The Burden of Hate: How Nonwork Discrimination Experienced During the COVID-19 Pandemic Impacts Asian American Employees.

The COVID-19 pandemic has been accompanied by a sharp increase in prejudice and discrimination targeting Asian Americans in the USA. Thus, in addition to the public health risks associated with the virus, exposure to discrimination poses a unique threat to the health and well-being of Asian Americans. Indeed, empirical evidence has documented the linkage between experiencing anti-Asian discrimination during the pandemic and health decrements among Asian Americans. The goal of this study was to expand that research to also consider the ways experiencing discrimination in a nonwork context may spill over to affect the general and job-related well-being of Asian American employees as well as the potential mitigating role of coworker compassion. Results from a sample of 311 Asian American employees demonstrated that experiencing nonwork discrimination was associated with decrements in physical health and increased depression and job-related exhaustion. Further, there were significant interactions between nonwork discrimination and coworker compassion for engagement, emotional exhaustion, and depressive symptoms such that nonwork discrimination was more strongly related to each outcome when coworker compassion was low. The findings from the current study suggest that experiences of racial derogation, even those that occur outside the workplace environment, are detrimental to the well-being of employees and that coworker compassion is a positive resource that may foster healthier and more inclusive work environments. Supplementary Information: The online version contains supplementary material available at 10.1007/s10869-022-09848-6.

Since the onset of the COVID-19 pandemic, there has been a notable rise in racial animus toward Asian Americans in the USA. Evidence suggests that such incidents have been widespread with anti-hate organizations documenting over 11,500 reports of anti-Asian discrimination since the beginning of the pandemic (STOP AAPI Hate, 2022) and surveys reporting that nearly one third of respondents have been targeted by racial slurs or jokes since the onset of the pandemic (Ruiz et al., 2021). Efforts to track negative sentiments toward Asian Americans have further revealed that negative tweets referencing Asian Americans increased by 68.4% during the pandemic, whereas the number of negative tweets referencing other racial groups remained constant (Nguyen et al., 2020).

This has created a double burden for the health and well-being of Asian Americans such that they are burdened first by the general health risks posed by COVID-19 and second by the stress associated with exposure to discrimination. Indeed, evidence has demonstrated that anti-Asian violence and discrimination experienced during the pandemic has adversely affected the psychological health of Asian Americans. More specifically, exposure to discrimination during the pandemic has been linked to increased symptoms of posttraumatic stress (Hahm et al., 2021), depression (Lee & Waters, 2021; Wu et al., 2020), and anxiety (Lee & Waters, 2021) as well as lowered subjective well-being (Yang et al., 2020).

Some studies have also begun to consider the ways in which the growing stigma against Asian Americans has infiltrated the workplace, with findings showing that people who attributed more blame for the COVID-19 pandemic to Asian Americans were also more likely to engage in hiring bias toward Asian applicants (Gardner et al., 2021) and that some leaders adopted stigmatizing language to refer to COVID-19 (Jun & Wu, 2021). While these studies shed light on the extent to which Asian Americans may be targeted in the workplace, they do not consider the consequences of discriminatory experiences originating outside of the work context. The goal of the current paper, therefore, was to draw on the literature on the Conservation of Resources (COR) theory (Hobfoll, 1989) and cross-domain spillover effects to posit that experiences of discrimination that occur outside of a work context may spillover to affect the job-related and general well-being of Asian American employees. We further consider how receiving compassionate responses from coworkers may help buffer the harm of this nonwork stressor.

In examining these research questions, the current study makes three primary contributions. First, we contribute to a small but growing body of research that considers the workplace impacts of experiencing mistreatment in one’s nonwork life. This research has expanded the boundaries of the mistreatment literature to show that there are cross-domain effects of experiencing mistreatment outside of work, such as decreased job performance (Lim & Tai, 2014) and increased counterproductive work behaviors (Bai et al., 2016). We add to that literature by considering how nonwork discrimination experienced during a time of pronounced national shifts in racial animus may disadvantage employees belonging to the targeted group. Second, we expand on the literature on the work–nonwork interface to recognize a need to examine nonwork stressors that originate outside of a family context. The predominant approach is to focus on family-related stressors, such conflicting demands between family responsibilities and work responsibilities. However, the current study illuminates that discrimination is another nonwork stressor worthy of study given its work implications. Finally, we examine the degree to which coworkers’ supportive and compassionate responses to anti-Asian discrimination buffer against the negative spillover effects of nonwork discrimination. Organizations may be aware of the increased need to support their Asian American employees during this time but may not know what intervention points to utilize to achieve this goal. The current study identifies whether encouraging compassionate responses among coworkers during times of unique hardship can help bolster the well-being of Asian American employees.

Anti-Asian Discrimination During COVID-19

Prejudice toward people of Asian descent residing in the USA has been documented for decades. Early examples from the nineteenth century include the passing of the Chinese Exclusion Act of 1882 as well as a string of notable attacks targeting Asian communities (e.g., the Chinese Massacre of 1871). Hostility toward Asian Americans has persisted into the present day, and recent estimates illustrate that substantial proportions of Asian Americans report being targeted by racial discrimination and/or microaggressions (Chae et al., 2008; Ong et al., 2013; Sue et al., 2009). Though it has long existed, anti-Asian sentiment has also sharply increased since the onset of the COVID-19 pandemic. This is illustrated by the 145% increase in anti-Asian hate crimes reported in 2020 (Center for Study of Hate & Extremism, 2021) as well as the 11,500 reported incidents of discrimination and violence targeting Asian Americans since March 2020 (STOP AAPI Hate, 2022). Further, recent surveys have found that over 81% of Asian Americans believe that violence against them has increased (Ruiz et al., 2021) and that nearly one third of Asian Americans have been personally targeted by discrimination during the COVID-19 pandemic (Artiga et al., 2021).

