Citation
Bureaucratic constraints and street level performance of emotional labor

Material Information

Title:
Bureaucratic constraints and street level performance of emotional labor
Creator:
Drury, Ida J.
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Doctorate ( Doctor of philosophy)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
School of Public Affairs, CU Denver
Degree Disciplines:
Public affairs
Committee Chair:
Guy, Mary E.
Committee Members:
Ronquillo, John
Hansberry, Jane
Mastracci, Sharon

Notes

Abstract:
Street level bureaucrats are the face of government in public service delivery. To facilitate interactions with citizens in pursuit of policy goals, they must engage in emotional labor. Emotional labor may be self-focused, controlling one’s own emotions, or it may be other focused, with the intent to assist others in emoting to facilitate change. This is especially true in human services environments, where street level bureaucrats utilize people changing technologies in their daily work. Street level bureaucrats operate in the context of organizations, and the culture and climate of the organization may impose bureaucratic constraints on their behavior to optimize accountability and efficiency. This dissertation examines the manner in which bureaucratic constraints influence emotional labor, specifically in the organizational context of public child welfare. Findings of this research indicate that bureaucratic constraints, specifically those concerning support by agency leadership for discretion and autonomy in light of a negative outcome and time or resource constraints, have detrimental impacts on the performance of emotional labor by the caseworkers in this study. These findings contribute to the literature on theories of autonomy and discretion in street level bureaucracy as well as on the practical setting of child welfare casework and public administration.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
Copyright Ida J. Drury . Permission granted to University of Colorado Denver to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item has the following downloads:


Full Text
BUREAUCRATIC CONSTRAINTS
AND STREET LEVEL PERFORMANCE OF EMOTIONAL LABOR
by
IDA J. DRURY
B.A.S.W., Wartburg College, 2001 M.S.W., St. Ambrose University, 2003
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Doctor of Philosophy Public Affairs Program
2018


2
This thesis for the Doctor of Philosophy degree by Ida J. Drury
has been approved for the Public Affairs Program By
Mary E. Guy, Chair John Ronquillo Jane Hansberry Sharon Mastracci
Date: May 18, 2019


3
Drury, Ida J. (Ph.D., School of Public Affairs)
Bureaucratic Constraints and Street Level Performance of Emotional Labor Thesis directed by Professor Mary E. Guy
ABSTRACT
Street level bureaucrats are the face of government in public service delivery. To facilitate interactions with citizens in pursuit of policy goals, they must engage in emotional labor. Emotional labor may be self-focused, controlling one’s own emotions, or it may be other focused, with the intent to assist others in emoting to facilitate change. This is especially true in human services environments, where street level bureaucrats utilize people changing technologies in their daily work. Street level bureaucrats operate in the context of organizations, and the culture and climate of the organization may impose bureaucratic constraints on their behavior to optimize accountability and efficiency. This dissertation examines the manner in which bureaucratic constraints influence emotional labor, specifically in the organizational context of public child welfare. Findings of this research indicate that bureaucratic constraints, specifically those concerning support by agency leadership for discretion and autonomy in light of a negative outcome and time or resource constraints, have detrimental impacts on the performance of emotional labor by the caseworkers in this study. These findings contribute to the literature on theories of autonomy and discretion in street level bureaucracy as well as on the practical setting of child welfare casework and public administration.
The form and content of this abstract are approved. I recommend its publication.
Approved: Mary E. Guy


4
ACKNOWLEDGEMENTS
I am grateful to many as I complete this milestone. I would not be in the position to add PhD to my name without the support of a village. When I think of all the shoulders I leaned on, coffee shops haunted, happy hours commiserated, and celebrations large and small along the way, I feel full. There are too many people to name in this document, so I highlight a few that rise to the forefront in my mind.
First, to James, I am grateful for your steadfast support and belief in my dreams. You filled and emptied the dishwasher, tread lightly through PhD student emotions, left me alone when I needed it, and drew close when I needed that, too. I am excited for our next chapter together, zusammen.
To Dr. Mary E. Guy, aka “PhD Student Whisperer,” I am so thankful that you saw this vision for me before I truly saw it for myself. Your guidance, wisdom, and sense of humor shepherded me through this process, and helped me remember the goal. The achieved goal is taking form in North Dakota. Thank you.
I am also grateful to educators and scholars along the way who instilled in me the desire to join the academy. A few: Dr. Susan Vallem, Dr. Katherine Van Blair, Dr. Jessica Sowa, the scholars at PA Theory Network, and my committee. Thanks for being dedicated, brilliant, and for sharing it with me as a student and colleague.
Then, my tribe: Maren, Annie, and Laura. You get it. Thank you for laughter, tears, fierce loyalty, texts, and friendship. I look forward to many more years of the same, despite the many miles and states between us.
Finally, I am grateful to my family and friends, near and far, who supported me even
when this endeavor was hard to understand.


5
TABLE OF CONTENTS
I. STATEMENT OI THE PROBLEM.........................................................7
Emotional Labor.............................................................11
Street-Level Bureaucracy Theory.............................................12
Impact of Organizational Culture on Street-level Bureaucrats................13
Institutional Context for Practical Emotion Work............................17
Research Questions..........................................................20
II. LITERATURE REVIEW..............................................................21
Emotional Labor.............................................................21
Street-Level Bureaucracy Theory.............................................26
Impact of Organizational Culture on Street-level Bureaucrats................29
Public Child Welfare as an Institutional Context for Practical Emotion Work.37
III. HYPOTHESES AND RESEARCH DESIGN................................................42
Hypotheses..................................................................43
Research Design.............................................................45
Operationalization of Theoretical Constructs................................48
Factor Analysis.............................................................55
IV. ANALYSIS AND FINDINGS.........................................................59
Missing Data................................................................60
Results.....................................................................60
Results of Hypotheses Testing...............................................70
V. DISCUSSION....................................................................72
Summary of Findings.........................................................72
Authentic Expression........................................................75
Deep Acting.................................................................76


6
False Face..............................................................78
Self-Efficacy...........................................................80
Intent to Quit..........................................................81
Limitations and Recommendations for Future Research.....................82
Conclusion..............................................................87
REFERENCES.....................................................................88
APPENDIX.......................................................................96


7
I. STATEMENT OF THE PROBLEM
Child welfare caseworkers operate daily on the front line as intermediaries between families and government. Citizen understanding of government is cultivated in these interactions, as are relationships that facilitate personal changes in service recipients to preserve or promote the safety of children. However, the experiences of a worker who must operate within organizational cultures and climates that impose bureaucratic constraints is often ignored in favor of the idea that autonomy and discretion on the front line is a never-ending fount that can only be quelled but never extinguished. Emotion work, though, may be more dependent on the organizational balance of constraint and discretion than more cognitive, process-oriented types of public service, and lack of attention to this practical reality may set up nearly untenable jobs for those serving the most vulnerable of citizens.
Child welfare caseworkers are certainly street-level bureaucrats in function and form, but are less frequently addressed as such in the extant public administration literature which favors police officers, welfare eligibility workers, and those charged with direct human service delivery. All these street-level bureaucrats are described as the connectors between everyday citizens and government. Face-to-face interactions with citizens through direct service delivery is necessary to facilitate coproduction for achieving public value. The hope for these interactions is to improve the lives of citizens, and in doing so, promote health and well-being of the community. As such, the work routines, choices, and strategies of street-level bureaucrats are so integral in this performance that research has concluded in many instances they are creators of public policy (Brodkin, 2011). Street-level bureaucrats work within an organizational context that applies both constraints and flexibility for the performance of these work tasks.


8
All service delivery technologies have both cognitive and emotive components. The cognitive is well known: it is described in the minimum qualifications of positions and outlined in performance evaluations. The invisible part of the work is the emotive component. Emotional labor occurs in two directions: It is both self-focused and other-focused. Self-focused refers to the management of one’s own emotional state. Other-focused refers to the sensing of the citizen’s emotional state, deciding how to respond, and adjusting one’s own tone of voice, countenance, and body language so as to affect the other person’s state. Self-focused and other-focused emotional labor work in synergy to create a context for service delivery and both have different implications for job satisfaction as well as for stress and burnout (Gelderen, 2016; Hsieh, Jin, & Guy, 2011). This balance may be impacted by constraints facing street-level bureaucrats, such as limitations to time and resources. Further, in the name of accountability to citizens and the organization, constraints may be imposed at the organizational level on the discretion and autonomy afforded to street-level bureaucrats.
Bureaucratic constraints are commonly understood as a framing characteristic of street-level activities (Lipsky, 1980/2010; Wilson, 2000). These constraints become the organizational context, referred to in the industrial psychology literature as organizational culture and climate. There exists considerable debate about the defining characteristics of organizational culture and climate, however two parts of empirical work in this area serve to operationalize bureaucratic constraints: rigidity of routines and rules and centrality of decision making. Both these areas describe the level of discretion in the face of formal and informal rules allowed for the street level bureaucrat in the performance of her/his work. Eligibility decisions or discretion for enforcement of rules are examples of the dependent values in these types of studies. For the purpose of this argument, bureaucratic constraints are operationalized as those imposed by the


9
organizational environment. These constraints are separate and apart from those outlined in the ‘red tape’ or policy literature (Bozeman, 2000). This distinction is supported by literature on street level bureaucracy, where behaviors of these individuals are governed by the unlimited autonomy and discretion.
What is less understood is the impact of this organizational culture on the performance of and choices surrounding emotional labor. Further, most studies and observation of emotional labor are focused on individual characteristics and performance rather than organizational context. Specifically, for the purpose of this research, the balance of other-focused and self-focused emotional labor are known to impact daily life of the street-level bureaucrat (and by default, then, citizens), but less is known about the impact of organizational constraints designed to shape and manage individual street-level behavior. Indeed, the literature often refers anecdotally to a logical relationship, but one has not been empirically explored. How then, does an individual emotional laborer experience bureaucratic constraint, and how does that experience shape their performance of emotional labor and the balance between other focused and self-focused management of emotions?
This is of concern in certain policy contexts which rely more heavily on other-focused emotional labor, as in programs designed to facilitate change in citizen behavior. Education, injury prevention, and protection of vulnerable populations are examples of these. Human services contexts are observed to focus on people changing technologies, but also include a mix of people sustaining and people processing technologies (Hasenfeld, 1983). All of these contexts to some degree demand coproduction of public values between the government agent and the citizen/service recipient (Bovaird, 2007). In people changing organizations, engagement with citizens to generate individual change is the primary technology employed. Indeed, there is


10
indication that this coproduction or engagement is fundamental to desired policy outcomes (Whitaker, 1980). Certainly, engagement between the practitioner and the citizen and his/her family is a form of emotional labor, and it becomes the primary ‘job’ of these practitioners to produce desired policy outcomes like public safety, children safety, and community preservation.
In these contexts, as in other contexts demanding high levels of emotional labor, practitioners are demonstrably gendered, with females carrying the majority in areas like child protective services, education, nursing, and mental health services (Guy & Newman, 2004). Conversely, constraints are often imposed from the top down, where executives are more often male (Riccucci, 2009). This gender divide might logically indicate a disconnect between the facilitation of emotive work and the organizational structures where it is performed. This tension leads to further need for exploration of the impact of these imposed constraints on the performance of emotional labor by direct practitioners at the bottom of this decidedly gendered hierarchy.
Just as the cognitive aspects of a job demand resources, emotion management demands, such as time, autonomy, and discretion, also exist. Constraints, by definition and consequence, lower the resources afforded. How can other-focused emotional labor occur in such situations? Likewise, is self-focused emotive regulation simply a matter of survival for the street-level bureaucrat? Finally, do program rules demand other-focused emotional labor while organizations simultaneously create constraints that render it untenable or emotionally exhausting for the street-level bureaucrat? This research may have practical implications for successful performance of emotional labor, including an impact on the understanding of workforce interventions that effectively guide practice and also create work-able and rewarding careers for street level bureaucrats. Turnover is of utmost cost and concern in these areas of practice, and in


11
the struggle to create accountability, it is possible that high turnover rates have become somewhat inevitable. This research also may impact the theoretical and commonplace picture of ‘unrestrained’ behaviors by street-level bureaucrats, posing instead a more nuanced picture of the organizational impact on emotional labor performance.
Emotional Labor
Operationally, emotion management can be empirically observed in several forms: 1) assisting clients in managing their emotions, 2) displaying many different emotions when interacting with others, some of which may be incongruent with the worker’s actual emotions, and 3) dealing with emotionally charged issues as a critical dimension of the job (Guy, Newman, & Mastracci, 2008). The performance of emotion work as a primary job function is referred to as emotional labor. Specifically, in public administration, street-level bureaucrats perform emotional labor to pursue policy goals.
The research on emotional labor is divided into a variety of contexts and scenarios. As mentioned earlier, it is recognized as gendered, so one focus is on the undervalued role of women in the performance of emotional labor. This is so much the case that gender has been used as a proxy for an agency’s emotional labor capacity (Meier, Mastracci, & Wilson, 2006). Research also identifies antecedents to emotional labor, including organizational demands on the types of emotional behavior required of the individual worker (Morris & Feldman, 1996).
Finally, a significant body of research also identifies personal characteristics impacting the street-level bureaucrat’s experience and performance of emotional management. These varied contexts create pathways to either job satisfaction or burnout.
There are moderating variables that impact the performance of emotional labor. For instance, the management of a client’s emotional state in balance with one’s own emotional state


12
may enhance a sense of control and indicate higher job satisfaction (Wharton, 1996). Personal efficacy at performing other-focused emotional labor coupled with enjoyment in working with people can also influence the degree to which burnout occurs. Overall job control, feedback, and opportunities to learn are related to lower incidence of burnout as related to emotional labor (Guy et al., 2008). Workers who report self-efficacy at emotional labor also have positively associated service outcomes for clients (Hsieh & Guy, 2008).
Deleterious effects of emotional labor occur most prominently when workers must fake their feelings, or constantly self-manage emotions to maintain a “false face” (Hsieh et al., 2011; Sloan, 2014). This is particularly so for women (Johnson & Spector, 2007). In situations of low autonomy or high constraints for job tasks, workers were also more likely to suffer from burnout and its antecedents: emotional exhaustion and job stress (Johnson & Spector, 2007). This makes sense given the relation of job satisfaction with self-efficacy, and is an initial glimpse into the potential for organizational culture to impact emotional labor performance.
Street-Level Bureaucracy Theory
Lipsky’s Street-Level Bureaucracy (1980/2010) provided a theoretical framework through which government agents may be viewed or understood. Metaphorically, Street-Level Bureaucracy Theory connected the vast body of scholarship in public administration with the realities of front line labor, work routines, and interactions in much the same way that street-level bureaucrats connect government with citizens. The primary characteristics of this theory can be grouped into five areas: 1) front line status, 2) people processing, 3) inherent discretion,
4) irreducible autonomy, and 5) ultimate policy making (Maynard-Moody & Portillo, 2010). All five areas of this theory have implications for the research questions under examination.


13
Due to the inherent discretion and irreducible autonomy afforded to the street-level bureaucrat, scholars and managers seek to describe and attain accountability. This takes shape in practice as performance management. Effective performance management often focuses on outputs, seeking to quantify the routines and structures at the street level (Behn, 2014). To tighten the control over this workforce, bureaucratic and organizational constraints are applied (Wilson, 2000). Emotional labor, on the other hand, is the least measured or acknowledged element of job performance (Meier et al., 2006). When should the street-level bureaucrat focus on the citizen relationship to cultivate change in the citizen’s life or circumstance? When does the street-level bureaucrat choose to mask frustration or irritation with their work by simply smiling their way through? In the name of accountability to agency and citizen, can constraints render the effective and sustainable practice of emotional labor impossible, or otherwise so challenging that direct practitioners either turnover or languish in their positions long after their effectiveness has been tapped? Given high levels of turnover and challenges in these policy areas, could this be a mechanism to address?
Impact of Organizational Culture on Street-level Bureaucrats
Industrial psychology and the study of bureaucratic behavior have common roots. At the turn of the twentieth century, with the industrial revolution in full swing, the American government had undergone significant changes since the drafting of the constitution. So too, the bureaucracy took on new character and federal, state, and local levels. Using knowledge gained from factories and shops, scholars sought to apply these principles in public administration. Lamenting the national losses from a lack of productivity and efficiency in the public sector, Frederick Taylor (1911) drew connections between successes in current industry and the practice of scientific management. Taylor recommended that administrative processes be examined for


14
ineffective components; perhaps, those rule-of-thumb actions passed down from generations before were unnecessary in the current context. Using the example of a bricklayer, Taylor suggested the introduction of simple apparatus might further eliminate superfluous actions and increase efficiency (1911). Taylor gained a great deal of government attention for these proposals, and lent support to the growing ‘scientific management’ movement in public administration, the seemingly practical roots of which remain to the current day.
While the bureaucracy blossomed, practitioners urged standardization of municipal management. Citing the standardization of typewriter keyboards, threading on water pipes, and admission requirements to institutions of higher education, Dr. Jesse Burks, then director of Municipal Research in Philadelphia, suggested that similar standardization could be applied to government processes and budgeting with successes ranging from saving money to saving life and limb (Burks, 1912). Similar bureaus of research emerged across the country, with the purpose of increasing efficiency of local government through scientific methods. With caution not often seen in latter day reform efforts, a New York Municipal Bureau of Research trustee commented,
Municipal research is a method, not a panacea. It aims not to make over either the man in office or the men who vote, but to give men as they are better methods of watching and judging what their public servants do (Hopkins, 1912, p.
244).
The overt focus on efficiency in public administration is apparent in much of the other formative literature as well. Scholars like Frank Goodnow and William Willoughby called for budgetary and structural reform in government, which was a reflection of the need for efficient use of the public trust and public monies. An underlying concern was that public servants would actively pursue pure self-interest instead of the interests of the people they represented, as


15
Goodnow had observed in the Italian system (1900). Similarly, Willoughby is credited for moving the field toward cementing scientific management principles as a foundation for efficient and effective public administration (Sayre, 1958). The first textbook on public administration further solidified scientific management as the dominant theme (White, 1926).
These early scholars codified rationality as synonymous with efficiency in the practice of public administration (Sayre, 1958). Consequently, the public service was expected to be dispassionate and impartial to the changing winds of the political environment so as to maintain effective and efficient performance. This is evident in emphasis on the oft-debated politics-administration dichotomy (Goodnow, 1900). Other scholars explored the application of scientific management principles as related to the structure of government, and, echoing Goodnow, emphasized that politics and administration could not be exercised simultaneously without producing inefficiency (Gulick & Urwick, 1937). In addition to the P/A dichotomy, the theoretical demands of the new bureaucracy also weighed on the importance of neutrality. Weber initially considered the work to be conducted, “sine ira et studio,” or without emotion.
However, at this same time, psychology experimenters engaged in an industrial study of the impact on illumination on work outputs at the Hawthorne Works of the Western Electric Company. Ironically, the lasting results of this study do not speak to the study’s original intent, which had been an increased understanding the relation of the work environment (essentially the lighting in the production area) and the relationship to the generation of outputs by workers. Instead, the experiment showed that regardless of the physical work environment, workers were most impacted by the ‘meaning’ they made out of their work in the context of the experiment (Roethlisberger, 1941). This new knowledge impacted the study of human resources in both industry and in government. The lead researcher commented on the implications of this new


16
discovery, “Too often we try to solve human problems with nonhuman tools and.. .in terms of nonhuman data” (Roethlisberger, 1941, p. 8). Certainly, if the work product is overtly physical and industrial in nature, and the producers need attention to their humanity, what can be expected of the impact of the work environment, or organizational context on the performance of emotional labor?
Theories of street-level bureaucracy are focused on the individual as well as the irreducible autonomy and discretion afforded members of this group. There is general consensus that this group can never be fully controlled by their organizational context. Ultimately, argues Lipsky, street-level bureaucrats are called to be accountable to clients rather than and sometimes in spite of their organization (p. 161, 1980/2010). This creates tension, then, between two theoretical frameworks of human behavior: that of the individual and that of the organization.
The practical observation of emotional labor provides a platform for this juncture. The study of emotional labor readily crosses these barriers, recognizing but not full examining the importance of organizational context on work where, “emotional performance.. .is bought and sold as a commodity,” (Mastracci, Newman, & Guy, 2006, p. 124).
Like emotional labor performance by the individual, the influence of organizational culture on client outcomes is well-established in the field of human services, including in substance use treatment (Ghose, 2009) and child welfare (Glisson, Green, & Williams, 2012). In the field of industrial psychology, the culture of an institution is linked to poor performance of practitioners, but experience of emotional labor goes uninspected. Some aspects of organizational culture include decided focus on bureaucratic constraints. Specifically, centrality, or the degree to which decisions are limited to higher ranking individuals in the hierarchy, as well as rigidity, the extent to which front line staff have discretion and flexibility over


