THE IMPACT OF UNETHICAL LEADER-REQUESTS ON EMPLOYEES’ ANGER, 

ANXIETY, AND FAMILY LIVES 

 

 

 

 

 

 

 

 

 

 

 

 

 

by 

 

FENG QIU 

 

 

 

 

 

 

 

 

 

 

 

 

 

A DISSERTATION 

 

Presented to the Department of Management 

and the Graduate School of the University of Oregon 

in partial fulfillment of the requirements 

for the degree of 

Doctor of Philosophy 

 

June 2020



 

 

ii 
 

DISSERTATION APPROVAL PAGE 

 

Student: Feng Qiu 

 

Title: The Impact of Unethical Leader-requests on Employees’ Anger, Anxiety, and 

Family Lives 

 

This dissertation has been accepted and approved in partial fulfillment of the 

requirements for the Doctor of Philosophy degree in the Department of Management by: 

 

David Wagner   Chair 

Keith Leavitt   Core Member 

Debra Shapiro   Core Member 

Jiao Zhang   Core Member 

Sanjay Srivastava  Institutional Representative 

 

and 

 

Kate Mondloch  Interim Vice Provost and Dean of the Graduate School 

 

Original approval signatures are on file with the University of Oregon Graduate School. 

 

Degree awarded June 2020. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



 

iii 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

© 2020 Feng Qiu 

This work is licensed under a Creative Commons 

Attribution-NonCommercial-NoDerivs (United States) License. 

 

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



 

iv 
 

 

DISSERTATION ABSTRACT 

 

Feng Qiu 

 

Doctor of Philosophy 

 

Department of Management 

 

June 2020 

 

Title: The Impact of Unethical Leader-requests on Employees’ Anger, Anxiety, and 

Family Lives 

 

 

This dissertation aims to explore the potential non-work consequences of 

unethical leader-requests. Specifically, it examines how unethical leader-requests can 

trigger anxiety and anger in employees, which in turn harmfully influence employees’ 

insomnia, emotional exhaustion at home, and interactions with family members. In 

addition, this dissertation examines whether employees’ moral identity and responsibility 

displacement propensity will serve as two moderators that affect the degree to which they 

emotionally and behaviorally respond to unethical leader-requests. A three-wave field 

survey, a laboratory experiment, and an experience sampling method study were 

conducted to collectively improve the internal and external validity of the findings. 

Overall, the findings suggest that employees feel anxious and angry when they are 

requested by their leaders to engage in unethical behavior and that the negative emotions 

can spill over to employees’ family domain to harmfully impact their family lives. 

Implications and future directions will be discussed. 

 

 

 

 

 



 

v 
 

 

CURRICULUM VITAE 

 

 

NAME OF AUTHOR: Feng Qiu 

 

 

GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED: 

 

University of Oregon, Eugene, OR 

Oregon State University, Corvallis, OR 

Wenzhou University, Wenzhou, China 

 

 

DEGREES AWARDED: 

 

Doctor of Philosophy in Management, 2020, University of Oregon 

Master of Business Administration, 2015, Oregon State University 

Bachelor of Business Administration, 2012, Wenzhou University 

 

 

AREAS OF SPECIAL INTEREST: 

 

Behavioral Ethics 

Unethical Leadership 

Whistleblowing 

 

 

PROFESSIONAL EXPERIENCE:  

 

Dissertation Research Fellow, University of Oregon, Eugene, OR, 2019-2020 

Course Instructor, University of Oregon, Eugene, OR, 2017-2019 

Research Assistant, University of Oregon, Eugene, OR, 2015-2019 

 

 

GRANTS, AWARDS, AND HONORS: 

 

 University of Oregon Dissertation Research Fellowship, 2019-2020 

Institute of Cognitive and Decision Sciences Dissertation Research Award, 2019 

 Robin and Roger Best Research Award, 2019 

Robin and Roger Best Research Award, 2018 

 Lundquist College of Business Ph.D. Fellowship, 2015-2019 

 

 

PUBLICATIONS: 

 



 

vi 
 

 

Liu, X., Liao, H., Derfler-Rozin, R., Zheng, X., Wee, E. X. M., & Qiu, F. (2020). 

In line and out of the box: How ethical leaders help offset the negative 

effect of morality on creativity. Journal of Applied Psychology. Advance 

online publication. https://doi.org/10.1037/apl0000489 

 

Leavitt, K., Qiu, F., & Shapiro, D. L. (2019). Using Electronic Confederates for 

Experimental Research in Organizational Science. Organizational 

Research Methods. https://doi.org/10.1177/1094428119889136 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

https://psycnet.apa.org/doi/10.1037/apl0000489
https://doi.org/10.1177/1094428119889136


 

vii 
 

 

ACKNOWLEDGMENTS 

I would like to sincerely acknowledge many individuals, without whom I would 

never complete my doctoral studies and become the person I am today. First, I owe an 

enormous debt to my advisor, David Wagner, whose insights, wisdom, and patience 

steered me through the rigors in this journey. He deserves credit as a large contributor to 

my achievements at several stages of my doctoral studies. Second, I would like to express 

my sincere gratitude to my committee members, Keith Leavitt, Debra Shapiro, Jiao 

Zhang, and Sanjay Srivastava, for their thoughtful and constructive comments that help 

me improve the quality of this dissertation. Also, I would like to thank the faculty of the 

Department of Management, the PhD program, the staff at Lundquist College of Business, 

and the staff at the Research Compliance Services office of the University of Oregon for 

their continued support.  

 

 

 

 

 

 

 

 

 

 

 



 

viii 
 

 

 

 

 

 

 

 

 

I dedicate this dissertation to my family. Thank my wife, Guanjun, for changing my life 

and for making who I am today. Thank my parents for always believing in me and for 

giving me the courage to face the storm in this journey. Finally, thank my two wonderful 

and lovely children, Austin and Lillis, for providing endless inspiration. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



 

ix 
 

 

TABLE OF CONTENTS 

Chapter Page 

 

CHAPTER I: INTRODUCTION ................................................................................. 1 

 Introduction ............................................................................................................ 1 

 The Overall Structure of the Dissertation .............................................................. 3 

 

CHAPTER II: LITERATURE REVIEW .................................................................... 5 

 Leaders’ Ethicality and Unethical Requests .......................................................... 5 

 Overview of Ethical Leadership ...................................................................... 5 

 Research on Ethical Leadership ....................................................................... 5 

 Overview of Unethical Leadership .................................................................. 7 

 Research on the Direct Form of Unethical Leadership .................................... 9 

 Research on the Indirect Form of Unethical Leadership ................................. 12 

 Unethical Leader-requests ............................................................................... 14 

 Cognitive Appraisal Theory, Anger, and Anxiety ................................................. 19 

Cognitive Appraisal Theory ............................................................................. 19 

Overview of Anger .......................................................................................... 22 

Research on Anger ........................................................................................... 23 

Overview of Anxiety........................................................................................ 26 

Research on Anxiety ........................................................................................ 27 

 Work-family Spillover ........................................................................................... 29 

Overview of Work-family Spillover Theories ................................................. 29 

 Sleep, Emotional Exhaustion, and Interactions with Family ................................. 32 



 

x 
 

 

Chapter Page 

 

Physiological, Mental, and Social Wellbeing .................................................. 32 

Employee Sleep ............................................................................................... 33 

Employee Emotional Exhaustion ..................................................................... 36 

Employee Negative Family Interactions .......................................................... 39 

 Moral Identity ........................................................................................................ 42 

Overview of Moral Identity ............................................................................. 42 

Research on Moral Identity .............................................................................. 44 

Integration of Moral Identity and Leaders’ Ethicality ..................................... 45 

 Moral Disengagement ............................................................................................ 47 

Overview of Moral Disengagement ................................................................. 47 

Moral Disengagement as a Behavioral Propensity .......................................... 49 

 

CHAPTER III: THEORETICAL MODEL AND HYPOTHESES ............................. 52 

 Cognitive Appraisal, Unethical Requests, Anxiety, and Anger ............................. 52 

 Spillover of Anxiety and Anger across Work-family Boundary ........................... 55 

 Moral Identity as a Moderator ............................................................................... 60 

 Responsibility Displacement Propensity as a Moderator ...................................... 64 

 

CHAPTER IV: METHODS ......................................................................................... 67 

 Study 1 (Field Survey) ........................................................................................... 67 

Population and Sample Selection..................................................................... 67 

Study Procedure ............................................................................................... 68 



 

xi 
 

 

Chapter Page 

 

Measures .......................................................................................................... 69 

Data Analysis Procedure and Strategy ............................................................. 73 

Confirmatory Factor Analysis (CFA) .............................................................. 73 

Study 1 Results ................................................................................................ 75 

Study 1 Discussion ........................................................................................... 84 

 Study 2 (Lab Experiment) ...................................................................................... 86 

Population and Sample Selection..................................................................... 86 

Experimental Procedure ................................................................................... 87 

Experimental Intervention ............................................................................... 89 

Measures .......................................................................................................... 92 

Data Analysis Procedure and Strategy ............................................................. 93 

Study 2 Results ................................................................................................ 93 

Study 2 Discussion ........................................................................................... 96 

 Study 3 (Experience Sampling Method Study) ..................................................... 96 

Population and Sample Selection..................................................................... 97 

Study Procedure ............................................................................................... 99 

Data Cleaning Process ..................................................................................... 99 

Measures .......................................................................................................... 101 

Data Analysis Procedure and Strategy ............................................................. 104 

Study 3 Results ................................................................................................ 105 

Study 3 Discussion ........................................................................................... 109 

 



 

xii 
 

 

Chapter Page 

 

CHAPTER V: DISCUSSION ...................................................................................... 112 

 Theoretical and Practical Implications................................................................... 112 

Consequences of Unethical Leader-requests ................................................... 112 

Harmfulness of Unethical Leadership ............................................................. 114 

Employee Wellbeing ........................................................................................ 115  

 Limitations and Remedies...................................................................................... 117 

The Experiment ................................................................................................ 117 

Proposed Experiment 2.1 ................................................................................. 121 

 Future Directions ................................................................................................... 125 

 Conclusion ............................................................................................................. 129 

REFERENCES CITED ................................................................................................ 130 

 

 

 

 

 

 

 

 

 

 

 



 

xiii 
 

 

LIST OF FIGURES 

Figure Page 

 

Figure 1. Theoretical model ......................................................................................... 3 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



 

xiv 
 

 

LIST OF TABLES 

Table Page 

 

Table 1. Confirmatory factor analyses for Study 1 variables ...................................... 74 

Table 2. Means, standard deviations, and correlations among Study 1 variables ....... 75 

Table 3. Mediation analyses for anxiety in Study 1 ..................................................... 76 

Table 4. Mediation analyses for anger in Study 1 ....................................................... 77 

Table 5. Indirect effects via anxiety contingent on moral identity in Study 1 ............. 79 

Table 6. Indirect effects via anger contingent on moral identity in Study 1 ................ 81 

Table 7. Indirect effects via anxiety contingent on responsibility displacement  

propensity in Study 1 ................................................................................................... 83 

 

Table 8. Indirect effects via anger contingent on responsibility displacement  

propensity in Study 1 ................................................................................................... 84 

 

Table 9. Variance components of null models for Level-1 variables in Study 3 ......... 105 

Table 10. Means, standard deviations, and correlations among Study 3 variables ..... 106 

Table 11. Results for anxiety and anger in Study 3 ..................................................... 107 

Table 12. Results for indirect effect on final dependent variables in Study 3 ............. 108 

 

 

 

 

 

 



 

1 
 

 

CHAPTER I: INTRODUCTION 

Introduction  

Leaders’ unethicality has become a growing social concern. In the business world, 

leaders frequently engage in unethical leader behavior (M. E. Brown & Mitchell, 2010). 

