scholarly attention because fights over the best strategies for pandemic mitigation or the ‘legitimate’ health behaviour might cause political polarization (Ares et al. 2021).

The paper presents the first detailed analysis of negative health classifications in Austria and Germany, two countries that had comparable strategies to tackle the health crises and implemented nationwide shutdowns and social distancing to prevent the spread of COVID-19 in 2020/21. The text is structured as follows: I first present a theoretical framework of classification struggles and conceptualise how negative health classifications are intertwined with pandemic mitigation practices. In the methods section, I describe the forms of analysis used (quantitative and text-based). I then report the results, a) new and traditional social cleavages created and exacerbated by individuals who strive to tackle medical crises by avoiding the 'disease carriers', and b) the polarization of social groups through practices of social exclusion. In the end, I discuss the findings, and conclude by noting the limitations and implications of the study.

Theoretical framework: classification struggles during a pandemic

The conceptual framework of the paper theorizes classification struggles during the COVID-19 pandemic by integrating the sociology of health with Pierre Bourdieu’s theory of classification. Influenced by philosophers such as Gaston Bachelard (1978), Ernst Cassirer (1994), Ludwig Wittgenstein (1984) and Maurice Merleau-Ponty (1976), Bourdieu’s epistemological approach is characterised by a relational logic that integrates subjective and objective dimensions of social phenomena as well as the individual and social structure (Bourdieu 1990; 2000). According to Pierre Bourdieu's social theory of social distinction, classifications are principles of division that function as ‘marks of distinction or stigma’ (Bourdieu 1982, p. 752, own translation). In Classification Struggles, his most recently published lectures on the topic at the Collège de France (1981–1982) that deepen his theoretical ideas of symbolic struggles in democratic societies that he developed in Distinction (1982) and The Logic of Practice (1990), Bourdieu (2018) reminds sociologists that classifying the world is one of the most basic social activities. In contrast to the anthropological analysis of collective mental structures (e.g., in the work of Émile Durkheim, Marcel Mauss and Claude Lévi-Strauss, Völker and Trinkhaus 2014), classifications of the self vis-à-vis selected Others can be viewed as practical classifications, such as insults and official nominations (Bourdieu 2018), in everyday life. As discussed by philosophers of science, social agents (unlike natural objects) interact with their classification (Hacking 2002). Confronted with the need for stigma management (Goffman 1963), individuals, for example, face the risk of social exclusion and respond by acting on their discrimination in a variety of ways. Complementing the interactionist approach to stigmatization, Bourdieu (1990) asserts that classifications are practices that relate the individual to the social order and are thus closely linked to social status, inequality and class struggle. In that respect, he emphasised the symbolic power of authorities to performatively enact classifications, particularly the symbolic power of the state and social scientists to institutionalize social classifications in society (Bourdieu 1985). Aadopting a phrase from Max Weber, he noted the state’s ‘monopoly of legitimate symbolic violence’ (Bourdieu 2018, p. 125). Bourdieu introduced the concept of 'symbolic violence' to analyse instances in which agents who are addressed by stigmatizing classifications internalise this misrecognition and subsequently apply negative classifications to themselves. However, social agents might also challenge and even oppose this monopoly (idib., p. 126), resulting in social classification struggles. Such a struggle, I argue, was taking place during the early phase of the COVID-19 pandemic.

Academic knowledge is also entangled in struggles over the production and reproduction of legitimate classifications, as are sociologists and their ‘objective’ classifications (Bourdieu 1988, 1992, 2004, 2018; Bourdieu et al. 1991). According to Bourdieu, reflexive sociology is a ‘science of classification’ (Bourdieu 2018, p. 70) that is characterised by its own engagement in classification struggles–an intellectual practice that I argue becomes even more pronounced in times of crisis. The COVID-19 pandemic was an important moment for public expert knowledge and while sociology was not among the disciplines that were most relevant to tackling the health crisis in the early days of the pandemic, sociological research gained attention as soon as the unanticipated negative consequences of health policies became obvious. However, in the case of Austria and Germany, the state failed to support academics and health professionals at the peak of the pandemic, resulting in rather common experiences of hostility against cautious public health experts and professionals (Whittle 2022). During pandemic crisis, it is challenging to reflect on public intellectual’s role in legitimizing (or not) classifications within and beyond academic networks while simultaneously ‘classifying the classifiers’ (Bourdieu 2018; Averett 2021). It is not a surprise, then, that literature on new emerging classifications in the pandemic is scarce (Economou 2021), with most of these focussing on the role that COVID-19 news and the media play in polarizing public opinion (Joshi and Swarnakar 2021; Dan and Dixon 2021; Chu et al. 2021). The underlying classification struggles have not yet been studied in detail.