17
bureaucratic rules and routines, may serve to operationalize bureaucratic constraints (Glisson et al., 2012). The research on organizational culture forms a significant basis for human resource approaches in public agencies and is focused on burnout reduction, gains in retention, and maintenance of job satisfaction for employees, but it ignores the literature on emotional labor as the primary work task.
In the context of child protection, the impact of the work is also characterized as burnout, as well as ‘compassion fatigue’ and ‘secondary trauma.’ Unfortunately, the majority of research treats the latter two terms as relatively inevitable, and does not fully explain organizations where practitioners persevere in the workforce for longer tenure and organizational cultures where teams demonstrate higher levels of job satisfaction and burnout avoidance despite similarities in everyday crisis or client groups. Essentially, these are deleterious after effects of the work, but do not take into account the job design itself. The research on emotional labor provides an interesting counterfactual for empirical study that is absent in these other constructs: while prolonged practice of high levels of emotional labor are related to burnout, emotional labor may also produce high levels of job satisfaction (Guy, Mastracci, Newman, & Maynard-Moody, 2010). As stated earlier, job satisfaction typically occurs when the performer feels self-efficacy toward emotional labor, and incurs a sense that their work is meaningful. Any attempt at empirically examining relationships outside the well-established pathways between the practice of emotional labor and burnout resulting from it must include the organizational culture of the performer.
Institutional Context for Practical Emotion Work
The policy context for this inquiry is child protective services and specifically, the portion of that system relying primarily on encounters with front line caseworkers for the


18
realization or maintenance of child safety. While much of the media, and also a good portion of many child welfare agencies, focus on the removal, placement, and permanency of children or on those children who die from maltreatment, this dissertation narrows to the front end of the child protection system, or the part most prevalently encountered by the public. Nationally, 3.2 million children and their families received an investigation or assessment for services in 2014. Of those children, approximately 1.3 million children received some sort of family preservation or safety-related service following the assessment or investigation. During the same time, 241,919 children (less than 8% of total children encountered) were removed for foster care services (US Department of Health and Human Services, Administration for Children Youth and Families, & Children’s Bureau, 2014). Due to limitations in reporting and definition, it is hard to pinpoint the number of children who die of maltreatment following system intervention, but the national report Child Maltreatment, 2014 has a count of 1580, making this occurrence a generally rare event.
The focus on removal and foster care of very vulnerable and traumatized children is not unwarranted, nor is the goal for better understanding situations for children who die as a result of child maltreatment following system involvement. However, it is important to note in cases of foster care and the expansive formal system of courts, that foster caregivers, guardians ad litem and court-appointed special advocates play important roles in the delivery of services alongside the individual child welfare caseworker. Similarly, the wide variety of circumstances, social contexts, and involved entities for children who die from maltreatment make it difficult to generalize a fully preventative response (Commission to Eliminate Child Abuse and Neglect Fatalities, 2016). While the community-based system of care for children at risk or experiencing low to moderate levels of child maltreatment in their home is also expansive, the public child


19
protection caseworker is often the only formal entity involved in these cases, and most often on a voluntary basis once a relatively low threshold for child safety has been determined. Typically, in-home services are focused on tertiary prevention of child maltreatment and success is often measured by rates of recurrence. However, these rates are not often tied to the practices designed to reduce the rates, but are rather pared with more tangible outputs such as time to initial response and thoroughness of assessment and investigative practices. This performance-oriented reality ignores the emotive components of child welfare work by claiming effectiveness can be measured quantitatively. However, consideration of the basic logic model of child welfare services indicates effectiveness is achieved when families alter their behavior to provide ongoing safety for their children. This change does not occur as a result of a well-timed, coordinated process but is instead fostered by the emotion work on the street-level.
It is unclear whether it is despite or due to the aforementioned lopsided focus of the child welfare system, but broad sweeping child welfare policy changes in the wake of child fatalities and/or scandals are often focused on the front end portion of the system (Gainsborough, 2010). One might postulate that these reactive policies might constrain front line caseworkers in the hope to further improve the system and subsequent outcomes for children and families. These constraints come in the form of performance regimes that elevate the importance of policy and procedure (e.g. PerformanceStat) in importance for outcomes such as trust, transparency, and responsiveness (Behn, 2014). Similarly, organizational leadership may apply constraints and limits on street-level discretion, favoring instead tools like actuarial risk assessment or the algorithms of predictive analytics to determine workflow and emphasis. Emotional labor, on the other hand, including the display rules of the technology or the logic of the street-level bureaucrat as change agent are subsequently ignored. How does the pursuit of accountability for


20
relatively rare systemic events impact the emotional labor desired on the front line for the majority of children and families impacted by the child protection system? For these street-level bureaucrats, emotional labor is their dominant work product, but do their jobs ever become highly difficult or untenable because of this quest for accountability?
Research Questions
In sum, the goal of this research is to further enliven the experience of street-level bureaucrats engaged in emotional labor within a defined organizational culture that includes bureaucratic constraints. This dissertation queries these questions: How is the context of bureaucratic constraint in organizational culture related to the performance of emotional labor by street-level bureaucrats? How is the performance of other-focused emotion management limited or expanded in situations of high bureaucratic constraint? Similarly, is self-focused emotion management simply survival for the street-level bureaucrat and a disguise for self-preservation? How can bureaucratic constraints contribute to the creation of difficult jobs for street-level bureaucrats, thus leading to high levels of emotional exhaustion, burnout and costly turnover?


21
II. LITERATURE REVIEW
The foundation for this research takes root from an enduring public administration examination of street-level behavior. Debates about the limits of practical rationality of these actors as distinct from economic man led to Simon’s observation that they lack the wits to fully understand the implications or outcomes of each act of discretion (Simon, 1997). This literature review explores street-level behaviors with this underlying assumption. Further, it highlights the institutional context of street-level bureaucrats that further bounds behavior. In particular, the review focuses on technologies that require greater degrees of emotion work. The choice to emote, to self-regulate, or to create conditions under which others may emote, is an act of discretion. As such, the literature sets the stage to explore the impact of bureaucratic constraints in this area. Finally, this review explores the policy context of child welfare, as current reforms cause work technologies to shift from people processing to people changing.
Emotional Labor
Emotional labor as a construct was first observed and described empirically by Hochschild in 1983. When observing flight attendants, she began to uncover the unique characteristics of this service industry, where as a condition of employment, individuals were charged with tending to the emotional environment of an airplane as their primary job function. Air travel is stressful, and airlines found that the smiling, unflappable, servile but confident, attendant (usually a female) made all the difference for the customer experience. Hochschild’s observations, which took a sociological bent, birthed an extensive body of research that spanned numerous industries, private and public. Further, this research also took place in a variety of disciplines, from sociology to industrial psychology, to public administration. The focus of this inquiry comes from the area of public administration, though as a construct, emotional labor, as


22
well as its antecedents and consequences, exhibits a great deal of stability across contexts, fields of study, and specific professions and job responsibilities (Hiilsheger & Schewe, 2011).
For the public servant, emotional labor is a characteristic of the job, and it occurs in pursuit of a policy goal (Guy et al., 2008; Mastracci, Guy, & Newman, 2011). Similar to Hochschild’s flight attendant, the public servant engaged in emotional labor is creating an emotional environment intended to elicit trust, satisfaction, and confidence from the public. Emotional labor takes place especially in those jobs which are crisis driven and emotion laden, such as 911 call takers, first responders, crisis response workers including domestic violence advocates, and human services employees like child welfare caseworkers and mental health providers.
Typically, emotional labor per se takes on two distinct forms, surface acting and deep acting. Both were described in Hoschchild’s early work based on a quite literal interpretation of the word ‘performance.’ Hoschchild turned to the theater, where a paradigmatic shift had occurred in dramaturgical methods. In early days, actors were instructed to manipulate their actions, inflections and emotive performances to reflect the intention of the script and director. This, then, is considered ‘surface acting’ in that it does not penetrate the actual felt experience of the actor. On the other hand, Stanislovski, a 19th century Russian actor and director, derived a different ‘system’ by which actors would attempt to experience the character as intended by the writer. In this manner, referred to as the Stanislovski System or Method Acting, the intended emotion is more genuinely felt by the actor and thus results in a more convincing, spontaneous, and authentic performance. This is what Hoschchild called “deep acting” (1983).
In a job, surface acting occurs when a worker’s true emotion is covered by an incongruent emotion according to display rules of the organization (Grandey, 2003). For


23
example, the organization might require a cashier to smile at customers during a transaction (Rafaeli & Sutton, 1990). Another characterization of surface acting is ‘false face’ (Guy et al., 2008). Typically, this type of ‘mask’ conserves emotive resources and involves less effort than deep acting (Gelderen, 2016; Kruml & Geddes, 2000). Conversely, deep acting involves a great deal of emotive effort, and occurs when the worker expresses emotion that s/he genuinely feels in an effort to bring about the type of situation or client experience desired by the organization.
In deep acting, the worker then behaves as if they are truly feeling the emotion required by the job, and this results in a more authentic performance for the service recipient. Empirical evidence suggests that service recipients are adept at discerning the difference between surface acting and deep acting, and may even respond more positively to deep acting by the service employee (Zhan, Wang, & Shi, 2015).
The performance of surface acting and deep acting are both important in the delivery of human services. However, there are implications for the balance of this work on burnout, job satisfaction, and retention of the employee. Often, research in this area utilizes Conservation of Resources (COR) theory to explore the manner in which emotional laborers select from a finite resource pool, and conserve efforts throughout the day (Gelderen, 2016; Grandey, Foo, Groth, & Goodwin, 2012). There exists a paradoxical relationship between the resources needed for different types of acting and the consequences of the performance. On one hand, more resources are required for deep acting, whereas surface acting ‘costs’ less to the worker. Different from ‘harder’ physical labor that results in more exhaustion from prolonged heavy lifting, with emotional labor, the deleterious impact of emotional labor comes from prolonged surface acting, even though the emotional resources required are less. Time and again, research indicates that the most detrimental consequences of emotional labor result from prolonged surface acting


24
(Chau, Dahling, Levy, & Diefendorff, 2009; Goodwin, Groth, & Frenkel, 2011; Hsieh et al., 2011; Roh, Moon, Yang, & Jung, 2016; Sloan, 2014).
Similarly, burnout in particular has been identified to result from prolonged surface acting, and appears to be less about the amount but rather on the ability of the worker to manage the associated emotional dissonance (Brotheridge & Grandey, 2002). This decrease in the authenticity of worker emotions on the job has also been linked beyond just consequences for the job, including decreases overall in Public Service Motivation (PSM), specifically with the distinct motivator of compassion (Hsieh, Yang, & Fu, 2012). Similarly, this inauthentic performance can have a negative impact on wellbeing, specifically depressed mood and depression symptoms (Erickson & Wharton, 1997). Increases in surface acting are also related to decreases in affective commitment to the job (Seery & Corrigall, 2009). Some research proposes a connection between emotional labor and Secondary Traumatic Stress (STS) in the field of child welfare (Caringi, Lawson, & Devlin, 2012).
The consequences of emotional labor may be impacted by the worker’s self-efficacy regarding the performance. Self-efficacy may be operationalized as, “the worker’s feeling of effectiveness in her role” (Sloan, p. 277, 2014). The positive impact of self-efficacy is noted empirically in several studies of service workers (Guy et al., 2008; Hsieh et al., 2011). Self-efficacy plays a moderating role in the relationship between surface acting and ‘selfestrangement’ or feelings of authenticity in the working role, particularly in situations where high emotional labor exists (Sloan, 2014).
From an organizational perspective, many studies examine a worker’s autonomy as a characteristic of the overarching context. Typically, autonomy is measured by responses on a 1-10 scale to similarly versions of the question about the level of control they have over how job


25
tasks are done (Wharton, 1993). Autonomy on the job is related to lower levels of inauthenticity (Erickson & Wharton, 1997). Similarly, emotional labor as a whole is significantly related to positive consequences in situations where individuals have high levels of job autonomy (Wharton, 1993).
Pugliesi (1999) discussed the constructs of deep and surface acting within the confines of the audiences, or subjects, of the labor. This slightly different bent delineates the categories as ‘self-focused’ and ‘other-focused’ emotional labor. This construction suits the study of emotional labor in human services in that the client focus is important as a work technology, but the maintenance of self and the focus on one’s own emotions has implication for the closely associated workforce issues. This is in contrast to some of the literature on bureaucratic ‘coping’ behaviors. For example, Tummers, Bekkers, Vink, & Musheno (2015), illustrate that a particular coping behavior observed in frontline workers is ‘moving toward clients’ to break and bend rules, or offer personal resources. This construction completely lacks an understanding of the emotive components of work and incorrectly reduces these street-level behaviors to a linear, rational, and cognitive framework. I argue in this research that ‘moving toward clients’ is an aspect of ‘other-focused’ emotional labor, and that the emotive foundations of this behavior cannot be ignored. This argument is supported by studies on the orientation of organizations, where client-oriented firms see frontline workers elevated in status due to their distinct knowledge of clients and longer relationships as compared to others in the organization (Lively, 2002). This client orientation as the emotion work of street-level bureaucrats is less understood in the literature, and specifically with regard to the ever-present constraints that form the institutional context where they are performed.


26
Street-Level Bureaucracy Theory
Lipsky’s work to uncover the experiences of street level bureaucrats provides further foundation for this study (1980/2010). At the street level of the bureaucratic hierarchy, citizens encounter the government with personal interactions and relationships with government agents. In many cases, these encounters are most tangible government-related experience. Each encounter may serve, alongside media venues and popular culture, to shape public values such as trust and confidence in government (Moore, 1995). Further, street level bureaucrats are afforded discretion and autonomy to conduct their work, which is guided by accountability to both the agencies and clients they serve. In human services organizations, street-level bureaucrats’ context is shaped by the technologies they deliver, whether people processing, people sustaining, people changing, or necessarily multi-technology in design (Hasenfeld, 1983). Their responses and behaviors to these varied contexts play an important role in public management literatures, specifically in personnel management strategies and performance monitoring and management. Typologies emerge according to organizational situations, and efforts to understand these more nuanced contexts also form an important backdrop to this work. This section explores each of these facets of street-level bureaucracy theory and applied research.
The front line status of street-level bureaucrats puts them face-to-face with citizens. Their focus may be in direct service delivery in law enforcement, social services, mental health, job counseling, or emergency services. They may serve as gatekeepers to citizens seeking access to services, as in entitlement benefits or medical insurance. In the last twenty years, citizens have been increasingly defined as consumers (Powell, Greener, Szmigin, Doheny, & Mills, 2010) or even co-producers of services (Bovaird, 2007). The front line, then, has rapidly become an arena where citizen satisfaction and engagement are tell-tales of policy effectiveness.


27
As ‘people processors’ (Prottas, 1979), activities at the front line take on additional weight. Personal, direct contacts characterize work designed to form both short- and long- term relationships with citizens (Maynard-Moody & Portillo, 2010). Ultimately, street-level bureaucrats are called to be distinctly ////-bureaucratic in the traditional, efficiency-focused, Taylor-esque impression elicited by the term ‘processing.’ Indeed, Lipsky considers the street-level bureaucrat as ‘the link between bureaucracy and democracy’ (p. 160, 1980/2010), further illustrating their need for flexibility and responsiveness. A balance must be maintained between duty to the agency and duty to the citizen.
Instead of mechanical repetition, then, individuals must exercise fluid discretion in everyday work at the street-level. This inherent discretion further shapes the work and the desired outcomes. In addition to the adaptations and strategies identified by Lipsky (1980/2010), street-level bureaucrats are known to go above and beyond in heroic services to some citizens, even at the expense of others or themselves (Maynard-Moody & Musheno, 2003). Others have gone ‘rogue’ in favor of advocating for a desired reality at cross-purpose with program or policy (O’Leary, 2006). As they pursue their understanding of justice, street-level bureaucrats exercise discretion to make choices (Kelly, 2005). The degree to which discretion should be limited or enabled at the street-level is an ongoing debate in public administration scholarship. These limits on discretion come in the form of policies, mandates, performance goals, and outcomes measures (Radin, 2006; Wilson, 2000).
Limits on discretion only go so far in limiting the flexibility afforded the street-level bureaucrat. This “irreducible autonomy” is part of the reality of front line work (Maynard-Moody & Portillo, 2010). It does not make sense that every decision, small, minor, or otherwise, would be observed and regulated by someone higher in the hierarchy, even in the name of


28
accountability. Further, the limited resources afforded to the street-level worker demand hard-to-measure choices for and against citizens: whom to help and whom to ignore (Calabresi & Bobbitt, 1978). Given that emotional labor can smooth interactions with citizens, it is important to understand the bounds, if any, of autonomy in the performance of emotion work, particularly as it applies to a worker’s own finite emotional resources. These organizational bounds are largely ignored in the literature, with the focus instead on bounds within the individual or according to the situation or client.
Irreducible discretion and autonomy are also highlighted when considering the final characteristic of street-level bureaucracy theory: ultimate policymaking. Full understanding of policy implementation remains a challenge in the field of public administration (Bardach, 1977; Pressman & Wildavsky, 1973; Robichau & Jr, 2009). The street-level bureaucrat is often the enforcer, gatekeeper, and/or vessel of service delivery. As such, the discretion and autonomy afforded these individuals leads to the conclusion that they are the ultimate policy makers (Brodkin, 2011; Meyers & Vorsanger, 2007). This commonly recognized idea of bottom-up implementation is not without influence, of course, from the top down. Managers, jurisdictional characteristics, and politicians have all been found to have influence, albeit limited in some cases, on actions at the street level (May & Winter, 2009).
Beyond mere influence, how can accountability be ensured for these actors? Scholars and public managers alike shape responses for this question. In repeated new emergence of management eras, from New Public Management to Managerialism, researchers explore different ideas of accountability (Behn, 2001; Hood, 1991; Radin, 2006). From these efforts emerge various strategies for performance management and ideas about administrative control (Lipsky, 2010). In the 30th anniversary edition of Street-Level Bureaucracy, Lipsky tackles this


29
idea on human services in his new chapter, “77ie Assault on Human Services: Bureaucratic
Control, Accountability, and the Fiscal Crisis .” In it, he argues that bureaucratic accountability
for street-level actors is ‘virtually impossible’ to attain due to high levels of discretion and
because accountability must simultaneously be to the agency and the client (2010, p. 159). While
Lipsky acknowledges that accountability at heart is the relationship between people and groups,
he does not explore the consequences of this emotive dilemma in great detail. Specifically, street
level bureaucrats are accountable for the performance of emotional labor. If Lipsky’s argument
holds true in the emotional realm, I argue that street level bureaucrats must weigh their emotion
work on this balance between the organization and the client. Perhaps the impossibility of this
task is the mechanism for the burnout Lipsky later observed:
Thus, generations of thoughtful and potentially self-sacrificing people are disarmed in their social purpose. They come to believe that it is impossible to find conditions conducive to good practice, and that the public agencies cannot be otherwise structured. Their choices appear to be to leave public service for other work or to resign themselves to routine processing of clients. . . (2010, p.186).
Impact of Organizational Culture on Street-level Bureaucrats
The impact of the organization as related to human behavior was also manifest in a formative, classic, parallel Friedrich/Finer debate. This debate turned largely on the understanding of human, specifically bureaucratic, behavior. The exchange began with Friedrich’s assertion that bureaucrats must perform in honor of their profession and use the responsibility entrusted to them to develop and apply specialized knowledge in the interest of the public (Friedrich, 1968/1940). Finer responded by asserting that mutual professional responsibility was not a sustainable practice, and that performance of bureaucrats must be preserved through structures that ensured enduring responsiveness (Finer, 1941). History reveals that neither of these scholars was entirely correct (Stewart, 1985). However, echoes of both sides