The rise in prejudice and discrimination targeting Asian Americans has been attributed to the tendency for people to associate the COVID-19 virus with Chinese Americans and more generally with Asian Americans, due to its believed origin in Wuhan, China. Despite recommendations from the World Health Organization against linking locations or ethnicities to diseases (Kopecki, 2020), key public figures have racialized the virus by giving it derogatory names that reference the virus’s geographic origin (e.g., Reja, 2021). These and other tactics (e.g., circulating sensationalized depictions of food markets, calling for the Chinese government to issue reparations) have led many to blame Chinese Americans or anyone who appears Chinese, for the spread of the virus (Scheimer & Chakrabarti, 2020). Importantly, the rise in prejudice has not only been limited to people of Chinese descent, and studies have documented that prejudice toward people of other Asian ethnicities has also grown (e.g., Ha et al., 2021).

Given the notable rise in anti-Asian sentiments in the USA, it is increasingly important to understand how Asian American employees are affected by exposure to discrimination. Recent research has demonstrated the detrimental effects of anti-Asian prejudice during the COVID-19 pandemic on Asian Americans’ general health and well-being (Lee & Waters, 2021; Yang et al., 2020), but these events, despite occurring outside of the work domain, may also have an impact on workplace outcomes. The goal of the current study, therefore, is to examine the general and job-related well-being consequences of experiencing nonwork racial discrimination during the COVID-19 pandemic as well as how coworker compassion may serve as a buffer that reduces these harmful effects. In the subsequent sections, we explicate why experienced nonwork discrimination may impair job-related, psychological, and physical functioning among Asian American employees.

Spillover Effects of Nonwork Discrimination

As summarized above, the growing racial animus has left Asian Americans vulnerable to experiencing discrimination and harassment in a variety of settings outside of the workplace. We draw on COR theory and cross-domain spillover effects to propose that experiencing anti-Asian discrimination in a nonwork setting may spill over to affect the emotional exhaustion and engagement of Asian American employees. While much of the research on the work–nonwork interface has focused on tensions that arise when demands in the two domains are incompatible, this body of work has also demonstrated that the boundaries between employees’ work and nonwork lives are permeable such that experiences that occur in one domain can cross over to affect outcomes in the other. For example, experiencing stress in one’s personal life can impair an employee’s ability to fulfill their job responsibilities (e.g., Leiter & Durup, 1996), and having positive nonwork experiences can improve employees’ moods, job satisfaction, and job engagement (e.g., Leavitt et al., 2019). Recent work has similarly begun to recognize that mistreatment experienced in the nonwork domain can have important consequences for employee outcomes. More specifically, empirical evidence supports the cross-domain effects of experiencing mistreatment from family members on both task performance and counterproductive work behavior (Bai et al., 2016; Lim & Tai, 2014). Other work also demonstrates that vicarious exposure to race-based events, including police killings of Black civilians, can spillover to harm employees (Leigh & Melwani, 2022). We extend this work to posit that nonwork racial discrimination may similarly spill over to affect employee outcomes.

We employ Conservation of Resources theory (COR) (Hobfoll, 1989; Hobfoll et al., 2018) and research on cross-domain spillover effects to develop our hypotheses regarding the work consequences of nonwork discrimination. COR theory most centrally posits that people are motivated to acquire, conserve, and restore their resources and that people experience stress following resource loss or the threat of resource loss (Hobfoll, 1989). Resources span several categories and can encompass objects, external conditions, personal characteristics, and energetical resources. When resources are lost or threatened, employees experience a range of negative outcomes associated with resource depletion, including increased emotional exhaustion and burnout (Ito & Brotheridge, 2003; Wright & Hobfoll, 2004) and reduced engagement (Bakker et al., 2007; Gorgievski & Hobfoll, 2008). Importantly, COR theory also maintains that stress can be transmitted across different domains such that resource losses in one domain can spill over into another domain (Hobfoll, 1989; Ragins et al., 2014).

While this has most commonly been studied in the context of family demands (and their associated resource gains or losses) spilling over into the workplace (e.g., Amstad et al., 2011; Grandey & Cropanzano, 1999), we extend this view of the work–nonwork interface to propose that discrimination experienced in nonwork contexts can also precipitate negative spillover effects into the workplace. We conceptualize nonwork discrimination as a stressor that may permeate the boundary between the nonwork and work domains and deplete one’s resources. That is, experienced nonwork discrimination consumes available energetical and cognitive resources by diverting attention and energy toward understanding, appraising, and coping with discrimination. Indeed, discrimination has long been conceptualized as a stressor that depletes cognitive, psychological, and physical resources and produces significant strain outcomes (Joseph et al., 2021; Walker et al., 2022). This conceptualization is further bolstered by recent work on mega-threats which posits that nonwork identity-related events can have notable spillover effects on the work domain (Leigh & Melwani, 2019; 2022). As such, nonwork discrimination experienced in relation to the mega-threat induced by the COVID-19 pandemic is expected to shape the ways employees behave and feel at work.