It is estimated that unethical leader behavior costs U.S. firms billions of dollars annually 

as a function of increased absenteeism, health care expense, diminished productivity, and 

costs associated with the lawsuit (Detert, Treviño, Burris, & Andiappan, 2007; Tepper, 

Duffy, Henle, & Lambert, 2006). News reports about corporate scandals often reveal that 

upper management engages in unethical behavior. Moreover, they even drag their 

employees into the mire by requesting their employees to commit transgressions. For 

example, the top managers of some infamous companies such as Enron (Healy & Palepu, 

2003) and Theranos (Carreyrou, 2018) have been found to request their employees to 

engage in extremely unethical business practices such as using deceptive accounting 

practices and even using problematic and fraudulent procedures to conduct blood tests for 

patients. More commonly, leaders often request their followers to engage in various more 

subtle forms of unethical behavior such as using unfair hiring process, overselling a 

product, and remaining silent about morally questionable behaviors. In management, this 

unethical leader behavior is identified as making unethical leader-requests which occurs 

when leaders request their subordinates to engage in morally questionable or unethical 

behavior (Desai & Kouchaki, 2017). By making unethical requests, leaders can fulfill 

their “sinful wish” without even directly engaging in the unethical behavior themselves 

due to their considerable influence and power over subordinates (French Jr & Raven, 

1959; Shapiro & Von Glinow, 2007), leading to the prevalence of unethical leader-



 

2 
 

 

requests at work. Despite their prevalence in the business world, unethical leader-requests 

have long been overlooked by organizational researchers. Research on unethical leader-

requests is still in its infancy, and more efforts are needed to establish a comprehensive 

understanding of its origins and consequences.  

This dissertation is dedicated to expanding this line of inquiry by exploring the 

potential non-work consequences of unethical leader-requests. Specifically, I aim to 

answer three important but under-investigated questions. First, given that people 

generally consider themselves just, virtuous, and moral (Tappin & McKay, 2017) and do 

not want to be a party to unethical behavior, how will employees emotionally respond to 

unethical leader-requests? Second, if unethical leader-requests arouse emotional 

responses in employees, how can those emotional responses spill over to non-work 

domains to influence their wellbeing-related state (e.g., emotional exhaustion) and 

wellbeing-related behavior (e.g., sleep, interactions with family)? Third, is every 

employee of whom an unethical request is made equally subject to the effects of unethical 

leader-requests?  

Drawing upon Lazarus’s (1991a, 1991b) cognitive appraisal theory of emotion 

and Staines’s (1980) spillover theory, this dissertation aims to investigate how unethical 

leader-requests can harmfully influence employees’ sleep, emotional exhaustion at home, 

and interactions with family members, three wellbeing-related aspects of employees’ 

personal lives that are critical to employees’ physical, mental, and social wellbeing, 

respectively (Lazarus & Folkman, 1984). In addition, I propose that employees’ anxiety 

and anger in response to unethical leader-requests explain the effects of unethical leader-

requests. Lastly, I examine whether employees’ moral identity and responsibility 



 

3 
 

 

displacement propensity will serve as two moderators that affect the degree to which they 

feel angry and anxious in response to unethical leader-requests, subsequently influencing 

the three wellbeing-related aspects (see Figure 1 for the theoretical model)1. 

 

 

Figure 1. Theoretical model. 

 

 

The Overall Structure of the Dissertation 

In Chapter II, I will first conduct a comprehensive and systematic literature 

review on the focal variables and the major theories upon which my dissertation will 

draw. The first purpose of the literature review is to define all key variables of interest. 

The second purpose is to provide both theoretical and empirical backgrounds for the 

research questions of this dissertation. To do so, I will elaborate on the most commonly 

used theoretical perspectives and summarize the main findings in the field of each focal 

 
1 Conflictual family interactions and family withdrawal in the model stand for conflictual interactions with 

family members and withdrawal from family interactions respectively. 



 

4 
 

 

variable. In Chapter III, I will discuss how the research questions of this dissertation and 

the relationships among focal variables are inspired and supported by existing literature. 

In addition, I will list the main hypotheses. In Chapter IV, I will describe a series of 

methods I used to test the hypotheses. The description will include the background 

knowledge of each method, the justification of the choice of each method, both 

advantages and limitations of the method, as well as detailed description of sample 

selection, procedure, measures, and data analysis strategy. In Chapter V, I will conclude 

by summarizing the theoretical insights of my findings and discussing the limitations and 

future directions.  

 

 

 

 

 

 

 

 

 

 

 

 

 

 



 

5 
 

 

CHAPTER II: LITERATURE REVIEW  

Leaders’ Ethicality and Unethical Requests 

 Overview of Ethical Leadership. Specific research on ethical leadership can be 

traced back to the early 2000s (Treviño, Brown, & Hartman, 2003; Trevino, Hartman, & 

Brown, 2000). The early research on ethical leadership focused on defining leader 

ethicality from a descriptive perspective. M. E. Brown, Treviño, and Harrison (2005) 

defined ethical leadership as “the demonstration of normatively appropriate conduct (e.g., 

honesty, trustworthiness, fairness, and care) through personal actions and interpersonal 

relationships, and the promotion of such conduct to followers through two-way 

communication, reinforcement, and decision-making” (p. 120). Treviño and her 

colleagues (2002) describe ethical leadership along two dimensions, namely moral 

person and moral manager (Trevino et al., 2000). The moral person dimension refers to 

the quality of being a moral person and reflects the extent to which the leader is seen as 

consistently moral in both personal and professional lives. The moral manager dimension 

refers to the extent to which the leader uses their power and status to model, promote, and 

maintain ethical behavior at work. In sum, “moral persons have a reputation for being fair 

and principled while moral managers set and communicate ethical standards and use 

rewards and punishments to ensure those standards are followed” (M. E. Brown & 

Mitchell, 2010, p. 584).  

Research on Ethical Leadership. In the past two decades, a tremendous amount 

of research endeavor has been exerted to explore a wide range of outcomes of ethical 

leadership (M. E. Brown & Treviño, 2006; M. E. Brown et al., 2005; D. M. Mayer, 

Kuenzi, Greenbaum, Bardes, & Salvador, 2009; Walumbwa & Schaubroeck, 2009). Most 



 

6 
 

 

research tends to use social learning theory (Bandura & Walters, 1977) and social 

exchange theory (Blau, 1964) to explain why ethical leadership can have a powerful 

impact.  

On the one hand, according to social learning theory (Bandura & Walters, 1977), 

people observe and model others’ behavior to learn how to interact and communicate 

through a socialization process—the “monkey see, monkey do” effect (Robinson & 

O'Leary-Kelly, 1998). In addition, social learning theory suggests that the likelihood of 

the internalization of behavioral patterns will increase as the credibility and legitimacy of 

the model increase. In other words, if the observed individual has high credibility, his/her 

actions will be more legitimate and appropriate to model. Leaders are usually perceived 

to be experienced, knowledgeable, and powerful, boosting their credibility and legitimacy 

in the eyes of followers and leading them to be an impactful role model.  

On the other hand, according to social exchange theory, people’s social and 

interpersonal behavior is characterized by an exchange of resources (Blau, 1964; Homans, 

1958; Thibaut & Kelley, 1959). Social exchange is rooted in the universal reciprocity 

norm (Gouldner, 1960) and is defined as “voluntary actions of individuals that are 

motivated by the returns they are expected to bring and typically in fact bring from 

others” (Blau, 1964, p. 91). The universal reciprocity norm represents a principal 

component of individuals’ moral codes that consist of two fundamental principles: 

“people should help those who have helped them, and people should not injure those who 

have helped them” (Gouldner, 1960, p. 171). It is well known that the pressure of social 

exchange greatly shapes and constrains interpersonal behavior and is so strong that “face 

is lost forever if a worthy return is not made” (Mauss, 1923, p. 41). Thus, when 



 

7 
 

 

employees believe their leaders have treated them in an ethical, fair, or just manner, they 

feel obligated to behave positively in return.  

Based on both social learning and social exchange theory, research on the 

outcomes of ethical leadership has been very fruitful. At the individual level, ethical 

leadership is found to positively predict a variety of desirable employee outcomes such as 

perceived ethical climate, job satisfaction, prosocial behavior, voice behavior, 

whistleblowing, extra effort on the job, perceived job autonomy, perceived task 

significance, and organizational commitment (M. E. Brown et al., 2005; D. M. Mayer, 

Nurmohamed, Treviño, Shapiro, & Schminke, 2013; Neubert, Carlson, Kacmar, Roberts, 

& Chonko, 2009; Piccolo, Greenbaum, Hartog, & Folger, 2010; Toor & Ofori, 2009; 

Walumbwa & Schaubroeck, 2009). At the group level, ethical leadership can promote 

organizational citizenship behavior and psychological safety and halt workplace deviant 

behaviors (D. M. Mayer et al., 2009; Walumbwa & Schaubroeck, 2009). With regard to 

the antecedents of ethical leadership, researchers have recently begun to investigate the 

origins of ethical leadership. For example, M. E. Brown and Treviño (2006) have 

proposed a series of predictors of ethical leadership, including situational (e.g., 

organizational climate) and individual predictors (e.g., individual characteristics). 