30
take shape in contemporary methods for performance management and administrative reform efforts, with ‘Finer-type’ strategies of rule setting and rule compliance measurements as ways of determining bureaucratic performance. ‘Friedrich-type’ methods of peer accountability and group processes of goal setting also emerged. Arguably, when seeking accountability for intractable problems, interventions focus on Finer’s method of limiting discretion and increasing rules. This is particularly true in the child welfare field, where multidisciplinary child fatality review teams across the nation review child deaths. Following reviews of child fatality cases across jurisdictions, these teams commonly recommend changes in the assessment process for child welfare, and specifically those surrounding assessment of safety and risk (Douglas & Cunningham, 2008). If the work were only ‘people processing’ and not at all about ‘people changing,’ this would remain logical, and the limits on street-level discretion would indeed resolve the problems at hand. This, however, is not the case for the former or the latter.
In the context of child welfare and work that is dominated by emotional labor, perhaps McGregor’s “Theory Y” management practices are more suitable (1957). Theory Y management was distinguished from “Theory X” management by its recognition and preferences for managing to higher level human needs, similar to the hierarchy of human motivation, as was emerging in psychology at the time (Maslow, 1943). The traditional Theory X, said McGregor, was inferior in that it focused on basic needs and utilized a carrot-and-stick approach to management. Like others at the time, he urged attention to the ‘human side of enterprise’ (McGregor, 1957).
The human side of enterprise is overtly relevant to human services organizations. The technologies employed by these organizations are primarily concerned with clients, patients, or citizens, as in the case of public administration. Hasenfeld sets forth a taxonomy of technologies


31
employed by human services that is useful for this inquiry. In short, three types of functional human service technologies are utilized in these organizations: 1) people processing, 2) people sustaining, and 3) people changing. Each technology has a distinct purpose, a certain level of demand for emotional labor by the street level bureaucrat, and different implications for the organization and those receiving services from the organization. To further complicate this taxonomy, the technologies employed by human services organizations are often intermingled, conflicting, and not based in current science (Hasenfeld, 1983; Sandfort, 2003)
People processing was of initial concern in early street level bureaucracy literature (Prottas, 1979) and in sociology (Hasenfeld, 1972). The primary purpose of this technology is to ‘confer on people a particular social label, social position, or status that will, in turn, produce a predetermined response from significant social groups or organizations’ (Hasenfeld, 1983, p. 135). This technology is at play in many public administration functions such as licensing and inspection, eligibility determinations for social security, and assessment and investigation of abuse and neglect of vulnerable populations. Even the lowly speeding ticket may be considered the employment of people processing. The people processing technology has implications for the citizen, who is granted a label and all the accompanying social implications of such. The processing role and subsequent citizen experience is explored in current day when considering ‘stop and frisk’ or investigatory police stops in policing (Epp, Maynard-Moody, & Haider-Markel, 2014). In each case, the outcome of the technology is a ‘disposition’ or assigned status intended to enable or disable client behavior.
From an organizational perspective, the locus of control for the agency is imparted by the boundaries it designs in the community or service population. Staff and client relationships are minimal, in that they are primarily focused on gathering applicable information and instruction


32
related to the boundaries developed. Typically, these are short term encounters designed with efficiency as a priority to prevent backlog (Hasenfeld, 1983). Bureaucratic, administrative controls may be of the most assistance in this area, as the technology is conducive to manualized delivery. Managers in this technology are concerned with outputs and dispositions. It is not surprising, perhaps, that performance management strategies that focus primarily on outputs emerged in agencies with substantial people processing roles [e.g. CompStat as implemented in policing and city management (Behn, 2003)].
People sustaining technologies are concerned with those situations where relatively low capacity for change by the individual has been determined (Hasenfeld, 1983). Examples of these technologies are nursing homes, long-term care facilities for the mentally ill, longer-term residential treatment for youth, and prisons where individuals carry out long sentences. These technologies are concerned primarily with what Hasenfeld calls attribute stabilization (1983), or achieving and maintaining a particular desirable status (e.g. safe from oneself or to others, fed, healthy, clothed, etc.). The most paternal of services, people sustaining technologies are concerned with custodial care, providing sustenance, and maintaining equitable allocation to similar individuals and situations. Similar to people processing, people sustaining technologies use threats and promises to control clients, and bureaucratic, administrative controls to adequately ensure staff compliance with policies and procedures. Typically, there is great insolation within the institution from outside pressures, and the tightest of control over service recipients.
Trends toward deinstitutionalization and its replacement, community-based care have reduced the usage of these types of human services organizations (Bachrach, 1983). Similarly, the The Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA), or


33
welfare reform, took long-term eligibility for financial benefits to alleviate poverty, which had once been considered people sustaining, and oriented it as a short term solution with a focus on employment (Blank, 2002; Hasenfeld, 1983). Similar changes occurred in child welfare, which will be discussed in greater depth in a bit. Suffice to say, these primarily people sustaining organizations have been effectively thrust into a situation where the emphasis is on for people changing technologies, which require significantly different organizational and practitioner characteristics and skills.
People-changing technologies, “in contrast to people-sustaining technologies.. .assume that clients have a significant capacity to improve and are amenable to change” (Hasenfeld,
1983, p. 140). This orients the organization and the practitioner toward the objective of partnering with clients to facilitate this change in the selected attribute(s). For example, in education, the focus is on the development of the student, and in mental health rehabilitation, the preparation of individuals to return to the community. In this context, managers must focus on the desired outcomes of planned change as result of intervention. Staff must be committed to the technology, according to Hasenfeld, and are most effective when they exhibit professional judgement, are recognized for their expertise, and are allotted legitimate autonomy (1983). The difficulty with this reality, however, echoes the dilemmas of accountability reflected in SLB literature. Ethnographic studies in the field reveal that the structuration of this technology takes on extensive variability, so much so that it renders the actual work of the organization ambiguous (Sandfort, 2003).
With that caveat, there exists little doubt that this effort to change attributes in clients takes on a considerably different tone than processing or sustaining. People changing requires a strong, dynamic, at times pro-longed, relationship between the street-level bureaucrat and the


34
client. This is particularly true when the magnitude of the change is great (Hasenfeld, 1983). I argue that to cultivate this relationship, emotional labor is required, and that people changing in general is primarily emotional labor as opposed to cognitive or physical labor. Hasenfeld acknowledges this shift as well when describing the evolution of human services delivery, “Ultimately, realizing organizational effectiveness depends on the quality of relations between workers and clients, we need to understand the environmental and organizational factors that shape them.. .because emotional work is a key element of these relations, we need a theoretical model to analyze it” (p. 4, (Hasenfeld, 1992).
Table 1.1. Hasenfeld's (1983) Typology of the Functions of Human Services Technologies
People Processing People Sustaining People Changing
Type of Product Altered status Attribute stabilization Attribute change
Core Activities Classification- disposition Custodial care; sustentation Planned change
Managerial Concern Disposition of product Acceptable allocation rules Demonstrative effectiveness
Staff-Client Relations Minimal Modest Extensive
Client Control Threats and promises Threats and promises; reinforcement control Persuasion; Reinforcement control
Staff Compliance Bureaucratic Bureaucratic Commitment
Human service organizations rarely fit into only one column of technology delivery, and often, these technologies are muddled in practice, scope, and definition (Sandfort, 2003). As noted earlier, policies and trends in governance may intermingle these technologies creating ambiguity for the SLB, lack of clarity for goals, and the characteristic “dilemmas of the


35
individual in public service” (Lipsky 1980/2010). Nonetheless, this typology serves an important purpose in highlighting the management and organizational characteristics conducive to different types of human services work, and in further amplifying the impact of these characteristics on the SLB. Further, the emphasis in this typology on the types of relationships (and thus, the levels and varieties of emotional labor) required for each technology will serve as a springboard for discussion of technologies in child welfare, the specific context for this dissertation.
Recognizing the human as the primary vessel for work technologies also helps examine the possible impact of jobs demanding a balance of other focused and self-focused emotional labor. Designing an organization where display rules are clear and also congruent with the general thoughts and feelings of the workforce in one area where recognizing the ‘human side of enterprise’ may be leveraged for emotional labor. For example, if those at the Hawthorne factory were to conduct their operation in a shop window, necessitating the public display of emotions that would build, say, trust in their brand or inspiration in the polity, the meaning each gained from work would be ‘written all over their faces,’ and those for whom this meaning was incongruous to the organizational display rules would not last long. In this manner, job satisfaction would be multiplied (or divided) by each’s genuine emotions for the work in concert with the organizational intent.
Public administration, or course, is not conducted behind glass. Metaphors aside, street level bureaucrats must interact with citizens, be responsive to needs and varied contexts, and maintain nimble levels of relational focus on the emotions and behaviors of citizens. The relational elements of accountability that Lipsky (2010) identified may be ignored at the expense of all involved, as can the requirement for relationship in the delivery of people-changing technologies. However, the scales may be also be tipped in support of other-focused emotion


36
work and the direction of relational accountability. Consider an example from Sweden, where national differences were identified from observations of safety inspectors in both the US and Sweden:
American inspections are designed more as formal searches for violations of regulations; Swedish inspections are designed more as informal, personal missions to give advice and information, establish friendship ties between inspector and inspected, and promote local labor-management cooperation. (Wilson, 2010, p. 296).
Emotional labor is the job task in the Swedish example, where the focus is on
establishing relationship, fulfilling personal mission, and assisting the client, or the ‘inspected’ in
managing emotions and relationships to accomplish workplace safety. This then, has more flavor
of the ‘people changing’ organization outlined by Hasenfeld (1983). In contrast, the observations
of American inspections seemed more concerned with people processing, or defining
status/violation of key safety regulations for the ‘inspected.’ Recall that people processing as a
technology requires ‘minimal’ staff-client relations (Hasenfeld, 1983), and thus, less emphasis on
emotive requirements of the job. Discretion in this case is much more about determining
compliance of the client, and controls are based on threats and rewards, or the carrots and sticks
of Theory X to achieve policy goals.
The organizational context of people changing organizations is a common area of study in industrial psychology. In general, this literature examines the impact of organizational culture on the delivery of human services, as well as the resulting client outcomes. While the differences and similarities of culture and climate as well as the merits of each are debated in the literature, this review considers organizational culture from a sociological perspective. Organizational culture is ‘the way things are done’ and encompasses the norms, values and resulting consequences for compliance and deviance (Schein, 2010; Verbeke, Volgering, & Hessels,
1998). Culture has been constructed to define the level of constraints at hand in an organization,


37
or the ‘rigidity’ of the organization (Glisson et al., 2008, 2012). Rigidity reflects the degree to which individuals may exercise discretion, where they must ask permission from superiors, or the flexibility in following rules, policies, and procedures in the agency.
In the area of children’s services, studies connect rigidity of culture to resulting outcomes for children (Glisson et al., 2012). This impact has been noted in acts of discretion as well, where organizational norms rather than needs of clients were found to be more influential in caseworker decisions about service and placement (Martin, Peters, & Glisson., 1998). Outside child services, discretion for emotion work was also found to be related to the organizational culture and climate of emergency room implementing a family-centered intervention with a strong emotional support component. (Hemmelgarn, Glisson, & Dukes, 2001). When organizations emphasized emotional support to families and patients as the primary mission (versus the more technical, life-saving functionality of the emergency room), workers were more likely implement emotion work as a primary part of their job. Similarly, workers also reported less negative impact from their work on their personal well-being in this type of culture (Hemmelgarn et al., 2001). These studies provide support for the inquiry at hand, but do not specifically address the rigidity of organizational culture as connected to the performance of emotion work.
Public Child Welfare as an Institutional Context for Practical Emotion Work
This study will use public child welfare as the institutional context for street-level workers engaged in emotion work. This setting is particularly relevant in the current era of constant reforms in child welfare systems, including debates about the organizational technologies best employed by child welfare agencies at the time of assessment and early intervention with families alleged to have maltreated their children. Historically, front line workers, or ‘investigators’ of child maltreatment have been structured to have a people


38
processing role. The primary work product for these investigators mirrors Hasenfeld’s people processing technology, in that the work was to seek out the reported family and child, conduct an assessment or investigation of the family situation, and determine whether or not child maltreatment occurred, usually with a preponderance of the evidence. Additionally, workers assess continuing danger to the child, and make the decision to remove the child weighed in light of situational variables. In many jurisdictions, dispositions, or ‘findings’ of child maltreatment determine the course of action for agency intervention, including access to on-going services. Further, a label was applied to the person found responsible for abuse or neglect, and that label is entered into a state central registry. In many states, potential employees can access central registries, with many states having rules about future employment in child or vulnerable adult serving agencies based on prior findings of child maltreatment.
This emphasis on triage for the child welfare investigator persisted until the 1990’s, paradoxically in response to a changing outcomes focus by the child welfare field. The onset of New Public Management (NPM) demanded a reexamination of the logic models and technologies of many human services agencies (Hasenfeld, 1992; Hood, 1995). In child welfare, the subjectivity of findings and their relative impotence for promoting change in parental behavior were identified as major difficulties (Kohl, Jonson-Reid, & Drake, 2009). A child welfare reform adopted steadily in the United States since that time is Differential Response. Differential response in child welfare provides that in cases of low and moderate risk for child safety, an alternative response may be used by the front-end caseworker. Rather than conducting an investigation to determine whether or not child maltreatment occurred, workers in an alternative response would instead assess child and family functioning to determine needs and strengths that might be reduced or augmented, respectively. The philosophical shift was related to a practical


39
belief that families might be more willing to engage with the child welfare worker in a less adversarial setting. Following initial assessment, the alternative response caseworker is instructed to seek family engagement in making necessary changes to reduce risk factors for future child maltreatment. This engagement is voluntary on the part of the family, according to model design (Merkel-Holguin, Kaplan, & Kwak, 2006).
Differential response as child welfare reform, then, effectively changed the technologies to be used by the street-level bureaucrat in an alternative response. Essentially, this change shifted the balance of the work from processing to changing, where the emphasis of the work is on relationship with clients as the primary intervention of the agency. Training modules for caseworkers new to the reform focus on this change in technology, they are told “Tow are the service,” meaning the change agent for the family. Further, a mantra of the reform is “Safety through engagement,” or the ability to achieve the desired outcome of child safety by developing family engagement and relationship in initial contacts.
Reactions to this reform in the child welfare system vary from wildly enthusiastic to cautionary, and may be reflective of the fundamental change in the service delivery technology. Misunderstandings abound as to the intent of such reform, and often focus on the need to hold caseworkers and families accountable. Critics cite the voluntary nature of these services as a primary issue. A particularly vocal critic of differential response comes from a law professor, whose stated concerns mirror those from public administration’s past: she advocates strongly for procedural fairness and court-oriented interventions rather than relationship-oriented approaches (Bartholet, 2014; Bertelli & Lynn, 2006). Despite criticism, advocates for Differential Response also abound, and agencies continue to expand and implement the practices. Further, in many


40
jurisdictions, there exists renewed emphasis on family engagement, no matter the type of initial response or stage of service (Winokur, Ellis, Drury, & Rogers, 2015).
Critics and advocates of child welfare reform often conflate the concept of family engagement with attitudes about family preservation and child safety. Indeed, caseworkers who express a proclivity for family preservation may exhibit differences in decision thresholds to take action in the course of casework (Dettlaff, Graham, Holzman, Baumann, & Fluke, 2015). Since these individual decision making thresholds occur in the context of varying organization culture and climate, as well as related to case and family characteristics, this area of study is considered an ‘ecology’ wherein nested levels of influence are understood to provide influence on decision making (Baumann et al., 2011).
Child welfare in a time of reform was selected due to the new emphasis on family engagement, or what I argue is a type of other-focused emotional labor. Successful performance of other-focused emotional labor results in increases in performance, positive service outcomes, and job satisfaction on the part of the performer. Arguably, these results might relate to increases in retention of child welfare services. Yet, for workers on the front end of the child welfare system, turnover remains an utmost concern. Nationally, the turnover rate for child protective services workers is 22%, or roughly six times the national average (American Public Human Services Association, 2005). The consequences of turnover are taxing on an already resource depleted system: retraining, abandoned caseloads, and significant impact on families served from discontinuity of care. Remaining colleagues must carry the bulk of this burden, which in turn may lead to more burnout and subsequent turnover (DePanfilis & Zlotnik, 2008).
Conversely, self-focused emotional labor, or the masking of felt emotions and surface acting, is related to burnout, emotional exhaustion, and turnover. When the technology on the


41
front end of the child welfare system changed from people processing to people changing with increased demands for emotion work, did the organizational contexts change with it? What is causing frontline workers to turn inward and burn out? Could this be a manifestation of self-preservation and conservation of emotional resources? Are bureaucratic constraints imposed for process accountability limiting the ability of street-level bureaucrats to perform emotion work? Based on the questions posed in this inquiry and this review of the literature, the next chapter delineates five hypotheses for investigation, outlines the operationalization of constructs, and describes the methods employed for testing the hypotheses.


42
III. HYPOTHESES AND RESEARCH DESIGN
This chapter provides a description of the research questions explored in the literature. The overarching question as well as the more specific sub questions, are addressed through five hypotheses that address the constructs of interest as operationalized and informed by the literature. As identified in chapter one, the overarching research question of this inquiry is: How is the context of bureaucratic constraint in organizational culture related to the performance of emotional labor by street-level bureaucrats? This question is examined using three sub questions: (1) How is the performance of other-focused emotion management limited or expanded in situations of high bureaucratic constraint? (2) How is self-focused emotion management survival for the street-level bureaucrat and a disguise for self-preservation? and (3) How can bureaucratic constraints contribute to the creation of job stress for street-level bureaucrats, thus leading to high levels of emotional exhaustion, burnout and costly turnover?
The policy and practice area is child welfare, where street level bureaucrats work with families identified in the child welfare system who meet the legislative and jurisdictional criterion for formal intervention, services, and out of home placement. Essentially, caseworker jobs are heavily weighted toward fostering non-adversarial relationships with families on their caseload to assess child safety, family needs and strengths, and to aid in delivering or arranging services to address family concerns. There also may be variability in emphasis on family engagement, family preservation, and child safety based on the stage of the case, the direction of the agency, or the needs of children, youth, and families. As noted in the review of literature, the success or performance of this group can be measured both federally and locally through the recurrence rates of families in the child protection system or of children in out of home placement due to safety concerns. Thus, this group also meets the one of the main aims of


43
agencies delivering people changing human services technologies: the need to foster relationships in an effort to cultivate changes in human behavior in pursuit of a policy goal. Hypotheses
The main research question for this dissertation examines the context of bureaucratic constraint in organizational culture and the relationship to the performance of emotional labor by street-level bureaucrats. The first sub question, designed to elicit answers regarding this context is: How is the performance of other-focused emotion management limited or expanded in situations of high bureaucratic constraint? When street-level bureaucrats experience depletion of resources, autonomy, and discretion, the literature notes that adaptive behavior is a possible consequence. In particular, ‘creaming’ behavior may occur, or strategies and work routines that focus on the easier or more expeditious routes of service delivery (Brodkin, 2011). There lies a key difference between other-focused and self-focused emotional management, in that other-focused emotional labor demands deep acting, and thus time and resources to effectively engage with clients and citizens, either in seeking contact, or establishing relationship. Self-focused emotional management, on the other hand, does not require relationship, and is a more efficient emotional choice for the street level bureaucrat. Further, it operates within the relative vacuum of an individual psyche, thus conserving resources and effort. Consequently, focusing on self as compared to the other can be considered the emotional equivalent of selecting the easier or more expeditious behavior in the face of constraints. This line of thinking lends to the following hypotheses:
HI: Bureaucratic constraints will be negatively related to authentic expression of emotion. H2: Bureaucratic constraints will be negatively related to deep acting.


44
Similarly, the second sub research question explores whether self-focused emotion management is simply survival for the street-level bureaucrat and a disguise for self-preservation? To further narrow street-level behaviors within self-focused emotional management, adaptive behavior might point to false face or surface acting as a primary strategy in the context of organizational constraint. That is, deep acting may not be possible in the time afforded, and may not be valued by the agency as much as basic display rules. Thus, the following hypothesis:
H3: Bureaucratic constraints will be positively related to ‘false face ’ strategies for self-focused emotion management.
Finally, at the practical level, a sub research question is: How can bureaucratic constraints contribute to the creation of job stress for street-level bureaucrats, thus leading to high levels of emotional exhaustion, burnout and costly turnover? The final two hypotheses point to an important crux, where the coexistence of organizational constraints implemented to shape street-level behavior for other-focused emotion work actually have the opposite impact. The consequence, then, is an untenable position for the street level bureaucrat. Self-efficacy for the worker centers on a belief that s/he is successful at emotion work (Heuven & Bakker, 2006). The following hypotheses highlight that autonomy and discretion might have finite lower bounds for street-level bureaucrats, where emotion work is focused on self-preservation and survival as opposed to service delivery. This work would then also be less satisfying, more burnout inducing, and might result in the intention to quit the work entirely. This becomes a ‘disservice’ to both the client citizen and the street level bureaucrat.
H4: Bureaucratic constraints will be negatively related to self-efficacy in the performance of emotion management.