To assess the potential spillover effects of nonwork discrimination, the current study focuses on the outcomes of emotional exhaustion and engagement given their close association with COR theory and energetical resources (Hobfoll et al., 2018). Beginning first with emotional exhaustion, this is a key facet of burnout which is characterized by feelings of being emotionally overextended and depleted (Maslach & Jackson, 1981). Experiencing nonwork discrimination places demands on available energetical resources through several processes, such as ruminating on the experience or attempting to understand why the discrimination occurred and how to reduce the risk of reoccurrence (e.g., Borders & Liang, 2011). The resource loss associated with these processes is expected to accumulate in emotional exhaustion given that a net loss of resources is proposed to be the core antecedent to states of burnout and exhaustion (Shirom, 1989). Supporting this idea, ample research has argued for and/or demonstrated a connection between resource loss and experienced burnout (Hobfoll & Freedy, 1993; Hobfoll & Shirom, 2000; Shirom, 2003; Westman et al., 2004). Moreover, experiencing discrimination has also been linked to increased emotional exhaustion (e.g., Hu et al., 2019; Thoroughgood et al., 2020; Trzebiatowski & Triana, 2020). We correspondingly examine anti-Asian discrimination as a potential nonwork stressor that may spillover to place demands on Asian American employees while on the job, resulting in the depletion of their emotional resources:

Hypothesis 1: Nonwork racial discrimination is positively associated with emotional exhaustion.

Employee engagement can be defined as “a multidimensional motivational concept reflecting the simultaneous investment of an individual’s physical, cognitive, and emotional energy in active, full work performance” (Rich et al., 2010, p. 619). As such, having high engagement is believed to be a positive experience for employees that reflects the degree to which they are psychologically present and therefore attentive to and absorbed in their work. Drawing on this conceptualization and on COR theory and spillover research, we propose that experienced nonwork discrimination will be associated with decreased employee engagement. As noted above, although our stressor originates outside of the work context, we anticipate that the demands associated with experiencing nonwork discrimination can spillover to the workplace by consuming needed energetical resources. The state of resource loss caused by experiencing nonwork discrimination can thus impair employee engagement through siphoning the energy needed to be fully and positively engaged in work. In support of this notion, prior research has shown a clear link between depleted psychological resources and disengagement (Demerouti et al., 2001; Lee & Ashforth, 1996). Empirical investigations have similarly upheld that there is a negative relationship between experiencing discrimination and employee engagement (James et al., 2013; Jones et al., 2009). Thus, when Asian American employees are exposed to anti-Asian discrimination, we posit that this will lead to diminished engagement in their job:

Hypothesis 2: Nonwork racial discrimination is negatively associated with job engagement.

Nonwork Discrimination and Well-Being

We also anticipate that nonwork discrimination will be associated with an increase in depressive and physical health symptoms. Research on discrimination has long identified the health decrements that accompany race-based mistreatment, arguing that exposure to racial discrimination is a stressor that proximally heightens physiological stress responses and allostatic load (Currie et al., 2019; Pascoe & Richman, 2009) which then precipitates mental and physical ill health (Carter et al., 2017; Gee, 2002; Lee & Ahn, 2011; Pascoe & Richman, 2009). Moreover, research examining anti-Asian discrimination that has specifically occurred during the COVID-19 pandemic comports with these findings, upholding that discrimination is associated with a variety of indicators of mental health, including posttraumatic stress symptoms (Hahm et al., 2021), symptoms of depression (Lee & Waters, 2021; Wu et al., 2020) and anxiety (Lee & Waters, 2021), and lowered subjective well-being (Yang et al., 2020). Racial discrimination during the pandemic has further been linked to increased physical symptoms and sleep difficulties (Lee & Waters, 2021). More broadly, experiencing discrimination in work and nonwork settings has similarly been linked to a host of psychological and physical health indicators (Dhanani et al., 2018; Schmitt et al., 2014). Drawing on this literature, we hypothesize that nonwork racial discrimination will be associated with increased depressive and physical health symptoms among Asian American employees:

Hypothesis 3: Nonwork racial discrimination is positively associated with depressive symptoms.

Hypothesis 4: Nonwork racial discrimination is positively associated with physical health symptoms.

The Role of Coworker Compassion

We finally examine coworker compassion as a potential remedy for the negative spillover effects of nonwork discrimination. Compassion is characterized by three primary components: (1) the recognition of and attention to suffering, (2) feelings of empathic concern, and (3) a motivation to alleviate suffering (Dutton et al., 2014; Strauss et al., 2016). We propose that compassion received from one’s coworkers is a form of coworker support that may serve as a buffer against the expected negative effects of experiencing anti-Asian discrimination. This proposition builds on our COR theory framework which posits that access to resources, such as those offered by compassionate coworkers, can facilitate the restoration of lost resources (Hobfoll, 1989). Indeed, COR theory proposes that resource gains can stem the negative consequences of resource losses by bolstering stress resistance among employees and alleviating the degree to which one’s overall resources are taxed (Hobfoll et al., 2018). Correspondingly, we conceptualize coworker compassion as an important social resource that may help replenish the energetical resources lost due to discrimination, thus mitigating the negative effects that accumulate from resource loss. That is, receiving supportive and compassionate responses from coworkers surrounding anti-Asian discrimination can signal a resource gain and activate stress resistance among employees.

Although compassion is seldom examined from the perspective of those who receive it (Dutton et al., 2014), prior research on social support aligns with our above rationale. This work has demonstrated that receiving support from others protects against the harm of experiencing a multitude of hardships. For example, social support has been shown to buffer against developing depression and other mental health consequences after severe negative events (Dalgard et al., 1995, 2006; Lakey et al., 2011), increase adaptation to illness and bolster physical health (Wallston et al., 1983), and promote and sustain engagement and motivation (Constable & Russell, 1986; Isaksson et al., 2007; Othman et al., 2013). Research has also demonstrated the capacity of social support to generally reduce strain, mitigate perceived stressors, and moderate the stressor–strain relationship (Viswesvaran et al., 1999).