Background: struggles over health precautions in Austria and Germany

In Germany and Austria, early pandemic mitigation policies and public academic disagreement inspired everyday practices of opposition that gradually polarized social relationships. Two lockdowns in which non-essential stores and services were closed were enforced in both countries for several weeks starting in March 2020 and again in November 2020. People had to adhere to strict contact restrictions and were obliged to wear a mask in public space. Official campaigns promoted the requirement of social distancing, e.g., by telling people to stay the distance of ‘a baby elephant apart’ (a one-meter distance rule for people living in different households). In both countries, ICU bed occupancy and hospitalization rates were core criteria for lockdown measures. Testing and vaccinations were not generally available until 2021, so only then were many of the preventive measures slowly withdrawn. During that time, opponents of COVID-19 health policies were reframing health risks in COVID-19 related 'fake-news', refusing to get vaccinated, and allying with right-wing forces, a problem that many countries were facing (Rooke 2021). A vivid scene of pandemic sceptics developed that threatened both researchers and health professionals on a regular basis (Whittle 2022). Similar to other countries, Austrian and German political authorities failed to accumulate enough symbolic power to control the national narrative of pandemic mitigation (Alexander and Smith 2020). As a result, anxiety, insecurity, a diversity of risk perceptions and fake news spread among the population (Copping 2022). People could not even agree on whether and how the most vulnerable should be protected. In Austria, a qualitative study on compliance found that people evaluated public health measures according to their complex legal, moral and social legitimacy in their everyday lives (Spahl et al. 2022). For example, people assessed everyday practices such as meeting friends in terms of collective responsibility and irresponsibility. They justified breaking the rules with good reasons—or were surprised by their own internalized overcompliance.

Applying Bourdieu’s social theory to pandemic health, Schneider-Kamp (2021) assumes that social distancing might affect the social and cultural capital necessary to preserve health and well-being. Snel et al. (2022) have recently shown how, indeed, social capital protects individuals against the mental health impact of the COVID-19 pandemic. For the current study, I hypothesize that social capital is affected by pandemic classification struggles (Hypothesis 1). Departing from the observation of everyday social conflicts over public health legitimacy in the Austrian and German society (Spahl et al. 2022), I assume that conflict over health precautions is associated with negative health classifications (Hypothesis 2). Prior research found that fear, risk perception and worries for the health of people with a chronic disease are associated with a higher compliance with health precautions (Taylor et al. 2020). I thus hypothesize that anxiety and subjective health concerns impact negative health classifications (Hypothesis 3). Drawing on early COVID-19 research (Taylor et al. 2020), I anticipate that the adoption of distancing norms is associated with intentional avoidance of groups that are perceived as carrying a high infection risk (Hypothesis 4). Vulnerable populations such as people with COPD were affected by anxieties and isolation (Yoeli 2021). In Germany, social participation of individuals with a high risk for a severe infection decreased until they were offered vaccination (Schröder et al. 2023). I thus assume that relative to individuals without health risks, those who belong to a high-risk group due to health conditions are more likely to experience negative health classifications (Hypothesis 5).

