Organizational trust in times of COVID-19: Hospitality employees’ affective responses to managers’ communication

Abstract

During a crisis, such as the COVID-19 pandemic, what managers communicate to their employees can greatly impact important organizational attitudes, such as organizational trust. There is, however, very little research focusing on the mechanisms explaining how managers’ messages during a crisis can influence employees’ organizational trust. To address this gap, the current study examined the role that emotions play in developing organizational trust using a 2 (following CDC norms vs. ignoring CDC norms) by 2 (employee focus vs. bottom-line focus) between-subjects factorial experiment, with COVID-19 as the context. The results showed that a manager’s communication that followed the CDC social norms made employees feel grateful, whereas communication that ignored CDC social norms enhanced fear and anger toward the organization. The feelings of gratefulness and fear influenced organizational trust. These results provide important theoretical and practical implications for understanding organizational trust during a crisis.

1. Introduction

Major negative events, such as terrorist attacks, natural disasters, and pandemics, like the COVID-19, can disrupt services, close businesses, and change the productions of service, thus creating a crisis for organizations (Morgeson et al., 2015). As a response to the COVID-19 pandemic, many hospitality organizations have been following and communicating the importance of applying the safety and health precautions from health experts, namely, the Centers for Disease Control and Prevention’s (CDC) guidelines for how to continue to operate during the COVID-19 outbreak (AHLA, 2020; NRA, 2020).

Unfortunately, there are also examples of businesses and managers who are not following these important guidelines, and, instead, are sending messages that go against what the CDC and other health officials are communicating. For instance, flight attendants from Delta Airlines received a corporate email asking them to "refrain from notifying other crew members on your own" and to "not post on social media about your health status" if they felt sick or were concerned they contracted COVID-19 (Kaufman, 2020). Multiple restaurants and bars also made the headlines for ignoring CDC health and safety guidelines, such as not requiring employees and customers to wear masks and failing to use social distancing protocols (Feldman, 2020; Himler, 2020; Pinsker, 2020).

How organizations react and communicate during a crisis can have a profound impact on their stakeholders. For instance, employees’ attitudes toward the organization (Bundy et al., 2017; Harvey and Haines, 2005) such as organizational trust (Koronis and Ponis, 2018) can be impacted depending how management decides to communicate. Communicating messages that defy the health and safety guidelines from experts and authorities can potentially lead to employees’ negative attitudes because, during times of crisis, employees often look at social norms when determining how to respond (Cialdini et al., 1991; Harvey and Haines, 2005). In fact, Hu et al. (2020), in their case study conducted with a Chinese restaurant, found that employees were aware of the COVID-19 threats, which contributed to the perceived utility of introduced safety measures to comply with government requirements. Perceptions about social norms are based on what others commonly do (descriptive norms) and what others frequently approve or disapprove (injunctive norms) (Cialdini et al., 1991). Therefore, employees are likely paying attention to how management is following social norms related to the crisis. For example, the CDC and other health authorities, as well as news coverage, have been consistent with the message of using face masks, social distancing, personal hygiene, and cleaning standards (AHLA, 2020; NRA, 2020).

Thus, what managers communicate in relation to social norms (i.e., follow CDC's safety recommendations vs. ignore CDC's safety recommendations) to employees during a crisis can have a major impact on important organizational attitudes, such as organizational trust. The literature on organizational trust has shown two important findings that have implications for managers’ messages in response to a crisis: (1) employees’ organizational trust influences important attitudes, such as organizational commitment, organizational citizenship behaviors, job satisfaction, and turnover intentions, and (2) managers have an impact on employees’ organizational trust (Li et al., 2012; Tourigny et al., 2019). Despite these findings, there is very little research focusing on the mechanisms explaining why managers’ messages during a crisis can influence employees’ organizational trust. One possible mechanism that could explain this dearth in the literature is the role emotions play in developing organizational trust (Lee and Selart, 2011).

Affective experiences related to work-life events can unleash positive or negative emotional reactions, and these emotions influence employees' attitudes and behaviors (Ashkanasy and Dorris, 2017; Weiss and Cropanzano, 1996). Hence, employees will have lower adverse affective reactions toward an organization that signals to be aligned with social norms. These affective reactions will then influence how much they trust the organization (Morrow et al., 2004). In addition, perceptions about organizational support can arise when employees observe that organizations care about their well-being. If an organization is supportive of its employees, it is expected that they would provide "sympathetic understanding and material aid" to deal with stressful situations (Eisenberger and Stinglhamber, 2011, p.58). Therefore, the current study also examined the focus of a message related to whether the organization is more concerned with its employees or its bottom-line.

The current study investigates how a manager communicates with employees regarding COVID-19. Specifically, we draw from affective events theory (AET) (Weiss and Cropanzano, 1996), organizational support theory (Eisenberger and Stinglhamber, 2011), and social norms (Cialdini et al., 1991) to understand how a manager's communication regarding the coronavirus (following CDC norms vs. ignoring CDC norms) and focus (employee focus vs. bottom line-focus) influences employees’ emotions and organizational trust. Specifically, the AET explains how one’s episodic emotions are contingent on one’s circumstances in which certain environmental features constitute the “affective events” that stimulate the development of emotions (Weiss and Cropanzano, 1996). To better understand the affective event (managers’ messages about the COVID-19 pandemic), social norms (i.e., what is common in a specific setting) should also be considered, as it has a strong and frequent impact on behavior (Cialdini et al., 1991; Schultz et al., 2018). In addition, the organizational support theory provides clarity about why employees tend to develop a general perception regarding the extent to which the organization appreciates employees’ contributions and cares for their happiness (Eisenberger et al., 2016).