Empirical endeavor has been taken to test those predictors. For example, researchers 

found that leaders who were agreeable and conscientious were more likely to display 

ethical leadership (Walumbwa & Schaubroeck, 2009).  

Overview of Unethical Leadership. Research on unethical leadership is strongly 

connected with research on ethical leadership, but this line of inquiry is less developed. In 

the following, I will give an overview of unethical leadership.  



 

8 
 

 

In comparison to research on ethical leadership, research on unethical leadership 

has been less systematic and more evolving in the sense that various terms have evolved 

in the literature. In the past three decades, researchers have uncovered various unethical 

leadership constructs such as tyrannical leadership (Ashforth, 1994), abusive supervision 

(Tepper, 2000), supervisor undermining (Duffy, Ganster, & Pagon, 2002), and toxic 

leadership (Lipman-Blumen, 2005). Recently, researchers have reached a consensus that 

unethical leaders are oppressive, abusive, manipulative, and undermining (Tepper, 2007). 

Integrating different perspectives about unethical leadership, M. E. Brown and Mitchell 

(2010) defined unethical leadership as “behaviors conducted and decisions made by 

organizational leaders that are illegal and/or violate moral standards, and those that 

impose processes and structures that promote unethical conduct by followers” (p. 588). 

Based on this definition, in this dissertation, I see unethical leadership as having two 

dimensions: 1) leader behaviors and decisions that violate ethical and/or illegal standards, 

and 2) those that encourage, induce, or suggest followers to engage in unethical behavior. 

The first dimension of unethical leadership is direct in the sense that it reflects leaders’ 

own unethical behaviors and decisions such as abusing or sexually exploiting followers. 

The second dimension is indirect in the sense that it reflects leaders’ behaviors and 

decisions that harness and embed unethical behavior by followers such as encouraging 

lying or cheating.  

It is important to note that neither the direct nor the indirect form of unethical 

leadership is simply the opposite of the ethical leadership (M. E. Brown et al., 2005) 

because a lack of ethical leadership (e.g., setting an example of how to do things the right 

way in terms of ethics, conducting his/her personal life in an ethical manner) does not 



 

9 
 

 

equate to or imply engagement in unethical leader behavior (M. E. Brown & Mitchell, 

2010; Greenbaum, Quade, & Bonner, 2015). In other words, a lack of ethical leadership 

can also mean that one is an ethically neutral leader (Treviño et al., 2003) or an amoral 

leader (Carroll, 1987). This thus provides theoretical support for the distinction between 

ethical leadership and unethical leadership. 

In the following, I will review research on both direct and indirect forms of 

unethical leadership in sequence.  

Research on the Direct Form of Unethical Leadership. In the past two decades, 

the consequences of the direct form of unethical leadership have gained a great amount of 

attention from researchers. Like those studying ethical leadership, unethical leadership 

scholars tend to use social learning theory (Bandura & Walters, 1977) and social 

exchange theory (Blau, 1964) to explain the harmfulness of unethical leadership. In 

addition to those, they have also been adopting self-regulatory perspective (Muraven, 

Tice, & Baumeister, 1998) and conservation of resources theory (COR; Hobfoll, 1989) to 

explain how unethical leadership can harm employees in not only the work domain (e.g., 

work performance) but also non-work domains (e.g., personal lives).  

According to the self-regulatory perspective, people need to draw from limited 

self-regulatory resources to purposefully engage in self-regulation (C. M. Barnes, 

Schaubroeck, Huth, & Ghumman, 2011). Research has described self-regulation 

metaphorically as a moral muscle that refrains people from engaging in questionable 

behaviors (Baumeister & Juola Exline, 1999). When the muscle loses some of its strength, 

people are inclined to engage in questionable behaviors (Muraven et al., 1998). Unethical 

leadership is taxing or depleting because it leads employees to feel victimized or 



 

10 
 

 

threatened and, in turn, impairs their finite self-regulatory resources (Bandura, 1991; 

Muraven et al., 1998; Thau, Aquino, & Poortvliet, 2007). As a result, unethical leadership 

impairs employees’ ability to self-regulate and resist the temptation of engaging 

questionable or unethical behavior. It is important to note that the ego-depletion view is 

recently under great debate (Lurquin & Miyake, 2017). While a series of recent meta-

analyses revealed that the effect of ego-depletion might not differ from zero (Carter, 

Kofler, Forster, & McCullough, 2015; Carter & McCullough, 2014), in a more recent 

study, Lurquin and Miyake (2017) concluded that “the critical evidence is unlikely to 

convince proponents that ego depletion does not exist, and the supporting evidence is 

unlikely to convince skeptics that ego depletion does exist” (p. 125). They also 

emphasized that “better empiricism and better theory are needed to move the field 

forward and find more conclusive answers to the question whether, when, and why ego 

depletion does (not) exist” (p. 125).  

By contrast, COR suggests that individuals strive to gain and maintain resources 

that help them achieve goals (Hobfoll, 1989). When individuals experience psychological 

stress, it can consume resources and causes a spill-over effect on employees’ work and 

non-work lives. Hence, stress caused by unethical leadership behavior (e.g., abusive 

supervision, sexual harassment) can negatively influence not only employees’ work lives 

but also their non-work lives (e.g., family).  

Drawing upon the above theoretical perspectives, research on the consequences of 

the direct form of unethical leadership has been very fruitful. Researchers have revealed 

that the direct form of unethical leadership negatively predicted a wide range of 

employee outcomes such as work performance (K. J. Harris, Kacmar, & Zivnuska, 2007; 



 

11 
 

 

Zellars, Tepper, & Duffy, 2002), work attitudes (Tepper, 2000), employees’ personal 

lives (D. Carlson, Ferguson, Hunter, & Whitten, 2012; Hoobler & Brass, 2006), and 

employees’ psychological well-being (Tepper, 2000; Tepper, Moss, Lockhart, & Carr, 

2007). It could also lead to employee deviance and unethical behavior at work (Mitchell 

& Ambrose, 2007; Tepper et al., 2009; Tepper, Henle, Lambert, Giacalone, & Duffy, 

2008; Thau, Bennett, Mitchell, & Marrs, 2009) and employee problem drinking 

(Bamberger & Bacharach, 2006). In addition, the direct form of unethical leadership has 

been found to serve as a boundary condition under which employees’ certain traits will 

lead them to engage in unethical behavior at work. For example, based on trait activation 

theory, a recent research showed that abusive supervision functioned as a trait activator 

for the positive link between employees’ Machiavellianism and their unethical behavior 

(Greenbaum, Hill, Mawritz, & Quade, 2017). It is worth noting that although great 

attention has placed on the consequences of the direct from of unethical leadership, the 

majority of the research to date has predominantly focused on the abusive form of 

unethical leadership, namely abusive supervision (Tepper, 2000), leaving other direct 

forms of unethical leadership largely overlooked or untested. 

With regard to the antecedents of the direct form of unethical leadership, both 

leader- and follower-level factors have been revealed. As for leader-level factors, 

research has shown that leaders with certain personality traits such as social dominance 

orientation (Son Hing, Bobocel, Zanna, & McBride, 2007) and Machiavellianism (Kiazad, 

Restubog, Zagenczyk, Kiewitz, & Tang, 2010) were more likely to engage in the direct 

forms of unethical leader behavior such as exploiting or abusing employees. In addition, 

leaders’ moral exclusion belief (Opotow, 1990) has been found to be a potential source of 



 

12 
 

 

unethical leadership; when leaders do not think moral rules apply to an employee, they 

are more likely to perceive abusive treatment as appropriate (Tepper, Moss, & Duffy, 

2011). Taking the self-regulatory perspective, Joosten, Van Dijke, Van Hiel, and De 

Cremer (2014) found that constant pressure caused leaders to engage in the direct form of 

unethical leader behavior such as saying something hurtful to someone. More recently, 

using an experimental experience sampling study, Foulk, Lanaj, Tu, Erez, and 

Archambeau (2018) showed that leaders who perceived a strong psychological power 

tended to be more abusive toward followers. Another recent study showed that ethical 

leadership could lead to abusive supervision through ego depletion and moral licensing 

processes (Lin, Ma, & Johnson, 2016). As for the follower-level factors, it is well known 

that employees’ performance is a strong predictor of abusive supervision (e.g., Tepper et 

al., 2011; Walter, Lam, Van Der Vegt, Huang, & Miao, 2015). Interestingly, while poor 

performers have been found to elicit abusive supervision, a recent study showed that high 

performers were also subject to abusive supervision. Specifically, Khan, Moss, 

Quratulain, and Hameed (2018) showed that having high performance could lead to 

abusive supervision as a function of the perceived threat to hierarchy. In addition, 

employees’ similarity with leaders (similarity in personality, values, and attitudes) has 

been found to negatively predict abusive supervision (Tepper et al., 2011).  

Research on the Indirect Form of Unethical Leadership. Although closely 

related to the direct form of unethical leadership, the indirect form of unethical leadership 

has received comparatively sparse attention. In the following, I will review research on 

the indirect form of unethical leadership. 



 

13 
 

 

The indirect form of unethical leadership happens when leaders enable or foster 

unethical behavior among employees rather than engage in unethical behavior themselves 

(M. E. Brown & Mitchell, 2010). Such a form of unethical leadership is so common that 

news reports frequently describe corporate scandals that are caused by it. In a review of 

corporate scandals in Fortune 100 corporations, Clement (2006) concluded that those 

transgressions are primarily caused by the encouragement given by top management and 

government officials. Leaders can promote unethical behavior among employees by 

rewarding desirable outcomes of unethical behavior, forgiving transgressors, and 

overlooking wrongdoing (Ashforth & Anand, 2003; Brief, Buttram, & Dukerich, 2001). 

In a qualitative study, Sims and Brinkman (2002) highlighted that leaders who value 

short-term results, model aggressive and Machiavellian behaviors, and promote like-

minded employees foster unethical behavior among employees.  

In an effort to develop a model that explains how corruption becomes normalized 

in an organization, Ashforth and Anand (2003) highlighted three processes through which 

collective corruption emerges: 1) institutionalization, the process through which corrupt 

practices become part of the routine; 2) rationalization, the process through which 

individuals rationalize or legitimatize their corrupt practices; 3) socialization, the process 

through which newcomers are taught to accept and engage in corrupt practices. “These 

three processes are mutually reinforcing and reciprocally interdependent; once 

established in an organization, the processes create a situation where corruption is 

practiced collectively by employees and may endure indefinitely” (Ashforth & Anand, 

2003, p. 3). According to them, leaders play a significant role in all three processes. 