45
H5: Bureaucratic constraints will be positively related to expressions by street level bureaucrats
that they intend to quit their position.
Research Design
This research utilized a data collection process that was part of a larger study of workforce needs and interventions in the United States. Due to the magnitude of the larger effort, that project obtained all data from sites across the country under protocols approved by multiple institutional review boards, including those of project partner universities as well as the individual state child welfare agencies involved. Under these protocols, participants in the overall study completed a survey containing the items of interest for this dissertation alongside a culture and climate measure discussed later in this chapter. These instruments were completed in local agencies during work hours, with the agreement of supervisors and agency leadership. For those individuals who were unavailable on the day of administration, the project offered an online opportunity to participate.
For the purpose of this research, staff of the larger project recoded the data to de-identify individuals as well as their specific agency, team, and supervisor prior to sending the data to the author. Upon receipt of the data set, data were cleaned to exclude 8 units of non-child welfare workers who participated in the culture and climate portion of the research in smaller agencies. Then, culture and climate data were merged with survey data on elements of interest for this study. Thirty-eight cases were dropped when they failed to merge on their unique identifier. This failure was due to a flaw in survey collection, where unique identifiers of individuals who participated on-line were not saved in the Qualtrix application. Finally, because of the nested nature of the analysis in this dissertation, units with only one participating worker were dropped.


46
The participants illustrated in Table 3.1 reflect the participants whose responses are included in this study.
Participants
The sample of child welfare caseworkers in this study (n=345) were employed at one of 17 agencies participating in a large workforce development initiative in one eastern state. These caseworkers were all part of child serving, case carrying units within their agency. At the time of survey administration, the larger project goal was to achieve as close to the population of those workers within the jurisdictions as possible, with the intent to capture sufficient information related to organizational culture and climate. The 17 participating local agencies nest within a state supervised, locally administered child welfare system. The 17 agencies range in size from Table 3.1. Sample
Agency # of units Total Caseworkers Original estimate Inclusion and Response Rate
A 6 18 27 67%
B 1 6 6 100%
C 4 20 29 69%
D 27 106 139 76%
E 2 8 10 80%
F 4 21 30 70%
G 1 3 13 23%
H 1 7 7 100%
I 3 8 25 32%
J 9 37 53 70%
K 2 16 16 100%
L 7 41 51 80%
M 1 8 8 100%
N 2 13 13 100%
0 1 3 4 75%
P 2 17 17 100%
Q 2 13 13 100%
Total 75 345 461 75%


47
four to 139 child welfare workers, and serve a variety of populations, including rural, mid-sized, and urban. Table 3.1 illustrates participation by agency, which varied from 100% to 23% for an overall response rate of 75%.
Participants ranged in age from 22-70 years old, with a mean of 39 years. Eighty-seven caseworkers stated their undergraduate degree was in social work, followed by 85 who reported their degree was in Psychology. Twenty-eight caseworkers reported they had earned an advanced degree in social work, and 50 reported advanced degrees ranging from criminal justice to psychology. Similar to national trends in child welfare casework, 87% of the sample reported identifying as women. Half of the caseworkers reported they were Caucasian/white, followed by
Table 3.2. Demographic Characteristics of Caseworkers
Frequency Percent
Gender (Missing=3)
Female 298 87.1
Male 44 12.9
Race (Missing=0)
African American/Black 129 37.4
Caucasian/White 188 54.5
Asian 5 1.4
Native American 1 0.3
Biracial/Multiracial 20 5.8
Othelr 2 0.6
Ethnicity (Missing=0)
Hispanic/Latino Origin 29 8.4
Undergraduate Major (Missing=14)
Counseling 5 1.4
Criminal Justice 18 5.2
Human Services 20 5.8
Other 61 17.7
Psychology 90 26.1
Social Work 99 28.7
Sociology 38 11


48
35% who identified as African American/Black. The rest reported race or ethnicity as Hispanic, biracial/multiracial, Native American, or Asian. A full report of demographics may be found in Table 3.2.
Operationalization of Theoretical Constructs
The performance of emotional labor is operationalized via two constructs: the management of one’s own emotive state (self-focused) and the management of the citizen’s emotive state (other-focused). Indicators of self-focused emotional labor include masking of true emotions by the caseworker, which is often referred to as ‘false face.’ This is related to behaving according to display rules of the public agency (e.g. “act professional toward clients”).
Indicators of other-focused emotional management are work to calm or deescalate clients and work to help others emote in a desirable manner according to desired outcomes (Guy et al., 2008). Deep acting occurs when the laborer truly feels the emotion necessary for a certain situation (e.g. unconditional positive regard as in humanistic psychotherapy). These constructs are organized in Figure 3.1 and correspond with individual measures.
Concept
Emotional Labor
Constructs
Other-focused
Indicators
Emotion
management
â– 
Deep acting: Helping others emote, relationship buidling
Pretending expresssion ('false face')
â– 
Inauthentic
expression
Figure 3.1 Operationalization of Emotional Management
Bureaucratic constraints may be operationalized via three constructs: limited resources, limited autonomy, and limited discretion (see figure 3.2). Limitations to resources are illustrated by concern over liability in the work, which is an individually perceived constraint, as is the inability to do a complete job in the time allotted. Limitations to autonomy are expressed in the


49
climate measures of stress in the OSC, with attention to role conflict and role overload, specifically. Finally, limited discretion is operationalized as organizational rigidity as indicated by formalization and centralization of decision-making and authority.
Concept Bureaucratic Constraints
. 1
Constructs A“y Limited Discretion
1
Indicators Role conflict Role overload Formalization Concerns for liability
Figure 3.2. Operationalization of Bureaucratic Constraint
Burnout, low self-efficacy and expressed intent to quit may all be consequences of stress in the work environment, and have strong relationships to turnover in the child welfare literature (Mor Barak, 2001)These concepts, constructs, and indicators are illustrated in Figure 3.3.
Concept Turnover of SLBs
i ;
Constructs Employee Employeee
expression experience

Indicators Intent to quit Low self-efficacy
Figure 3.3. Operationalization of Turnover Intent Measures
The research design elicits understanding of the key constructs as manifested in the study sample. Given the primary research questions as related to the dependent variable of emotional labor performance, the survey included selected items from the Guy, Newman, Mastracci (GNM) Emotional Labor Questionnaire (Guy et al., 2008). Researchers have administered the


50
GNM internationally as a measure of emotional labor performance in public service. The survey included four constructs: Pretending Expression/False Face, Authentic Expression, Emotional Management, and Deep Acting. Constructs and known alpha coefficients were obtained from a larger pilot of these scales.
Pretending Expression/False Face captures self-focused emotional labor to maintain display rules. On a scale of one to seven, with one being ‘strongly disagree’ and seven being ‘strongly agree,’ respondents rate themselves on items “I hide my true feelings so as to appear pleasant at work,” “In my job I act confident and self-assured regardless of how I actually feel,” and “I wear a “mask” in order to deal with clients/customers in an appropriate way.” The original alpha coefficient for this measure is .746 (Guy et al., 2008). For this study, the alpha coefficient was .619.
Authentic Expression addresses the manner in which workers are able to emote without constraint in their daily work. On a scale of one to seven, with one being ‘strongly disagree’ and seven being ‘strongly agree,’ respondents rate themselves on items “I let my true feelings show when working with clients/customers,” “It is easier for me to show my true feelings than to pretend,” and “I am good at expressing how I feel.” The original alpha coefficient for this measure is .638 (Guy et al., 2008). For this study, the alpha coefficient was .647.
Emotion Management addresses other-focused emotion management, where workers express the importance of a skill for assisting others in emoting according to the professional requirements of their work. On a scale of one to seven, with one being ‘strongly disagree’ and seven being ‘strongly agree,’ respondents rate themselves on items “I am good at getting people to calm down,” “Dealing with emotionally charged issues is an important part of my job,” and


51
“In my job I am good at dealing with emotional issues.” The original alpha coefficient for this measure is .710 (Guy et al., 2008). For this study, the alpha coefficient was .814.
Deep Acting is measured by items that center around the effort necessary to sincerely feel the emotions required in work. On a scale of one to seven, with one being ‘strongly disagree’ and seven being ‘strongly agree,’ respondents rate themselves on items, “I try to actually experience the emotions that I must show to clients/customers,” “I work hard to actually feel the emotions that I need to show to clients/customers” and “I work at developing the feelings inside of me that I need to show to clients/customers.” There was not an existing alpha coefficient at the time of administration for this construct. For this study, the alpha coefficient was .842.
Self efficacy is related to performance of work tasks and people changing technology. The survey contained a scale derived from a state training system, and contained child welfare casework related successes. Caseworkers responded to five items on a Likert scale of one through five, with one representing Never and five representing Always. The items for this scale were “How often do you feel you are successful in helping families make positive changes in their lives?” “How often do you feel you are successful in helping children and families achieve permanency?” “How often do you feel you are successful in helping children and families remain safe?” “How often do you feel your efforts have contributed to the physical, mental, and educational well-being of children and families?” and “How often do you receive thanks or expressions of gratitude from children or families you have worked with?” Following item reduction, Cronbach’s alpha for this scale was .827.
Intent to quit was measured through five items items distributed throughout the survey instrument. Each were scaled on 1 to 7 scales that ranged in choices related to the specific question (e.g. not at all likely to very likely and terrible (chances) to excellent (chances)) and


52
included “I will look for a new job OUTSIDE the agency in the next year,” and “I will probably look for a new job in the next year.” Looking for a new job is related to actual turnover, and is often used as a proxy in the absence of longitudinal data (Auerbach, Schudrich, Lawrence, Claiborne, & McGowan, 2014). After item reduction, Chronbach’s alpha for all the items was .948.
Key organizational elements were measured through the Organizational Social Context (OSC) Questionnaire (Glisson et al., 2012). The OSC is a nationally normed instrument in both child welfare and children’s mental health. The OSC captures a variety of organizational culture and climate constructs, but this research is concerned with those that exemplify bureaucratic constraint. Rigidity and role stress were of primary concern.
Rigidity within the unit, as measured by the subscales offormalization and centralization (seven items each, measured on a Likert scale ranging from strongly agree to strongly disagree), examines the culture where caseworkers conduct their work. Formalization measures the need for procedural compliance in everyday work, where creativity is restricted in favor of a step-by-step process. Items on formalization include, “The same steps must be followed in every piece of work.” Centralization items measure deference in decision making to higher level management within the agency. Items include, “I have to ask a supervisor or coordinator before I do almost anything.” Alpha coefficients for these scales are .61 for centralization and .69 for formalization (Glisson et al., 2012).
Stress within the unit, as measured by the subscales of role conflict and role overload, captures the degree to which caseworkers understand their positions and roles and can operate freely within them. The OSC measures stress as the overall climate within a unit, as opposed to traditional measures of stress at the individual level. The seven items on role conflict, or being


53
torn between two disparate functions of the job, include, “Interests of clients are often replaced by bureaucratic concerns, e.g. paperwork.” Role overload deals with time and resource constraints, where the seven items include, “The amount of work I have to do keeps me from doing a good job.” Both stress and rigidity are aggregated at the unit level and normed according to prior administration in 81 child welfare agencies in the United States. This produces a unit-level T score for each overall concept and construct. The alpha scores for these items are .83 for role conflict and .82 for role overload (Glisson et al., 2012).
On an individual level, the survey assessed caseworker factors related to unique aspects of child welfare work using portions of a general staff survey designed for caseworkers (Dettlaff et al., 2015). Specifically, scales related to caseworker concerns for liability, time constraints related to workload, and an index of decision-making thresholds were included in the survey.
Concerns for liability are caseworker perceptions that he or she may be disciplined or fired if a child on one of their cases is harmed over the course of or after the agency intervention. It also examines worry over media attention and attention to others that have been disciplined or fired in the course of their work. This real possibility is a constant backdrop to the high stakes and high accountability realm of child protection work. However, the scale also includes items that speak to the caseworker’s perceived level of support from their direct supervisor or overall agency leadership if the worst were to occur, including the likelihood that the agency would conduct a thorough investigation of the situation prior to assigning blame to the caseworker. After conducting item reduction, the items on this scale exhibited a Cronbach’s alpha of .828.
The variables developed for this analysis are displayed in Table 3.3, with their corresponding alpha coefficients.


54
Table 3.3. Variables Used in the Analyses1
False Face
I hide my true feelings so as to appear pleasant at work In my job I act confident and self-assured regardless of how I actually feel I wear a "mask" in order to deal with families and children in an appropriate way
Authentic Expression
I let my true feelings show when working with families and children It is easier for me to show my true feelings than to pretend I am good at expressing how I feel.
Emotion Management
I am good at getting people to calm down
Dealing with emotionally charged issues is an important part of my job In my job, I am good at dealing with Emotional Issues*
Deep Acting
I try to actually experience the emotions that I must show to families and children
I work hard to actually feel the emotions that I need to show to families and children
I work at developing the feelings in side of me that I need to show families and children
Concerns for Liability
I believe I have had adequate training to help me make the right decision about the safety and well-being of my clients*
I am worried that one of my cases may draw media attention.*
I have known caseworkers that have been disciplined or fired because of real or perceived mistakes.*
If a child in one of my cases is harmed, I believe the agency will conduct a thorough investigation into what happened before assigning blame I know my supervisor will be supportive of me and the decisions I made if a child is harmed in one of my cases
I know my leadership will be supportive of me and the decision I made if a child is harmed in one of my cases
Success/Self Efficacy
How often do you feel you are successful in helping families make positive changes in their lives
Cronbach's
Alpha
0.619
0.647
0.814
0.842
0.828
0.832


55
Table 3.3 cont’d
How often do you feel you are successful in helping children and families achieve permanency
How often do you feel you are successful in helping children and families remain safe
How often do you feel your efforts have contributed to the physical, mental, and educational well-being of children and families How often do you receive thanks or expressions of gratitude from children and families you have worked with?*
Intent to Quit
I will probably look for a new job in the next year* N/A
How often do you think about quitting your job?*
What are the chances that you will leave your current job for a job OUTSIDE the agency sometime in the next six months?*
What are the chances that you will leave your current job for a job OUTSIDE
the agency sometime in the next year?_________________________________________________
*Asterisks designate items excluded from analysis following dimension reduction.
'The OSC items are proprietary, but a full description including reliability and validity analysis of factors is available in Glisson et al., 2012.
Factor Analysis
In SPSS, construct validity was established using Principal Components Analysis (PCA) with Varimax rotation with Kaiser normalization. This approach was appropriate for this data set given the ordinal nature of the items, the size of the sample, and established suitability for data reduction of all items. Each set of items was non-significant using Bartlett’s test of sphericity. There were also no significant outliers in this data set.
Emotional labor factors loaded on originally intended constructs, similar to past analysis using these items. The exception to this was “I am good at expressing how I feel,” which loaded on self-efficacy at .558, which was lower than all other items in that factor (.885, .752, and .669,


56
respectively). This item loaded lightly on authentic expression, at .286. Consequently, this item was dropped from analysis.
Concerns for liability loaded into two separate factors: worry and support. For the worry category, Cronbach’s alpha was quite low at .484. Conversely, the three items loading on agency support for worker discretion and fairness of determining accountability in light of a child being harmed exhibited a Cronbach’s alpha of .828. Consequently, all items loading on the ‘worry’ factor were dropped from analysis.
Items in the success/self-efficacy scale did not divide into factors. However, the item “How often do you receive thanks or expressions of gratitude from children and families you have worked with?” had a considerably lower value after rotation than the other items in the scale. Consequently, this item was dropped from the analysis.
Finally, the items for Intent to Quit were analyzed. The questions were quite similar and all items loaded on one factor. After rotation, items ranged from .684 and .914. After dropping the two lowest items, the remaining items exhibited a Cronbach’s alpha of .948. Given the high correlation of all the items, the most predictive of variance was selected, “What are the chances that you will leave your current job for a job OUTSIDE the agency sometime in the next year?” This variable was somewhat skewed, with most respondents stating they did not intend to quit their position or that the chances were “so-so.” This variable was transformed into a binomial distribution, with 1-4 being 0 and 5-7 representing at least some desire to quit ones job in the subsequent six months.
Analytic Technique
Each of the five hypotheses were examined using either Hierarchical Linear Modeling (HLM) or Ordinary Least Squares (OLS) regression. Because survey participants were clustered


57
in units, and the variables of rigidity, role conflict, and role overload were measured at the unit level, data are clustered in such a manner that regression intercepts may also be clustered. This clustering violates assumptions of linear regression and can lead to erroneous conclusions should OLS be utilized in those instances (Bryk & Raudenbush, 1992). Consequently, each dependent variable was first tested using a null hypothesis to see if its interclass correlation coefficient (ICC), or variance within units, was sufficient to warrant the use of HLM. If the data were sufficiently clustered, hypotheses were tested using a homogeneous and heterogeneous full random coefficients model, where both level one and level two contain independent variables [e.g. rigidity at the unit (level 2) and liability (level 1) as potential independent variables related to false face acting]. For those dependent variables without sufficient variance by unit, OLS was used to test each hypothesis. To address these hypotheses, thirteen variables were developed using data from an organizational culture and climate survey and a survey containing the variables of interest for this inquiry. See Figure 3.4 for an illustration of the analysis plan for these data, with level one and level two independent variables (bureaucratic constraints) and the dependent variables in each hypothesis.
Figure 3.4 Variables for Analysis
Variable Name Description Uses by Analysis
Unit This is the de-identified unit number for each unit. Grouping variable for HLM
Social Work Major Indicates whether or not the participant has an undergraduate major in social work. Control variable
Gender Indicates whether participant identifies as a woman. Control variable
Support Identifies the level to which participant anticipates they might receive support from their agency if a child were harmed on their caseload. Independent variable in all hypotheses


58
Table 3.4 cont’d
Rigidity (centralization & formalization) A t-score reflecting the unit’s perception of centralization and formalization in their agency. Independent variable in all hypotheses
Role Overload (Stress subdomain) A t-score reflecting the unit’s perception of time constraints and overload in the work. Independent variable in all hypotheses
Role Conflict (Stress subdomain) A t-score reflecting the unit’s concern with incongruent pressures in their work. Independent variable in all hypotheses
Authentic Expression Mean degree to which participant is authentic with emotions at work. Dependent variable in HI.
Deep Acting Mean degree to which participant engages in efforts to feel emotions required by job. Dependent variable in H2.
False Face Mean degree to which participant uses false face acting Dependent variable in H3.
Emotion Management Participants mean of perception that they are efficacious at other-focused emotional management. Dependent variable in H4.
Success Mean degree to which participant perceived they are successful in achieving people changing policy goals in child welfare. Dependent variable in H4.
Intent to Quit Binomial variable that indicates whether participant gives any indication of seeking new employment within the subsequent six months. Dependent variable in H5.


59
IV. ANALYSIS AND FINDINGS
This chapter reports results of hypothesis testing for the five hypotheses in this study. The five hypotheses are designed to address the overarching research question: How is the context of bureaucratic constraint in organizational culture related to the performance of emotional labor by street-level bureaucrats? Addressing this research question was accomplished by examining three sub questions: (1) How is the performance of other-focused emotion management limited or expanded in situations of high bureaucratic constraint? (2) Similarly, is self-focused emotion management simply survival for the street-level bureaucrat and a disguise for self-preservation? and, (3) How can bureaucratic constraints contribute to the creation of job stress for street-level bureaucrats, thus leading to emotional exhaustion, burnout and turnover? Table 4.01 reports each hypothesis alongside its corresponding research question.
Table 4.01. Research question and corresponding hypotheses
Primary Research Question: How is the context of bureaucratic constraint in organizational culture related to the performance of emotional labor by street-level bureaucrats?
Sub research question: How is the performance of other-focused emotion management limited or expanded in situations of high bureaucratic constraint? HI: Bureaucratic constraints will be negatively related to authentic expression of emotion. H2: Bureaucratic constraints will be negatively related to deep acting.
Sub research question: Is self-focused emotion management simply survival for the street-level bureaucrat and a disguise for self-preservation? H3: Bureaucratic constraints will be positively related to ‘false face’ strategies for self-focused emotion management.
Sub research question: How can bureaucratic constraints contribute to the creation of job stress for street-level bureaucrats, thus leading to high levels of emotional exhaustion, burnout and costly turnover? H4: Bureaucratic constraints will be negatively related to self-efficacy in the performance of emotion management. H5: Bureaucratic constraints will be positively related to expression by the street level bureaucrat that they intend to quit their position.