Methods and data

This article draws on empirical data from an exploratory study on the transformation of social relationships during the pandemic, which was conducted in Austria and Germany from March 2020 through November 2021. When analysing data of a preliminary survey conducted in April 2020, the research team encountered new stigmatizing classifications during the pandemic. The pilot survey provided initial evidence of a loss of social relationships: social exclusion appeared to be linked to health classifications, that is, either anxiety about catching the virus or disagreement over pandemic mitigation strategies. A second survey with refined questions was then developed for people age 14 and older and conducted during the second nationwide shutdown from 9 November to 9 December, 2020 (see the appendix for detailed wording and translations of selected items into English). After respondents provided informed consent, the survey collected anonymous data on experiences of avoidant behaviour, contact termination and social exclusion. The study adhered to the ethical principles of the Declaration of Helsinki and was approved by the Research Ethics Commission of the author’s university. Participation in the study was voluntary. The survey was implemented in SoSci and the public survey link was distributed via major daily newspapers and radio stations in Austria and Germany. Table 1 shows the sample demographics that differ from the general population in the following respects: Compared to official statistics and despite its wide public circulation, women, young respondents and respondents with higher education were overrepresented in the study.

Table 1 Characteristics of the quantitative sample (%)
Full size table

In this paper, I analyse the empirical data on negative health classifications in two steps: a) quantitative analysis of survey data using SPSS and b) qualitative analysis of text-based survey data.

a) In addition to presenting descriptive statistics on the avoidance of allegedly contagious groups, the perception of being avoided due to belonging to a high-risk group, and experiences of social exclusion in the pandemic, I conduct binary logistic regression models to examine how negative health classifications differ by respondents’ sociodemographic characteristics (gender, age, education, grandparents in the same household, occupational status) and pandemic-specific characteristics (risk groups, health concerns, compliance, pandemic conflicts).

Dependent variable

To assess negative health classifications, I asked respondents: 'Do you avoid specific groups of people because you are afraid to contract COVID-19 from them?' Answers were recoded: no = 0, sometimes + often = 1. ‘If you avoid specific groups: Which and why?’ (Open text).

Independent variables

Five covariates are included in the regression analysis: belonging to a high-risk group (chronic disease (COVID-19 related precondition) or diagnosed mental illness (self-reported)), subjective health concerns, adoption of distancing norms, pandemic social conflicts (experiences of exclusion and conflict avoidance). To assess participants' internalized sense of social distancing, I asked, ‘Have you developed a new sense of distance between people, for example, an inner alarm system when people get closer than one meter to you?’ Experiences of social exclusion were measured using the question, 'Have you experienced exclusion because of your attitude toward pandemic mitigation? If yes, in what respect?' To assess pandemic conflict avoidance, participants were asked to respond to the statement 'During the pandemic, have you ever allowed more physical closeness than you actually wanted in order to avoid a conflict (e.g., a hug as a greeting, too much closeness in the exercise of your job, …)?' For measures and question wording, see the appendix.

In respect of loss of social capital, respondents were asked, 'Did you lose contact with important confidants during the pandemic?' and 'Have you intentionally ended relationships with friends, family members, partners or colleagues because of the pandemic?' In each case, respondents were asked to describe their experiences in more detail.

b) Thematic analysis is a well-established method of theorizing empirical data that is often used with the aim of complementing quantitative findings with a rich account of qualitative data (Braun et al. 2021; Nowell et al. 2017). In methodological respect, it has to be noted that classifications are also created during coding in which the complexity of the qualitative data is compressed, yet ideally not just into any classes, but classes that are also differentiated by social agents (Bourdieu 2018, p. 24). To do this, however, one must not apply formal codes repetitively to any given set of data, but must proceed reflexively to discover the criteria or schemes that divide the social world (ibid.). In this study and based on inductive, that is, data-driven coding of participants’ responses to open-ended questions, I identified emerging patterns which represent shared schemes of social division and classification. First, to investigate which groups avoidant respondents consider 'risky' and consequently classify as contagious, I analysed text-based explanations of which groups they believe are particularly infectious and why. Second, I examined respondents’ comments about whether they have experienced being avoided themselves, and if so, what their experience was like. Third, I analysed experiences of social exclusion after disagreement over pandemic prevention strategies. After generating initial codes for each topic, I developed a coding framework that was based on inductive coding. In this process, I worked through the entire data set and refined the provisional codes. After that, all statements were coded and, eventually, after revision of subthemes, categorized into the observed coherent themes of negative health classifications (Braun et al. 2021). Typical example statements were then selected to illustrate the detailed analysis of the themes in this paper.