Drawing on the above rationale and addressing prior calls to examine the effects of compassion on those who receive it, we theorize that perceived compassion from coworkers will play an important role in buffering the effects of experienced anti-Asian discrimination on emotional exhaustion, engagement, depressive symptoms, and physical health symptoms:

Hypothesis 5: The relationship between nonwork racial discrimination and (a) emotional exhaustion, (b) engagement, (c) depressive symptoms, and (d) physical health will be moderated by coworker compassion such that the relationships will be weaker when coworker compassion is high.

Method

Participants and Procedure

Data were collected from participants who self-identified as Asian American, were employed full- or part-time, and resided within the USA. Participants were recruited through Qualtrics Panels, which is a third-party market research firm that distributes surveys to eligible participants. Data were collected at two time points in April 2021. The two surveys were separated by approximately 1 week to reduce concerns about common method variance. The Time 1 survey assessed participants’ experiences of discrimination in nonwork settings since the onset of the COVID-19 pandemic, their perceptions of coworker compassion, and three control variables (i.e., discrimination experienced at work, COVID-19 anxiety, trait negative affectivity). At Time 2, participants were asked to report their emotional exhaustion, job engagement, depressive symptoms, and physical health symptoms. Each survey asked participants to reflect on the incidents of anti-Asian racism and violence that occurred since the onset of the pandemic and to respond with that timeframe in mind.

A total of 401 participants who met our eligibility criteria completed the survey at Time 1 and 311 provided data again at Time 2. Of the 311 participants who completed both surveys, approximately half (50.2%) identified as women, and the average age was 33.57 (SD = 8.66). The majority of participants (78.8%) were employed full time, and participants worked an average of 36.73 h per week. Participants reported a variety of educational backgrounds with roughly half reporting a bachelor’s degree (47.9%), 24.8% reporting a master’s degree, and 5.8% reporting a PhD or higher. Participants also worked in a variety of occupations (e.g., healthcare, information technology, retail, education). t tests were conducted to compare participants who completed the Time 2 survey to those who did not, and results showed no significant differences in age, gender, work hours, COVID-19 anxiety, trait negative affectivity, experienced workplace discrimination, experienced nonwork discrimination, or coworker compassion, which suggests no evidence of nonresponse bias for the variables used in our analyses.

Nonwork Discrimination (T1)

Participants were asked to report their experiences of racial discrimination in everyday life that occurred since the onset of the COVID-19 pandemic. We assessed these experiences using the Everyday Discrimination Scale (Williams et al., 1997) (α = 0.96), which contains 9 items that ask about discrimination that occurred in a variety of routinely encountered interactions and contexts. An example item is “You were treated with less courtesy than other people were.” Given that the items could potentially assess discrimination that occurred in any interpersonal context, the instructions asked participants to only reflect on and report their experiences of discrimination that occurred in nonwork settings. Responses were given on a 6-point scale from never to almost every day.

Coworker Compassion (T1)

Coworker compassion was measured using a 5-item scale developed for this study. We chose to develop a measure because we wanted to capture the extent to which coworkers demonstrated compassion that was specific to the unique negative experiences Asian Americans have encountered during the COVID-19 pandemic rather than more general forms of compassion. Furthermore, compassion scales tend to focus on the tendency of an actor to be compassionate rather than on the experience of receiving compassion (Dutton et al., 2014; Strauss et al., 2016). We developed items to capture three core aspect of compassion: (1) the recognition of suffering (e.g., “My coworkers understand how I’m feeling after incidents of anti-Asian racism and violence”), (2) empathy toward suffering (e.g., “My coworkers express empathy after incidents of anti-Asian racism and violence occur”), and (3) motivation or action to relieve suffering (e.g., “My co-workers are supportive after incidents of anti-Asian racism and violence occur”). Developed items were also consistent with prior conceptualizations of the prosocial ways coworkers can respond toward targets of discrimination (see Dhanani and LaPalme, (2019) for review) which include demonstrating empathy, checking in on the target, and otherwise offering general emotional support to the target.

We generated an initial set of six items using the guidelines provided by Hinkin, (1998). We removed one item (i.e., “My coworkers really care about social justice”) because it did not reflect specific coworker behaviors/actions directed toward the participant. The retained items are shown in Table S1 (in supplemental materials), and the response scale ranged from 1 (strongly disagree) to 5 (strongly agree). The retained items were subjected to an exploratory factor analysis using principal axis factoring. Results indicated a single factor with an eigenvalue above 1 (eigen value = 3.50) and the factor explained 69.94% of the variance. Item loadings are shown in Table S1, and all loadings exceeded 0.73. We next assessed the internal consistency reliability of the scale which fell above acceptable cutoffs (α = 0.89). We finally conducted a confirmatory factor analysis to confirm the factor structure of the measure. Results for the CFA (Table S1) indicated acceptable fit (CFI = 0.95; TLI = 0.90; SRMR = 0.04; RMSEA = 0.17), and all factor loadings exceeded 0.72. Hinkin, (1998) also recommends replicating the CFA and reliability estimate in an independent sample. To do so, data were collected from a second sample of 193 Asian American employees. The CFA from the additional sample also demonstrated acceptable fit (CFI = 0.99; TLI = 0.97; SRMR = 0.02; RMSEA = 0.11), factor loadings (> 0.79), and reliability (α = 0.92). We also replicated the CFA within the subset of participants who only completed the Time 1 survey and similarly found acceptable fit (CFI = 0.98; TLI = 0.96; SRMR = 0.02; RMSEA = 0.12), factor loadings (> 0.72), and reliability (α = 0.90). Convergent and discriminant validity were assessed by examining the relationships between our developed measure and measures of coworker empathy (Batson et al., 1981) and organizational support specific to anti-Asian discrimination. Our measure had a correlation of 0.632 (p < 0.001) with coworker empathy, demonstrating that these constructs are related but not identical. We further found a correlation of 0.481 (p < 0.001) between the coworker compassion and organizational support measures.