This study makes several contributions to understanding how communication during times of crisis is of high importance. First, although crisis management in hospitality research has evolved considerably (Israeli et al., 2011), most studies focus on strategies to overcome those crises and on financial results (Arampatzi et al., 2015; Poria et al., 2014). Employees play a central role in service delivery and quality, and, therefore, organizational success (Dawson and Abbott, 2011; Elsharnouby and Elbanna, 2020; Pfeffer, 1995). Surprisingly, little research has been done related to how hospitality organizations assist their employees during crises (Poria et al., 2014; Hu et al., 2020). Considering the global impacts of COVID-19 in the hospitality industry and many other crises that hospitality businesses face (e.g., natural disasters, terrorism), knowing how to properly communicate about a crisis with employees to lower adverse affective reactions and build trust is paramount. Second, uncertainties and risks related to work status and health can inevitably cause negative emotions (Hu et al., 2020; Mao et al., 2020). During challenging times like COVID-19, employees look for guidance and support, which will mold their emotions toward their organizations. It is expected that such emotions will have an impact on employees’ attitudes and behaviors (Weiss and Cropanzano, 1996). Still, organizational research has neglected the essential role of emotions (Troth et al., 2018). This study provides a unique perspective on how a manager’s communication during crises affects employees’ gratitude, fear, and anger - all crucial emotions related to work outcomes.

Third, little research has focused on how managers can influence employees’ organizational trust (Legood et al., 2016), despite the fact that research suggests that emotion-related events at work can have important attitudinal consequences (Ashkanasy and Dorris, 2017; Forgas and George, 2001) and that trust is often based on how one feels (Ozyilmaz et al., 2018; Schwarz and Clore, 2003). Yet, research has not examined how managers’ communication during a crisis can influence organizational trust via employees’ emotions. This is a critical gap in the literature, considering that organizational trust plays a vital role as an antecedent of important organizational attitudes, increased cooperation, and reduced conflict (Dirks and Ferrin, 2002; Tourigny et al., 2019).

2. Literature review

2.1. CDC guidelines as social norms

In light of the COVID-19 pandemic, hospitality industry associations are taking steps to protect employees as much as possible. For example, the American Hotel and Lodging Association (2020) and the National Restaurant Association (2020) are utilizing standards set by the CDC for keeping businesses sanitary and ensuring employees’ peace of mind. By following the standards set by governmental task forces, most businesses opt to follow what instructions and warnings have been set. It has been shown repeatedly by multiple studies that people tend to follow what is thought to be the descriptive or social norm (Nolan, 2017). Social norms can influence individuals, in an implicit or explicit way, by reminding them of the social values of society (Nolan, 2017).

For example, according to one social experiment studying how social norms affect littering rates, the research found that people tended not to litter when it was possible to infer that littering was against the norm (Cialdini et al., 1991). Another example can be found in a series of studies examining how social norms influence hotel energy conservation. Hotels using appeals, such as signs in guest rooms that said “the majority of guests in this room reuse their towels,” led to more guests reusing their linens and towels than guests in rooms that did not signal social norms (Goldstein et al., 2008).

In another study on social norms, the study issued each participating household information comparing their energy usage to those who lived around them. Families who saw they had been using more electricity than their more conservative neighbors tended to drastically lower the amount they used (Schultz et al., 2007). This research on social norms can be analogized to current preventative measures against COVID-19, in that the CDC guidelines are becoming the norms for workplace safety and health (Czeisler et al., 2020). For example, CDC safeguard guidelines, such as the use of sneeze guards, face masks, social distancing, and personal hygiene when interacting in social areas, can be perceived as an effective social norm because establishments using these precautionary measures do not have the same rate of outbreaks (Czeisler et al., 2020). On the other hand, employees of establishments that do not use these methods are more likely to resent management for not doing everything they could to protect those who work there from the spread of the virus, and, in turn, become less positive about their safety.

2.2. Emotional reactions to leader messages

Emotion involves experiential components after the subjective assessment of meaning and implication of daily events. The subjective emotional experience is typically directed at some person, object, organization, or event (Ashkanasy et al., 2016). The organizational support theory posits that employees personify their workplace and develop emotions toward the organization (Eisenberger and Stinglhamber, 2011). Accordingly, employees develop emotions directed toward the organization based on the messages sent by their managers (Waytz and Young, 2012).

According to the law of emotion (Ashforth and Kreiner, 2002), people get used to familiar situations and generate less and less affective reactions. As the COVID-19 crisis is out of the scope of habituation and comfort, employees particularly attend to the treatment they receive from the organization, such as the message sent on behalf of the organization regarding the crisis handling. Accordingly, the COVID-19 pandemic poses an unusual and uncertain context that elicits various emotions toward organizations among employees. Managers’ messages are interpreted as being an embodiment of the organizational attitude (Eisenberger et al., 2010). As a result, employees may generate different emotional responses toward the organization in the receipt of managers’ messages. Drawing on the AET (Weiss and Cropanzano, 1996), this paper concerns managers’ messages sent to employees from the organization about the COVID-19 pandemic as an important affective event, which, in turn, exerts positive or negative impacts on employees’ different episodic emotions, such as anger, fear, and gratitude identified in the current study.

2.2.1. Gratitude

Gratitude toward the organization can be defined as an employee’s feeling of appreciation toward the organization because of its intentional admirable gesture that is of value to the employee and costly to the organization (Waters, 2012). Gratitude has been identified as an essential contributor to individual happiness (e.g., Watkins et al., 2003) and extra-role job behavior (Ford et al., 2018).

According to the multilevel model of gratitude in organizations, one gratitude initiative at the organizational level that can elicit the episodic gratitude is the recognition of helpful actions from the organization (Fehr et al., 2017). With the COVID-19 pandemic, employees attend to what an organization says or does as a response to this unusual situation. An organization’s message that follows the CDC guidelines tends to have positive impacts on employees’ experience of gratitude toward the organization. In this situation, the core characteristics of gratitude can be observed: first, the benefactor (the organization) intentionally chooses to follow the CDC recommendations to better protect employees; second, following CDC recommendations may be costly to the organization. For example, employees having symptoms related to COVID-19 do not need to work; third, through the organization’s following CDC norms, the recipient generates the feeling of being valued, understood, and cared for by the benefactor (Fehr et al., 2017). Taken together, gratitude is more likely to emerge when employees receive an organization’s message following the CDC norms, compared to a message that ignores the CDC norms.