Leaders do not have to perform corrupt practices in person to serve a role model of 



 

14 
 

 

corruption. Whether intentionally or not, they can facilitate it by rewarding, condoning, 

or neglecting corruption, and authorize the corruption formally or informally. In sum, 

their theory provides an important theoretical insight for the research on the indirect form 

of unethical leadership.  

 Although it has received both conceptual and empirical support, unlike the direct 

form of unethical leadership, the empirical support for the indirect form of unethical 

leadership has been limited, greatly constraining our understanding of the overall 

harmfulness of unethical leadership. Fortunately, emerging research has started to place 

more empirical attention on the indirect form of unethical leadership. As a prevalent type 

of indirect unethical leader behavior, making unethical leader-requests has recently 

drawn an increasing amount of attention. In the following, I will elaborate on making 

unethical leader-requests as an important indirect form of unethical leadership.  

Unethical Leader-requests. Consistent with existing definitions of other types of 

unethical leader behavior (e.g., abusive supervision), making unethical leader-requests 

can be defined as an employees’ perception of the extent to which their supervisors ask 

them to engage in unethical behavior. This could happen, for example, when supervisors 

ask employees to stay silent while they engage in questionable conduct, when supervisors 

ask employees to misrepresent facts to make them look good, when supervisors ask 

employees to treat some people disrespectfully, and when supervisors ask employees to 

do tasks that involve lying to others (Desai & Kouchaki, 2017).  

Surprisingly, as an indirect form of unethical leader behavior that occurs 

ubiquitously, making unethical leader-requests has been largely overlooked by 



 

15 
 

 

organizational researchers2. However, the topic has long been attended by social 

psychologists. In social psychology, inquiries that are relevant to unethical leader-

requests ascend to as early as the 1960s. In 1961, a U.S. psychologist, Stanley Milgram, 

ran an experiment to examine justifications for acts of genocide by those accused of war 

crimes during World War II. The experiment has become one of the most well-known 

experiments in the field of psychology and other relevant disciplines. Milgram began to 

conduct the experiment in July 1961, a year after the trial of Adolf Eichmann in 

Jerusalem. The main purpose of the experiment was to see how far people would go in 

following orders given by an authority figure (e.g., leader) if the order had something to 

do with harming or even killing others (S. Milgram, 1974). In the experiment, 

participants were made to believe that they were interacting with another participant (i.e., 

the learner) in the other room who did not actually exist (unbeknown to the participants). 

Participants were told that the learner had been asked to learn a list of word pairs. Their 

“jobs” were to test how well the learner could recall those paired words and to give an 

electric shock every time when the leaner made mistakes. The level of shock would 

increase after every mistake. In the experiment, when the shock achieved a high level 

 
2 It is worth noting that in this dissertation, making unethical leader-requests will be seen as a construct that 

is distinct from ethical leadership for the following reasons. As mentioned earlier, Mesdaghinia, Lewis, and 

Eisenberger (2019) conceptualized Leaders’ Immorality-encouragement (LIE) as an employee’s perception 

that his/her supervisor encourages him/her to engage in unethical behavior to benefit the organization 

and/or its members. In their study, they theoretically and empirically differentiated LIE from ethical 

leadership. In doing so, they found that none of the ethical leadership items were rated by the raters as 

consistent with the definition of LIE and that the CFA analyses also showed that LIE is distinct from 

ethical leadership. The common ground between making unethical leader-requests and LIE is that they both 

involve asking and encouraging employees to engage in unethical behavior. The difference lies in the 

intention of giving the requests. By definition, LIE is conducted with the intention to benefit the 

organization and/or its members, and hence, the intention behind LIE is prosocial or altruistic. In contrast, 

making unethical leader-requests is a broader construct and does not specify the motivation behind it. In 

other words, the motivation behind unethical leader-requests could be prosocial or antisocial. Hence, 

logically, LIE should relate to ethical leadership more closely than do making unethical leader-requests 

because LIE always involves a prosocial intention, but unethical leader-requests do not. From this 

perspective, if LIE has been shown to be distinct from ethical leadership, making unethical leader-requests 

is also likely to be distinct from ethical leadership. 



 

16 
 

 

(e.g., 300 volts), participants tended to refuse to give the shock. However, an 

experimenter (i.e., a confederate) in the same room with participants would press and 

request participants to give the shock. The findings of the experiment shocked the 

world—65% of participants continued to the highest level of 450 volts, and 100% of the 

participants continued to 300 volts. Milgram concluded that people tend to obey orders 

given by an authority figure, even when the orders are unethical in the sense that it 

involves harming or killing another person.  

S. Milgram (1974) raised an agency theory to explain the findings in the 

experiment. He proposed that people have two different states when facing social 

situations: 1) the autonomous state and 2) the agentic state. The autonomous state refers 

to the situation when individuals can direct and take responsibility for their own 

behaviors. The agentic state refers to the situation when individuals behave as agents for 

another person’s will; that is, they allow others to direct their behaviors by following the 

orders and attributing the responsibility to those directing their behaviors. According to 

Milgram, two premises need to be met for an individual to enter the agentic state. First, 

the person who gives orders needs to be seen as legitimate, a viewpoint that is consistent 

with social learning theory (Bandura & Walters, 1977). Second, the individual must 

believe that the person giving orders will take responsibility for the consequences of 

his/her behaviors in obedience to the orders. The above proposals have been verified by a 

series of follow-up studies done by Milgram. One of the studies showed that when 

participants were reminded that they would be responsible for what they did, very few of 

them would follow the order.  



 

17 
 

 

The Milgram experiment demonstrated that people are inclined to obey the 

commands by an authority figure even when those commands are unethical even though 

they tend to have strong emotional responses when they are asked to do so. In 

organizations, leaders are authority figures, and hence, the findings of the Milgram 

experiment have a huge implication on the research on how people will emotionally and 

behaviorally react to the unethical leader-requests in organizations. However, research on 

unethical leader-requests is still nascent. Until very recently, organizational researchers 

began to pay close attention to unethical leader-requests.  

As for the research on the predictors of unethical leader-requests, to date, very 

few studies specifically examine when leaders are more or less likely to request their 

followers to engage in unethical behavior. One of the few studies found that when 

followers display moral symbols such as words, images, and objects that have moral 

implications, leaders are less likely to raise unethical requests (Desai & Kouchaki, 2017). 

With regard to the consequence of unethical leader-requests, in an effort to expand our 

understanding of why employees engage in unethical behavior that benefits the 

organization and/or its members, a behavior called unethical pro-organizational behavior 

(UPB; Umphress & Bingham, 2011; Umphress, Bingham, & Mitchell, 2010), 

Mesdaghinia et al. (2019) examine whether leaders’ unethical requests in the form of 

Leaders’ Immorality-encouragement (LIE) can cause more UPB. They defined LIE as an 

employee’s perception that his/her supervisor encourages him/her to engage in unethical 

behavior to benefit the organization and/or its members. Using both field study and lab 

experiment, they found that LIE was positively related to UPB, and the positive 

relationship was strengthened by leader-member exchange relationship (LMX; Graen, 



 

18 
 

 

Novak, & Sommerkamp, 1982). In addition to moral consequences, researchers have 

found that unethical leader-requests can cause performance consequences. Specifically, 

leaders’ unethical requests have been found to reduce the employees’ job performance by 

arousing anxiety and impairing employees’ intrinsic job motivation (I. H. Smith, 

Kouchaki, & Wareham, 2013).  

 Although organizational researchers have realized the significance of studying 

unethical leader-requests and started paying increasing attention to both antecedents and 

consequences of unethical leader-requests, what we have learned by far has been limited 

in the mutual relationships between unethical leader-requests and work domain aspects 

(e.g., displaying moral symbols at work, unethical behavior at work, job motivation and 

performance), without regard to the link between unethical leader-requests and non-work 

domain aspects. Work-family scholars have long emphasized that a comprehensive study 

of the organizational phenomena needs to include a work-family interface (e.g., Dubin, 

1973). Thus, merely focusing on unethical leader-requests’ workplace antecedents and 

outcomes is myopic and even unrealistic, limiting our understanding of this prevalent 

phenomenon. Therefore, it will be meaningful for me to examine how unethical leader-

requests connect to employees’ non-work domains such as family. To address this 

intellectual gap, in this dissertation, I intend to investigate the family outcomes of 

unethical leader-requests by taking an emotion spillover perspective. Specifically, I argue 

that two emotional responses (i.e., anxiety and anger), in reaction to unethical leader-

requests, can explain how unethical leader-requests readily erode employees’ family lives. 

To provide both theoretical and empirical foundations to explain the psychological bridge 

between unethical leader-requests and employees’ family lives, in the following, I will 



 

19 
 

 

review cognitive appraisal theory which will provide an overarching theoretical 

foundation for this dissertation as well as research on anxiety and anger which are 

proposed to be two key emotional responses to unethical leader-requests in this 

dissertation.   

Cognitive Appraisal Theory, Anger, and Anxiety 

Cognitive Appraisal Theory. Emotion is an interface between an organism and its 

environment (Mulligan & Scherer, 2012). Emotion episodes involve a series of 

components, including cognitive appraisal of the emotional stimuli (e.g., event, object, 

memory), subjective feeling, physiological responses, and expressive behavior (Scherer 

& Moors, 2019). As one of the most important components, cognitive appraisal has 

received considerable attention from researchers in a wide range of disciplines including 

but not limited to psychology, management, marketing, and education. Specific studies 

using cognitive appraisal approach to explain people’s emotional experience in response 

to emotional stimuli started in the 1980s and flourished in the early 1990s. The 

popularization of cognitive appraisal approach is a result of its well-known capability of 

explaining the subtle nuances of emotion. Specifically, the approach aims to predict 

which specific emotions (i.e., discrete emotions) should be evoked by a given emotional 

stimulus as well as how elicited emotions affect subsequent behavior. During the period 

between the 1980s and 1990s, a number of appraisal theories were proposed by 

researchers (e.g., Lazarus, 1991b; Ortony, Clore, & Collins, 1988; Roseman, Spindel, & 

Jose, 1990; C. A. Smith & Ellsworth, 1987). Of those different forms of appraisal 

theories, Lazarus’s (1991a, 1991b) cognitive appraisal theory is one of the most adopted 

appraisal theories to explain coping responses to stressful situations. The cognitive 



 

20 
 

 

appraisal theory was popularized by Richard Lazarus and colleagues (Lazarus, 1966, 

1991b; Lazarus & Folkman, 1984) and has been seen as a promising avenue to study 

emotions in business research contexts (Johnson & Stewart, 2005). In the following, I 

will focus on reviewing Lazarus’s (1991a, 1991b) cognitive appraisal theory as well as its 

application in the management literature.  