Employee Engagement (T2)

We assessed participants’ engagement using Rich et al., (2010) 18-item scale. This scale measures three facets of engagement: physical engagement (6 items; example, “I devote a lot of energy to my job”), emotional engagement (6 items; example, “I feel positive about my job”), and cognitive engagement (6 items, “At work, I concentrate on my job”). Items were rated on a 5-point scale ranging from strongly disagree to strongly agree. Prior work has supported combining the three facets of engagement into a single, combined scale, and the reliability calculated across all 18 items was 0.91 in our sample. As with the other Time 2 measures, the instructions asked participants to report on their level of engagement during the time since the previous survey was completed.

Emotional Exhaustion (T2)

Participants reported their emotional exhaustion using the exhaustion subscale of the Maslach’s Burnout Inventory (Maslach and Jackson, (1981) (α = 0.92).1 This 8-item scale measures the degree to which people feel emotionally drained by their work, and a sample item is “I feel used up at the end of the workday.” We eliminated one item from the original scale which asks about working with patients given that our sample was diverse in their occupational backgrounds and may not work directly with patients or customers. The response scale is 1 (strongly disagree) to 5 (strongly agree).

Depressive Symptoms (T2)

We used the Patient Health Questionnaire (PHQ-9) (Kroenke et al., 2001) to assess depressive symptoms. This scale was designed to measure general symptoms of depression, including a loss of interest, feelings of hopelessness, and decrements in energy. Participants were asked to rate the extent to which they have been bothered by any of these symptoms since the Time 1 survey on a scale from 0 (not at all) to 3 (nearly every day). The item which assesses suicidality was removed to minimize the risk of participant distress. The reliability for the remaining 8 items was 0.90.

Physical Health Symptoms (T2)

Physical health symptoms were measured with Schat et al., (2005) Physical Health Questionnaire, which asks participants to indicate the frequency with which they have experienced physical health symptoms or illness. Examples include headaches, sleep disturbances, and stomach upsets, and responses are given on a 7-point frequency scale ranging from 1 (not at all) to 7 (all of the time). The final three items which assess experiences of colds and respiratory infections were removed given the short time frame during which symptoms were measured (i.e., approximately 1 week). The shortened 11-item scale demonstrated adequate reliability (α = 0.89).

Control Variables (T1)

COVID-19 has affected work experiences in important ways that may also affect employees’ well-being and job outcomes. That is, people may be experiencing anxiety about potentially contracting COVID-19 which could affect their overall well-being. We therefore controlled for COVID-19 anxiety using the Fear of COVID-19 Scale (Ahorsu et al., 2020) example item, “I am afraid of the novel coronavirus”; α = 0.90). Further, because discrimination may have also occurred at work, we also measured and controlled for racial/ethnic harassment experienced at work since the onset of the COVID-19 pandemic using Schneider et al.’s Ethnic Harassment Experiences Questionnaire (Schneider et al., 2000) (α = 0.96). Further, given that trait negative affectivity has been implicated as a potential confound in relationships between discrimination and outcomes (e.g., Dhanani et al., 2018), we controlled for trait NA in our analyses (PANAS) (Watson et al., 1988) (α = 0.94).

Analyses

Hypotheses were tested using regression models in which participant demographics (i.e., age, gender, work hours, work modality), COVID-19 anxiety, work discrimination, and trait negative affectivity were entered in step 1; nonwork discrimination and coworker compassion were entered in step 2; and the interaction between nonwork discrimination and compassion was entered in step 3. Separate models were tested for each outcome variable (i.e., job engagement, emotional exhaustion, depressive symptoms, and physical health symptoms). Continuous predictor variables and interaction terms were mean-centered to reduce concerns about multicollinearity. Significant interactions were probed by calculating the simple slopes for the relationship between nonwork discrimination and the relevant outcome variable at one standard deviation above and below the mean for coworker compassion (e.g., Hayes & Matthes, 2009).

Results

Means, standard deviations, and intercorrelations for all study variables are presented in Table 1, and results for the regression analyses are shown in Table 2. The first model, which examined emotional exhaustion, indicated a significant relationship between nonwork discrimination and emotional exhaustion (b = 0.22, p = 0.003, ΔR2 = 0.029), supporting Hypothesis 1. Further, consistent with Hypothesis 5a, the interaction between coworker compassion and nonwork discrimination was also significant (b =  − 0.10, p = 0.047, ΔR2 = 0.010). Tests of the simple slopes (Fig. 1) showed a significant positive relationship between perceived discrimination and emotional exhaustion when coworker compassion was low (b = 0.30, p = 0.001). However, when coworker compassion was high, this relationship was nonsignificant (b = 0.12, p = 0.172).