Empirical findings: struggles over negative health classifications

Practical insults: avoiding the ‘contagious'

Negative health classifications were widespread during the second COVID-19 shutdown in 2020. One of the most prominent schemes was classifying others as 'contagious'. Among the study participants, 13% reported avoiding contact with certain populations 'frequently', and another 28% reported doing so 'sometimes' because they were afraid to contract COVID-19 from them.

Who is intentionally avoiding allegedly contagious groups of people? To answer that question, I present the results of the binary logistic regression model (χ2 = 233.52, p < 0.001, R2 = 0.22) for avoidance of social groups during the pandemic (see Table 2). The strongest covariate of avoidant behaviour is an internalised sense of social distancing. People who react to physical closeness of strangers with an alarmist feeling have significantly higher odd of avoidance than other respondents, supporting Hypothesis 4 that negative health classifications are associated with the internalisation of distancing norms. Consistent with Hypothesis 3, subjective health concerns and anxieties significantly increased the likelihood of avoidance. Respondents with a high score on the scale of worries about their own health have higher odd of avoidance. Other statistically significant covariates of stigmatizing behaviour are self-reported experiences of social exclusion due to respondents' attitude towards the pandemic and conflict avoidance when someone is not complying with distancing requirements, both supporting Hypothesis 2 on the implications of pandemic conflict on negative health classifications. Being unemployed and being retired are significantly associated with higher odds of avoiding allegedly ‘contagious’ groups. Prior research revealed that pandemic policy measures had a negative impact on unemployment (Morris et al. 2023) and in Germany, negative classifications of unemployment were associated with the health and well-being of those who are disadvantaged on the labour market (Krug et al. 2019). Also retired individuals might belong to a risk group. Interestingly, neither belonging to a high-risk group nor worrying about other people's health were significantly associated with avoidant behaviour. Hypothesis 5 has thus been rejected. Studies in public health found an inconsistent association of age and cautious health behaviour in the pandemic (Choi et al. 2022). Thus, further studies need to be undertaken to understand how unemployment and retirement are associated with negative health classifications in the COVID-19 pandemic. Except for the predictions for those in health-based high-risk groups, the data support the theoretical assumptions.

Table 2 Results of the binary logistic regression of avoidance on selected independent variables. B (SE) and odds ratios
Full size table

Of 621 participants who avoided contact at least sometimes, 480 elaborated on which groups and the reasons for their avoidant behaviour in a text-based answer. I identified at least nine groups that respondents classified 'contagious', including children, health professionals, homeless people, and people violating distancing regulations. Table 3 offers an overview of these groups and examples of the explanations respondents gave for avoiding them. All answers were translated from German to English by the author.

Table 3 Negative health classifications of ‘contagious’ groups (Coding of text-based answers, only respondents who indicated yes)
Full size table

Respondents generally viewed non-compliance, such as not wearing a mask or not keeping enough distance, as a symbolic marker of infection risk. Avoiding a specific individual who does not wear a mask is different from avoiding an entire social group despite concrete precautions and behaviour adopted by members of that group. While some respondents reported personally observing non-compliance, most attributed non-compliance to all members of certain groups. For example, respondents commented, ‘Immigrants (…) do not seem to fully understand’ prevention concepts or ‘The disadvantaged (…) do not follow the rules’. From a Bourdieusian point of view, these symbolic insults of being collectively classified as non-compliant and contagious mostly exacerbated existing social inequalities.

However, and in contrast to the common assumption in stigma research that pandemic discrimination is primarily directed against long-term minoritized populations or infected individuals, my research reveals widespread under-recognized formations of new social hierarchies in public health crisis (Joshi and Swarnaker 2021). At least three groups emerged during the thematic analysis that were not previously long-term minoritized groups: (1) individuals who have a hedonistic or extremely social lifestyle that involves partying and meeting a larger network of friends, (2) young people and people who are in contact with children and adolescents (like parents and teachers), and (3) individuals who refuse vaccination and do not comply with public health prevention measures. At least at the beginning of the pandemic, the state sanctioned non-compliant individuals and thus authorised their negative classification.