2.2.2. Anger

Anger refers to “a negatively-valenced affect that arises from the blockage of movement toward a desired goal” (Carver and Harmon-Jones, 2009, p. 184). In particular, anger is a typical emotional reaction to affective events that involves a perceived deviation from or violation of standards, norms, and expectations (Mascolo et al., 2000). The experience of anger implies a moral assessment and is directed toward the entity that is perceived to act in defiance of norms and standards. This paper hypothesizes that an organization’s message that follows the CDC norms is likely to reduce employees’ anger toward the organization. The literature on psychological contract posits that employees and organizations have some unwritten agreements and expectations from each party (Suazo et al., 2009). For instance, employees expect their employers to care for their safety, health, and well-being. Accordingly, when an organization’s message is not in accordance with the CDC norms, employees may think that it is failing to carry out the psychological contract to tend to their well-being, thus leading to a perception of a psychological contract breach. The violation of the psychological contract has been found to elicit the experience of anger toward the blameworthy entity (Robinson and Wolfe Morrison, 2000).

2.2.3. Fear

Fear is an unpleasant emotion that arises from “perceived risk or danger, whether real or not, or a feeling of extreme dislike to some conditions/objects…” (De Lara, 2006, p. 582). In particular, appraisal of weak control and high uncertainty of the situation gives more impetus to the experience of fear (Lerner and Keltner, 2001). In areas where particular expertise is required, such as law, medicine, and science, individuals naturally choose to follow the opinions of the professionals and experts. COVID-19 poses an emergent and uncertain situation where recommendations from specialists are valued and heeded by the public, as they offer essential guidance under the situation of uncertainty and danger. As employees endorse the CDC recommendations and consider them to be aligned with social norms to cope with the COVID-19, organizations’ messages that follow the CDC guidelines are more likely to alleviate employees’ fear compared to organizations’ messages that ignore the CDC guidelines. Taken together, the following hypothesis is proposed:

H1

A manager’s message that follows the CDC norms, rather than ignoring the CDC norms, have a negative effect on (ai) anger and (aii) fear and a (b) positive effect on employees’ gratitude.

2.3. Message focus as a moderator

In addition to the influence of CDC guidelines, employees’ emotions toward the organization are also a function of whether a manager’s message is employee-focused or business-focused. The COVID-19 outbreak provides an example in which employees pay particular attention to treatment received from the organization, as they are in need of support and comfort. Meanwhile, organizations are under the pressure of achieving business goals in maintaining profitability. Therefore, organizations can choose to be employee-focused by expanding, for instance, sick leave duration and adjusting work schedules. Organizations can also be business-focused by holding the organization's interests above some stakeholders's interests, focusing on bottom-line outcomes to the neglect of other priorities (Greenbaum et al., 2012). More radically, the business-focused message may give employees the impression that the organization devotes itself to business goals at the sacrifice of employees’ safety and health. In comparison, employee-focused messages embody an organization’s determination to support and protect employees in a time of danger and uncertainty, thereby constituting a critical organizational support aspect. Based on the organizational support theory (Eisenberger and Stinglhamber, 2011), we propose that, compared to managers’ messages that are business-focused, employee-focused messages are likely to have an interaction effect with a manager’s message that follows the CDC norms, influencing employees’ gratitude toward the organization and reducing employees’ anger and fear. Kurtessis et al. (2017) found that leaders who demonstrate concern for employees’ well-being, namely the leader consideration, were more positively related to employees’ perceived organizational support, compared to leaders who only convey job-related messages.

The prerequisite of experiencing gratitude is twofold: 1) the organization (benefactor) needs to demonstrate behaviors that are taken for the sake of employees, and 2) employees need to recognize the good deeds of the organization and appreciate this behavior (Fehr et al., 2017). Employees are more likely to recognize the affective event (e.g., manager’s employee-focused message) that elicits gratitude, as the COVID-19 pandemic goes beyond the scope of typical situations where one may overlook the organizational support. For example, Wang et al. (2020) found that employees developed gratitude when errors occur, as error occurrence is considered as a situation above the ordinary. On the contrary, a manager’s message that highlights business profitability might not create a gratitude-eliciting event, as the message does not prioritize employees’ interests.

Anger, on the other hand, is elicited by the perceived harm and directed toward the entity that exerts detrimental treatment (Carver and Harmon-Jones, 2009). Based on the psychological contract and organizational support literature, employees have certain expectations from their workplace, among which attending to employees’ safety and health is of great importance (Kurtessis et al., 2017). Compared to a manager’s message that focuses on employees’ well-being, managers’ business-focused messages can be considered as a psychological contract breach, as the message reflects the organization’s profitability-driven mindset (e.g., Robinson and Wolfe Morrison, 2000). Also, the business-focused message disregards employees’ expectations that the organization should care for their health and safety at the time of the COVID-19 outbreak. Therefore, a manager’s business-focused message may trigger employees’ anger, whereas an employee-focused message appeases employees’ anger. Ford et al. (2018) found that organizational support has a strong negative connection to employees’ chronic anger toward the organization.

The COVID-19 pandemic also poses a fearful situation where employees feel a direct threat to their health, safety, or even lives. A business-focused message indicates that employees need to devote themselves to their role at work in order to achieve the the organization's goals. This message also implies that the organization’s profitability goals should take precedence over employees’ well-being. As a result, employees may naturally generate the emotion of fear because of the perceived threat and danger (Lebel, 2016). In comparison, a manager’s message that is employee-focused sets a tone that employees’ safety and health is the priority of the organization, which tends to alleviate employees’ fear. Taken together, the following hypothesis is formed:

H2

A manager’s message focus (employees vs. business bottom-line) has an interaction effect on the relationship between a manager’s message (following vs. ignoring CDC norms) and emotions, such that (ai) anger and (aii) fear are negatively impacted, and (b) gratitude is positively impacted.