Table 1

Means, standard deviations, and intercorrelations for study variables

	M	SD	1	2	3	4	5	6	7	8	9	10	11	12	13
1. Age	33.57	8.66
2. Gender	1.50	.50	− .22*
3. Work hours	36.73	14.20	.29*	− .15*
4. COVID-19 anxiety	2.95	1.02	.08	− .01	− .03	.90
5. Negative affectivity	2.34	.98	− .14*	− .02	− .06	.61*	.94
6. Work modality	1.57	.50	.01	− .03	.03	.03	.00
7. Work discrimination	1.66	1.07	− .06	− .06	− .04	.29*	.45*	.01	.96
8. Nonwork discrimination	1.94	1.17	− .10	− .02	− .03	.31*	.49*	− .01	.84*	.96
9. Coworker compassion	3.33	.93	.07	− .09	− .09	.24*	.10	− .02	.10	.06	.89
10. Employee engagement	3.78	.71	.11	− .04	.10	.11*	-.03	− .06	.01	− .03	.14*	.91
11. Emotional exhaustion	3.26	.93	− .22*	.11*	.01	.30*	.42*	− .02	.12*	.22*	− .06	− .16*	.92
12. Depressive symptoms	1.91	.65	− .18*	.12*	− .07	.31*	.54*	− .03	.22*	.31*	.01	− .15*	.61*	.90
13. Physical health symptoms	3.07	1.09	− .12*	.15*	− .05	.31*	.47*	− .01	.23*	.32*	.08	− .11*	.49*	.66*	.89

N = 311; gender is coded 1 = male, 2 = female; work modality is coded 1 = working in person and 2 = working remotely; reliability coefficients are presented on the diagonal

* p < .05

Table 2

Nonwork discrimination and coworker compassion predicting employee well-being outcomes

	Employee engagement			Emotional exhaustion			Depressive symptoms			Physical health Symptoms
Variable	B	SE	(Δ)R2	B	SE	(Δ)R2	B	SE	(Δ)R2	B	SE	(Δ)R2
Step 1
Gender	− .04	.08		.17	.10		.14*	.07		.32**	.11
Age	.01	.01		− .02*	.01		− .01	.00		.00	.01
Work hours	.00	.00		.01	.00		.00	.00		.00	.00
COVID-19 anxiety	.14**	.05		.12*	.06		− .01	.04		.04	.07
Negative affectivity	− .12*	.06		.36***	.07		.37***	.05		.48***	.08
Work modality	− .08	.08		− .05	.10		− .02	.06		− .02	.11
Work discrimination	.03	.04		− .10	.05		− .02	.03		.04	.06
			.041			.239			.317			.249
Step 2
Nonwork discrimination	− .06	.07		.22**	.08		.13*	.05		.23**	.09
Coworker compassion	.08	.05		− .08	.05		− .02	.04		.06	.06
			.013			.029			.016			.019
Step 3
Nonwork discrimination × Coworker compassion	.10*	.04		− .10*	.05		− .07*	.03		− .03	.06
			.015			.010			.010			.001

N = 306 for the models predicting employee engagement, emotional exhaustion, and physical health symptoms; 299 for the model predicting depressive symptoms; gender is coded such that 1 = male and 2 = female; work modality coded such that 1 = working in person and 2 = working remotely; all continuous predictor variables are mean centered

* p < .05, ** p < .01, *** p < .001

Fig. 1

Interaction between nonwork discrimination and coworker compassion on emotional exhaustion

The second model examined the relationship between nonwork discrimination and engagement and results showed that, after accounting for the control variables, perceived discrimination was not significantly related to engagement (b =  − 0.06, p = 0.376, ΔR2 = 0.013). This does not support Hypothesis 2. That finding was qualified by a significant interaction between nonwork discrimination and coworker compassion (b = 0.10, p = 0.027, ΔR2 = 0.015), supporting Hypothesis 5b. Results (Fig. 2) for the simple slopes revealed a marginally significant negative relationship between discrimination and engagement when coworker compassion was low (b =  − 0.14, p = 0.073) and a nonsignificant positive relationship when coworker compassion was high (b = 0.04, p = 0.597).

Fig. 2

Interaction between nonwork discrimination and coworker compassion on employee engagement

The model predicting depressive symptoms suggested a significant positive relationship between nonwork discrimination and depressive symptoms (b = 0.13, p = 0.012, ΔR2 = 0.016) and a significant interaction between nonwork discrimination and coworker compassion (b =  − 0.07, p = 0.039, ΔR2 = 0.010), in support of Hypotheses 3 and 5c. A probe of this interaction (Fig. 3) revealed that nonwork discrimination was significantly positively related to depressive symptoms when coworker compassion was low (b = 0.18, p = 0.003) but had a nonsignificant relationship when coworker compassion was high (b = 0.06, p = 0.367).

Fig. 3

Interaction of nonwork discrimination and coworker compassion on depressive symptoms

Finally, in support of Hypothesis 4, results for physical health symptoms revealed that nonwork discrimination significantly predicted physical health symptoms (b = 0.23, p = 0.008, ΔR2 = 0.019). However, counter to Hypothesis 5d, the interaction between perceived discrimination and coworker compassion was not significant (b =  − 0.03, p = 0.652, ΔR2 = 0.001). Together, our findings suggest that experiences of nonwork discrimination negatively impact general well-being and also spillover to affect work-related outcomes among Asian American employees but that these effects were only present when coworker compassion was low.

Discussion

The current paper examined the effects of experiencing anti-Asian discrimination in nonwork contexts during the COVID-19 pandemic on the general and job-related well-being of Asian American employees as well as how coworker compassion moderated these relationships. Results indicated that experienced nonwork discrimination had negative consequences for Asian Americans’ general mental and physical health symptoms. Experiences of nonwork discrimination similarly appeared to spillover to affect work-related exhaustion. Furthermore, significant interactions were found between experienced nonwork discrimination and coworker compassion for engagement, emotional exhaustion, and depressive symptoms such that these outcomes were exacerbated when coworker compassion was low. The implications of these findings for theory and practice are discussed below.

