Moral identity and morality-as-cooperation

As it was previously mentioned, morality in the era of the pandemic has been prominent and is worth of further empirical investigation. In our multi-country study, we conceptualized morality as an individual trait, as well as a cooperative group-relevant variable.

To begin with, moral identity, reflecting the individualized conceptualization of morality, refers to the subjective importance of morality to someone’s personal identity, typically represented as a trait-based difference between people or as a trait in the form of self-schema [3]. Moral identity is composed by a private “having” component (internalization) and a public “doing” component (symbolization). Internalization concerns the degree of centrality of moral traits for someone’s self-concept, while symbolization has more to do with the degree of reflection of these moral traits in public action and choices taking place in certain social settings [4]. From a motivational perspective, internalization relates to seeking for maintaining self-consistency between identity and behavior. On the other hand, symbolization is motivated by goals of self-presentation and recognition of the self in terms of a social entity that seeks for self-affirming feedback from other people [5]. To date, research mainly focuses on the internalization aspect, calling for more empirical attention to the symbolization–“doing”–aspect of moral identity [5,6].

Concluding, symbolization and internalization act as synergetic in people’s moral reasoning and decision making. Additionally, Colby and Damon [7] suggested that what differentiates moral people from other people is the extent to which they perceive their own sense of morality and personal goals as united. Extending this argument, Frimer and Walker [8] proposed a reconciliation model, which relies on moral identity and describes moral identity as entailing agency-related motives (e.g. self-interest) and communion-related motives (e.g. other-oriented morality, cooperation, prosocial behavior). Such motives are supposed to act as synergetic and not as competing and, eventually, the integration of personal and communion moral goals provide a powerful source of commitment to action.

Morality-as-cooperation, in this context, is exemplified in community-based and societal challenges, where moral actions are needed. Morality-as-cooperation, indeed, identifies problems and societal challenges, which can be solved via exhibiting cooperative behavior, such as reciprocation or help towards the group [9]. In light of mutualism as a prerequisite for morality-as-cooperation behaviors [10], there are situations, which beg for cooperative behaviors and where individuals benefit each other by jointly and cooperatively working with others. Coordinating to achieve a mutual advantageous outcome requests joint coalitions and efforts to compete with threats and uncertainties [11].

Conspiracy theories and moral concerns

During the pandemic, adherence to health protective behaviors and compliance with public health policies has become a morally expected issue with relevant moral concerns, when public health protection rules are violated [12]. Prioritizing public health and responding to public health risks also raised multiple moral dilemmas that have become the subject of extensive debates [13]. To this end, we aimed to consider morality-relevant variables in empirical investigation of the topical debate on conspiracy theories that have flourished during the pandemic and the degree to which morality variables mediate the relationship between endorsement of COVID-19 conspiracy theories and adherence to public health behaviors and support for equivalent policies.

Conspiracy theories share a common theme, in the way that they rely on negative and traumatic events conceptualized and explained as intentional and plotted plans from powerful elites against populations [14]. These “hidden plans” often entail a morality component. Indeed, Leone et al. [15] posit that conspiracy theories entail a moral evaluation component, on the grounds that traumatic and frightful events originate from malevolent elites [16] and, thus, trigger a moral reasoning relevant to the actual understanding of conspiracy theories [17]. Farinelli [18] has described conspiracy theories as “morality tales based on archetypal narratives about right versus wrong, good versus evil” (p.4). This conceptualization of conspiracy theories can reflect the degree to which people, who are attentive to morality issues may also believe in conspiracy theories which, in turn, appeal to people’s moral perceptions, values and judgements.

Moral concerns may be relevant to people’s individual and also collective well-being, referring to individual and group-foundations respectively [19]. Moral foundations concerning people’s group membership relate more strongly with endorsement of conspiracy beliefs than moral foundations concerning people’s individual moral traits and judgements [15]. This claim lies on the nature of conspiracy beliefs per se as a reflection of an “us-versus-them” mentality. This mentality is often anchored in events against nations, religious, and other large groups as victims of such conspiracy theories [20] that are supposed to attack the populations. Leone et al. [15] have suggested that the group perspective of conspiracy theories activates a group identity perspective when people think in terms of their binding to the group and not in terms of individual binding. However, following the aforementioned conceptualizations of moral identity and morality-as-cooperation, especially in light of the Reconciliation Model, as dimensions that can complement each other, we hypothesize that there will be a significant positive association between both morality-as-cooperation and moral identity with conspiracy beliefs about COVID-19 (H1a). The threats that are entailed in conspiracy theories are supposed to trigger a defensive way of life on the part of the people, reflected in their moral principles of self-responsibility and cooperation to overcome the impact of such threats. This explains the expected positive direction of the expected association.