How did people perceive being classified and avoided by others? 13% of the study participants noticed avoidant behaviour and reported that they thought people stay away from them because they belong to a high-risk group or profession. Of the 190 people who perceived being avoided, 132 elaborated on their experiences of social exclusion in a text-based answer. Most of these answers were from those in professions that have frequent contact to patients or clients, such as medical personal, teachers, police officers, social workers, and similar occupational groups who noticed avoidant behaviour from friends, family, colleagues, and even healthcare providers. The most common experiences were that people no longer wanted to meet them or be in the same room with them and/or did not invite them to meetings. A respondent who worked in the healthcare sector reported that he was publicly named and stigmatized as someone who might spread the virus. Parents and teachers reported similar experiences involving dehumanising terminology. One woman, for example, said she received the ‘nickname: virus (because I work with kids)’ and remarked that people avoided her, were not sitting at the same table, and bullied her (28, FFootnote 1). A sex worker (40, F) felt that the prohibition on sex workFootnote 2 stigmatized her as an ‘epidemiologically particularly dangerous super-spreader’. These examples illustrate that avoidant behaviour is not limited to healthcare workers as assumed by previous literature on the topic. Rather, individuals from a range of professions reported experiencing stigmatization as members of a contagious group.

The findings provide evidence that respondents not only exhibit avoidant behaviour, but also notice—and sometimes even approve of—negative health classifications (cf. Zolnikov and Furio 2021). For example, a psychotherapist (30, F) wrote that it ‘was communicated to me directly and appreciatively and I find it completely appropriate’ that ‘some friends do not want to meet me in person, because I represent a greater risk than people who have been working only in their home office for a long time’. Often resulting in voluntary self-exclusion (see below), this internalization and individual responsibilization of risk management can be interpreted as one form of symbolic violence.

Interestingly, and contrary to the public distancing discourse that claimed to protect the most vulnerable, only 20 individuals who described being avoided belonged to a health-based high-risk group. One respondent (59, F) noted that people told her she 'already belongs to a risk group' and avoided her due to her age—even though she did not consider herself to belong to a COVID-specific risk group. The results illustrate the fundamental disagreement over who is at risk and who should be avoided to reduce the risk of contracting the disease.

Illegitimate pandemic practices: sanctioning the ‘irresponsible’ and those who are considered ‘too anxious’

How do people negotiate the legitimacy of preventive practices in their social network? In addition to the theme of risk evaluation, the study found that classification struggles indicate which pandemic behaviour is considered illegitimate and how that evaluation affects social capital.

During the second lockdown, people observed—sometimes anxiously—that tensions had increased among the population and/or in their social network. One respondent explained, ‘The population is very divided because of these enormous fears, and some people's emotions boil up more quickly’ (32, F). Among respondents, 26% indicated they were isolated by others in their social network based on their attitude towards pandemic policies. Of the 389 respondents who reported experiencing exclusion, 304 described their experiences in the open-ended answer section in more detail. Most related their exclusion to disagreement over compliance and distancing during the pandemic. Respondents reported that the ‘situation tends to be polarized (pro- / anti-government action) and people try to force any differentiated opinion into this classification’ (32, M), resulting in the application of new classifications in their social relationships: Specifically, the pandemic Other was classified as either too careless or too anxious. As shown in the analysis of avoidant behaviour, respondents classified non-compliant individuals as 'irresponsible' and a health risk. It is not surprising, then, that respondents who were critical of pandemic prevention policies and expressed a ‘different opinion’ reported experiencing social exclusion. They were, for example, ‘being instantly labelled a Corona denier’ (42, M) or ‘conspiracy theorist’ (48, F)—classifications they rejected. Instead, they classified themselves as being ‘critical’ and ‘silenced’ by overly anxious people (49, M). A mother, for example, who expressed ‘different opinions’ about children wearing masks at school in a parents' WhatsApp group for her daughter's class was criticized for transmitting ‘Querdenker’ newsFootnote 3—an accusation she perceived as a ‘hysterical reaction’ (37, F). These study participants described experiencing breaks in contact because people were not willing to listen to their ‘critical thoughts’ on pandemic policies and political rhetoric.