Theoretical Implications

Findings from the current study contribute to the literature on spillover effects and, more specifically, our understanding of how mistreatment that originates in nonwork contexts affects employees, in several ways. First, in accordance with spillover theory, we found that nonwork discrimination is related to work-specific (i.e., emotional exhaustion) and general well-being outcomes. The literature has only recently begun to explore how mistreatment that occurs in the nonwork domain may affect employees and, to date, studies have primarily examined mistreatment experienced within one’s family (e.g., Bai et al., 2016; Lim & Tai, 2014). The current study expands on that work to recognize how mistreatment experienced in other nonwork contexts may similarly spill over into the workplace. Indeed, nonwork discrimination was found to relate to emotional exhaustion experienced at work; however, there was no direct relationship between nonwork discrimination and engagement. This may be because, as proposed by the job demands-resources model, engagement is primarily determined by available resources rather than demands (Bakker et al., 2005; Demerouti et al., 2001). Taken together, though, our findings demonstrate the importance of expanding our conceptualizations of the spillover effects of mistreatment to include mistreatment experienced in non-family contexts.

This also draws attention to the broader tendency to focus on family-related stressors when considering the work–nonwork interface, even outside of the context of mistreatment. We argue that theoretical models and empirical examinations of the work–nonwork interface would benefit from incorporating a more comprehensive set of experiences originating in the nonwork domain. Discrimination may be particularly important to consider given that it may uniquely disadvantage groups of employees who are more likely to experience discrimination both on and off the job. Subsequent research should continue to examine the ways that nonwork discrimination can harm employees, the work-related factors that may exacerbate or buffer that harm, and what theoretical mechanisms explain how the effects of nonwork discrimination are transmitted to the workplace. The need to understand the effects of nonwork discrimination is perhaps even more pressing as the rise of remote work has further blurred the boundaries of the work–nonwork interface. Remote work may thus make it even easier for the harm of nonwork discrimination to seep into the workplace, and subsequent research efforts should seek to understand how work modality may serve as a boundary condition of nonwork discrimination spillover.

Relatedly, conceptual work attempting to understand inclusion and potential inequities among employees would also benefit from incorporating nonwork stressors that disproportionately affect specific subgroups of employees. It is possible that, even when workplace environments are equitable, employees belonging to traditionally minoritized groups may face disadvantages due to events occurring in their nonwork lives. Models that attempt to capture sources of inequity may therefore need to expand beyond the borders of the work environment to also account for nonwork conditions and experiences that may differentially affect such employees.

Our study further contributes to the understanding of how coworkers can meaningfully shape employees’ experiences. General coworker support has been found to have widespread positive effects for employees and to mitigate the harm of a number of job stressors and negative workplace experiences (e.g., Bakker et al., 2010; Viswesvaran et al., 1999). Our findings extend that work to suggest that coworkers’ compassionate responses following discrimination may also help combat the harm of discriminatory experiences. This may argue for the utility of studying specific forms of coworker support in addition to more general perceptions of support. However, we did not find that coworker compassion offered the expected buffering effect for physical health symptoms. It is possible that somatic responses to discrimination are more difficult to combat than other outcomes. That is, similar to challenge stressors which still elicit negative health consequences despite being appraised positively (Mazzola & Disselhorst, 2019; Podsakoff et al., 2007), receiving support may help with the appraisal process and terminate ruminative thinking that results in the depletion of attentional and energetical resources. Yet, these benefits may not terminate the relatively automatic physiological response to stressors that accompanies discrimination. This remains speculative given that our data cannot test this proposition and future work is needed to confirm the relationship between coworker compassion and physical health symptoms.

Finally, our findings also have implications for our understanding of Asian Americans’ experiences in work and nonwork contexts. Asian Americans are often viewed as the “model minority,” and this belief is often accompanied by the assumption that Asian Americans are accepted by society and free of the barriers that other racial/ethnic groups experience (Chou & Feagin, 2010; Lai, 2013). Contributing to a growing body of work challenging the myth of the model minority, our study demonstrates that Asian Americans are indeed vulnerable to experiencing negative treatment and the associated downstream consequences, leading to disadvantages within organizations. These findings suggest a need to combat beliefs in the model minority myth and to take efforts to better include Asian Americans within diversity, equity, and inclusion efforts. The need to do so was particularly heightened by the COVID-19 pandemic, but mistreatment targeting Asian Americans will persist past the pandemic and organizations should be ready to address it.

Practical Implications

Results of the present study can also inform organizational practice. Organizations should engage in efforts to promote the well-being of their racial/ethnic minority employees in general, but our findings suggest the need to do so may be heightened during periods of rising racial animus. This aligns with recent work on mega-threats (Leigh & Melwani, 2019; 2022). Though the management literature has only begun to recognize the importance of mega-threats, our findings underscore the ramifications of such threats as well as the need for organizational action in the presence of them. Organizations would correspondingly benefit from mobilizing resources during times in which mega-threats are present.

Encouragingly, our findings suggest that fostering coworker compassion amid times of widespread racial tensions may be one way through which organizations can bolster employee well-being and buffer against the negative outcomes of discrimination. To date, effective interventions for reducing the harm of discrimination have been somewhat elusive, and coworker compassion offers a potential fruitful and low-cost intervention method that organizations could employ. To capitalize on the benefits of coworker compassion, though, organizations need to devote effort to creating environments that encourage compassionate coworker responses. As one pathway, established inclusive climates may help to encourage coworker compassion during difficult times. A key facet of inclusive climates is the integration of differences, which “reflects collective expectations and norms regarding the openness with which employees can enact and engage core aspects of their self-concept… without suffering unwanted consequences” (Nishii, 2013, p. 1756). This open expression of one’s identity may enable employees to discuss their experiences of racial derogation, and the lack of fear of negative consequences will further allow them to seek the support and compassion of others.