Morality, beliefs and conspiracy theories and health protective behaviors

While several conspiracy-based explanations of various events can fascinate people across time and be harmless [21], most conspiracy theories relevant to public health crises can be dangerous and harmful for individual and societal well-being and resilience [22,23]. These types of conspiracy theories include misinformation and disinformation related to COVID-19 and cast doubt on the existence of the virus per se, in turn making people who endorse such conspiracy theories less likely to comply with public health measures [24].

Concerning moral foundations, health choices, and practices of protection of public health during the pandemic exemplify a choice of individuals to depend on their moral attitude to protect themselves and others. Containing pandemic-related behaviors as actions that people need to take to tackle the spread of the virus, morality is one key factor that plays an important role in determining the extent to which people will engage, or not, in such actions. Indeed, individuals are supposed to be motivated to take actions, depending on their willingness and intention to defend and protect sacred values, such as their quality of life and health which, in turn, are threatened [25]. People, who endorse conspiracy theories rarely make health-related choices recommended by authorities. Pummerer et al. [26] confirm findings according to which believing in conspiracy theories decreases governmental support and adoption of health-related containing behaviors such as physical distancing. Uscinski and Parent [27] have showed that people high in conspiracy beliefs were less likely to engage in cooperative and altruistic behaviors or donation practices. More recent findings come from a study by Imhoff and Lamberty [28] on conspiracy worldviews of COVID-19 and engagement with pandemic preventive and hygienic behaviors. In this study, conspiracy theories that described the COVID-19 pandemic in terms of a hoax, were more strongly associated with reduced containment pandemic-related behavior (e.g., increased hygiene behaviors, handwashing, physical distance maintenance). This finding was observed in the United Kingdom and the United States.

Further, it has been suggested that when morality is central to people’s self-concept, then moral judgements of actions, such as intention to adopt hygienic measures, are translated into actual behavior [29]; here, scholars have shown that moral judgement and empathy for significant others is a key predictor of engagement with preventive pandemic-related behaviors. Additionally, Pagliaro et al. [30] have evidenced that discretional COVID-19 related behaviors are exhibited when they are prescribed by their very moral essence to foster the collective well-being and welfare of the community amidst the pandemic crisis. We, thus, hypothesize that conspiracy beliefs will be negatively associated with pandemic-related behaviors and policy support, but this association will be positively mediated by moral identity and morality-as-cooperation (H1b).

Overall, with our study we aim to show how both political and non-political underpinnings of conspiracy theories about COVID-19 can lead to public health containment behaviors and support of measures. The contribution of our study lies on identifying the associations between conspiracy beliefs and health-related behaviors and policy support via two levels of morality: individual and group-based morality, examining how such associations occur across 67 countries that differ in the way they implemented measures and contained the spread of the virus. Up until now, very few studies take both the individual and group dimension of morality at the same into account [15,31].

Rationale of the study

In the present study, we used morality-as-cooperation to measure the cooperative component of morality that ties individuals to their living community. Additionally, we used a trait-based measure of moral identity [32] to discover the traits that compose people’s unique moral identity as part of individuals’ self-concept rather than their cooperation with community group members. Moral identity is conceptualized as an individual trait-based tendency to consider morality as central to individual self-concept and sense of self-consistency on moral action and personal identity [5]. Formation of moral identity is, thus, linked with individual characteristics and contexts for moral actions [33]. When individuals integrate morality-relevant values into their self-concept, then a moral self-identity arises [34]. On the other hand, morality-as-cooperation conceptualizes morality in terms of a group-focused behavior and cultural solutions to issues of cooperation and conflict that may occur in people’s social life. Morality-as-cooperation is based on interactions between groups and people, characterized by mutual coordination, social exchanges or division and disputes [9,35]. Morality-as-cooperation mechanism motivates people towards altruistic and cooperative behaviors, as well as opportunities to evaluate behaviors of others [36]. To the extent that moral identity can be conceptualized as a foundation for comprehending moral agency in groups and organizations [37], we suggest that moral identity and morality-as-cooperation complement each other. Moral identity, by definition, entails an interactive, cooperative component too, since individuals with moral identity develop a sense of personal responsibility, considering themselves responsible for their actions in particular contexts. Therefore, they act either proactively or reactively and always in congruence between their judgement of the context and their actions for the benefit of themselves and others, who are present in a situation at hand [38].

Participants

In April 2020, a research team from New York University launched a call for participation in a survey, using various social media to collect data across the globe on psychological and social factors that might be related to responses to COVID-19 pandemic, with public health support as the key outcome. We firstly created a survey in English (see supplementary material) and then we sent it to each team for back translation. The survey secured ethical approval from the ethics board at the University of Kent and was conducted in accordance with the Code of Professional Ethics of the British Psychological Society, the Hellenic Psychological Society, the Danish Psychological Association, and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Written informed consent was obtained from all individual participants included in the study (see, supplementary material: https://osf.io/y7ckt/). Multi-level models were used for dataset collection and analysis. We report how sample size was determined, all data excluded from the analysis (if any), all manipulations, and all measures in our survey.