2.4. Organizational trust as an outcome

Trust is often defined as “a psychological state that comprises the intention to accept vulnerability based upon positive expectations of the intentions or behavior of another” (Rousseau et al., 1998, p. 395). Trust can manifest in reference to different targets (e.g., coworkers, managers, organizations) and units of analysis (i.e., individual-level or group-level trust) (Wildman et al., 2012). The current study focuses on organizational trust, which is the extent to which employees trust their organization for three significant reasons. First, organizational trust plays a vital role as an antecedent of organizational attitudes, such as intentions to quit, organizational commitment, employee engagement, organizational citizenship behaviors, employee cynicism, and job satisfaction (Archimi et al., 2018; Dirks and Ferrin, 2002; Hough et al., 2015; Tourigny et al., 2019). For example, in a study of U.S. hotel employees, (Yoon et al., 2016) found that organizational citizenship behaviors were predicted by the extent to which hotel employees trusted their organization. Top et al. (2015) found that organizational trust is a significant predictor of overall organizational commitment. Second, because the definition of trust underscores the importance of vulnerability in an employee-employer relationship, trust is particularly important in times of crisis and uncertainty (Gustafsson et al., 2020). Third, organizational trust is highly influenced by the actions, behaviors, and communication from leaders, including managers (DeConinck, 2010). For example, Dai et al. (2013) found that hotel employees’ organizational trust was influenced by their managers' leadership style (e.g., transformational and transactional leadership style). Mazzei and Ravazzani (2015) found that internal communication strategies significantly influenced employees’ perception of trust. Linking the literature on organizational trust and crisis communication, Fuoli et al. (2017) indicated that communication approaches (denial vs. apology) have different impacts on reestablishment of employees’ organizational trust. Thus, research clearly shows that employees’ organizational trust is important for the success of organizations, and that managers have an impact on how much employees trust their organization. In spite of previous findings, it is still unknown how organizational trust in the context of crisis is a function of leader communication via emotional underlying mechanisms.

2.5. Mediation effects of employee emotions between managers’ communication and organizational trust

Research on why managers’ messages during a time of crisis influence employees’ organizational trust is scant. One possible mechanism that could explain this dearth in the literature is the role emotions play in developing organizational trust (Lee and Selart, 2011). Schoorman et al. (2007, p. 348–349) indicated that “……affective responses influence how people evaluate their level of trust in another party.” For example, Engdahl and Lidskog (2014) went beyond the rationality-oriented trust and explored the emotional antecedents of trust. Similarly, Robbins (2016) found that the link between trustworthiness and trust is mediated by other-praising emotions (e.g., admiration, gratitude). Another study with five experiments showed that emotions can influence trust, even when the emotion is not related to the target of trust (Dunn and Schweitzer, 2005). Specifically, the experiment elicited anger or gratitude by having the participant describe a situation that either made them angry or grateful, which then led to lower trust (when anger was elicited) or higher trust (when gratitude was elicited) toward targets (e.g., coworkers). Interestingly, the elicited emotions were not only related to the targets of trust, but also to a general feeling of trusting, suggesting that trust judgments are influenced by the emotion and the valence of emotion (i.e., positive versus negative emotions) one feels. Smith et al. (2014) indicated that positive emotions can trigger adaptational cognitions aligned with the goal achievement, such as perseverance, commitment, and sense of duty. This inference was empirically affirmed in a study conducted by Tong and Jia (2017), which revealed significant correlations between different positive emotions and cognitive appraisals. Individuals feeling negative emotions (e.g., anger) were less trusting, and individuals feeling positive emotions (e.g., gratitude) were more trusting. These results suggest that if managers elicit emotions among employees, the type of emotions—negative or positive—can affect trust judgments, such as organizational trust.

The underlying mechanism linking emotions to trust is based on research that suggests that emotion-related events at work can have important attitudinal consequences (Forgas and George, 2001). The affective infusion model (AIM) holds that emotions intervene in the cognitive processing, which is in charge of the formation of attitudes (Forgas, 1995; Mao et al., 2018). In other words, one’s job attitudes are partially determined by the affective states of that person that can infuse the cognitive processing. In a similar vein, the affect-as-information model states that individuals often use their feelings to make judgments (Zadra and Clore, 2011; Schwarz and Clore, 2003). For example, when asked if one should trust a target (e.g., an organization), the individual will think and ask, “how do I feel about it?” The valence of their feelings (i.e., positive versus negative emotions) is used to inform their judgments. Specific negative emotions, such as anger and fear, and positive emotions, such as gratitude, will result in different cognitive appraisals (Clore and Schnall, 2005). The reason for the differences in negative and positive emotions on judgments is that they provide different information to individuals, such that negative emotions like anger and fear signal harm (Lebel, 2017), whereas positive emotions like gratitude signal safety and prosocial motives (Drążkowski et al., 2017).

In addition to the effects that negative and positive emotions can have on judgments of trust, research also shows that the valence of emotions on trust judgments are more likely to occur in times of uncertainty when heuristic processing is more likely to be used to make judgments and decisions (Mikels et al., 2011). A crisis like the COVID-19 pandemic is an example of an event full of uncertainty because it is unprecedented, authorities are unsure how long it will last, and the effect on consumers’ travel and leisure is still unknown. Thus, the literature suggests that the emotions elicited by a manager’s message during a time of crises affect employees’ organizational trust because emotions are used as heuristic to make judgments. Therefore, as shown in Fig. 1 , the current study also hypothesized:

H3

Emotions have an effect on organizational trust, such that (ai) anger and (aii) fear have a negative effect, and (b) gratitude has a positive effect on organizational trust.

H4

Emotions have a mediation effect between the interaction effect of a manager’s message (following vs. ignoring CDC norms) and a manager’s message focus (employees vs. business bottom-line) and organizational trust, such that (ai) anger and (aii) fear have a negative effect, and (b) gratitude has a positive effect on organizational trust.

Fig. 1.

Conceptual model.