Indeed, cautious respondents recounted ending contact with others (temporarily or permanently) for exactly these reasons. For example, one commented: ‘An acquaintance now produces pro conspiracy theory videos. That's why I broke off contact.’ (42, F) Participants reported that they tried to limit contact, or at least avoid talking to 'critical' friends about the pandemic.

Given the analysis of negative health classifications, it is somewhat surprising that the most frequent and intense experiences of social exclusion were reported not by non-compliant individuals but rather by cautious individuals. Only 15% of the respondents who reported a loss of social contact experienced exclusion due to a 'critical' attitude, while 54% felt excluded because of their compliance with distancing measures. Respondents who were afraid to contract COVID-19 and who asked friends, family and colleagues to conform to pandemic regulations sometimes experienced harsh social sanctions in their social networks, including disputes, ridicule, derogatory comments ‘intended to be funny’, hostility, social isolation and breaking off contact, name-calling, and physical attacks in public space. One woman, for example, said she ‘was verbally attacked’, with the other person claiming ‘that the pandemic was harmless and just a power play of politics and that I should not abide by the restrictions. I was spat at on the tram when I politely asked a young woman to use a mask’ (53, F).

A variety of negative classifications were applied to cautious individuals, among them popularised diagnosis of mental disorders. Participants reported that they had been ‘judged as excessive or hysterical’ (62, F) and 'a hypochondriac' (37, F) or ‘crazy’ (25, N-B) because of their cautious attitudes. Respondents also recalled experiencing contact termination in rural areas where distancing measures were largely ignored as well as social exclusion at the workplace, such as being excluded from conversations. Some study participants even discussed being excluded in terms of voluntary ‘self-exclusion’ (31, F) (for example, from meetings) if they followed pandemic measures more strictly than others in their social network, or if they were working at schools or hospitals and were afraid to spread the virus to friends and family. ‘I don't want to be in the position of having to justify myself for careless behaviour in the event of a possible Covid infection (of others, the author). I don't want to have to justify it to myself either’, stated a respondent (37, F).

The findings suggest that social sanctions pressure agents into adopting practices that are deemed legitimate within a specific social group and context, illustrating the struggles over the social worth of critique of public health policies among the population and the impact that policy compliance might have for their social status (Dionne and Turkmen 2020; Inman et al. 2021). During the peak of the pandemic, social agents accepted or actively enforced the break of social bonds if they disagreed over the right prevention practices. Of the respondents who experienced exclusion due to their policy-related attitude, one-quarter lost contact with an important confidant and one in seven had deliberately ended social relationships (see Fig. 1). Consistent with Hypothesis 1, this type of rearrangement of social ties was significantly more likely among people who experienced pandemic-related exclusion in their networks than among people who did not experience disagreement over mitigation policies [lost contact: χ2(2) = 33.51, p < 0.001; termination of contact: χ2(2) = 17.28, p < 0.001].

Fig. 1
figure 1

Loss of social capital depending on pandemic-related experiences of exclusion (n = 1482)

Full size image

Even if conflicts have been one important driver of a loss of social capital in the pandemic, it is important to take the complexity of social relationships into account. Subjective explanations of a loss of contact to close friends and family range from the inability to meet in person and engage in collective leisure activities to having grown apart during the pandemic. To some, relationships of trust were already strained before the pandemic which led to an accelerated readiness to break off contact as soon as disagreement over the legitimate ‘responsible’ practices became apparent.

These examples illustrate what Bourdieu calls the ‘classification work in everyday life’ (Bourdieu 1985, p. 19; translation by the author), asserting that agent’s symbolic strategies to enforce their world view are limited in their power as long as they remain on the level of insults and abusive language and are not authorised as the legitimate position of the representatives of the state. Negative classifications might still establish a feeling of group membership that result in a significant distance to other social groups.