For the purpose of this study, we used data collected by an international team of scholars as part of the International Collaboration on the Social & Moral Psychology of COVID-19 project (see https://icsmp-covid19.netlify.app/about.html). Each team was allowed to include additional items after the main survey under their own institutional protocol. Each research team across 67 countries was assigned to collect data from a number of at least 500 participants, a sample which should be representative with respect to gender and age, in their own country or territory, but not all countries achieved a recruitment of representative samples. Out of the 67 countries joined the project, representative samples were recruited in 28 countries, convenience samples were recruited in 36 countries, and both types of sampling were recruited in 3 countries. All authors of this study were a part of this international collaboration.

A sample of N = 51,404 individuals overall across 69 countries participated in our survey. Following the inclusion criteria, participants needed to be 18 years and older, and give informed consent. Raw data obtained from all research collaborators were cleaned, in order to exclude any duplicate answers, as well as those participants, who were younger than 18 years or older than 100 years. Next, we excluded data from two participants from Puerto Rico and 313 participants recruited from the UAE, where it was difficult to establish nationality of participants. This resulted in a total sample of 51,089 participants. For the subsequent analyses of this paper, we also excluded participants who had missing data on all the key variables of our interest. Therefore, we were left with (N = 49,965) participants from 67 countries, where 50.9% were females, 44% males, 0.3% others, and 4.8% unreported. Participants were over 18 years of age and younger than 100 years of age (M = 43.8, SD = 16.05). Countries from all continents (except for Antarctica) took part in our study. Some are overrepresented (e.g., Europe, Americas) while others are underrepresented (e.g., Africa, Middle East). For further information about the distribution of education, country of residence and other demographics across 67 countries, see the complete dataset in our supplementary material here: https://osf.io/tfsza/.

All the analyses reported in our paper were repeated after controlling for differences in methods of sample recruitment. In our analysis, no coefficients that differed as a function of sampling procedure were encountered, which would compromise and/or alter the reported main effects.

Measures

We used three measures of public health support [31]. We used a Spatial Limiting Distancing 5-item scale (Cronbach’s α = 0.74), to measure people’s maintenance of spatial distance and reduction of physical contact, with items such as “During the days of the coronavirus (COVID-19) pandemic, I have been staying at home as much as practically possible”. Furthermore, we used a Physical Hygiene 5-item scale (Cronbach’s α = 0.72), to measure people’s adoption of health-related behaviors and improvement of their physical hygiene, with items such as “During the days of the coronavirus (COVID-19) pandemic, I have been washing my hands longer than usual”. Finally, we used a Policy Support 5-item scale (Cronbach’s α = 0.81), to measure people’s support towards implementation of health-protective policies and measure as responses to the pandemic, with items such as “During the days of the coronavirus (COVID-19) pandemic, I have been in favor of closing all schools and universities”. We used an 11-point “slider scale with three labels: 0 = “strongly disagree”, 50 = “neither agree nor disagree”, 100 = “strongly agree”. These labels were re-coded to a scale ranged from 0 (strongly disagree) to 10 (strongly agree).

Additionally, we used a 4-item COVID-19 Conspiracy Beliefs scale (Cronbach’s α = 0.79), to measure people’s endorsement of conspiracy theories about the origin and the causes of the pandemic, with items such as “The coronavirus (COVID-19) is a bioweapon engineered by scientists.” As it was mentioned in the introduction, in this scale, we included not only political-related items, but also items relevant to the scientific community, interest social groups not necessarily political, as well as global economic issues. A scale from 0 (strongly disagree) to 10 (strongly agree) was used. Engaging government, science, interest social groups, global institutions and an explanation of the virus as a hoax in measurement of conspiracy beliefs helped us avoid effects of political attitudes and ideological preferences on the associations we examine.

Lastly, we measured morality using two scales: First, we used a 7-item Morality-as-Cooperation scale [9] (Cronbach’s α = 0.77), to measure different cooperative moral behaviors such as helping groups. Participants were instructed to think when they decide whether something is right or wrong, to rate the extent to which a number of considerations is relevant to their thinking. Example consideration was “Whether or not someone worked to unite a community”. A scale from 0 (strongly disagree) to 10 (strongly agree) was used.