3. Methodology

3.1. Sample and procedures

The target was employees working in the food and beverage and lodging segments in the U.S. Such segments were selected because the COVID-19 pandemic has highly impacted them. Restaurants and hotels that have decided to stay open had to drastically change operations, including measures related to safety and sanitation aspects, hours of operation, and employment policies and benefits. Amazon Mechanical Turk (MTurk) was used to collect data. Samples obtained from MTurk are considered appropriate for research focusing on employees (Barger et al., 2011). Other advantages of using MTurk include: responses are considered as reliable as responses collected via traditional techniques, and respondents are usually more demographically diverse than typical internet samples (Buhrmester et al., 2011).

Respondents were paid $1.00 for their anonymous participation. To ensure data quality, screening questions and duplicated I.P. address checks were conducted. Respondents who answered in less than two minutes and extreme outliers were excluded. Participation was restricted to workers with a high approval rate (95%). The final sample consisted of 240 participants. The average respondent’s age was 34 years (SD = 9.94); 63.3% were males, 35.4% females, and 1.3% preferred not to answer. When asked where they worked within the last year, 42.9% answered the lodging industry, and 57.1% worked in the food and beverage industry; 60% had supervisory/management positions, and 40% had non-supervisory/management positions. Approximately 65% identified as Caucasian American, 17.5% as Black or African American, 10% as Asian American, 6.3% as Latino(a)/Hispanic, and 1.2% as other. When asked to describe their current job situation, 58.8% were still employed and currently working, 32.9% were furloughed, and 8.3% lost their job within the last three months because of the COVID-19 pandemic.

3.2. Research design

A 2 (following CDC norms vs. ignoring CDC norms) by 2 (employee focus vs. bottom-line focus) between-subjects factorial experimental design was conducted. Prior being randomly assigned to one of the four experimental conditions (following CDC norms + employee focus; n = 55, following CDC norms + business focus, n = 61; ignoring CDC norms + employee focus; n = 58, ignoring CDC norms + business focus, n = 66;), participants were asked to read and answer the consent form and screening questions. After, all participants were asked to imagine that it was the early days of the COVID-19 outbreak and that they were working for another company similar to their current company. They were told that their jobs had not been affected by the COVID-19 pandemic and that they had just arrived for their work shift. Then, they were asked to imagine that their manager had scheduled a staffing meeting to discuss the COVID-19 pandemic. Right after, respondents would see their randomly assigned message and then answer questions regarding this study variables, manipulation check questions, and demographic questions. In each experimental condition, the same tone and layout were used (see Appendix A).

The following CDC norms message was created based on the Interim Guidance for Business and Employers Responding to Coronavirus Disease 2019, which mentions that employers should encourage employees who have symptoms to stay home (Centers for Disease Control and Prevention CDC, 2020). The ignoring CDC norms message was created based on testimonies made by employees working during the pandemic (Branch, 2020; Kaufman, 2020). The employee and business focus messages were created based on the main principles of the Organizational Support Theory (focus on employees’ well-being - Eisenberger and Stinglhamber, 2011) and Bottom-line Mentality (focus on obtaining bottom-line outcomes to the neglect of other priorities – Greenbaum et al., 2012). All conditions were evaluated separately by each author to ensure each topic's adequacy and consistency regarding tone.

3.3. Measures

Anger and Gratitude. Anger and gratitude were measured with four items, each from Ford et al. (2018). Participants were asked to indicate to what extent they felt the following emotions toward the organization after hearing the manager’s message: mad, angry, enraged, furious, gratitude, thankful, appreciative, and grateful. Items were measured with a five-point Likert-type scale (1= none at all, 5= a great deal). Anger’s reliability was 0.92. Gratitude’s reliability was 0.91.

Fear. Fear was measured with four items, each from Harmon-Jones, Bastian, and Harmon-Jones (2016). Participants were asked to indicate to what extent they felt each of the following emotions toward the organization after hearing the manager’s message: terror, scared, fear, panic. The reliability was 0.89.

Organizational trust. Organizational trust was measured with three items from Rawlins (2008). A sample item is: “After hearing that message, I am willing to let the organization make decisions for me.” Items were measured with a five-point Likert-type scale (1= strongly disagree, 5= strongly agree). The reliability was 0.79 (See Appendix B for all items and factor loadings).

Realism check. Realism was verified with two items developed by Dabholkar (1994): “It was easy imagining myself in the scenario situation” and “The scenario situation was realistic.” Items were measured with a seven-point Likert-type scale (1= strongly disagree, 7= strongly agree).

Manipulation check. Participants were asked to recall, after reading the manager’s message, if the company followed health authority recommendations, like the CDC’s, to keep social distancing when feeling symptoms. In addition, participants were asked if the company's focus was to be profitable and, in a separate question, if the company cared more about their employees than about the financial bottom-line. Items were measured with a five-point Likert-type scale (1= strongly disagree, 5= strongly agree).

4. Results

4.1. Preliminary analysis and manipulation checks

A confirmatory factor analysis (CFA) using AMOS version 26 was conducted. One item from the organizational trust construct was deleted due to low factor loading. The deleted item (loading = .299) presented a negative statement. After deleting one item from organizational trust, the measurement model fit the data satisfactorily (χ² = 153.991, df = 70, CFI = 0.97, TLI = 0.96, GFI = 0.91, RMSEA = 0.07). Factor loadings ranged from 0.74 to 0.91 (p < .001). As showed in Table 1 , the average variance extracted (AVE) from all variables was above the 0.50 threshold, confirming convergent validity (Hair et al., 2016). The square root of AVE for all variables was higher than the inter-correlations between two constructs of interest, confirming discriminant validity (Fornell and Larcker, 1981). The C.R. exceeded the recommended 0.70 threshold for all variables (Fornell and Larcker, 1981).

Table 1.

Descriptive statistics and associated model measurements.

Constructs	# items	M (S.D.)	C.R.	AVE	1	2	3	4
1. Anger	4	2.48 (1.19)	0.91	0.72	0.85
2. Gratitude	4	2.99 (1.21)	0.91	0.73	−0.25	0.85
3. Fear	4	2.57 (1.14)	0.90	0.68	0.81	−0.09	0.83
4. Org. trust	2	3.27 (1.08)	0.80	0.67	−0.21	0.80	0.02	0.82

Note. The square root of AVE is along the diagonal in bold. All correlations (p < 0.01).