Discussion and conclusion

In everyday practices of engaging with and evaluating other people during the COVID-19 pandemic, the emerging classifications are structured by the binary scheme of either ‘critical’ or ‘hysterical’, based on ascribed non-/compliance. These new health classifications form the foundation of social proximity and distance in pandemic times and have practical implications for health policy. Early anti-stigma programs are important for challenging stereotypes (WHO 2020) and addressing social exclusion during the pandemic before discrimination feeds into overall polarization. However, interventions based on pandemic stigma research might be limited because its primary level of analysis is focused on the individual and their use of stigmatizing labels in interactions that are analysed by a disengaged researcher. Drawing on practice theory, the exploratory study aims to contribute to the development of a systematic research agenda that includes negative classifications and classification struggles in the analysis of health policy implications in the future. In Bourdieu’s work, classification struggles are first and foremost a social practice characterised by uncertainty and crisis. When social structures are unstable, as was the case during the pandemic, the established common sense can become questionable (Bourdieu 1985, 2018). Agent’s classifying practices are thus entangled with social transformation (Bourdieu 1988). As Bourdieu (2019) has reminded us, classifications are the result of social struggles that are influenced by the state’s power to institutionalise (social and academic) classifications. During the pandemic, the empirical data show that negative health classifications were applied not only to non-compliant critics of pandemic regulations but also to cautious and compliant individuals, reflecting the conflicts over the right way to deal with the health crisis on a political level. Further research is needed to investigate why behaviour change to reduce disease transmission has become the ultimate fulcrum of new negative health classifications. A possible explanation might be that in Austria and Germany, the state did not sufficiently support compliant individuals and professionals at the height of the pandemic. The analysed examples of rejection of negative classifications such as belonging to a ‘risk group’, being a ‘Corona denier’ or acting ‘hysterical’ show how their application is only of temporary legitimacy—and how agents challenge established classifications.

Furthermore, the findings point to the fact that the analysed classification struggles affect the social capital and social status of groups who hold different attitudes on health policies.

Public health professionals and many members of the public regard social distancing as one of the most important pandemic mitigation strategies. Yet distancing can easily turn into avoiding certain groups of people, with negative classifications sometimes transforming social distance into social exclusion. Essential workers were not only burdened by increased workloads, but also suffered from social isolation during the pandemic. The study's findings show that the participants in the study, mostly women with high educational qualifications, consider many social groups—children, young people, the homeless, addicted people, people with mental health problems, immigrants, men, rich people, poor people—more contagious than others, irrespective of individuals’ compliance with public health measures such as wearing a mask, testing, or vaccination status. It illustrates the reproduction of intersectional inequalities during the pandemic. Cautious individuals who internalised social distancing requirements were more likely to ensure their safety by intentionally avoiding specific groups than other people, even more so if they have subjective health concerns and anxieties. That result aligns with findings from prior survey research (Taylor et al. 2020) that found that people who have high levels of pandemic stress and suffer from excessive anxieties during the pandemic avoid healthcare workers.

Being ‘critical’ or ‘cautious’ during a public health crisis is thus a practical and affective bodily process that is approving social divisions as directly legitimate. In that respect, the practice theoretical approach of this paper goes beyond stigma research by contributing to an understanding of the social dynamic of pandemic exclusion in its larger context. Even though data from a non-probability sample offer limited insight into classification struggles in society and can not be used for statistical generalisation, non-representativity is less problematic for the analysis of open answers. They offer initial evidence of widespread negative classifications emerging during the COVID-19 pandemic that future comparative research should explore in their implications for inequalities in other policy contexts.

Bourdieu's theory is also suitable for expanding stigma research in another respect. In pandemic societies, experts are engaging in struggles over the legitimate worldview more actively and with more symbolic power than they did before the pandemic. Against the background of reflexive sociology, pandemic-specific schemes of perception, thinking and evaluation are found among both social agents and academic agents: New health classifications are mobilized within the contemporary field of forces, pressuring social groups to position themselves on one or the other end of the spectrum. That requirement extends to sociologists, psychologists, public health experts and intellectuals in general. Classifications are not abstract concepts. Rather, they mobilize social groups. In the future, such precarious involvement in classification struggles must be considered more systematically in the social theory of health–not only to provide additional interesting insights into the social dynamics of health classification, but also to take seriously the epistemic responsibility for academic classifications and their reception during a time of health crisis. Understanding the complex interplay between authorised and everyday practices of health classification in pandemic societies is an important prerequisite for successfully tackling public health crises.