According to Lazarus (1991a, 1991b), the emotion elicitation process consists of 

two stages: 1) primary appraisal stage and 2) secondary appraisal stage. In the first stage, 

individuals evaluate whether an emotional stimulus is “goal-relevant” and whether it 

affects their well-being. This initial evaluation is gross in the sense that it is made based 

on whether the stimulus furthers or thwarts their personal goals and values (Cropanzano, 

James, & Citera, 1993; Frijda, 1993; H. M. Weiss & Cropanzano, 1996). In other words, 

if a stimulus is evaluated as beneficial for achieving a goal, general positive emotions 

result, and if it is evaluated as harmful, general negative emotions result. The primary 

appraisal mainly involves valence and intensity of the emotion and happens immediately 

and automatically; that is, the process could be unconscious. Importantly, it does not 

determine which discrete emotions individuals will experience. Following the primary 

stage, the second stage of appraisal is evoked quickly.  

The second stage is an interpretive meaning analysis stage (C. A. Smith & Pope, 

1992) which leads individuals to experience particular discrete emotions such as 

happiness, excitement, sadness, anger, anxiety, shame, and pride. In the second stage, 

individuals will evaluate the three dimensions of the emotional stimulus in terms of 

attribution of causes, potential for coping, and certainty of outcome. Specifically, 

attribution of causes dimension pertains to the extent to which we should blame (in the 



 

21 
 

 

case of negative emotions) or credit (in the case of positive emotions) self, another, or no 

one. The potential for coping dimension is associated with whether we can influence the 

environment for changes. Lastly, the certainty of outcome dimension pertains to the 

belief that outcomes will be desirable or undesirable. By evaluating the three dimensions 

of the emotional stimulus, individuals can simultaneously experience different discrete 

emotions. For example, one may feel happy for his/her brother’s success in academic 

performance while feeling envious. One may feel excited about receiving an offer from a 

high-paying company while feeling anxious and stressed about the performance pressure 

after entering the company.  

As a promising avenue to study the elicitation process of discrete emotions in 

business contexts (Johnson & Stewart, 2005), cognitive appraisal theory has gained 

popularization in the management literature. Research has adopted it to explain various 

emotional situations in organizations such as injustice, abusive supervision, work-related 

identifications, and role conflict (Conroy, Becker, & Menges, 2017; Lian et al., 2014; 

Perrewé et al., 2004; H. M. Weiss, Suckow, & Cropanzano, 1999). For example, based 

on Lazarus’s (1991a, 1991b) framework, Perrewé et al. (2004) found that perceived role 

conflict positively predicted strain, and the relationship was weakened when one’s 

political skill was strong. More recently, drawing from the same framework, Lian et al. 

(2014) found that abusive supervision led subordinates to react hostilely and, in turn, 

resulted in supervisor-direct aggression. 

Given the original purpose of cognitive appraisal theory—to explain individuals’ 

emotional responses to stressful situations—organizational researchers are inclined to use 

it to explain employees’ negative emotions such as anxiety and anger in response to 



 

22 
 

 

stressful organizational situations. In line with this, this dissertation is mainly interested 

in two negative discrete emotions—anxiety and anger—as reactions to unethical leader-

requests. To provide a foundation for these two emotions, in the following, I will review 

historical perspectives of anxiety and anger along with organizational research on these 

two emotions.  

Overview of Anger. Long accepted as one of eight basic emotions of human 

beings, anger is highly recognizable, and even infants can reliably recognize its 

expression and experience (Ekman, 1992). Attempts to define and conceptualize the role 

of anger in human interaction ascend to Aristotle’s time (J. Barnes, 1984). However, a 

consensus in terms of the definition and conceptualization of anger has not been fully met 

yet. Psychologists have highlighted anger as a “fuzzy concept” in the sense that we know 

when we see and feel it, but it is challenging for us to define it (Russell & Fehr, 1994). 

Recently, organizational researchers have tried to draw a firmer definition of anger. In 

their work of reviewing research on anger in organizations, D. E. Gibson and Callister 

(2010) defined anger as “an emotion that involves an appraisal of responsibility for 

wrongdoing by another person or entity and often includes the goal of correcting the 

perceived wrong” (p. 68). They defined anger around three critical components: 1) Anger 

is a discrete emotion which is characterized by a unique expression, specific 

physiological responses (i.e., increased heart rate and blood pressure; Stark, Walter, 

Schienle, & Vaitl, 2005), and a limited set of source events; 2) Anger is a social emotion 

in the sense that it is evoked by perceived others’ wrongdoing (e.g., incivility, unfairness, 

injustice, unethicality, goal interference) and also directed toward others (Averill, 1983); 

3) Anger can be both trait-like and state-like. Trait anger is a stable dispositional 



 

23 
 

 

tendency to experience state anger in terms of frequency and intensity (Spielberger, 1991) 

while state anger is a temporary emotional state “consisting of feelings ranging from 

irritation to intense rage, physiological and cognitive reactions, behavioral tendencies, 

and observable verbal and motor behaviors” (D. E. Gibson & Callister, 2010, p. 68).  

Research on Anger. As a common discrete negative emotion, anger has been 

studied by scholars from a wide range of disciplines including but not limited to 

psychology, marketing, education, sociology, and management. In the following, I will 

mainly review research on anger on two aspects: 1) the antecedents of anger and 2) the 

consequences of anger. The review will mainly focus on research that views anger as a 

short-term emotion rather than a trait. In addition, the review will mainly focus on 

organizational research on anger while taking perspectives from other disciplines as 

supporting information.  

 A variety of workplace sources have been found to be responsible for anger at 

work. These sources can be generally grouped into three categories: 1) unfairness and 

injustice, 2) interpersonal conflict, and 3) goal interference. There is a large body of 

research drawing on fairness and justice perspectives to examine the antecedents of anger 

at work. For example, based on Adams’ (1965) equity theory that emphasizes the 

perception of inequity, researchers have found that unjust of unfair treatments by others 

(e.g., supervisors) can lead the employees to feel angry and experience physiological 

states associated with anger (Cropanzano, Weiss, Suckow, & Grandey, 2000; Domagalski 

& Steelman, 2005; Fitness, 2000; Harlos & Pinder, 2000). Relatedly, research has 

revealed that employees also feel angry when experiencing incivility (i.e., “acting rudely 

or discourteously, without regard for others, in violation of norms for respect in social 



 

24 
 

 

interactions"; Andersson & Pearson, 1999, p. 455), disrespect, and condescension at work 

(Domagalski & Steelman, 2005; Fitness, 2000; Grandey, Tam, & Brauburger, 2002).  

 Interpersonal conflict is a natural and important part of the workplace and occurs 

when parties perceive that they have different values, attitudes, viewpoints, and goals 

(Jehn, 1997). The conflict is thus associated with a sense of threat. From an evolutionary 

perspective, anger is an instinctive reaction to reduce or completely remove perceived 

threats (Tracy, 2014). Anger has also been seen as the only self-protective emotion that 

motivates individuals to actively address perceived threats (Berkowitz & Harmon-Jones, 

2004; Johnston & Glasford, 2014). Accordingly, anger has been argued to be the most 

prominent and pervasive emotion among all emotions that arise in interpersonal conflict 

(Allred, 1999). These viewpoints regarding the relevance of anger to the interpersonal 

conflict have been supported both conceptually and empirically. For example, 

conceptually, based on attribution theory (Kelley, 1967), Allred (1999) proposed that 

interpersonal conflict can lead to anger and retaliation when the person believes another 

party is responsible for the act. Empirically, in a work of identifying sources of 

workplace anger and aggression, Glomb (2002) found that 80% of the respondents 

indicated interpersonal conflict at work as one of the primary sources.  

 Researchers have also identified goal interference (i.e., blockage for one’s 

execution of plans or attainment of goals) as a primary source of anger (Shaver, Schwartz, 

Kirson, & O'connor, 1987). Such a perspective has a long history and can be traced to the 

1930s. Dollard, Miller, Doob, Mowrer, and Sears (1939) proposed that a person will 

respond aggressively when his/her goal or plan is interrupted or hindered (i.e., 

frustration-aggression model). Researchers later criticized the one-sided view of this 



 

25 
 

 

proposal and confirmed that anger mediated the link between frustration and aggression 

(Berkowitz, 1993). In line with this perspective, another study found that when 

employees’ personal and organizational goals were constrained by situational factors, a 

feeling of anger resulted, subsequently leading them to engage in counterproductive 

workplace behaviors (Chen & Spector, 1992).  

 Indeed, anger comes with consequences, both negative and positive. As for 

negative consequences, anger has been found to positively associated with a wide range 

of undesirable outcomes such as increased blood pressure, heightened chance of heart 

diseases, decreased job satisfaction, harmful organizational climates, increased incivility, 

unethical behaviors, and even aggressive and violent behaviors (Andersson & Pearson, 

1999; Aquino, Douglas, & Martinko, 2004; Begley, 1994; Fox & Spector, 1999; Glomb, 

2002; Mitchell, Baer, Ambrose, Folger, & Palmer, 2018). Researchers have also revealed 

a series of desirable consequences of anger, which speaks to the socially functional aspect 

of anger. For example, anger can motivate protective physiological changes and 

behaviors to address perceived threats such as experienced injustice and inequity (Bies, 

1987; Frijda, 1986). Furthermore, anger expression has been found to help individuals 

clarify and signal boundaries for others’ proper decisions and behaviors, refraining others 

from socially undesirable manners (Tafrate, Kassinove, & Dundin, 2002). Researchers 

have also argued that strategically expressing anger can sometimes lead to desirable 

negotiation outcomes that benefit the anger expresser (D. E. Gibson & Schroeder, 2002).  

In the above, I have given a review of anger and relevant conceptual and 

empirical research. Similar to anger, anxiety is also a negative and high arousal emotion. 