Next, we used a 10-item Moral Identity scale [32], to measure people’s self-identification based on moral prosocial attributes. This scale consisted of two subscales; moral identity symbolization and moral identity internalization. Participants first read 9 characteristics (e.g. caring, compassionate etc.) that might describe a person, who could be themselves or it could be someone else. Next, participants had to visualize in their mind the kind of person who has these characteristics. Participants were asked to imagine how that person would think, feel, and act. When they had a clear image of what this person would be like, participants were asked to rate statements such as “I am actively involved in activities that communicate to others that I have these characteristics” (symbolization subscale; Cronbach’s alpha = 0.75) and “Being someone who has these characteristics is an important part of who I am” (internalization subscale; Cronbach’s alpha = 0.68) (see Appendix A, for the full list of items in symbolization and internalization dimensions). A scale from 0 (strongly disagree) to 10 (strongly agree) was used.

Recent research which has also used this data and originated from the same international collaboration project (ICSMP) have used advanced machine-learning algorithms, established measurement equivalence of the moral identity internalization and symbolization sub-scales across the 67 countries included in the data set [39]. The original paper of Van Bavel et al. [31] reported a two-factor model (Internalization and Symbolization), with acceptable internal consistency. This two-factor structure was confirmed in our subsequent machine-learning analysis [39] after correlating residuals of items 8 and 9, and 4 and 7. (CFI = 0.939, RMSEA = 0.077, 95% CI [0.070, 0.084], SRMR = 0.067, Cronbach’s alpha_{internalization} = 0.68, Cronbach’s alpha_{symbolization} = 0.75).

Covariates

In all analyses, we controlled for: (a) participants’ age, gender, employment status and living area (urban-suburban or rural); (b) participants’ levels of national identification, measured with two items (I identify as [nationality]; Being a [nationality] is an important reflection of who I am.); (c) participants’ political ideology, measured in one item, where participants had to rate their political views in a scale from 0 to 10, where 0 indicated very left-leaning views and 10 indicated very right-leaning views. The full list of scales, along with the dataset in SPSS and a CSV form can be found here: https://osf.io/y7ckt/.

Policy-making impact of our research

Practitioners and policy makers can benefit from our research by acquiring information about people’s moral foundations that can mediate their inclination to endorse conspiracy beliefs with a subsequent effect on adopting necessary health-related behaviors to reduce the spread of the virus. In addition to this, national leaders can obtain useful information by this study, in order to get an overview of how they can effectively formulate and tailor political messages to appeal to people’s moral principles. Once people perceive political messages and decisions as moral ones, then they would be more likely to willingly adopt containing health-related behaviors and reject any conspiracy theories that explain the origin and spread of the virus. Indeed, people’s individual and collective health-related choices are put under moral constraints and raise important moral questions [50], potentially depending on what moral foundations politicians appeal to and whether they achieve to make people reject the conspiracy explanations of the origin and spread of the virus.

Tackling conspiracy information–known also with the popular term ‘infodemic’–during the COVID-19 pandemic, constitutes another one important task for governments and politicians. From a political communication perspective, conspiracy theories about COVID-19 often debunk political messages, distort public perception of the virus and, eventually, weaken the credibility of the source of such political information and guidelines [51]. Conspiracy theories about the pandemic have also reduced the perceived effectiveness of political messages and recommendations of health containment behaviors [52]. Thus, perceived morality constitutes an important dimension for policy-makers to take into account so that to reduce any distortion of information about the pandemic, enhancing public well-being and encouraging people’s engagement with health recommendations [53]. By considering the mediating role of morality–either in an individual or a group, cooperative level–in the association between beliefs in conspiracy theories about COVID-19 and the actual engagement with health containment behaviors and policy support, politicians can tailor their political messages to communicate strategically structured narratives advancing public health, appealing to individual and collective welfare ensuring, at the same time, the communication of comprehensive and accessible health information.

Overall, the COVID-19 pandemic constitutes a global challenge for governments and countries, as well as for the people whose lives have been significantly affected. In a digital era, when fake news, conspiracy theories and rumors circulate and spread impressively fast, the pandemic features an additional challenge: the identification of the actual effects of such conspiracies on important dimensions such as people’s moral judgements and behavioral change, factors which are crucial to determine the spread of the virus and the future of public health.

Appendix A: Symbolization and Internalization Items of the Moral Identity Scale (Aquino & Reed, 2002).

Symbolization Items

• I often wear clothes that identify me as having these characteristics.
• The types of things I do in my spare time (e.g., hobbies) clearly identify me as having these characteristics.
• The kinds of books and magazines that I read identify me as having these characteristics.
• The fact that I have these characteristics is communicated to others by my membership in certain organizations.
• I am actively involved in activities that communicate to others that I have these characteristics.

Internalization Items

• It would make me feel good to be a person who has these characteristics.
• Being someone who has these characteristics is an important part of who I am.
• I would be ashamed to be a person who had these characteristics.
• Having these characteristics is not really important to me.
• I strongly desire to have these characteristics.