60
Missing Data
Prior to running additional analysis, a missing value analysis in SPSS was conducted to look for missing data and to understand whether to proceed as if the missing data occurred randomly in this sample. Forty-five cases were missing data on key items. Little’s Missing Cases at Random (MCAR) test indicated a non-significant finding (p=0.861), allowing the analysis to proceed under the assumption that these data might have been randomly omitted. The remaining analyses were conducting using list wise deletion of cases with missing values. Table 4.02 presents missing data in this analysis. List wise deletion resulted in a reduction in units from 75
to 73.
Table 4.02. Missing Data Pattern Analysis
Variable False Face Emotion Management Support SW Major Intent to Quit
Missing 7
10
5
6
5
*Patterns with less than 1% cases (3 or fewer) are not shown. Results
The variables related to the performance of emotional labor were authentic expression, deep acting, false face, and emotion management. Univariate statistics are displayed in Table 4.03. Of note is the low variability on the emotion management factor. No one in the sample answered “Strongly disagree” or “disagree” to the items “I am good at getting people to calm down,” “Dealing with emotionally charged issues is an important part of my job,” or “In my job I am good at dealing with emotional issues.” The mean is relatively high at 5.62 and the standard deviation is small, at .866. The skewed nature of this dependent variable and lack of variability makes it unsuitable for hypothesis testing using parametric tests. Because there are two variables


61
that address self-efficacy, the success variable, or the extent to which participants indicated they experience success in their jobs, was used instead. The success at efficacy for people changing technologies variable was also skewed, indicating that all respondents averaged at least a little bit of success in people changing with regard to child safety, permanency, and well-being. The skew was less than emotion management scores however, so the variable was retained in the analysis. Table 4.03. Univariate statistics, Dependent Variables
Min Max Mean Std. Deviation Skewness Kurtosis
False Face TOO 7.00 3.8600 1.22301 .110 -.309
Authentic Expression 1.00 7.00 4.0633 1.26873 .046 -.542
Emotion Management 3.00 7.00 5.6200 .86643 -.577 .078
Deep Acting 1.00 7.00 3.9422 1.17938 -.099 .066
Success 2.00 5.00 3.7533 .56846 .035 -.080
The final dependent variable in the analysis was related to the intention of respondents to quit their position, a proxy for turnover behavior in the workforce. After transforming to a binomial distribution, the data indicate that 37% percent or 111 caseworkers in the sample expressed a ‘so-so’ or greater chance of quitting their job for one outside the agency within one year.
Four variables served to operationalize bureaucratic constraints: concern for liability, rigidity (including centralization and formalization), role conflict, and role overload. The latter three are t-scores as related to normed data from 81 child welfare agencies nationally. As such, the t-scores center near the 50% percentile, with the units in this analysis trending slightly higher on rigidity and slightly lower on role overload than the norm. Concern for liability to the caseworker in the event of harm to a child ranged from a lot (1) to very little (7) (these items were reverse coded), with the mean (3.31) slightly favoring confidence in terms of support by the


62
agency, leadership, and supervisors. Univariate statistics for these independent variables may be found in Table 4.04.
Table 4.04. Univariate statistics for independent variables
Min Max Mean Std. Deviation Skewness Kurtosis
Rigidity 32.61 84.81 57.6163 1.22301 11.49965 .399
Role Conflict 29.94 72.36 50.7639 1.26873 10.21078 -.134
Role Overload 13.04 79.56 43.9378 .86643 13.33511 -.426
Liability 1.00 7.00 3.3178 1.17938 1.30279 .435
All variables were placed in a Pearson’s Correlation matrix to explore bivariate relationships. This matrix is reported in table 4.05. The coefficients were all lower than r = .329, with the exception of role conflict and role overload, which was significant (r = .510, p < .01). This is not surprising given that they exist in the same domain (stress) of the culture/climate instrument, though this does flag some concern related to discriminant validity for these two constructs. The choice was made, however, to exclude exhaustion under the stress domain given that it more closely reflects issues related to burnout than bureaucratic constraint. The hope is that teasing out the unique contributions of each construct might occur through regression analysis. Using a two-tailed test of significance, social work majors and women (control variables) were negatively correlated to Deep Acting (r = -. 142,p < .05 and r = -.\AA,p< .05, respectively). Women were also correlated negatively to authentic expression {r = -.\\5,p< .05). Success at people changing behavior was negatively correlated with false face (r = -.173, p < .01), but positively correlated with authentic expression (r = .175, p < .01), emotion management (r = .282, p < .01). Conversely, concern for liability was positively related to false face acting {r = .\\A,p< .05), but negatively related to authentic expression (r = -.237,p < .01),


Table 4.05. Correlation Matrix
1. SW 2. Gen 3. FF 4. Auth 5. EmoM 6. DeepA 7. Quit 8. Succ 9. Liab 10. Rigid 11. Rcon 12. Rover
1. Social Work Major 1.00
2. Gender 0.09 1.00
3. False Face -0.06 -0.06 1.00
4. Authentic Expression 0.04 -.115* -.236** 1.00
5. Emotion Management 0.02 -0.04 0.02 .222** 1.00
6. Deep Acting -.142* -.144* 0.09 .132* .212** 1.00
7. Intent to Quit -0.01 0.02 .252** -0.08 -0.03 0.00 1.00
8. Success 0.05 0.09 _ 173** .175** .282** 0.09 -0.09 1.00
9. Liability -0.02 0.10 .114* -.237** -.235** - .158** .329** -.123* 1.00
10. Rigidity -0.04 0.00 0.05 -0.04 0.03 -0.04 .150** 0.05 .174** 1.00
11. Role Conflict 0.01 0.06 .207** -0.06 0.04 0.06 .239** -0.10 .213** .157** 1.00
12. Role Overload -0.01 -0.02 .186** -0.03 0.05 0.10 0.08 -.148* -0.05 -.134* .510** 1.00
*. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed).


emotion management (r = -.235,p< .01), and deep acting (r = -.158,/) < .05). Intent to quit was positively related to liability (r = .329, p < .01), rigidity (r = .150,p < .01), and role conflict (r = .239, p < .01).
Hypothesis Testing
Hypothesis One. The first hypothesis tested was HI: Bureaucratic constraints will be negatively related to authentic expression of emotion. The dependent variable for this test is Authentic Expression. Because the variable is the mean of a combination of scores, it is continuous and close to normally distributed. Using HLM Version 7.03, a null model was utilized to test whether there is a unit (level 2) effect on the dependent variable (level 1) intercept. This test guides the suitability of HLM as a regression model for this hypothesis. The test employed is Full Information Maximum Likelihood (FIML) estimators given that unit sizes vary considerably in this sample. The estimation of variance components in the model was not significant (p = .254) meaning that after other variables in the null model were controlled, there was no residual between-groups variance from the random effects of the Unit on the dependent variable Authentic Expression (level 1). HLM was not the appropriate estimation method for this hypothesis. Consequently, OLS regression was conducted using IBM SPSS Statistics, version 25. The models, including controls, are illustrated in Table 4.06.
The first model, including just the control variables of social work major and gender, indicated a statistically significant relationship for gender (p < .05), wherein those individuals identifying as women were less likely to exhibit authentic expression. Adding liability in Model 2 reduced that relationship and it was no longer significant. However, liability, or the indication of low confidence in agency, leadership, and supervisor to fairly assess blame for child harm was significant atp = < .001. Finally, model three, including all the variables for bureaucratic


65
constraint, also exhibited significance for liability (p = < .01). The models did not identify significant relationships with the other independent variables. Hypothesis one was partially accepted.
Table 4.06. Regression Results of Bureaucratic Constraints on Authentic Expression
Model 1 Model 2 Model 3
Social Work Major 0.051 0.780 0.042
Gender -.120* -.097 -0.099
Liability -.227*** -0.234**
Rigidity -0.016
Role Conflict 0.033
Role Overload -0.065
R-Squared 0.016 0.067 0.069
Adjusted R-Squared 0.009 0.057 0.050
F 2.398 7 044*** 3.646**
a. Dependent Variable: Authentic Expression
b. Standardized coefficients are presented in the table
c. *Significant at p < .05 **Significant at p < .01 ***Significant at p < .001
Hypothesis Two: Next, testing was conducted for H2: Bureaucratic constraints will be negatively related to deep acting. The dependent variable for this test is Deep Acting. Because this variable is the mean of a combination of scores, it is continuous, and close to normally distributed. Using HLM Version 7.03, a null model was utilized to test whether there is a unit (level 2) effect on the dependent variable (level 1) intercept. This test guides the suitability of HLM as a regression model for this hypothesis. The test employed was Full Information Maximum Likelihood (FIML) estimators given that unit sizes vary considerably in this sample. The estimation of variance components in the model were not significant (p = .473) meaning that after other variables in the null model are controlled, there is no residual between-groups variance from the random effects of the Unit on the dependent variable Deep Acting (level 1). HLM is not the appropriate estimation method for this hypothesis. Consequently, OLS regression


66
was conducted using IBM SPSS Statistics, version 25. These models, including controls, are illustrated in Table 4.07.
The first model, which included just controls of social work major and women, exhibited a significant relationship to both variables, with both social work majors and women negatively related to the performance of deep acting (p = < 05). This continued throughout the other two models. Model two, which including just liability in addition to the controls, indicated a significant negative relationship (p = <001) to the performance of deep acting. Finally, model three, including all the variables for bureaucratic constraint, also exhibited significance for liability (p = < .01). The models did not identify significant relationships with the other independent variables. Hypothesis two was partially accepted.
Table 4.07. Bureaucratic Constraints on Deep Acting
Model 1 Model 2 Model 3
Social Work Major -.129* -.134* -.135*
Gender -.131* -.116* -.119*
Liability _ 249** .16i**
Rigidity -.023
Role Conflict .088
Role Overload .041
R-Squared .037 .059 .072
Adjusted R-Squared .031 .049 .053
F 5.735** 6.088*** 3.846***
a. Dependent Variable: Deep Acting
b. Standardized coefficients are presented in the table
c. *Significant at p < .05 **Significant at p < .01 ***Significant at p < .001
Hypothesis Three. Next, H3: Bureaucratic constraints will be positively related to ‘false face’ strategies for self-focused emotion management, was tested. The dependent variable for this test was False Face. Because this variable is the mean of a combination of scores, it is continuous, and close to normally distributed. Using HLM Version 7.03, a null model was


67
utilized to test whether there is a unit (level 2) effect on the dependent variable (level 1) intercept. This test guides the suitability of HLM as a regression model for this hypothesis. The test employed was Full Information Maximum Likelihood (FIML) estimators given that unit sizes vary considerably in this sample. In the case of this null hypothesis, the final estimation of variance components was mildly significant at/? = .022, which is less thanp = .05. Consequently, HLM was employed to estimate coefficients for this hypothesis. Results of this testing are in Table 4.08.
Table 4.08. HLM Analysis of False Face
Model Variable Coefficient SE t-ratio df p-Value
Random effects only
Constant 3.849** 0.08 47.822 73 <.001
Unit Variance .123
Residual Variance 1.36
Chi-square 99.231* 73 .022
ICC .082
Unit Level and Caseworker-level fixed effects
Unit Constant 2.39 .471 5.079 70 <.001
Rigidity .003 .007 .422 70 .674
Role Conflict .015 .008 1.902 70 .061
Role Overload .006* .006 2.067 70 .042
Social Work
Caseworker Major -.013 .157 -.82 223 .413
Gender -.259 .189 -1.373 223 .171
Liability .091 .054 1.679 223 .094
Unit Variance .05
Residual Variance 1.34
Chi-Square 84.309 70 .117
*p<05 **p<001
The full model including both unit and caseworker independent variables and controls indicated a statistically significant relationship with role overload on the performance of false face or surface acting. Role overload captures the degree to which participants indicated their


68
strain in doing all the parts of the job demanded of them. None of the other independent variables were significant. The model comparison test was significant (p < .01) with a Chi-Square of 18.57. Hypothesis three was partially accepted.
Hypothesis Four. Next to be tested was H4: Bureaucratic constraints will be negatively related to self-efficacy in the performance of emotion management. The dependent variable for this hypothesis is Success. Because this variable is the mean of a combination of scores, it is continuous, and close to normally distributed. Using HLM Version 7.03, a null model was utilized to test whether there is a unit (level 2) effect on the dependent variable (level 1) intercept. This test guides the suitability of HLM as a regression model for this hypothesis. The test employed was Full Information Maximum Likelihood (FIML) estimators given that unit sizes vary considerably in this sample. The estimation of variance components in the model were not significant (p = >.500) meaning that after other variables in the null model are controlled, there is no residual between-groups variance from the random effects of the Unit on the dependent variable Success (level 1). HLM was not the appropriate estimation method for this hypothesis. Consequently, OLS regression was conducted using IBM SPSS Statistics, version 25. These models, including controls, are illustrated in Table 4.09.
Model one, which included just the control variables, did not exhibit statistical significance. Model two, which included just liability in addition to the controls, indicated a significant negative relationship (p = < 05) to self-efficacy in people changing technologies. Finally, model three, including all the variables for bureaucratic constraint, also exhibited significance for liability (p = < .05). The models did not identify significant relationships with the other independent variables. Hypothesis four was partially accepted.


69
Table 4.09. Bureaucratic constraints on self-efficacy
Model 1 Model 2 Model 3
Social Work Major -.039 .036 .036
Gender -.087 .101 0.1
Liability -.132* -.147*
Rigidity .064
Role Conflict -.014
Role Overload -.136
R-Squared .010 .027 .054
Adjusted R-Squared .003 .017 .034
F 1.461 2.749* 2.770*
a. Dependent Variable: Self Efficacy
b. Standardized coefficients are presented in the table
c. *Significant at p < .05 **Significant at p < .01 ***Significant at p < .001
Hypothesis Five. Finally, testing was conducted of H5: Bureaucratic constraints will be positively related to expression by the street level bureaucrat that they intend to quit their position. The dependent variable for this hypothesis is Intention to Quit. Because this variable is has 0/1 Bernoulli distribution, estimation will be conducted by using a logit link function to transform the variable. Using HLM Version 7.03, a null model was utilized to test whether there is a unit (level 2) effect on the dependent variable (level 1) intercept. This test guides the suitability of HLM as a regression model for this hypothesis. The test employed is Full Information Maximum Likelihood (FIML) estimators given that unit sizes vary considerably in this sample. In the case of this null hypothesis, the final estimation of variance components is significant at/? = .006, which is less thanp = .01. Consequently, HLM was employed to estimate coefficients for this hypothesis. The final estimation of fixed effects, with a population average model with robust standard errors is illustrated in Table 4.11.
The full model including both unit and caseworker independent variables and controls indicated a statistically significant relationship with concerns for liability positively impacting a


70
caseworker’s intent to quit their job. None of the other independent variables were significant. Hypothesis five was partially accepted.
Table 4.11. HLM Analysis of Intent to Quit
Model Variable Coefficient SE t-ratio df p-Value
Random effects only
Constant - 517*** .148 -3.490 73 <.001
Unit Variance .468* .255 73 .006
Chi-square 107.147
Unit Uevel and Caseworker-level fixed effects
Unit Constant -5 J44*** 1.18 -4.35 70 <.001
Rigidity .016 .012 1.329 70 .188
Role Conflict .035 .018 1.938 70 .057
Role Overload .005 .013 .447 70 .656
Social Work
Caseworker Major .069 .311 -.223 223 .824
Gender -.102 .446 -.229 223 .819
Uiability 502*** .110 4.559 223 <.001
Unit Variance .170
Chi-Square 87.823
*p<05 ***p<001
Results of Hypotheses Testing
All five hypotheses were partially supported by this analysis, with at least one independent variable significantly modeling an influence on each dependent variable in each hypothesis. The results of this hypotheses testing are illustrated in 4.12. The implications of these results and their relation to the research questions of this study are contained in Chapter Five.


71
Table 4.12 Summary of Hypothesis Testing
Hypothesis Results
HI Bureaucratic constraints will be negatively related to authentic expression of emotion. Partially supported
H2 Bureaucratic constraints will be negatively related to deep acting. Partially supported
H3 Bureaucratic constraints will be positively related to ‘false face’ strategies for self-focused emotion management. Partially supported
H4 Bureaucratic constraints will be negatively related to self-efficacy in the performance of emotion management. Partially supported
H5 Bureaucratic constraints will be positively related to expression by the street level bureaucrat that they intend to quit their position. Partially supported


72
V. DISCUSSION
This chapter summarizes the results section in relation to the overarching research question, the sub research questions, and the hypotheses proposed in Chapter 3. Each dependent variable for the five hypotheses (authentic expression, deep acting, false face, self-efficacy, and intent to quit) are reviewed alongside the extant literature to compare and contrast the results of this study. Where these results deviate from theoretical positions in the literature, this chapter reviews differences and develops meaning as applicable. Where practical implications exist for public administration, each will be delineated as takeaways or areas of attention in the realm of practice. Limitations of the study are considered, including threats to internal and external validity, study design, sampling, and other factors that impact the generalizability of these results. Finally, the chapter concludes with suggestions for further study and outlines a research agenda related to this inquiry, with attention to establishing the importance of this line of research to both practice and theory.
Summary of Findings
Table 5.1 contains a summary of the overarching research question, the sub research question, and hypotheses, alongside significant findings. For the first sub research question, which queried how the performance of other-focused emotion management limited or expanded in situations of high bureaucratic constraint, two hypotheses were tested. The first, that bureaucratic constraints are negatively related to authentic expression of emotion, was partially accepted. In this sample, increases in worry by caseworkers that they will not be supported by their agency in the event of a child being harmed on their caseload (liability) are negatively related to authentic expression of emotion. The second hypothesis, that bureaucratic constraints are negatively related to deep acting was also partially supported. Having majored in social


73
work, identifying as a woman, and increases in worry by caseworkers that they will not be supported by their agency in the event of a child being harmed on their caseload (liability), are negatively related to the performance and effort related to deep acting.
The second research question of how self-focused emotion management might be simply survival for the street-level bureaucrat and a disguise for self-preservation, was answered using one hypothesis. This hypothesis, that bureaucratic constraints are positively related to ‘false face’ strategies for self-focused emotion management was partially supported by the analysis.
Increases in expressions of role overload such as feeling constrained by time and resources are positively related to false face strategies for self-focused emotion management.
The third research question queried as to how bureaucratic constraints contribute to the creation of job stress for street-level bureaucrats, thus leading to high levels of emotional exhaustion, burnout and costly turnover? This research question was addressed by two hypotheses. First, it was hypothesized that bureaucratic constraints are negatively related to self-efficacy in the performance of emotion management. This hypothesis was partially supported. Analysis indicated that increases in worry by caseworkers that they will not be supported by their agency in the event of a child being harmed on their caseload (liability) are negatively related to their experience of self-efficacy, or ability to successfully assist families and children in achieving necessary changes to promote safety, permanency, and well-being. Finally, it was hypothesized that bureaucratic constraints will be positively related to expression by the street level bureaucrat that they intend to quit their position. This hypothesis was partially supported. Increases in worry by caseworkers that they will not be supported by their agency in the event of a child being harmed on their caseload (liability) increased the odds that they express an intention to seek work outside their agency in the subsequent year.


Table 5.1. Summary of Findings
Primary Research Question: How is the context of bureaucratic constraint in organizational culture related to the performance of emotional labor by street-level bureaucrats?_______________________________________________________________________________________
Sub Research Question Hypothesis Findings
How is the performance of other-focused emotion management limited or expanded in situations of high bureaucratic constraint? HI Bureaucratic constraints will be negatively related to authentic expression of emotion. Increases in worry by caseworkers that they will not be supported by their agency in the event of a child being harmed on their caseload (liability) are negatively related to authentic expression of emotion.
H2 Bureaucratic constraints will be negatively related to deep acting. Having majored in social work, identifying as a woman, and increases in worry by caseworkers that they will not be supported by their agency in the event of a child being harmed on their caseload (liability) are negatively related to deep acting.
How is self-focused emotion management simply survival for the street-level bureaucrat and a disguise for self-preservation? H3 Bureaucratic constraints will be positively related to ‘false face’ strategies for self-focused emotion management. Increases in expressions of role overload such as feeling constrained by time and resources are positively related to false face strategies for self-focused emotion management.
How can bureaucratic constraints contribute to the creation of job stress for street-level bureaucrats, thus leading to high levels of emotional exhaustion, burnout and costly turnover? H4 Bureaucratic constraints will be negatively related to self-efficacy in the performance of emotion management. Increases in worry by caseworkers that they will be not be supported by their agency in the event of a child being harmed on their caseload (liability) are negatively related to their experience of self-efficacy, or ability to successfully assist families and children in achieving necessary changes to promote safety, permanency, and well-being.
H5 Bureaucratic constraints will be positively related to expression by the street level bureaucrat that they intend to quit their position. Increases in worry by caseworkers that they will not be supported by their agency in the event of a child being harmed on their caseload (liability) increase the odds that they will express an intention to seek work outside their agency in the subsequent year.