In this dissertation, I argue that anxiety is another negative emotional response to 



 

26 
 

 

unethical leader-requests. In the following, I will provide a review of anxiety as well as 

relevant research. 

Overview of Anxiety. Serving as an emotion that alerts us to potential threats and 

allows us to appropriately evaluate and respond to them, anxiety is “a state of distress 

and/or physiological arousal in reaction to stimuli including novel situations and the 

potential for undesirable outcomes” (Brooks & Schweitzer, 2011, p. 44). According to 

the affective circumplex model, anxiety is an emotion with negative valence or 

unpleasantness and a high level of activation (Russell, 1980). By definition, anxiety can 

arise when individuals confront threatening situations that have the potential for 

undesirable outcomes (Gray, 1991). For example, students will feel anxious when 

teachers give a random in-class quiz, workers will feel anxious when a deadline is 

approaching, and investigators will feel anxious when they see the downfall of the price 

of a stock they own. Anxiety is not a single feeling, but a mixed feeling consists of fear, 

frustration, stress, tension, worry, apprehension, and nervousness (Gray, 1991), and it is 

usually accompanied with an intense sense of uncertainty and of a lack of control 

(Raghunathan & Pham, 1999). Common physiological responses of anxiety include 

muscular tension, fatigue, restlessness, and concentration deficiency (American 

Psychiatric Association, 2013). Common behavioral responses to anxiety include 

rumination, somatization, and pacing back and forth (Seligman, Walker, & Rosenhan, 

2001).  

Unlike anger that motivates people to actively address the threatening situations 

(Berkowitz & Harmon-Jones, 2004; Johnston & Glasford, 2014), anxiety is passive and 

aversive by nature in the sense that it can motivate individuals to escape, withdraw, or 



 

27 
 

 

flee from sources that have provoked anxiety (Nesse & Marks, 1994). However, like 

anger, anxiety is also a self-protective emotion because from the evolutionary perspective, 

the defensive responses provoked by anxiety help to prevent people from further harm by 

driving them to stay away from the anxiety-producing situations rather than actively 

eliminate or remove those situations. Like anger, anxiety can also be a trait. Researchers 

have identified anxiety as a long-term stable dispositional characteristic that determines 

how strongly and frequently individuals feel anxious in response to anxiety-producing 

situations (Endler & Kocovski, 2001). In addition, the trait anxiety has been found to 

contribute to various mental disorders such as anxiety disorders and depression (Hettema, 

2008; Sandi & Richter-Levin, 2009).  

Research on Anxiety. Anxiety has long received attention from researchers from 

various disciplines such as psychology, marketing, education, sociology, and 

management. In the following, I will mainly review research on anxiety in terms of 1) the 

antecedents of anxiety and 2) the consequences of anxiety. The review will mainly focus 

on research that views anxiety as a short-term emotion rather than a trait. In addition, the 

review will mainly focus on organizational research on anxiety while taking perspectives 

from other disciplines as supporting information.  

Previous research on the antecedents of anxiety can be approximately grouped 

into two categories: 1) internal sources and 2) external sources. As for internal sources of 

anxiety, Hochschild (1983) emphasized that individuals’ sense of self may be threatened 

when their own behavior is inconsistent with what they believe they should behave 

because the need to maintain self-consistency is one of the most important human needs. 

The threat to self can produce anxiety (Lazarus, 1991a). In line with the self-consistency 



 

28 
 

 

perspective, researchers have revealed that when people engaged in inauthentic behaviors 

such as lying and surface acting (e.g., suppressing one’s emotion to display unfelt 

emotions), they tended to feel anxious (Ennis, Vrij, & Chance, 2008; Tomura, 2009; D. T. 

Wagner, Barnes, & Scott, 2014).  

With regard to external sources of anxiety, research has indicated a series of 

work-relevant stressors that are closely related to employees’ anxiety. For example, Doby 

and Caplan (1995) found that work stressors threatening employees’ reputation (e.g., 

negative feedback from leaders) caused anxiety at both work and home. Totterdell, Wood, 

and Wall (2006) conducted a one-week daily diary study and showed that participants 

reported a higher level of anxiety during weeks when work demands were high versus 

low. In addition, Perrewé et al. (2004) showed that role conflict was positively related to 

anxiety, and political skills mitigated the positive relationship.  

As for the consequences of anxiety, a large body of research in occupational 

health has revealed a positive link between anxiety and stress (e.g., Fay & Sonnentag, 

2002; Ford, Cerasoli, Higgins, & Decesare, 2011; Gomes, Faria, & Gonçalves, 2013). 

This line of inquiry highlights the negative impact of anxiety on employees’ 

physiological and mental health. Another line of inquiry focuses on how anxiety can 

influence employees’ workplace outcomes such as work performance (e.g., Haslam, 

Atkinson, Brown, & Haslam, 2005; M. K. Jones, Latreille, & Sloane, 2016). For example, 

drawing from conservation of resource theory, McCarthy, Trougakos, and Cheng (2016) 

found that anxiety harmfully influenced work performance as a function of emotional 

exhaustion, and such effect was attenuated by employees’ social exchange relationship 

with their supervisors and coworkers. In addition, anxiety has been found to positively 



 

29 
 

 

predict undesirable workplace outcomes such as turnover intentions (Jensen, Patel, & 

Messersmith, 2013) and the adoption of inferior negotiation strategies (Brooks & 

Schweitzer, 2011).  

Work-family Spillover 

In the last section, I review research on anxiety and anger, two proposed key 

emotional responses to unethical leader-requests in this dissertation. The next question is 

how these two emotions spill over to employees’ family domain. To answer this question, 

I will draw upon perspectives in the work-family spillover literature. In this section, I will 

provide a high-level overview of the major theories and perspectives about both negative 

and positive work-family spillovers.  

Overview of Work-family Spillover Theories. Staines’s (1980) spillover theory 

highlights that work experiences (e.g., affect, attitudes, values) can readily spill over to 

the family domain and vice versa. In line with Staines’s spillover theory, later research of 

this inquiry tends to agree that spillover is a major linkage between work and family (e.g., 

Pleck, 1995). Empirical evidence has generally supported the spillover theory (e.g., 

Williams & Alliger, 1994). As the spillover perspective evolved and advanced, research 

has identified and differentiated two distinct types of spillovers, namely positive spillover 

and negative spillover.  

The first type of spillover represents positive spillover between work and family 

and has become the major theoretical framework upon which the work-family enrichment 

literature draws (Greenhaus & Powell, 2006). According to Greenhaus and Powell (2006), 

“work-family enrichment happens when work experiences improve the quality of family 

life, and family-work enrichment happens when family experience improves the quality 



 

30 
 

 

of work life” (p. 73). Greenhaus and Powell (2006) highlighted two paths through which 

one domain can facilitate another domain. The first one is the instrumental path; that is, 

resources (e.g., energy, self-esteem, skills, knowledge, social capital) generated in one 

domain directly transfer to another domain and then facilitate performance in the domain. 

Research has validated the theorization of the first path. For example, Wayne, Randel, 

and Stevens (2006) showed that both family instrumental and emotional support 

positively predicted work-family enrichment. More recently, drawing from self-

determination theory, Menges, Tussing, Wihler, and Grant (2017) found that family 

motivation could compensate for low intrinsic motivation at work and, ultimately, 

improved work performance. The second one is the affective path; that is, resources 

generated in one domain can lead to positive affect, which ultimately facilitates 

performance in another domain. The second path has also received empirical support. For 

example, using a 2-week daily diary study, Leavitt, Barnes, Watkins, and Wagner (2019) 

found that sexual activities on a given evening led to positive affect at work the next 

morning, which in turn increased job satisfaction and job engagement.  

The second type of spillover involves negative spillover between work and family 

and serves as an essential conceptual foundation for the work-family conflict literature 

(Greenhaus & Beutell, 1985; Small & Riley, 1990). In the seminal work, Greenhaus and 

Beutell (1985) referred to the work-family conflict as “a form of inter-role conflict in 

which the role pressures from the work and family domains are mutually incompatible in 

some respect” (p. 77). Three types of work-family conflicts were identified, namely time-

based conflict (i.e., engaging in family or work domain activities reduces the time 

available for another domain), strain-based conflict (i.e., stress derived from one domain 



 

31 
 

 

reduces the ability to perform in another domain), and behavior-based conflict (i.e., the 

behavioral pattern of a role disables functional behavioral adjustment for another role). 

All three types of conflicts have received an extensive amount of empirical support. 

Research has found that the number of children was negatively associated with 

employees’ ability to satisfy work demands (Bedeian, Burke, & Moffett, 1988; Kelly & 

Voydanoff, 1985). Relatedly, researchers also found that individuals had young children 

at home were more likely to experience work-family conflict as a function of the reduced 

time available for work (Burke, Weir, & DuWors Jr, 1979; Grandey & Cropanzano, 

1999). In addition, unpleasant work experience in the form of abusive supervision has 

been found to lead abused employees to engage in family undermining behaviors (e.g., 

taking negative work emotions out on family members, giving negative evaluations 

directed toward family members) as a function of displaced aggression (Hoobler & Brass, 

2006).  

In this dissertation, I am mainly interested in negative spillover, and I examine 

whether as unpleasant work experience, unethical leader-requests can spill over to 

employees’ family domain to influence their family domain outcomes. Specifically, I 

focus on three family domain outcomes—sleep, emotional exhaustion at home, and 

interactions with family. While the work-family scholars have emphasized the work-

family conflict as a consequence of strain spillover, time conflict, and role 

incompatibility (D. S. Carlson, Kacmar, & Williams, 2000; Greenhaus & Beutell, 1985), 

my dissertation goes beyond work-family conflict because I do not focus on general 

workplace stressors which lead to generalized strain or conflict at home. While most 

research on negative spillover focuses on examining family stress as a consequence of 



 

32 
 

 

general job stressors (Westman & Vinokur, 1998), this dissertation focuses on a more 

specific job stressor that is likely to be ignored by leaders—unethical leader-requests. 

Furthermore, I focus on specific stressful family outcomes that are critical to employees’ 

physiological, mental, and social wellbeing (Lazarus & Folkman, 1984). Hence, my 

dissertation helps to provide a more nuanced angle to look into how discrete and adverse 

workplace stressors can contaminate specific aspects of employees’ family lives. As such, 

the findings of my investigation could be particularly worrisome. Given that making 

unethical leader-requests is a comparatively morally subtle and fuzzy form of unethical 

leader behavior in comparison to others such as abusive supervision, leaders are more 

unlikely to envision the harmfulness of unethical leader-requests to the employee who is 

requested to engage in unethical behavior. The perception of the harmlessness of 

unethical leader-requests is dangerous because it promotes unethical leader-requests 

which can cause great harm to employees’ lives without receiving leaders’ attention. 