The concern of common method bias was mitigated by following Podsakoff et al. (2012) procedures, such as the items’ order being counterbalanced, respondents’ confidentiality being warranted, and different rating anchors being used. The four-factor model conducted through CFA had a better model fit when compared to the three-, two-, and single-factor models (χ² = 1246.90, df = 76, Δ χ² = 1092.913, p < 0.001, CFI = 0.54, TLI = 0.45, GFI = 0.46, RMSEA = 0.25). Moreover, Harmon’s single factor score was performed. The total variance explained by all variables was 37.55% (under the 50% threshold), providing additional support that common method bias is not a serious threat in this study.

To verify manipulation effectiveness, two steps were conducted. First, realism was evaluated with two items from Dabholkar (1994). The items’ means indicated that the scenario was seen as realistic (M = 5.63), and that respondents did not have difficulty imagining themselves in the proposed situation (M = 5.67). Second, using a t-test, manipulation checks revealed that participants in the follow CDC norms condition rated the firm’s recommendations to keep social distancing when feeling symptoms significantly higher (M = 4.01, SD = .97, t= -5.562, p < 0.001) than those in the ignore condition (M = 3.19, SD = 1.27). Participants in the employee focus condition rated the firm’s care about employees significantly higher (M = 3.37, SD = 1.25, t= -2.25, p = 0.02) than those in the business focus condition (M = 2.98, SD = 1.44). Participants in the business’ focus condition rated the firm’s profitability as being the company's main focus significantly higher (M = 3.92, SD = 1.21, t = 4.58, p < 0.001) than those in the employees’ focus condition (M = 3.19, SD = 1.24).

4.2. Test of hypotheses

Process on SPSS version 26 (Model 7) with a bootstrap function extracting 5000 samples for the analysis (95% CI) was used to test the conceptual model. As shown in Table 2 , the manager’s message following CDC norms had a significant negative effect on employees’ anger (b= -0.81, CI [-1.19, -0.42]) and fear (b= -0.49, CI [-0.87, -0.11]), in support of Hypotheses 1ai and 1aii, and a significant positive effect on employees’ gratitude (b = 0.56, CI [0.14, 0.98]), in support of Hypothesis 1b. The interaction effect between follow vs ignoring CDC norms and employee vs. business focus did not have an effect on anger (b = 0.17, CI [-0.39, 0.74]), fear (b = 0.16, CI [-0.40, 0.74]), nor on gratitude (b= -0.50, CI [-0.65, 0.53]). Thus, Hypothesis 2 was rejected. While anger (b= -0.11, CI [-0.25, 0.03]) did not have a significant effect on organizational trust, fear did have a significant effect (b = 0.17, CI [0.02, 0.32]), but not the negative effect that was hypothesized, therefore Hypotheses 3ai and 3aii were rejected. Gratitude had a significant positive effect on organizational trust (b = 0.61, CI [0.52, 0.70]), in support of Hypothesis 3b.

Table 2.

Main effects.

Hypotheses	Effects	Boot SE	Boot LLCI 95%	Boot ULCI 95%
H1ai: CDC → anger	−0.81	0.19	−1.19	−0.42
H1aii: CDC → fear	−0.49	0.20	−0.87	−0.11
H1b: CDC → gratitude	0.56	0.21	0.14	0.98
H2ai: CDC x focus → anger	0.17	0.29	−0.39	0.74
H2aii: CDC x focus → fear	0.16	0.29	−0.40	0.74
H2b: CDC x focus → gratitude	−0.05	0.30	−0.65	0.53
H3ai: anger → org. trust	−0.11	0.07	−0.25	0.03
H3aii: fear → org. trust	0.17	0.08	0.02	0.32
H3b: gratitude → org. trust	0.61	0.05	0.52	0.70

Note. CDC = manager’s message that follows the CDC norms, org. trust = organizational trust.

Next, the conditional indirect effects of the following vs. ignoring the CDC norms message on organizational trust through the mediators at the values of the organization’s focus (employees vs. business bottom-line) were tested. Table 3 shows the conditional moderated mediation effects.

Table 3.

Conditional indirect effects.

Hypotheses	Focus	Effects	Boot SE	Boot LLCI 95%	Boot ULCI 95%
H4ai	Business	0.09	0.06	−0.02	0.22
H4ai	Employees	0.07	0.05	−0.02	0.18
H4aii	Business	−0.08	0.05	−0.19	−0.01
H4aii	Employees	−0.05	0.05	−0.17	0.02
H4b	Business	0.34	0.13	0.09	0.60
H4b	Employees	0.31	0.13	0.06	0.57

The manager’s message focus did not moderate the influence of following vs. ignoring the CDC norms message on organizational trust through anger. Thus, Hypothesis 4ai was rejected. When the manager’s message focused on business bottom-line (b= -0.08 CI [-0.19, -0.01]), the influence of following vs. ignoring the CDC norms message on organizational trust through fear was negative, in support for Hypothesis 4aii. When the manager’s message focused on either business bottom-line (b = 0.34, CI [0.09, 0.60]) or employees (b = 0.31, CI [0.06, 0.57]), the influence of following vs. ignoring the CDC norms message on organizational trust through gratitude was positive, in support for Hypothesis 4b. None of the three indexes of moderated mediation were significant, indicating that the two conditional indirect effects of the moderator have no statistical differences (Hayes, 2015), which is aligned with the findings in the Hypothesis 4b, on which both messages had similar effects. The R² values indicate that 13% of the variance in anger, 5% of the variance in fear, 8% of the variance in gratitude, and 51% of the variance in organizational trust can be explained from the relationships with other constructs in the model.

Respondents’ demographics (e.g., gender, age, employment status, industry segment) were dummy-coded and entered in the model as a covariate. Results demonstrated that respondents’ demographics did not affect this study’s variables, except for employment status. Losing a job because of the COVID-19 negatively impacted respondents’ gratitude toward the organization (b= -0.87 CI [-1.29, -0.42]).