Authentic Expression
Authenticity of emotional display is empirically linked to worker well-being, including depression, depressed mood, and secondary traumatic stress (Caringi et al., 2012; Erickson & Wharton, 1997; Grandey et al., 2012). A worker’s concern over a lack of support by her agency, leadership, or supervisor in the event of harm to a child under her purview is related to decreases in feelings of authenticity in the current study. This is consistent with multiple studies that consistently report that limits on autonomy, or control over one’s work, are related to inauthenticity and subsequent mental health outcomes as expressed by workers in the service and manufacturing industries (Erickson & Wharton, 1997; Hochschild, 2003). While the operationalization of autonomy was more nuanced in this study, to account for the unique facets of the bureaucracy within which child welfare work is performed, the implications are similar. It is logical to assume that if one expresses considerable worry about outcomes that are for the most part largely not under her control, this can be construed as decreased autonomy of the worker. As discussed in the literature review, significant harm or death of a child on one’s caseload following agency intervention is nationally a relatively rare event as compared to the daily work tasks of child welfare caseworkers. Similarly, direct connections of worker behavior to child harm often oversimplify very complex family, community, and social dynamics. Nonetheless, worry that her agency will not provide a fair and balanced investigation of the facts surrounding such cases may also be considered a constraint on worker discretion, as worker and supervisor decision-making throughout the case is the focus of these types of internal investigations. While accountability under such circumstances is certainly justified and is a key component of the responsibilities assumed by child welfare workers, the relation of these constraints to authenticity of emotional displays is an important finding.


76
It is interesting in this case to consider the study participants’ overall assertion that they are good at emotional management and efficacious in its delivery. No one in the sample indicated disagreement that emotional management is a key component of their job in child welfare casework, or that they felt competent at calming others and maintaining agency display rules. The sample also demonstrated prior affinity for interacting with and dealing with people through their careers by obtaining bachelor’s degrees in either social work, psychology, or other related fields such as human services, criminal justice, or sociology. Even so, in this study, feelings of authenticity diminished alongside constraints imposed by caseworker’s organizations and the everyday realities of their work.
Deep Acting
Deep acting is the core of other-focused emotion work, and as I argue in this dissertation, it is integral to people changing technology, or family engagement, as in the child welfare literature. Deep acting is other-focused, and facilitates the change process by allowing relationships to develop (Hasenfeld, 1983; Pugliesi, 1999). Change occurs, then, through the citizen’s relationship to the street level bureaucrat. Thus, deep acting as relationship building can be considered instrumental to child protection work that seeks to change parent and caregiver behaviors in the interest of child safety, permanency, and well-being. Client-oriented firms in business, for example, recognize this dynamic and elevate the position of frontline workers to leverage their intricate knowledge of the client experience and their relationships (Lively, 2002).
Given the primacy of deep acting to the work of child welfare, it is concerning to observe the impact found in this study by bureaucratic constraint on performance of deep acting by caseworkers. Again, the elevation of concern that one will not receive support from one’s agency, leadership, or supervisor in the event of child harm was negatively related to the


77
caseworker’s performance of deep acting with clients. Similar to authenticity, deep acting is related to outcomes for the workforce, including job satisfaction, mediation of burnout, and wellbeing (Brotheridge & Grandey, 2002; Grandey, 2000). Also of concern, however, are possible impacts on outcomes for children, youth and families who receive services from the child welfare system through the conduit of the caseworker. In customer service situations, customers differentially assess their satisfaction with the interaction according to whether the service worker was surface (false face) or deep acting. This was compounded when the service recipient accurately detected which strategy (false face or deep acting) was being utilized, which occurred with some frequency (Groth, Hennig-Thurau, & Walsh, 2009). Similarly, concerns related to emotional contagion may point to impacts on service recipients in emotionally intense situations where intervening caseworkers do not work to feel the emotions they are to express in carrying out child welfare responsibilities and working toward facilitating change in detrimental behaviors, though these have not been fully investigated in child welfare work (Hatfield, Cacioppo, & Rapson, 1992). In this case, we again see worry for accountability impinging on the emotion work of the job.
With regard to deep acting, two surprising factors emerged as influential to caseworker performance of deep acting. Having an undergraduate major in social work and identifying as a woman both exhibited negative impacts on performance of emotional labor. This is interesting, because some studies utilize the proportion of women employed in an organization as a proxy for the degree of emotional labor involved (Meier et al., 2006). Similar, given the emphasis in undergraduate social work curricula on the employment of people changing technologies and the importance of relationship, it is puzzling that in the context of the agencies in this sample, both social workers and women employ less deep acting in their work. The setting, of course, exhibits


78
a large proportion of women (83% in this sample) with a high number of colleagues who have entered the workforce with the expectation that people changing technologies may be employed, based on their undergraduate education. While further analysis is warranted and necessary regarding this finding, one may speculate that those with strong proclivities toward emotion work might exhibit repressive behaviors in favor of conserving valuable emotive resources, as we might find in studies based on conservation of resources (COR) theory (Gelderen, 2016; Grandey et al., 2012).
False Face
False face, or surface acting, is the operationalization of self-focused emotion work. False face strategies conserve emotional resources allowing the street level bureaucrat to “grin and bear it” through daily work. One item that characterizes false face is particularly poignant, where caseworkers assessed their tendency to “wear a ‘mask’ in order to deal with families and children in an appropriate way.” In this study, wearing a false face increased as related to role overload within the unit. Role overload is a characteristic of bureaucratic constraint, where time and resources are limited. The example item provided in the literature for role overload states, “The amount of work I have to do keeps me from doing a good job.” Again, role overload as a construct makes up the climate in the agency, and the variable is a normed t-score from across a sample of 81 child welfare agencies nationwide. This means false face occurs more to individuals who work in agencies that are above the norm in terms of their constraints on time and resources (Calabresi & Bobbitt, 1978). Role overload points to lack of autonomy in the job, with limits to time and resources. This is congruent with the literature, in that false face or surface acting is about doing the job while adhering to display rules, even when the ‘heart’ is in a


79
different place due to outside circumstances (Gelderen, 2016; Guy et al., 2008; Hochschild,
2003; Kruml & Geddes, 2000).
Similar to deep acting, the reality of these emotional ‘performances’ may be detected by
the service recipient, and false face has been observed empirically to determine response of the
customer or client to the service worker (Zhan et al., 2015). The implications in the applied field
of child welfare, then, become clear. William Schwartz, a prominent social work educator and
author of numerous foundational textbooks writes about the role of the social worker,
“The process by which the client constructs [her] experience is not one the worker creates; [she] simply enters, and leaves... [she] is an incident in the life of [her] client. Thus the worker should ask [herself]; what kind of incident will I represent.. How do I enter the process, do what I have to do, and then leave?”
(Lee, 2001, p. 186).
Indeed, what incident might be represented by the intervention of a child welfare caseworker who is preoccupied with self-focused emotional labor due to constraints that impose enormous pressure and burden? This emotional restraint is deleterious to relationship building, and people changing technologies in human services require extensive relationship and emotional engagement for their efficacy (Hasenfeld, 1983).
Of note, false face acting behaviors in this study varied significantly by unit, to the point where statistical analysis not grouped at the unit level might have falsely reported trends in, and influences on, performance. Similarly, role overload is not an individual perception, but rather the mean of perceptions of all the caseworkers in a unit. Organizational climate, then, is the unit of analysis in this relationship, regardless of individual unit members’ perceptions on overload. It takes little effort to imagine a frenetic scene, wherein individuals in a unit are scrambling to accomplish the important work of child welfare, unable to assist team members, and unable to slow down for reflection and critical thinking. The consequences of this unit level reality in the


80
office are reflected in restricted emotion in the field by caseworkers interacting with families and children. Practically speaking, it seems the mission of child welfare might be at cross-purposes with this scenario, and as noted above, this is most likely not lost on those citizens receiving interventions and services from the agency.
Self-Efficacy
The caseworkers in this sample emphasized the importance of emotion management in the practice of child welfare and their skill in performing these necessary tasks (“I am good at getting people to calm down”). To achieve some variability in analysis, I used a compilation of items representing caseworker reflections of success in the people changing technology of child welfare (child safety, permanency and well-being). For this measure of self-efficacy, participants reflected on their success for impacting children and families according to the policy goals of child welfare as set forth in Federal law. These goals also represent categories where agencies measure caseworker performance (e.g. recurrence of maltreatment or time to permanency for children in foster care). Arguably, these categories are of primary importance to the work of child welfare. In this sample of caseworkers, however, participants expressed that when they experienced worry that if a child is harmed on their caseload they will not be supported by their agency, leadership, or supervisor, their experiences of self-efficacy in their job decreased. This observation mirrors the impact on authentic expression and deep acting.
In this finding, two areas of the work are again at cross-purposes. Caseworkers who feel enormous accountability for the safety of children on their caseload, and who worry about their agency supporting their discretion and autonomy if that were to happen, at the same time report they feel less effective at meeting overall goals in child welfare. There are two ways to consider this observation in light of the literature: first, the bureaucratic burden must be lightened to free


81
up autonomy to practice people changing technologies including emotional labor to pursue policy goals, or that the measures of achievement of goals and mission for these caseworkers are out of balance. The first reflects the debate in the field related to how child welfare work should be conducted, and the disproportionate attention to relatively rare events, particularly in light of child welfare scandals and ad hoc policy development (Gainsborough, 2010). The second possibility drives at issues related to performance management. Child safety, child permanency, and child well-being are the expressed goals of child welfare agencies in the United States, but the mechanisms for achieving them are varied and the metrics for their measurement can be distorted in pursuit of management strategies. Often performance management metrics reflect a linear line of cause and effect, or inputs (adherence to policy) and outputs (child safety, permanency, and well-being) without including the unpredictable (and in most cases uncontrollable) dynamics of human behavior (Radin, 1998). Thus, street level bureaucrats must make a complicated assessment of their achievement of goals and performance, within the reality that their agencies grant them a degree of discretion and autonomy in their work, but their impact is ultimately determined by the choices of individual caregivers and families, over which they have little perceived or actual control. This is decidedly not a linear situation, and the bureaucratic constraint in this equation contains within it the fear that agencies, leadership, and supervisors may judge casework in a linear manner despite this reality.
Intent to Quit
Turnover is a prevalent issue in child welfare organizations. The most recent national study, which is now quite dated, estimated that turnover rates for child welfare agencies is 22% per year (American Public Human Services Association, 2005). Local, state, and national jurisdictions spend considerable time, effort and money to address this issue. In this study, the


82
odds of preliminary thinking by caseworkers about quitting their position increased, once again, in relationship to concerns that they might not be supported by their agency, leadership or supervisor in the event of a child being harmed on their caseload. This finding is unique in the literature related to turnover in child welfare systems specifically, which typically targets burnout, job dissatisfaction, low commitment to the work and the organization, stress, and lack of social support as antecedents to turnover (Lizano & Mor Barak, 2012; Mor Barak, 2001). While this phenomenon does mirror social support, it specifically operationalizes restraints on autonomy and discretion, and encompasses a portion of the accountability child welfare workers feel for children and families on their caseloads. It is important to note that self-efficacy is tied in the literature to job satisfaction, which is subsequently tied to intent to quit, so more research on these linkages, mediators, and moderators is warranted in light of this finding. Similarly, high frequency of false face acting is also empirically tied to burnout and job satisfaction, specifically in public services like child welfare (Hsieh et al., 2011).
Limitations and Recommendations for Future Research
As with all inquiries, there are limitations to this study that impede interpretation and generalizability of findings. Because of the critical nature of emotion work in public service and specifically in child welfare, these limitations provide outline of areas for future research and inquiry. Recommendations for future research directly follow limitations in this section. The limitations of this study may be described in the following areas: sample strategy, measurement strategies including self report via survey, cross-sectional design, and solely quantitative analysis. Each limitation is outlined in this section with attention to its description, explanation, potential impacts on findings and conclusions, and suggestions for on-going attention to the research questions.


83
The first limitation is the sample strategy. The sample for this inquiry was one of convenience, as part of a larger project related to workforce development in child welfare. While care was taken to access large portions of the population of child serving caseworkers in each of the seventeen jurisdictions, the jurisdictions themselves were selected based on a variety of factors outside the purview of this study. These factors included: interest in the larger workforce project, lack of participation in other initiatives within the state, availability and willingness of leadership to participate in the larger initiative and overall perception that there were workforce needs to address within the agency. These factors could introduce bias at least in terms of the population sampled. It is possible that the agencies that self-selected into the study already had some indication in the jurisdiction that there were difficulties with retention and turnover, or that there were concerns related to organizational culture and climate. One possible study mitigation of this potential bias was the normalization of variables related to organizational culture and climate. Bias at the unit level may have been reduced by basing these scores on norms across child welfare and child serving agencies. On the other hand, however, there was no such normalization with the individually based variables within the study, which were based on self reported perceptions of individuals within the agency. Norms do not exist in the same manner for these variables.
Similarly, a second limitation of this study is use of survey methodology to understand the experiences of the sample. Self-report certainly captures important aspects of experiences of street level bureaucrats, but such methodologies may introduce bias. For example, study participants might have over- or under- estimated their experiences upon hearing about the workforce initiative designed to improve their work life. While participants were assured of their confidentiality through informed consent, it is difficult to ascertain sincerity of response or the


84
match of response with the reality of the work. The survey, attached in Appendix 1, was developed for a national research study on building evidence for effective interventions in the child welfare workforce. To assist in this aim, the attached survey, which is quite lengthy, was administered alongside the Organizational Social Context (OSC), which contains 105 items. Participants filled out the survey and the OSC (not necessarily in that order) typically in one sitting, introducing the possibility of survey fatigue with items that occur later in either of the tools. While the survey does contains items that measure social desirability (e.g. “I never litter,” “I always return items that I borrow from others”), these items were not analyzed as part of this study.
The design of this study is cross-sectional, whereas the organizational life and the experience of child welfare caseworkers is dynamic. The choice for the cross-sectional nature of this study was based on time constraints as well as logistic feasibility. There is not analysis of contextual information pertaining to study participants such as recent or significant events that impacted the workforce. Thus, survey responses and perceptions of factors might be based on events immediately preceding the administration of the survey and OSC rather than on the general experience of the caseworker.
Finally, analysis and conclusions are based solely on quantitative analysis of survey items. While useful, these items only capture a portion of the lived experience of street-level bureaucrats. Several conclusions in the prior sections point specifically to this limitation, including that related to lower levels of performance of deep acting by women and social work majors, and the last finding related to caseworker intention to quit her/his job. In particular, the variable related to concern for liability and support by agency, leadership, and supervisors cannot fully be understood through statistical analysis. For example, how does this bureaucratic


85
constraint play out in the agency? Are caseworker worries that they will lack support during difficult times based on watching this play out with peers? Are they based on some part of the culture and climate that was not explored through this study? While still significant in finding, much more could be understood through qualitative methods.
The preceding limitations, in addition to the conclusions and findings in this study contribute to the development of a research agenda that continues to address this area of public administration. This research agenda contains three parts: (1) continued analysis of this data set, (2) replication of this design paired with a mixed methods approach, and (3) longitudinal designs that measure the dynamics in this study over time and in the context of dynamically changing child welfare agencies.
The first approach, continued analysis of this data set, draws on the rich data still unexplored in this extensive survey design. Attention to social desirability factors (discussed earlier) and analysis of skip patterns or answer patterns if evident might further inform the limitation related to survey fatigue from the length of the instrument. Further, the survey contains several cross-examinations of related concepts that might enliven the analysis conducted here. These concepts and constructs include personality through a version of the Big Five Inventory (BFI) (John & Srivastava, 1999) as well as measurements of burnout through the Maslach Burnout Inventory (MBI) (Maslach, Schaufeli, & Leiter, 2001). Further, scales are included that measure relationships with co-workers, supervisory and social support, empathic orientation, coping and resilience characteristics, time constraints for completion of paperwork and family engagement, characteristics of workload including caseload size, and overall orientation toward family preservation as compared to child safety (Dettlaff et al., 2015; Graham, Dettlaff,
Baumann, & Fluke, 2015).


86
The next opportunity for continued research on these questions is the replication of the design of this study outside the context of the larger workforce study, with attention to triangulating qualitative data with the quantitative results. Specifically, more detailed explanation of deep acting by the caseworkers who score themselves on either side of the spectrum (high disagree or highly agree) might assist in characterization that sheds light on how deep acting plays out in practice with children and families, and assist in understanding how those dynamics might impact people changing technologies. Further, semi-structured interviews might create space to examine organizational dynamics that lend to the worry described in the main findings of this study. How do caseworkers receive signals about support for incidents that have likely not yet occurred? How overt is the messaging in their unit or organization, and for those that perceive less worry in this area, how are they being treated that reassures them? What types of incentives exist related to display rules for ‘high performers’ of emotional labor, and what level of awareness to workers have about resulting responses from children and families for either deep acting or false face (surface acting)? This level of triangulation also introduces approaches to concerns related to performance management, with the ability to examine subsequent administrative data and client level outcomes, as well as target analysis that includes families and children who experience these workers in the field.
Longitudinal research designs, especially when paired with a mixed methods approach, also present opportunities for continued inquiry. Specifically of interest based on findings in this study is the degree to which certain individuals might enter the field with proclivity toward the performance of emotional labor and specifically deep acting, but who later may resort to surface acting for self-preservation and/or in the race of constraints on time and resources related to their working role. While longitudinal designs are time and investment heavy, “longitudinal” designs


87
focusing on the child welfare workforce, where there are high and frequent rates of turnover for such large portions, might shorten the time needed for such inquiry (Mor Barak, 2001). Conclusion
This chapter contained a summary of the findings in this dissertation, alongside an examination of the extant literature in relation to these findings. It then concluded with limitations of the overall study as well as recommendations for further research. In sum, some bureaucratic constraints were related to impacts on street level performance of emotional labor, perceptions of self-efficacy in achievement of policy outcomes, and overall intention of street level bureaucrats’ to stay in their job. In a child welfare setting, efforts to limit autonomy and discretion may result in restrained emotion, which could impact the deployment of people changing technologies and overall outcomes of the agency. More research is needed to understand the organizational and individual characteristics that mediate and moderate this dynamic, and theory must consider the degree to which autonomy and discretion may be restricted in organizational contexts.


88
REFERENCES
American Public Human Services Association. (2005). Report from the child welfare workforce study. Washington, DC. Retrieved from
http://www.theprofessionalmatrix.com/docsAVorkforceReport2005.pdf
Auerbach, C., Schudrich, W. Z., Lawrence, C. K., Claiborne, N., & McGowan, B. G. (2014). Predicting Turnover. Research on Social Work Practice, 24(3), 349-355. http://doi.org/10.1177/1049731513494021
Bachrach, L. L. (1983). An overview of deinstitutionalization. New Directions for Mental Health Services, 1983(11), 5-14. http://doi.org/10.1002/yd.23319831703
Bardach, E. (1977). The implementation game: What happens after a bill becomes a law. Cambridge, MA: MIT Press.
Bartholet, E. (2014). Differential response: A dangerous experiment in child welfare. Florida State University Law Review, 42, 573.
Baumann, D., Dalgleish, L., Fluke, J., & Kern, H. (2011). The decision-making ecology.
Behn, R. (2001). Rethinking democratic accountability. Washington, DC: Brookings Institution Press.
Behn, R. (2003). Why measure performance? Different purposes require different measures. Public Administration Review, 63(5), 586-606. http://doi.org/10.llll/1540-6210.00322
Behn, R. (2014). PerformanceStat Potential: a Leadership Strategy for Producing Results. Brookings Institution Press.
Bertelli, A. M., & Lynn, L. E. (2006). Madison’s managers: Public administration and the constitution. Baltimore, MD: Johns Hopkins University Press.
Blank, R. M. (2002). Evaluating Welfare Reform in the United States. Journal of Economic Literature, 40(4), 1105-1166. http://doi.org/10.1257/002205102762203576
Bovaird, T. (2007). Beyond engagement and participation: User and community coproduction of public services. Public Administration Review, 67(5), 846-860. http://doi.org/10.1111/j. 1540-6210.2007.00773 .x
Bozeman, B. (2000). Bureaucracy and red tape. Upper Saddle River, NJ: Prentice Hall.
Brodkin, E. Z. (2011). Policy work: Street-level organizations under new managerialism. Journal of Public Administration Research and Theory, 27(Supplement 2), i253-i277. http://doi.org/10.1093/jopart/muq093
Brotheridge, C., & Grandey, A. (2002). Emotional labor and burnout: Comparing two perspectives of “people work.” Journal of Vocational Behavior, 60(1), 17-39. http://doi.org/10.1006/jvbe.2001.1815
Bryk, A., & Raudenbush, S. (1992). Hierarchical Linear Models: Analysis and Data Analysis Methods. London, UK: SAGE Publications.