Sleep, Emotional Exhaustion, and Interactions with Family  

In this section, I will first review Lazarus and Folkman’s (1984) viewpoint 

regarding the three dimensions of employee wellbeing. Then, I will review research on 

sleep, emotional exhaustion, and family interactions. Lastly, I will elaborate on how 

previous and recent research adopts a work-family spillover lens to explore the role of the 

three focal outcome variables of this dissertation, sleep, emotional exhaustion at home, 

and interactions with family, in the work-family interference process.  

Physiological, Mental, and Social Wellbeing. Employee wellbeing lies in the 

heart of employee job satisfaction (Sousa-Poza & Sousa-Poza, 2000) and has been found 

to be a key determinant of workplace behavior (e.g., Wright & Bonett, 2007). As a result, 



 

33 
 

 

in the past two decades, great attention has been directed to the antecedents of employee 

wellbeing (Cotton & Hart, 2003). However, while a great number of studies have focused 

on specific aspects of employee physiological and mental wellbeing such as sleep 

(Barling, 2016), eating (Y. Liu et al., 2017), and burnout (Maslach, Schaufeli, & Leiter, 

2001), little research has paid attention to employee social wellbeing, not to mention 

simultaneously examining all three aspects of employee wellbeing. Employee wellbeing 

scholars have long advocated that in order to show a complete picture of employee 

wellbeing, researchers need to include all three aspects of employee wellbeing—

physiological, mental, and social (Lazarus & Folkman, 1984). However, researchers tend 

to “emphasize one specific aspect of employee wellbeing without regard to others” 

(Lazarus & Folkman, 1984, p. 182). Guided by Lazarus and Folkman (1984), in this 

dissertation, I will simultaneously examine all three aspects of employee wellbeing at 

home. In so doing, I select sleep, emotional exhaustion at home, and negative family 

interactions as focal outcome variables to represent physiological, mental, and social 

wellbeing respectively.  

In the following, I will review research on sleep, emotional exhaustion, and 

family interactions in sequence. The review will mainly contain organizational research 

while taking perspectives from other disciplines as supporting information.   

 Employee Sleep. People spend one-third of their time at work and another third of 

their time on sleep. Sleep plays a critical role in important human functioning such as 

learning and memory (U. Wagner, Gais, Haider, Verleger, & Born, 2004; Walker & 

Stickgold, 2006) and immune system responsiveness (Irwin, 2015). The relationship 

between work and sleep has long been recognized, but the research on their relationship 



 

34 
 

 

flourished only recently. About a decade ago, organizational researchers started to pay 

increasing attention to the interdependence between work and sleep. As a result, in the 

past decade, researchers have adopted a wide range of theories and approaches to uncover 

various mechanisms through which work experience influences sleep and vice versa, and 

an extensive body of research has been conducted to provide rich empirical evidence for 

the connection between work and sleep. The following review will mainly center on two 

aspects of work-sleep interdependence: 1) how sleep experience can spill over to the 

work domain to affect workplace outcomes (e.g., work performance) and 2) how work 

experience (e.g., leaders’ behavior) can spill over to the family domain to affect sleep.  

Given that more than 25% of workers in the U.S. are involved in some forms of 

shift work (Alterman, Luckhaupt, Dahlhamer, Ward, & Calvert, 2013) which is closely 

related to employee sleep problems (Åkerstedt, 2003; Drake, Roehrs, Richardson, Walsh, 

& Roth, 2004), researchers in this area of inquiry has been particularly eager to study 

how sleep problems detrimentally influence employees’ workplace outcomes. Research 

over the past decade has shown that sleep deprivation is a strong predictor of various 

workplace outcomes. For example, C. M. Barnes and Wagner (2009) found that sleep 

quality was a strong predictor of workplace injuries. Specifically, via two novel studies 

using archival data, they found that sleep deprivation caused by daylight saving time led 

to more workplace injuries. Based on the self-regulatory perspective, C. M. Barnes et al. 

(2011) found that sleep deprivation could reduce one’s self-control resource and, in turn, 

lead employees to engage in more unethical behavior. D. T. Wagner, Barnes, Lim, and 

Ferris (2012) found that loss of sleep could positively predict cyberloafing which occurs 

when employees use office computers for something unrelated to their jobs. C. M. 



 

35 
 

 

Barnes, Lucianetti, Bhave, and Christian (2015) found that leaders’ lack of sleep could 

lead to more abusive supervision. C. M. Barnes, Guarana, Nauman, and Kong (2016) 

found that leaders’ sleep deprivation could reduce their emotional labor performance and, 

in turn, negatively affect subordinate’s perception of charismatic leadership. In addition, 

they found that subordinates’ sleep quality could also influence their perception of 

leaders’ charisma as a function of reduced positive affect. More recently, a relevant study 

by Guarana and Barnes (2017) showed that both leaders’ and followers’ lack of sleep 

negatively influenced the perception of relationship quality between them, and hostility 

explained the effects. In sum, research on how sleep can influence workplace outcomes 

has been very fruitful.  

 Due to a great number of negative workplace outcomes caused by poor sleep, 

increasing attention has been directed at stressful work experience that can impair 

employees’ sleep. Research has identified a great number of work stressors related to 

poor sleep. Those stressors can be grouped into two categories: 1) task stressors and 2) 

non-task stressors. The first group of stressors are relevant to job task per se and can be 

manifested as high job demands. There is an extensive body of empirical evidence 

supporting the positive link between job demands such as workload, job rules, and time 

pressure, and sleep problems (e.g., Berset, Elfering, Lüthy, Lüthi, & Semmer, 2011; 

Winwood & Lushington, 2006). Research findings generally support that high job 

demands will lead to difficulty of falling asleep, maintaining sleep, and waking up 

(Åkerstedt et al., 2002; Knudsen, Ducharme, & Roman, 2007). For example, as one of 

the common work norms or rules in service relevant jobs, emotional labor has been found 

to be a novel indicator of insomnia (D. T. Wagner et al., 2014). In a meta-analysis 



 

36 
 

 

looking into the overall effect of job demands on poor sleep, Nixon, Mazzola, Bauer, 

Krueger, and Spector (2011) revealed a positive correlation between job demands and 

sleep disturbances, providing solid evidence for the relevance of task stressors to sleep.  

 The second group of stressors is relevant to the work but not to the job task itself. 

Non-task stressors have received as much empirical attention as task stressors. One of the 

most salient non-task stressors is social stressors. Social stressors exist in any type of 

social structure, and given the social nature of the organizational structure, social 

stressors will inevitably emerge in any organization. Recognizing the impactful role of 

social stressors in organizations, researchers have identified a series of social stressors 

such as interpersonal conflicts, the experience of workplace bullying, the experience of 

inequity or injustice, and the experience of abusive supervision as the indicators of sleep 

problems (Greenberg, 2006; Hietapakka et al., 2013; Niedhammer et al., 2009; Winwood 

& Lushington, 2006). Another non-task stressor is derived from employees’ concerns 

about their careers. Little research has attended to this perspective. One of the very few 

studies showed that job insecurity negatively predicted sleep quality (Burgard & Ailshire, 

2009). A more recent study showed that being over-qualified could also predict sleep 

problems (Stenfors, Hanson, Oxenstierna, Theorell, & Nilsson, 2013). Lastly, research 

has also revealed that employees’ own decisions or behaviors can cause sleep problems. 

A very recent study by Yuan, Barnes, and Li (2018) showed that engaging in 

counterproductive work behavior could lead the actors to lose sleep (i.e., insomnia) as a 

function of moral deficits and a heightened level of rumination.  

 Employee Emotional Exhaustion. Research on specific topics of emotional 

exhaustion started in the 1970s, but at the time, there was not a systematic framework for 



 

37 
 

 

the inquiry of emotional exhaustion, as Schuler (1980) noted that little had been known 

about stress and emotional exhaustion at work. In the 1980s, this line of inquiry moved 

from the periphery to the mainstream of organizational research, and a tremendous 

amount of attention was placed on the importance of emotional exhaustion. Nowadays, 

employee emotional exhaustion is still a major topic of interest in the management 

literature. Historically, organizational research on emotional exhaustion has been 

predominately guided by Maslach’s and Jackson’s three-dimension conceptualization of 

burnout (Maslach, 1982; Maslach, Jackson, Leiter, Schaufeli, & Schwab, 1986; Maslach 

& Leiter, 1997). According to the model, burnout comprises three interdependent or 

interrelated dimensions: emotional exhaustion, depersonalization, and diminished 

personal accomplishment.  

As one of the critical dimensions, emotional exhaustion is defined as a chronic 

state of physical and emotional depletion that is caused by an overly heavy workload and 

continuous hassles (Shirom, 1989; Zohar, 1997). It is a feeling of physical fatigue and of 

being psychologically drained. Although all three dimensions of burnout are important, 

researchers generally agree that emotional exhaustion is the key dimension of burnout 

(Cordes & Dougherty, 1993; Gaines & Jermier, 1983; Wright & Bonett, 1997; Zohar, 

1997). Such consensus has been evidenced by both theoretical and empirical research. 

Shirom (1989) concluded that emotional exhaustion characterized by physical and 

psychological depletion lies in the “heart” of burnout. Furthermore, in a field study, Lee 

and Ashforth (1993) found that emotional exhaustion was a central mechanism in the 

burnout process. Since the 1980s, emotional exhaustion has been a critical topic in 

organizational research and has been seen as having great implications for employees’ 



 

38 
 

 

workplace behavior and overall wellbeing (Cherniss, 1993; Cordes & Dougherty, 1993; 

Kahill, 1988; Maslach, 1982). Given the importance of emotional exhaustion, researchers 

have explored both outcomes and antecedents of emotional exhaustion.  