5. Discussion

5.1. Theoretical implications

This study’s main theoretical contribution is to provide evidence that employees’ emotions, namely anger, fear, and gratitude, can be influenced by how managers communicate in response to a crisis. Specifically, considering the COVID-19 pandemic, following the CDC health and safety recommendations enhanced gratitude feelings and reduced negative feelings of fear and anger toward the organization. This is consistent with the theoretical foundations of social norms (Cialdini et al., 1991) and AET (Weiss and Cropanzano, 1996). Communication about COVID-19 that followed the social norms made employees feel grateful and appreciative of organizational actions. Communicating the opposite, going against the norms, made employees feel less positive and probably confused about what they know and hear regarding what other companies are doing or should be doing during the pandemic. Ignoring the CDC norms enhanced fear and made employees angry toward the organization. The exposure to negative events is accompanied by negative affective reactions that, in turn, can lead to negative job attitudes. These findings help to expand the understanding of the roles played by hospitality organizations during crises and their effects on employees (Hu et al., 2020; Poria et al., 2014). To our knowledge, this is one of the first studies in hospitality to employ an experimental design to investigate the impact of a crisis on employees’ emotional responses.

Gratitude predicted organizational trust, suggesting that emotions play a role in organizational attitudes, as emotions are used as heuristic to make judgments (Mikels et al., 2011). Although fear also had a significant influence on organizational trust, it had a positive effect, which was not hypothesized. One possible explanation for such effect can be connected to how individuals perceive risks. Authorities are still unsure about how long the pandemic will last and when businesses will be able to operate as they did in the past. There are still many uncertainties about the future, and, thus, it is just natural that employees have feelings of fear, an emotion that arises from perceived risk. Trust itself can be understood as a “behavioral intention to take risk” (Mayer and Gavin, 2005, p.874), meaning that employees are taking risks while devoting their trust towards an organization. Thus, trusting an organization during uncertain times like the COVID-19 pandemic is still possible, although fear is present.

Anger did not have a significant negative effect on organizational trust as hypothesized. Considering all the uncertainties about the future that the COVID-19 pandemic brought and that organizations at the stage when data was collected were still learning about how to deal with the situation, employees that heard the ignoring CDC message were angrier toward the organization. Yet, they might also have been more forgiving. Employees could have perceived the organization as unintentionally failing to fulfill its implied obligations, which could explain the non-significant effect of anger on organizational trust (Robinson and Wolfe Morrison, 2000).

In addition, this study has implications for organizational support theory, as little attention has been paid to its influence on emotions (Ford et al., 2018). The interaction effect between message focus (employees vs. bottom-line focus) with the CDC norms message (following vs. ignoring CDC norms) on employees’ emotions was unexpectedly not significant. While investigating the moderated mediation effects, some effects were found to be significant, which helped us to understand under which circumstances the interaction effect influenced the results. Gratitude had a positive mediation effect between the manager’s message and organizational trust when the message had a business or employee focus. It was expected that such moderated mediation effect would occur for the employee focus message, consistent with recent findings that organizational support is a strong predictor of gratitude (Ford et al., 2018). However, when the message had a business focus, a positive effect also occurred. Considering the economic consequences connected to the COVID-19 pandemic (e.g., the number of hospitality organizations that had to cease operations, furlough, and lay off), employees could have seen the focus on being profitable not as negatively. Having a bottom-line mentality means to “treat every situation as if the bottom-line is the only relevant outcome” (Greenbaum et al., 2012, p. 343). Employees could have seen this as a way to keep the operations running, which turns to be beneficial to employees, from an employment perspective, at least in the short term. Such a stance could also explain why anger did not have a mediation effect between the manager’s message and organizational trust independent of the message focus. Fear, on its turn, had a negative mediation effect between the manager’s message and organizational trust when the message had a business focus. This result is consistent with the affect-as-information model (Clore et al., 2001): when employees processed that the manager had a focus exclusively on bottom-line outcomes and was not considering employees’ safety and well-being, then their fear enhanced, leading to reduced feelings of trust in the organization. Considering the findings about the message focus moderation effects, and since bottom-line mentality has been associated with social undermining behavior (Greenbaum et al., 2012), and unethical pro-leader behavior (Mesdaghinia et al., 2019), but also with positive outcomes such as more focused work efforts (Babalola et al., 2020), more targeted research is needed to determine its influence on employees’ emotions.

Schoorman et al., (2007) incorporated emotions in the integrative model of organizational trust and called for more empirical investigations on how emotions can influence organizational trust. The current study examined the mediating role emotions played in the relationship between managers’ messages and employees’ organizational trust. The results are aligned with previous studies, which hold that trust is not strictly rational and cognitive, but can be influenced by affective elements (e.g., Dunn and Schweitzer, 2005).

5.2. Practical implications

During crises, and especially during a global crisis such as the COVID-19 pandemic, establishing and maintaining trust among employees in times of crisis is critical to the survival and success of hospitality organizations. Gatling et al. (2017, p. 18) indicated that “…the success of hospitality business depends on the ability of leaders to forge trustful relationships with followers, whom leaders rely on to be honest, give additional discretionary effort, and be committed to organizational goals.” In particular, service production and delivery are often a result of collective work. The effectiveness of this collective work is contingent on the quality of employees’ attitude toward the organization. An organization perceived as trustworthy by employees can promote positive emotions and cognitions and motivate prosocial behaviors, which, eventually, benefit customers and organizations (Lee et al., 2013). As demonstrated in this research, during a time of health crisis, managers should highlight official health-related norms (e.g., CDC, World Health Organization) and how the organization is supportive of its employees in order to mitigate negative emotions, such as fear and anger, promote positive emotions, such as gratitude, and influence work attitudes and behaviors. Specifically, this study found that organizational trust can be influenced by how managers communicate with their employees through the mechanisms of emotions. Assessing the situation, following official guidelines, and making sure that all leaders understand the plan and communicate it properly to employees are steps that every tourism and hospitality organization must take. To keep up to date with how the pandemic is evolving, owners and managers could, for instance, access official health websites and focus on specific guidelines that address their industry concerns. For instance, both the National Restaurants Association and the American Hotel and Lodging Association have a COVID-19 information and resource center on their webpages.