89
Burks, J. (1912). Efficiency standards in municipal management. National Municipal Review, 7(3), 364-371.
Calabresi, G., & Bobbitt, P. (1978). Tragic choices. New York, NY: WW Norton and Company.
Caringi, J., Lawson, H., & Devlin, M. (2012). Planning for emotional labor and secondary
traumatic stress - Google Scholar. Journal of Family Strengths, 72(1), Article 11. Retrieved from https://0-scholar-google-
com. skyline.ucdenver.edu/scholar?q=planning+for+emotional+labor+and+secondary+trau matic+stress&btnG=&hl=en&as_sdt=0%2C6
Chau, S. L., Dahling, J. J., Levy, P. E., & Diefendorff, J. M. (2009). A predictive study of emotional labor and turnover. Journal of Organizational Behavior, 30(8), 1151-1163. http://doi.org/10.1002/job.617
Commission to Eliminate Child Abuse and Neglect Fatalities. (2016). Within our reach: A national strategy to eliminate child abuse and neglect fatalities. Retrieved from http://www.acf.hhs.gov/programs/cb/resource/cecanf-fmal-report
DePanfilis, D., & Zlotnik, J. (2008). Retention of front-line staff in child welfare: A systematic review of research. Children and Youth Services Review. Retrieved from http://0-www. sciencedirect. com. skyline.ucdenver. edu/ science/article/pii/SO 190740908000091
Dettlaff, A. J., Christopher Graham, J., Holzman, J., Baumann, D. J., & Fluke, J. D. (2015). Development of an instrument to understand the child protective services decision-making process, with a focus on placement decisions. Child Abuse and Neglect. http://doi.Org/10.1016/j.chiabu.2015.04.007
Douglas, E. M., & Cunningham, J. M. (2008). Recommendations from child fatality review teams: results of a US nationwide exploratory study concerning maltreatment fatalities and social service delivery. Child Abuse Review, 77(5), 331-351. http://doi.org/10.1002/car.1044
Epp, C., Maynard-Moody, S., & Haider-Markel, D. (2014). Pulled over: How police stops define race and citizenship.
Erickson, R. J., & Wharton, A. (1997). Inauthenticity and Depression: Assessing the
Consequences of Interactive Service Work. Work and Occupations, 24(2), 188-213. http://doi.org/10.1177/0730888497024002004
Finer, H. (1941). Administrative responsibility in democratic government. Public Administration Review, 7(4), 335-350.
Friedrich, C. (1968). Public policy and the nature of administrative responsibility. In F. Rourke (Ed.), Bureaucratic power in national politics (3rd ed., pp. 316-26). Boston: Little Brown.
Gainsborough, J. (2010). Scandalous politics: Childwelfare policy in the states - Google Scholar. Washington, DC: Georgetown University Press.
Gelderen, B. van. (2016). Emotional labor among police officers: a diary study relating strain, emotional labor, and service performance. The International Journal of Human Resource Management, 1-28.


90
Ghose, T. (2009). Linking Organizational Factors to Substance Abuse Treatment Outcomes: Multilevel Correlates of Treatment Effectivenes. In Y. Hasenfeld (Ed.), Human Services as Complex Organizations (2nd ed., pp. 429-451). Los Angeles, CA: University of California Press.
Glisson, C., Green, P., & Williams, N. J. (2012). Assessing the Organizational Social Context (OSC) of child welfare systems: implications for research and practice. Child Abuse & Neglect, 36(9), 621-32. http://doi.Org/10.1016/j.chiabu.2012.06.002
Glisson, C., Landsverk, J., Schoenwald, S., Kelleher, K., Hoagwood, K. E., Mayberg, S., ... Health, T. R. N. on Y. M. (2008). Assessing the Organizational Social Context (OSC) of Mental Health Services: Implications for Research and Practice. Administration and Policy in Mental Health and Mental Health Services Research, 35(1-2), 98-113. http://doi.org/10.1007/sl0488-007-0148-5
Goodnow, F. (1900). Politics and administration: A study in government. New York, NY: Macmillan.
Goodwin, R., Groth, M., & Frenkel, S. (2011). Relationships between emotional labor, job performance, and turnover. Journal of Vocational Behavior, 79(2), 538—548.
Graham, J. C., Dettlaff, A., Baumann, D., & Fluke, J. (2015). The Decision Making Ecology of placing a child into foster care: A structural equation model. Child Abuse & Neglect. http://doi.Org/10.1016/j.chiabu.2015.02.020
Grandey, A. (2000). Emotional regulation in the workplace: A new way to conceptualize emotional labor. Journal of Occupational Health Psychology, 5(1), 95.
Grandey, A. (2003). When the show must go: Surface acting and deep acting as determinents of emotional exhaustian and peer-rated service delivery. Academy of Management Journal, 46(1), 86-96. http://doi.org/10.2307/30040678
Grandey, A., Foo, S. C., Groth, M., & Goodwin, R. (2012). Free to be you and me: a climate of authenticity alleviates burnout from emotional labor. Journal of Occupational Health Psychology, 77(1), 1-14.
Groth, M., Hennig-Thurau, T., & Walsh, G. (2009). Customer reactions to emotional labor: The roles of employee acting strategies and customer detection accuracy. The Academy of Management Journal, 52(5), 958-974.
Gulick, L., & Urwick, L. (1937). Papers in the science of administration. New York, NY: Institute of Public Administration.
Guy, M., Mastracci, S., Newman, M., & Maynard-Moody, S. (2010). Emotional Labor in the Human Service Organization. In Y. Hasenfeld (Ed.), Human Services as Complex Organizations (pp. 291-309). Thousand Oaks, CA.
Guy, M., & Newman, M. (2004). Women’s Jobs, Men’s Jobs: Sex Segregation and Emotional Labor. Public Administration Review, 64(3), 289-298. http://doi.Org/10.l 111/j. 1540-6210.2004.00373.x


91
Guy, M., Newman, M., & Mastracci, S. (2008). Emotional labor: Putting the service in public service. New York, NY: M.E. Sharpe.
Hasenfeld, Y. (1972). People processing organizations: An exchange approach. Sociological Review, 37, 256-263.
Hasenfeld, Y. (1983). Human Service Organizations. Englewood Cliffs, NJ: Prentice Hall.
Hasenfeld, Y. (1992). Human services as complex organizations. Newbury Park, CA: Sage.
Hatfield, E., Cacioppo, J. T., & Rapson, R. L. (1992). Primitive emotional contagion. In M.
Clark (Ed.), Review of Personality and Social Psychology: Emotion and Social Behavior (pp. 151-177). Newbury Park, CA: Sage.
Hemmelgarn, A. L., Glisson, C., & Dukes, D. (2001). Emergency Room Culture and the
Emotional Support Component of Family-Centered Care. Children’s Health Care, 30(2), 93-110. http://doi.org/10.1207/S15326888CHC3002_2
Heuven, E., & Bakker, A. (2006). The role of self-efficacy in performing emotion work. Journal of Vocational Behavior, 69(2), 222-235.
Hochschild, A. (2003). The managed heart: Commercialization of human feeling. Retrieved from
https://books. google. com/books?hl=en&lr=&id=UiSsLXeERF4C&oi=fnd&pg=PP2&dq=th e+managed+heart&ots=NVf2WGhlJF&sig=jJ5MA2qgxhvCfDrxWTxlzRU2kic
Hood, C. (1991). A public management for all seasons? Public Administration, 69( 1), 3-19.
Hood, C. (1995). The “New Public Management” in the 1980s: Variations on a theme. Accounting, Organizations and Society, 20(2), 93-109.
Hopkins, G. (1912). The New York bureau of municipal research. Annals of the American Academy of Political and Social Science, 77(1912), 235-244.
Hsieh, C.-W., & Guy, M. (2008). Performance outcomes: The relationship between managing the “heart” and managing client satisfaction. Review of Public Personnel Administration, 29(1), 41-57.
Hsieh, C.-W., Jin, M., & Guy, M. (2011). Consequences of work-related emotions: Analysis of a cross-section of public service workers. The American Review of Public Administration. Retrieved from http://0-
arp.sagepub.com.skyline.ucdenver.edu/content/early/2011/07/27/0275074010396078.abstra
ct
Hsieh, C.-W., Yang, K., & Fu, K.-J. (2012). Motivational Bases and Emotional Labor: Assessing the Impact of Public Service Motivation. Public Administration Review, 72(2), 241-251. http://doi.org/10.1111/j. 1540-6210.2011.02499.x
Hiilsheger, U. R., & Schewe, A. F. (2011). On the costs and benefits of emotional labor: A metaanalysis of three decades of research. Journal of Occupational Health Psychology, 16(3), 361-389. http://doi.org/10.1037/a0022876


92
John, O., & Srivastava, S. (1999). The big-five trait taxonomy: History, measurement, and theoretical perspectives. In L. Pervin & O. John (Eds.), Handbook of personality: Theory and research, Volume 2 (pp. 102-138). New York, NY: Guilford Press.
Johnson, H., & Spector, P. (2007). Service with a smile: Do emotional intelligence, gender, and autonomy moderate the emotional labor process? Journal of Occupational Health Psychology, 72(4), 319.
Kelly, J. M. (2005). The dilemma of the unsatisfied customer in a market model of public
administration. Public Administration Review, 65(1), 76-84. http://doi.org/10. Ill 1/j. 1540-6210.2005.00432.x
Kohl, P. L., Jonson-Reid, M., & Drake, B. (2009). Time to Leave Substantiation Behind: Findings From A National Probability Study. Child Maltreatment, 74(1), 17-26. http://doi.org/10.1177/1077559508326030
Kruml, S. M., & Geddes, D. (2000). Exploring the dimensions of emotional labor. Management Communication Quarterly, 74(1), 8-49.
Lee, J. A. B. (2001). The empowerment approach to social work practice: Building the beloved community. New York, NY: Columbia University Press.
Lipsky, M. (2010). Street-level bureaucracy (30th Anniv). New York, NY: Russell Sage Foundation.
Lively, K. J. (2002). Client Contact and Emotional Labor: Upsetting the Balance and Evening the Field. Work and Occupations, 29(2), 198-225. http://doi.org/10.1177/0730888402029002004
Lizano, E. L., & Mor Barak, M. E. (2012). Workplace demands and resources as antecedents of job burnout among public child welfare workers: A longitudinal study. Children and Youth Services Review, 34(9), 1769-1776.
Martin, L. M., Peters, C. L., & Glisson., C. (1998). Factors Affecting Case Management
Recommendations for Children Entering State Custody. Social Service Review, 72(4), 521— 544. http://doi.org/10.1086/515777
Maslach, C., Schaufeli, W., & Leiter, M. (2001). Job burnout. Annual Review of Psychology, 52, 397-422.
Maslow, A. H. (1943). A theory of human motivation. Psychological Review, 50(4), 370-396. http://doi.org/10.1037/h0054346
Mastracci, S., Guy, M., & Newman, M. (2011). Emotional labor and crisis response: Working on the razor’s edge. Armonk, NY: ME Sharpe.
Mastracci, S., Newman, M., & Guy, M. (2006). Appraising emotion work: Determining whether emotional labor Is valued in government jobs. The American Review of Public Administration, 36(2), 123-138. Retrieved from http://0-arp. sagepub. com. skyline, ucdenver. edu/content/3 6/2/123. short


93
May, P. J., & Winter, S. C. (2009). Politicians, Managers, and Street-Level Bureaucrats: Influences on Policy Implementation. Journal of Public Administration Research and Theory, 19(3), 453-476. http://doi.org/10.1093/jopart/mum030
Maynard-Moody, S., & Musheno, M. (2003). Cops, teachers, counselors: Stories from the front lines of public service. Ann Arbor, MI: University of Michigan.
Maynard-Moody, S., & Portillo, S. (2010). Street-level bureaucracy theory. In R. Durant (Ed.), The Oxford handbook of American bureaucracy (pp. 252-277). New York, NY: Oxford University Press.
McGregor, D. (1957). The human side of enterprise. Management Review, 46(11), 22-34.
Meier, K., Mastracci, S., & Wilson, K. (2006). Gender and emotional labor in public
organizations: An empirical examination of the link to performance. Public Administration Review, 66(6), 899-909. Retrieved from http://0-
onlinelibrary.wiley.com.skyline.ucdenver.edu/doi/10.1111/j. 1540-6210.2006.00657.x/full
Merkel-Holguin, L., Kaplan, C., & Kwak, A. (2006). National study on differential response in child welfare. Washington, DC. Retrieved from https://0-scholar-google-com. skyline.ucdenver.edu/scholar?cluster=1498212247039603311&hl=en&as_sdt=0,6
Meyers, M., & Vorsanger, S. (2007). Street-level bureaucrats and the implemenation of public policy. In B. Peters & J. Pierre (Eds.), Handbook of public admininstration.
Moore, M. (1995). Creating public value: Strategic management in government. Boston, MA: Harvard University Press.
Mor Barak, M. E. (2001). Antecedents to Retention and Turnover among Child Welfare, Social Work, and Other Human Service Employees: What Can We Learn from Past Research? A Review and Metanalysis. The Social Service Review, 75(4), 625-661.
Morris, J. A., & Feldman, D. C. (1996). The dimensions, antecedents, and consequences of emotional labor. Academy of Management Review, 21(4), 986-1010. http://doi.org/10.5465/AMR.1996.9704071861
O’Leary, R. (2006). The ethics of dissent: Managing guerrilla government. Washington, DC: CQ Press.
Powell, M., Greener, I., Szmigin, I., Doheny, S., & Mills, N. (2010). Broadening the Focus of Public Service Consumerism. Public Management Review, 12(3), 323-339. http://doi.org/10.1080/14719030903286615
Pressman, J., & Wildavsky, A. (1973). Implementation: How great expectations in Washington are dashed in Oakland: or, why it’s amazing that federal programs work at all, this being a saga of the Economic Development Administration as told by two sympathetic observers who seek to build morals . Berkeley, CA: University of California Press.
Prottas, J. (1979). People processing: The street-level bureaucrat in public service bureaucracies. Lexington, MA: Lexington Books.
Pugliesi, K. (1999). The consequences of emotional labor: Effects on work stress, job satisfaction, and well-being. Motivation and Emotion, 23(2), 125-154.


94
Radin, B. (1998). The Government Performance and Results Act (GPRA): Hydra-headed monster or flexible management tool? Public Administration Review, 55(4), 307-316.
Radin, B. (2006). Challenging the performance movement: Accountability, complexity, and democratic values. Washington, DC: Georgetown University Press.
Rafaeli, A., & Sutton, R. I. (1990). Busy stores and demanding customers: How do they effect the display of positive emotion? Academy of Management Journal, 33(3), 623-637. http://doi.org/10.2307/256584
Riccucci, N. (2009). The Pursuit of Social Equity in the Federal Government: A Road Less Traveled? Public Administration Review, 69(3), 373-382. http://doi.org/10. Ill 1/j. 1540-6210.2009.01984.x
Robichau, R., & Jr, L. L. (2009). The implementation of public policy: Still the missing link. Policy Studies Journal. Retrieved from http://0-
onlinelibrary.wiley.com.skyline.ucdenver.edu/doi/10.1111/j. 1541-0072.2008.00293.x/full
Roethlisberger, F. (1941). Management and morale. Cambridge, MA: Harvard University Press.
Roh, C.-Y., Moon, M. J., Yang, S.-B., & Jung, K. (2016). Linking Emotional Labor, Public Service Motivation, and Job Satisfaction: Social Workers in Health Care Settings. Social Work in Public Health, 37(2), 43-57. http://doi.org/10.1080/19371918.2015.1087904
Sandfort, J. R. (2003). Exploring The Structuration Of Technology Within Human Service Organizations. Administration & Society, 34(6), 605-631. http://doi.org/10.1177/0095399702239167
Sayre, W. (1958). Premises of public administration: Past and emerging. Public Administration Review, 18(2), 102-105.
Schein, E. (2010). Organizational culture and leadership (4th ed.). San Francisco, CA: Jossey-Bass.
Seery, B. L., & Corrigall, E. A. (2009). Emotional labor: links to work attitudes and emotional exhaustion. Journal of Managerial Psychology, 24(8), 797-813. http ://doi .org/10.1108/02683940910996806
Simon, H. (1997). Administrative behavior: A study of decision-making processes in administrative organizations (4th ed.). New York, NY: The Free Press.
Sloan, M. (2014). The consequences of emotional labor for public sector workers and the
mitigating role of self-efficacy. The American Review of Public Administration. Retrieved from http://0-arp.sagepub.com.skyline.ucdenver.cdu/content/44/3/274.shoit
Stewart, D. (1985). Professionalism vs. democracy: Friedrich vs. Finer revisited. Public Administration Quarterly, 9(1), 13-25.
Taylor, F. (1911). The principles of scientific management. New York and London: Harper Brothers.


95
Tummers, L., Bekkers, V., Vink, E., & Musheno, M. (2015). Coping during public service delivery: A conceptualization and systematic review of the literature. Journal of Public Administration Research and Theory, 25(4), 1099-1126. http://doi.org/10.1093/jopart/muu056
US Department of Health and Human Services, Administration for Children Youth and Families, & Children’s Bureau. (2014). Child Maltreatment 2014.
Verbeke, W., Volgering, M., & Hessels, M. (1998). Exploring the Conceptual Expansion within the Field of Organizational Behaviour: Organizational Climate and Organizational Culture. Journal of Management Studies, 55(3), 303-329. http://doi.org/10.llll/1467-6486.00095
Wharton, A. (1993). The Affective Consequences of Service Work: Managing Emotions on the Job. Work and Occupations, 20(2), 205-232. http://doi.org/10.1177/0730888493020002004
Wharton, A. (1996). Service with a smile: Understanding the consequences of emotional labor.
In C. MacDonald & C. Sirianni (Eds.), Working in the service society (pp. 91-112). Philadelphia, PA: Temple University Press.
Whitaker, G. (1980). Coproduction: Citizen participation in service delivery. Public Administration Review, 40(3), 240-246.
White, L. (1926). Introduction to the study of public administration. New York: Macmillan.
Wilson, J. (2000). Bureaucracy: What government agencies do and why they do it. New York, NY: Basic Books.
Winokur, M., Ellis, R., Drury, I., & Rogers, J. (2015). Answering the big questions about
differential response in Colorado: Safety and cost outcomes from a randomized control trial. Child Abuse and Neglect, 39(98).
Zhan, Y., Wang, M., & Shi, J. (2015). Interpersonal Process of Emotional Labor: The Role of Negative and Positive Customer Treatment. Personnel Psychology, 69(3), 525-557. http://doi.org/10. Ill 1/peps. 12114


96
APPENDIX
ID_________ Today’s Date______________________________________________________
Indicate your undergraduate degree major_____________________Year Graduated_______
From what institution did you receive your undergraduate degree?________________________
Are you a graduate of a Title IV-E Child Welfare Stipend Program (i.e., a program which paid for college in return for work in the field of child welfare which also included coursework on child maltreatment and a field placement in child welfare)?
â–¡ Yes ___No
â–¡ If YES, what year________________
Do you have an advanced degree? If so what type of advanced degree do you hold? Check all that apply
â–¡ MA/MS__________________(in what discipline)
â–¡ MSWorMSSW
â–¡ MEd.
â–¡ MPH
â–¡ M.Div.
â–¡ Masters in Nursing
â–¡ JD
â–¡ Ed. D.
EH Ph.D.__________________(in what discipline)
EH Other_________________________(please specify)
Did you complete an MSW Stipend Program through VDSS or in another state?
â–¡ Yes____________________From where? _______-If yes, what year? ____No
Are you of Hispanic/Latina/o Origin?
â–¡ Yes ___No
What is your self-identified race?
EH African American/Black
â–¡ Caucasian/White EH Asian
â–¡ Native American EH Biracial/Multiracial
EH Other (please specify):____________________________
What year were you born?: ________________