 In terms of the outcomes of emotional exhaustion, researchers have long been 

interested in the negative outcomes of emotional exhaustion, and a large body of research 

has shown that it can lead to a series of undesirable workplace outcomes such as reduced 

work performance (Halbesleben & Bowler, 2007; Halbesleben & Buckley, 2004; 

Maslach, 1982; Wright & Bonett, 1997), reduced commitment and job satisfaction 

(Cordes & Dougherty, 1993; Demerouti, Bakker, Nachreiner, & Schaufeli, 2001; Green, 

Walkey, & Taylor, 1991; Lee & Ashforth, 1996), increased actual turnover or turnover 

intention (Jackson, Schwab, & Schuler, 1986; Wright & Cropanzano, 1998), impaired 

work attitudes (Leiter & Maslach, 1988; Wolpin, Burke, & Greenglass, 1991), and 

increased counterproductive work behavior (Quattrochi-Tubin, Jones, & Breedlove, 

1982). More recently, Halbesleben and Bowler (2007) found that emotional exhaustion 

influenced three different motivations (i.e., achievement striving, status striving, and 

communion striving) and, in turn, influenced employees’ in-role performance and 

organizational citizenship behavior. Van Jaarsveld, Walker, and Skarlicki (2010) found 

that emotional exhaustion was positively related to employee incivility toward customers. 

In sum, emotional exhaustion has been generally found to have a negative impact on 

organizational functioning.  

As compared to the outcomes of emotional exhaustion, at least an equal amount 

of attention has been placed on the causes of emotional exhaustion. As highlighted in the 

definition of emotional exhaustion (Shirom, 1989; Zohar, 1997), the excessive workload 



 

39 
 

 

has been seen as the most direct predictor of emotional exhaustion (Maslach et al., 2001). 

Accordingly, research has provided rich evidence for this insight (e.g., Maslach & Leiter, 

1997; Schaufeli & Bakker, 2004). As this line of inquiry advanced and evolved, 

researchers have revealed an extensive number of antecedents of emotional exhaustion 

including but not limited to employees’ personal characteristics such as personality, self-

esteem, gender, and education (Maslach et al., 2001), employees’ workplace behavior 

such as emotional labor (Grandey, 2003; Martínez-Iñigo, Totterdell, Alcover, & Holman, 

2007), peers’ behavior such as workplace bullying by coworkers (Sa & Fleming, 2008) 

and peer-support (Peterson, Bergström, Samuelsson, Åsberg, & Nygren, 2008), and 

leaders’ behavior such as abusive supervision (Wu & Hu, 2009), ethical leadership 

(Zheng et al., 2015), and goal-focused leadership (Perry, Witt, Penney, & Atwater, 2010). 

Overall, research on this area of inquiry has been very fruitful, providing us a solid 

foundation to understand the factors that can increase or reduce employees’ emotional 

exhaustion.  

Employee Negative Family Interactions. Historically, research on family 

interactions is sociological and can be traced back to as early as the 1920s (Gottman, 

2013). At that time, researchers started to explore the link between sexual activities of 

couples and marital satisfaction. As research in this area evolved, researchers started to 

explore other factors associated with family interactions such as economic status and 

personality. Specific research on the relationship between work and family interactions 

started in the period from the 1970s to the 1980s. Early research in this area has 

documented that positive and healthy interactions with family are not always easy to 

achieve among employees, and researchers generally agree that unpleasant or stressful 



 

40 
 

 

work experience on a given day can impair their quality of interactions with family later 

that day through negative spillover (e.g., Bolger, DeLongis, Kessler, & Wethington, 

1989). Research on negative marital interactions generally examines four dimensions: 1) 

defensiveness, 2) conflict, 3) stubbornness, and 4) withdrawal (Gottman & Krokoff, 

1989; R. L. Weiss & Summers, 1983). Therefore, in light of this, in this dissertation, I 

define negative family interactions as family interactions that harm the relationship 

between family members. The forms of negative family interactions include being 

defensive to family members, being stubborn toward family members, conflicting with 

family members, or withdrawing from family interactions. In addition, based on research 

on the negative and unsupportive social interaction literature (e.g., Ingram, Betz, Mindes, 

Schmitt, & Smith, 2001), negative family interactions can be seen as interpersonal 

behavior among family members that produce physiological hurt or unpleasant 

psychological feelings such as distress, sadness, isolation, and rejection.  

 In the past four decades, voluminous studies have provided evidence for the link 

between work experience and negative family interactions. Although there are several 

different forms of negative family interactions, existing research tends to focus on two 

forms of negative family interactions: 1) conflictual interactions with family members 

and 2) withdrawal from family interactions.3  

With regard to the first form, research has shown that work stressors can lead 

employees to engage in conflictual family interactions. For example, Bolger et al. (1989) 

found that having a stressful day at work led to arguments and tense interactions at home. 

Relatedly, research showed that parent-child tension was greater on Monday when work 

 
3 In the following content, conflictual family interactions with family members and withdrawal from family 

interactions respectively will be referred to as conflictual family interactions and family withdrawal. 



 

41 
 

 

stress is more salient (Almeida & McDonald, 1998). Another study showed that fathers 

were more likely to have conflicts with their children on workdays with excessive 

workload than on days with a typical workload (Almeida, Wethington, & Chandler, 

1999). Research has also indicated specific work stressors that can lead to conflictual 

family interactions. For example, Hoobler and Brass (2006) found that employees abused 

by their supervisors were more likely to undermine family members (e.g., giving negative 

labels) after going home as a function of displaced aggression. 

In terms of the second form, research has shown various work experience that can 

lead to employees’ family withdrawal. For example, Repetti (1989) found that an 

increased daily workload was associated with increased marital withdrawal at home 

manifested by reduced employees’ involvement or interest in interacting with their 

spouses. Similarly, drawing from resource allocation theory which suggests that 

individuals tend to conserve their resource lost when a certain resource is stretched thin, 

Harrison and Wagner (2016) found that creative behavior at work could negatively 

predict the time spent with spouses at home. Relatedly, Leavitt, Barnes, et al. (2019) 

found that daily work-related stress could spill over to the family domain and, in turn, 

reduce the probability of having sex—an important marital interaction for maintaining 

healthy marriage relationship—later that night. Furthermore, through a moral licensing 

lens (Benoit Monin & Miller, 2001), Li, Mai, and Bagger (2017) found that engaging in 

prosocial behavior at work led to a reduced provision of family support at home. Of all 

the forms of negative family interactions, conflictual interactions and family withdrawal 

are the most common reasons for the breakdown of family relationships (e.g., divorce; 

Chang, 2004; Gigy & Kelly, 1993; Hawkins, Willoughby, & Doherty, 2012). For 



 

42 
 

 

example, in a survey of 437 men and women who had divorced, having conflict and 

growing apart were rated as two top reasons for divorce (Gigy & Kelly, 1993). 

Conflictual family interactions and family withdrawal contribute to family conflict and 

growing apart, respectively. Hence, it is important for us to pay great attention to 

conflictual family interactions and family withdrawal. In light of this, this dissertation 

will focus on these two forms of negative family interactions.   

Moral Identity  

Overview of Moral Identity. An identity refers to a person’s self-conception or 

self-definition (Erikson, 1964). Moral identity is a specific type of identity that reflects 

the moral aspects of a person, and individuals differ in the strength of such identity 

(Aquino & Reed, 2002; Lapsley & Lasky, 2001). Moral identity is defined as “the mental 

representation of one’s moral character held internally as a cognitive schema and 

expressed to others externally through one’s actions” (Winterich, Aquino, Mittal, & 

Swartz, 2013, p. 759). First proposed by Aquino and Reed (2002), moral identity has 

been widely adopted to explain people’s own moral decisions and behaviors as well as 

their judgment and reaction toward others’ moral decisions and behaviors. In their 

seminal work, Aquino and Reed (2002) described moral identity along two dimensions: 

moral identity internalization which is conceptualized as the private dimension and moral 

identity symbolization which is conceptualized as the public dimension.  

Moral identity internalization is rooted in a self-consistency perspective (Blasi, 

1984) and is associated with self-monitoring. As the private dimension of moral identity, 

identity internalization reflects moral centrality and indicates the extent to which a person 

is chronically accessible to moral-related knowledge of his/her moral characters such as 



 

43 
 

 

moral traits, goals, and behaviors within the working self-concepts (Aquino, Freeman, 

Reed, Lim, & Felps, 2009; Winterich et al., 2013). In other words, identity internalization 

answers if moral self-schemas preexist within the working self-concepts without 

considering situational factors. Hence, for those high in identity internalization, moral 

self-schemas are firmly “preinstalled.” Conversely, individuals low in identity 

internalization are less constrained by moral self-schemas, and thus, are more “morally 

flexible” regarding their moral decisions and behaviors.  

Unlike moral identity internalization, moral identity symbolization is rooted in a 

symbolic-interactionist perspective (Mead, 1934; O'brien, 2006) and is related to the 

recognition of self as a social entity (Goffman, 1959). As the public dimension of moral 

identity, identity symbolization indicates the extent to which a person tends to engage in 

the public display of visible activities (e.g., prosocial behavior) to convey one’s moral 

characters, and a higher identity symbolization represents a stronger willingness to do so 

(Aquino & Reed, 2002). Individuals high in identity symbolization are conceptualized as 

especially active in conveying their commitment to certain moral goals and ideals to 

others. By contrast, individuals low in identity symbolization are less likely to engage in 

publicly visible moral behaviors.   

Together, identity internalization and identity symbolization dimensions reflect a 

person’s moral identity (D. M. Mayer, Aquino, Greenbaum, & Kuenzi, 2012). Although 

both of them are closely associated with a person’s moral self-concept, it has been argued 

that because moral behavior is seen as mostly driven internally (Bandura, 1999; Rest, 

1986; Reynolds & Ceranic, 2007), the internalization dimension should more directly 

reflect the core definition of moral identity and should be a stronger predictor of moral 



 

44 
 

 

behaviors than the symbolization dimension (Reynolds & Ceranic, 2007), and empirical 

findings generally support the argument (Aquino & Reed, 2002; Aquino, Reed II, Thau, 

& Freeman, 2007; I. Reed & Aquino, 2003; Reynolds & Ceranic, 2007).  

Research on Moral Identity. There is a large body of empirical research on moral 

identity. Research has generally focused on two aspects: 1) direct consequences of moral 

identity, and 2) moderating effects of moral identity.  

Extensive research has shown that moral identity is a critical predictor of 

individuals’ moral decisions and behaviors. Moral identity has been found to directly 

motivate a variety of ethical or prosocial behaviors such as volunteering, willingness to 

minimize harm, and charitable giving (Aquino & Reed, 2002; I. Reed & Aquino, 2003; 

Reynolds & Ceranic, 2007). It has also been found to e