Many workers are filing complaints with the federal government, as employers have been failing to protect their employees; examples of this include failing to provide masks and gloves, ignoring social distancing guidelines, and having their employees work with people that are showing symptoms (Rocheleau, 2020; Whoriskey et al., 2020). The COVID-19 pandemic shattered individuals emotionally worldwide. For hospitality and tourism employees, the effects are magnified by the nature of hospitality work (e.g., not having the option to work from home, constant customer contact). Thus, properly communicating, following official guidelines, and providing support to employees during crises can also help to mitigate OSHA complaints. In addition, organizations must ‘walk the talk’ and make efforts and changes to follow guidelines regarding social distancing and personal protection, having procedures in place in case employees show symptoms, and so on, as directed by the CDC's Interim Guidance for Business and Employers Responding to Coronavirus Disease (Centers for Disease Control and Prevention CDC, 2020).

It is essential to highlight that not considering hospitality employees’ health and well-being as a priority can have profound negative consequences in the long-term, not only for employees (i.e., contracting the virus, mental health issues) but also for the organization itself (i.e., having to deal with sick leaves and a bad reputation). Indeed, corporate social responsibility and satisfaction with corporate COVID-19 responses have been recently found to increase hospitality employees’ psychological capital (Mao et al., 2020). Organizations have a social responsibility for their employees. Hospitality organizations also have a social responsibility for the survival of their businesses. Thus, organizations should strive to do both, protecting not only employees, but all stakeholders involved in the process.

5.3. Limitations and future research

There are limitations to this study that should be noted. First, other factors, such as exposure to media coverage, the development of the virus itself (i.e., number of cases increasing in the U.S.), and the economic consequences of the pandemic (e.g., organizations closing or laying-off and furloughing employees) at the moment the data was collected could have impacted the findings. This can explain, for instance, why respondents who lost their jobs because of the COVID-19 pandemic evaluated gratitude towards the company more negatively. For this study, the data collection time was ideal in that sense, as the primary goal was to capture a real picture of employees’ emotional response during the COVID-19 pandemic. Since salience can be a source of stress affecting psychological processes (Kastenmüller et al., 2014), future studies could try to replicate the findings of this study by collecting responses in another moment, for instance, when the number of cases starts to decline in the country and organizations start to operate normally. Such measures, including a manipulated salience, can help understand how much external factors can influence employees’ emotional responses to how organizations are dealing with crises.

Second, the geographical location where the data was collected could have also impacted the findings. Countries worldwide had different responses to the COVID-19 pandemic, which were very much embedded in cultural values. Comparing results across respondents from different countries could also reveal interesting findings. Third, although the manipulation regarding message focus (employees vs. bottom-line) is theoretically sound and passed the manipulation tests, it still provided mixed findings. In addition, the COVID-19 pandemic itself is unprecedented, which is different, for instance, from natural disasters, which can sometimes be predicted and safety measures are already known. Thus, it is expected that different results can be found regarding individual emotional responses. Future research should further develop the message focus manipulation. Researchers can conduct a qualitative study to better understand how bottom-line mentality and focus on organizational support impact employees, specifically during crises.

Appendix A. Scenarios

Instructions

Imagine that it is the early days of the Covid-19 (coronavirus) outbreak.

You work for another company that is similar to your current company.

Your job at this other company has not been affected by the coronavirus outbreak.

You have just arrived for your work shift.

Now imagine that your manager has scheduled a staffing meeting to discuss the coronavirus situation. He opens the meeting saying the following:

Then they read one of the four 2 × 2 conditions.

		Focus
		Employee	Bottom-line
Guidelines	Follow CDC norms	These are unprecedented days for all of us. Our number one priority now is to make sure that we are doing everything to keep our employees safe and healthy, as we care about your well-being. After this meeting, I will explain all the resources available to you, such as our new expanded sick leave policy. If you feel sick or you are experiencing any symptoms related to Covid-19, please do NOT report to work. We will contact you to ensure you stay home to recover.	These are unprecedented days for all of us. Our number one priority now is to make sure we are profitable, as we care about the business’ bottom-line. After this meeting, I will show you the profitability results for this quarter. If you feel sick or you are experiencing any symptoms related to Covid-19, please do NOT report to work. We will contact you to ensure you stay home to recover.
	Ignore CDC norms	These are unprecedented days for all of us. Our number one priority now is to make sure that we are doing everything to keep our employees safe and healthy, as we care about your well-being. After this meeting, I will explain all the resources available to you, such as our new expanded sick leave policy. If you feel sick or you are experiencing any symptoms related Covid-19, please do NOT tell your peers and do NOT post on social media. Don’t cause unnecessary panic. We will contact you to evaluate the situation.	These are unprecedented days for all of us. Our number one priority now is to make sure we are profitable, as we care about the business’ bottom-line. After this meeting, I will show you the profitability results for this quarter. If you feel sick or you are experiencing any symptoms related Covid-19, please do NOT tell your peers and do NOT post on social media. Don’t cause unnecessary panic. We will contact you to evaluate the situation.

Appendix B. Measures used in the study

Measures	Standardized loading	Cronbach’s alpha
Anger (Ford et al., 2018)		0.92
Mad	0.91
Angry	0.89
Enraged	0.78
Furious	0.81
Gratitude (Ford et al., 2018)		0.91
Gratitude	0.77
Thankful	0.86
Appreciative	0.90
Grateful	0.88
Fear (Harmon-Jones et al., 2016)		0.89
Fear	0.80
Terror	0.81
Scared	0.87
Panic	0.83
Organizational trust (Rawlins, 2008)		0.79
After hearing that message, I’m willing to let the organization make decisions for me	0.74
After hearing that message I trust the organization to take care of me	0.88