Yet, despite these benefits, organizations should not simply rely on their long-standing inclusive climates and may need to engage in more targeted efforts to respond to times of acute racial stress. There are a number of ways that organizations can take direct action to foster compassion among employees, including encouraging supervisors to model compassionate behaviors (Wayne et al., 1997), being mindful of the nonwork experiences employees may be facing and how they may manifest as barriers in the workplace, and encouraging and initiating open and respectful communication about ongoing events, which may facilitate support seeking behaviors. Organizations can also directly communicate about the importance of coworker compassion and engage in conversations surrounding ways that coworkers can effectively respond when employees disclose experiences of discrimination. Investing in facilitating a positive organizational climate might also help amplify the positive effects of coworker compassion given findings that resources are more effective in environments that encourage their use (Hobfoll et al., 2018). The benefits of fostering organizational climates that encourage coworker compassion following exposure to discrimination may thus be multifaceted.

Limitations and Future Directions

Although this study offers important insight into how experiences of anti-Asian discrimination impact the health and well-being of Asian American employees, there are limitations to note. First, our study asked participants to recall experiences of discrimination that have occurred since the beginning of the pandemic. A potential limitation of this methodology is that asking participants to engage in retrospective recall may lead to issues of forgetting or misattributions of events that occurred (Schwarz & Sudman, 1994). As such, research in this area may benefit from alternative methodologies such as daily diary studies that allow participants to report discrimination events with reduced recall times.

Additionally, although nonwork discrimination and well-being outcomes were measured at different times, the design of our study does not enable us to draw causal conclusions about the relationships examined in the current study. Indeed, the time lag between our two measurement periods was 1 week, which may not have been sufficient for establishing temporal precedence. However, our primary purpose in including a time lag was to reduce concerns about common method variance, such as transient error, and a lag of 1 week may have been appropriate for that purpose. Future research would benefit from more longitudinal examinations of the long-term health and well-being consequences of nonwork discrimination.

Another limitation to note is that our data were collected during the COVID-19 pandemic, which caused significant shifts in the work landscape for many employees. For example, some employees transitioned into online environments where coworker compassion may take different forms than support offered in person. These shifting workplace conditions may impact the generalizability of the study results to a non-pandemic setting. Next, we also focused on experiences of general discriminatory behaviors rather than unique manifestations of discrimination related to the COVID-19 pandemic (e.g., being blamed for the pandemic). We took this approach because there are often unclear motives behind acts of discrimination and many general forms of discrimination could result from the increased prejudice spurred by the pandemic. Only measuring unique manifestations could thus fail to capture the full range of Asian American employees’ experiences. However, understanding the consequences of COVID-19-specific discrimination is also worthy of attention, and subsequent research should devote attention to this topic.

Moreover, the effect sizes associated with the interaction terms in our models were relatively modest which might suggest they have little practical value. However, recent research has challenged the interpretation of effect sizes associated with the interaction term as an indicator of the practical significance of the effect (Vancouver et al., 2021). As demonstrated in this work, effect sizes can substantially underestimate the role of the interactive relationship, and the authors instead recommend focusing on the statistical significance rather than the effect size estimates.

A final limitation in our study is that our analyses did not examine potential differences in the experiences of, or well-being consequences felt by, employees of different Asian ethnicities. Due to the believed geographic origin of the COVID-19 virus, employees of Chinese descent and other East Asian ethnicities may have experienced higher levels of discrimination during the pandemic as compared to other groups, such as South Asian employees. This group of employees may similarly experience stronger decrements in health in response to discrimination than employees of other Asian ethnicities. As the experiences of Asian American employees are not monolithic, we encourage future research to examine differences in discrimination experiences for employees of various Asian ethnicities.

Despite these limitations, our study highlights numerous avenues for future research. One important future direction is to understand what types of coworker compassion are most beneficial for reducing negative outcomes related to discrimination. Support from coworkers can come in many different forms, and the quality and types of relationships held between employees may meaningfully impact the degree to which coworker compassion can mitigate negative outcomes. Additionally, building on our findings for coworker compassion, researchers should also examine what home- or family-related resources may also impact spillover into the workplace. Finally, our study also produced an unexpected result wherein our analyses did not show a significant interaction between nonwork discrimination and coworker compassion for physical health. Future research should investigate why coworker compassion buffered harm for mental health outcomes but not physical health as well as what other job resources may provide benefits for physical health.

Conclusion

This study examined the spillover effects of nonwork discrimination experienced during the COVID-19 pandemic on Asian American employees’ well-being and workplace outcomes and whether coworker compassion mitigated those negative consequences. Results suggest that experiences of discrimination in the nonwork domain can spillover to affect work-related exhaustion, engagement, and mental and physical health symptoms, particularly when coworker compassion is low. This underscores the need for organizational awareness regarding the cross-domain effects of nonwork discrimination. Finally, this study also suggests that resources within the work environment can provide important benefits for employees who are attempting to cope with negative nonwork experiences. Scholars and practitioners seeking to establish inclusive workplaces should consider these findings to gain a more holistic understanding of the barriers facing Asian American employees and how such employees may be supported within the workplace.

Below is the link to the electronic supplementary material.

Supplementary file1 (DOCX 14 KB)