97
Gender ____Male ___Female _________Transgender ________Gender nonconforming
Over the course of your career working in child welfare, how much experience have you had serving as a caseworker managing the following types of cases? (Please do not count experience supervising others.)
l(None at all) 2(A little) 3(A moderate amount)) 4(A lot) 5(A great deal)
Intake 1 2 3 4 5
Investigations 1 2 3 4 5
Family Assessments 1 2 3 4 5
CPS Ongoing 1 2 3 4 5
Foster Care 1 2 3 4 5
Adoption 1 2 3 4 5
Prevention 1 2 3 4 5
Check one: In our county/locality we define a “case” as familv or a child
Which types of cases are on your current caseload? (Please check all that apply.)
__Intake (hotline)
__Investigations
__Family Assessments
__CPS Ongoing
__Foster Care
__Adoption
__Prevention
__Other____________________________________________
Thinking about your current caseload, how many of your current cases reflect the following case types? (Please enter 0 for those case types you do not currently serve.)
Type of case Number of cases
Intake (hotline) ____
Investigations ____
Family Assessments ____
CPS Ongoing ____
Foster Care ____
Adoption ____
Prevention ____
Total number of cases ____
Average hours you have worked on your job per week in the past 6 months (not just what were paid for but
actual hours it takes you to do your job each week): __39 hours or less ____40-49 _____50-59 ______60-69
70+


98
Circle the number that corresponds to your level of agreement with each statement
1 (Strongly Disagree) 2(Disagree) 3(Neutral) -/(Agree) 5(Strongly Agree)
Tilings I learn are useful 12 3 4
It is easy for me to use what I know in new situations. 1 2 3 4
It does not matter if I do well or poorly, I improve after finding out 1 2 3 4
People I care about think I should try to know more or do tilings better 1 2 3 4
My friends will disapprove if I do not strive to improve myself 1 2 3 4
People that mean a lot to me expect me to learn. 1 2 3 4
When I have time to spare, I try to do something better 1 2 3 4
When I am bored, I improve my skills. 1 2 3 4
Through learning I am more able to do my job 1 2 3 4
Learning helps me to do better in what I do. 1 2 3 4
Using the transcription service will improve my job performance 1 2 3 4
Using the transcription service will increase my productivity. 1 2 3 4
Using the transcription service will enhance my effectiveness 1 2 3 4
I think the transcription service will be useful. 1 2 3 4
The amount of time I spend on paperwork will decrease after I have mastered the transcription service. 1 2 3 4
I will have more time for contact with families and children after I have mastered the transcription service. 1 2 3 4
Learning to use the transcription service will be easy for me 1 2 3 4
I will find it easy to get the transcription service to do what I desire. 1 2 3 4
I will find the transcription service easy to use 1 2 3 4
I intend to use the transcription service frequently. 1 2 3 4
I believe I will have the ability to transcribe notes after client visits. 1 2 3 4
I believe the transcribed notes will be accurate. 1 2 3 4
I believe I will have the ability to upload the transcribed notes into the case management system. 1 2 3 4
I believe I will have the ability to use the transcribed notes to complete assessments, plans, case updates on progress, etc. 1 2 3 4
I am satisfied with my job. 1 2 3 4
My supervisor genuinely cares about me. 1 2 3 4
My supervisor gives me help when I need it. 1 2 3 4
My supervisor supports me in difficult case situations. 1 2 3 4
My supervisor helps me learn and improve. 1 2 3 4
My supervisor values and seriously considers my opinions in case decision making. 1 2 3 4
My supervisor helps me prevent and address burnout. 1 2 3 4
I have tender, concerned feelings for people less fortunate than me. 1 2 3 4
I don’t feel very sorry for other people when they are having problems. 1 2 3 4
When I see someone being taken advantage of, I feel kind of protective towards them. 1 2 3 4
Other people’s misfortunes do not usually disturb me a great deal. 1 2 3 4
When I see someone being treated unfairly, I don’t feel very much pity for them. 1 2 3 4
I am quite touched by tilings I see happen. 1 2 3 4
I would describe myself as pretty soft-hearted person. 1 2 3 4
Co-workers in my unit professionally share and learn from one another 1 2 3 4
Co-workers in my unit share work experiences with each other to improve the effectiveness of client services. 1 2 3 4
Co-workers in my unit encourage each other to exercise professional judgment when making decisions. 1 2 3 4
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5


99
Circle the number that corresponds to your level of agreement with each statement
1 (Strongly Disagree) 2(Disagree) 3(Neutral) -/(Agree) 5(Strongly Agree)
Co-workers in my unit are willing to provide support and assist each
other when problems arise. 1 2 3 4 5
Co-workers in my unit accept support from their colleagues. 1 2 3 4 5
My supervisor assists me in setting and assessing long-tenn case goals. 1 2 3 4 5
My supervisor encourages creative solutions. 1 2 3 4 5
My supervisor is knowledgeable about effective ways to work with children and families. 1 2 3 4 5
My supervisor demonstrates leadership. 1 2 3 4 5
My supervisor teaches me the skills I need in this job. My supervisor requires that I use standards (i.e., criteria) to address 1 2 3 4 5
case decisions. 1 2 3 4 5
In uncertain times, I usually expect the best. 1 2 3 4 5
If something can go wrong for me, it will. 1 2 3 4 5
I’m always optimistic about my future. 1 2 3 4 5
I hardly ever expect tilings to go my way. 1 2 3 4 5
I rarely count on good tilings to happen to me. 1 2 3 4 5
Overall, I expect more good tilings to happen to me than bad. 1 2 3 4 5
I see myself as extraverted, enthusiastic. 1 2 3 4 5
I see myself as critical, quarrelsome. 1 2 3 4 5
I see myself as dependable, self-disciplined. 1 2 3 4 5
I see myself as anxious, easily upset. 1 2 3 4 5
I see myself as open to new experiences, complex. 1 2 3 4 5
I see myself as reserved, quiet. 1 2 3 4 5
I see myself as sympathetic, warm. 1 2 3 4 5
I see myself as disorganized, careless. 1 2 3 4 5
I see myself as calm, emotionally stable 1 2 3 4 5
I see myself as conventional, uncreative. 1 2 3 4 5
I often think about quitting my job The primary reason I stay in this job is the families and children 1 2 3 4 5
I work with 1 2 3 4 5
The primary reason I stay in this job is that I enjoy protection and safety work The primary reason I stay in this job is that I like to work for 1 2 3 4 5
this agency (my employer) 1 2 3 4 5
Indicate how you have felt over the past month
I tend to bounce back quickly after hard times 1 2 3 4 5
I have a hard time making it through stressful events. 1 2 3 4 5
I does not take me long to recover from a stressful event 1 2 3 4 5
It is hard for me to snap back when something bad happens. 1 2 3 4 5
I usually come through difficult times with little trouble. 1 2 3 4 5
I tend to take a long time to get over set-backs in my life. 1 2 3 4 5


100
Circle the number that corresponds to your level of agreement with each statement
1 (Strongly disagree) 2 (Disagree) 3(Somewhat disagree) 4(Neither agree nor disagree) 5 (Somewhat agree) 6 (Agree) 7 (Strongly agree)
Even working overtime, I cannot finish all of my work 1 2 3 4 5 6
My caseload is too high 1 I have too many cases to do a good job, yet I am expected 2 3 4 5 6
to do so. 1 2 3 4 5 6
I cannot spend enough time with my clients 1 It is difficult for me to keep up with agency policies and 2 3 4 5 6
guidelines 1 2 3 4 5 6
I will probably look for a new job in the next year. 1 2 3 4 5 6
I believe I have had adequate training to help me make
the right decision about the safety and well-being of my clients 1 2 3 4 5 6
I am worried that one of my cases may draw media attention 1 I have known caseworkers that have been disciplined or fired 2 3 4 5 6
because of real or perceived mistakes 1 2 3 4 5 6
If a child in one of my cases is harmed, I worry that I will be disciplined or fired 1 If a child in one of my cases is harmed, I believe the agency will conduct a thorough investigation into what a happened 2 3 4 5 6
before assigning blame 1 2 3 4 5 6
I know my supervisor will be supportive of me and the decisions I made if a child is harmed in one of my cases 1 I know my leadership will be supportive of me and the 2 3 4 5 6
decision I made if a child is harmed in one of my cases 1 2 3 4 5 6
I hide my true feelings so as to appear pleasant at work. 1 In my job I act confident and self-assured regardless of how 2 3 4 5 6
I actually feel. 1 2 3 4 5 6
I wear a “mask” in order to deal with families and children in an appropriate way. 1 I let my true feelings show when working with families 2 3 4 5 6
and children. 1 2 3 4 5 6
It is easier for me to show my true feelings than to pretend. 1 2 3 4 5 6
I am good at expressing how I feel. 1 2 3 4 5 6
I am good at getting people to calm down. 1 Dealing with emotionally charged issues is an important part 2 3 4 5 6
of my job. 1 2 3 4 5 6
In my job I am good at dealing with emotional issues. 1 I try to actually experience the emotions that I must show to 2 3 4 5 6
families and children 1 2 3 4 5 6
I work hard to actually feel the emotions that I need to show to families and children 1 2 3 4 5 6
I work at developing the feelings inside of me that I need to show to
families and children 1 2 3 4 5 6
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7


Full Text

PAGE 1

BUREAUCRATIC CONSTRAINTS AND STREET LEVEL PERFORMANCE OF EMOTIONAL LABOR by IDA J. DRURY B.A.S.W., Wartburg College, 2001 M.S.W., St. Ambrose University, 2003 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Doctor of Philosophy Public Affairs Program 201 8

PAGE 2

2 This thesis for the Doctor of Philosophy degree by Ida J. Drury has been approved for the Public Affairs Program By Mary E. Guy, C hair John Ronquillo Jane Hansberry Sharon Mastracci Date: May 18, 2019

PAGE 3

3 Drury, Ida J. (Ph.D., School of Public Affairs) Bureaucratic Constraints and Street Level Performance of Emotional Labor Thesis directed by Professor Mary E. Guy ABSTRACT Street lev el bureaucrats are the face of government in public service delivery . To facilitate interactions with citizens in pursuit of policy goals, they must engage in emotional labor. Emotional labor may be self e other focused, with the intent to assist others in em oting to facilitate change . This is especially true in human services environments, where street level bureaucrats utilize people changing technologies in their daily work. Street level bureaucrats ope rate in the context of organizations, and the culture and climate of the organization may impose bureaucratic constraints on their behavior to optimize accountability and efficiency. This dissertation examines the manner in which bureaucratic constraints i nfluence emotional labor, specifically in the organizational context of public child welfare. Findings of this research indicate that bureaucratic constraints, specifically those concerning support by agency leadership for discretion and autonomy in light of a negative outcome and time or resource constraints, have detrimental impacts on the performance of emotional labor by the caseworkers in this study. These findings contribute to the literature on theories of autonomy and discretion in street level bure aucracy as well as on the practical setting of child welfare casework and public administration. The form and content of this abstract are approved. I recommend its publication. Approved: Mary E. Guy

PAGE 4

4 ACKNOWLEDGEMENTS I am grateful to many as I complet e this milestone. I would not be in the position to add PhD to my name without the support of a village. When I think of all the shoulders I leaned on, coffee shops haunted, happy hours commiserated, and celebrations large and small along the way, I feel f ull. There are too many people to name in this document, so I highlight a few that rise to the forefront in my mind. First, to James, I am grateful for your steadfast support and belief in my dreams. You filled and emptied the dishwasher, tread lightly t hrough PhD student emotions, left me alone when I needed it, and drew close when I needed that, too. I a m excited for our next chapter together, zusammen . vision for me be fore I truly saw it for myself. Your guidance, wisdom, and sense of humor shepherded me through this process, and helped me remember the goal. The achieved goal is taking form in North Dakota. Thank you. I am also grateful to educators and scholars along the way who instilled in me the desire to join the academy. A few: Dr. Susan Vallem, Dr. Katherine Van Blair, Dr. Jess ica Sowa, the scholars at PA Theory Network , and my committee. Thanks for being dedicated, brilliant, and for sharing it with me as a stu dent and colleague. Then, my tribe: Maren, Annie, and Laura. You get it . Thank you for laughter, tears, fierce loyalty, texts, and friendship. I look forward to many more years of the same, despite the many miles and states between us. Finally, I am gra teful to my family and f riends , near and far , who supported me even when this endeavor was hard to understand.

PAGE 5

5 TABLE OF CONTENTS I. STATEMENT OF THE PROBLEM ................................ ................................ ............................ 7 Emotional Labor ................................ ................................ ................................ ............... 11 Street Level Bureaucracy Theory ................................ ................................ ..................... 12 Impact of Organizational Culture on Street level Bureaucrats ................................ ......... 13 I nstitutional Context for Practical Emotion Work ................................ ............................ 17 Research Questions ................................ ................................ ................................ ........... 20 II. LITERATURE REVIEW ................................ ................................ ................................ .......... 21 Emotional Labor ................................ ................................ ................................ ............... 21 Street Level Bureaucracy Theory ................................ ................................ ..................... 26 Impact of Organizational Culture on Street level Bureaucrats ................................ ......... 29 Public Child Welfare as an Institutional Context for Practical Emotion Work ................ 37 III. HYPOTHESES AND RESEARCH DESIGN ................................ ................................ ......... 42 Hypotheses ................................ ................................ ................................ ........................ 43 Research Design ................................ ................................ ................................ ............... 45 Operationalization of Theoretical Constructs ................................ ................................ ... 48 Factor Analysis ................................ ................................ ................................ ................. 55 IV. ANALYSIS AND FINDINGS ................................ ................................ ................................ 59 Missing Data ................................ ................................ ................................ ..................... 60 Results ................................ ................................ ................................ ............................... 60 Results of Hypotheses Testing ................................ ................................ .......................... 70 V. DISCUSSION ................................ ................................ ................................ ........................... 72 Summary of Findings ................................ ................................ ................................ ........ 72 Authentic Expression ................................ ................................ ................................ ........ 75 Deep Acting ................................ ................................ ................................ ...................... 76

PAGE 6

6 False Face ................................ ................................ ................................ ......................... 78 Self Efficacy ................................ ................................ ................................ ..................... 80 Intent to Quit ................................ ................................ ................................ ..................... 81 Limitations and Recommendations for Future Research ................................ .................. 82 Conclusion ................................ ................................ ................................ ........................ 87 REFERENCES ................................ ................................ ................................ .............................. 88 APPENDIX ................................ ................................ ................................ ................................ .... 96

PAGE 7

7 I. ST ATEMENT OF THE PROBLEM Child welfare caseworkers o perate daily on the front line as intermediaries between families and government. Citizen understanding of government is cultivated in these interactions, as are relationships that facilitate personal chang es in service recipients to preserve or promote the safety of children . However, the experience s of a worker who must operate within organizational cultures and climates that impose bureaucratic constraints is often ignored in favor of the idea that autono my and discretion on the front line is a never ending fount that can only be quelled but never extinguished. Emotion work, though, may be more dependent on the organizational balance of constraint and discretion than more cognitive, process oriented types of public service, and lack of attention to this practical reality may set up nearly untenable jobs for those serving the most vulnerable of citizens. Child welfare caseworkers are certainly street level bureaucrats in function and form, but are less freq uently addressed as such in the extant public administration literature which favors police officers, welfare eligibility workers, and those charged wit h direct human service delivery. All these street level bureaucrats are described as the connectors betw een everyday citizens and government. Face to face interactions with citizens through direct service delivery is necessary to facilitate coproduction for achieving public value. The hope for these interactions is to improve the lives of citizens, and in do ing so, promote health and well being of the community. As such, the work routines, choices, and strategies of street level bureaucrats are so integral in this performance that research has concluded in many instances they are creators of public policy (Brodkin, 2011) . Street level bureaucrats work within an organizational context that applies both constraints and flexibility for the performance of these work tasks.

PAGE 8

8 All service delivery technologies have both cognitive and emotive components. The cognitive is well known: it is described in the minimum qualifications of positions and ou tlined in performance evaluations. The invisible part of the work is the emotive component. Emotional labor occurs in two directions: It is both self focused and other focused. Self focused refers to focus ed refers to the sensing of the focused and other focused emotional labor work in syner gy to create a context for service delivery and both have different implications for job satisfaction as well as for stress and burnout (Gelderen, 2016; Hsieh, Jin , & Guy, 2011) . This balance may be impacted by constraints facing street level bureaucrats, such as limitations to time and resources . Further, in the name of accountability to citizens and the organization, constraints may be imposed at the organization al level on the discretion and autonomy afforded to street level bureaucrats. Bureaucratic constraints are commonly understood as a framing characteristic of street level activities (Lipsky, 1980/ 2010; Wilson, 2000) . These constraints beco me the organizational context, referred to in the industrial psychology literature as organizational culture and climate. There exists considerable debate about the defining characteristics of organizational culture and climate, however two parts of empiri cal work in this area serve to operationalize bureaucratic constraints: rigidity of routines and rules and centrality of decision making. Both these areas describe the level of discretion in the face of formal and informal rules allowed for the street leve l bureaucrat in the performance of her/his work. Eligibility decisions or discretion for enforcement of rules are examples of the dependent values in these types of studies. For the purpose of this argument, bureaucratic constraints are operationalized as those imposed by the

PAGE 9

9 organizational environment. These constraints are separate and apart from those outlined in the (Bozeman, 2000) . This distinction is supported by literature on street level bureaucracy, where behaviors of these individuals are governed by the unlimited autonomy and discretion. What is less understood is the impact of this organizational culture on the performance of and choices surrounding emotional labor. Further, most studies and observation of emotional labor are focused on individual characteristics and performance rather than organizational context. Specifically, for the purpose of this research, the balance of other focused and self focused emotional labor are known to impact daily life of the street level bureaucrat (and by default, then, citizens) , but less is known about the impact of organizational constraints designed to shape and manage individual street level behavior. Indeed, the literature often refers anecdotally to a logical relationship, but one has not been empirically explored. How then , does an individual emotional laborer experience bureaucratic constraint, and how does that experience shape their performance of emotional labor and the balance between other focused and self focused management of emotions? This is of concern in certain policy contexts which rely more heavily on other focused emotional labor, as in programs designed to facilitate change in citizen behavior. Education, injury prevention, and protection of vulnerable populations are examples of these. Human services contex ts are observed to focus on people changing technologies, but also include a mix of people sustaining and people processing technologies (Hasenfeld, 1983) . All of these contexts to some degree demand coproduction of public values between the government agent and the citizen/service recipient (Bovaird, 2007) . In people changing organizations, engagement with citizens to generate individual change is the primary technology employed. Indeed, there is

PAGE 10

10 indication that this coproduction or engagement is fundamental to desired policy outcomes (Whitaker, 1980) . Certainly, engagement between the practitioner and the citizen and his/her fami produce desired policy outcomes like public safety, children safety, and community preservation. In these contexts, as in other contexts demanding high levels of em otional labor, practitioners are demonstrably gendered, with females carrying the majority in areas like child protective services, education, nursing, and mental health services (Guy & Newman, 2004) . Conversely, constraints are often imposed from the top down, where executives are more often male (Riccucci, 2009) . This gender divide might logically indicate a disconnect between the facilitation of emotive work and the organizational structures where it is performed. This tension leads to further need for exploration of the impact of these imposed constraints on the performance of emotional labor by direct practitioners at the bottom of this decidedly gendered hierarchy. Just as the cognitive aspects of a job demand resources, emotion management demands, such as time, autonomy, and discretion, also exist. Constraints, by definition and consequence, lower the resources afforded. How can other focused emotional labor occur in such situations? Likewise, i s self focused emoti ve regulation simply a matter of survival for the street level bureaucrat? Finally, do program rules demand other focused emotional labor while organizations simultaneously create constraints that render it untenable or emotionally exha usting for the street level bureaucrat? This research may have practical implications for successful performance of emotional labor, including an impact on the understanding of workforce interventions that effectively guide practice and also create work ab le and rewarding careers for street level bureaucrats. Turnover is of utmost cost and concern in these areas of practice, and in

PAGE 11

11 the struggle to create accountability, it is possible that high turnover rates have become somewhat inevitable. This research a lso may impact the theoretical and commonplace picture of level bureaucrats, posing instead a more nuanced picture of the organizational impact on emotional labor performance. Emotional Labor Operationally, emotion manag ement can be empirically observed in several forms: 1) assisting clients in managing their emotions, 2) displaying many different emotions when and 3) dealing with emotionally charged issues as a critical dimension of the job (Guy, Newman, & Mastracci, 2008) . The performance of emotion work as a primary job function is referred to as emotional labor. Specifically, in public administration, street level bureaucrats perform emotional labor to pursue policy goals. The research on emotional labor is divide d into a variety of contexts and scenarios. As mentioned earlier, it is recognized as gendered, so one focus is on the undervalued role of women in the performance of emotional labor. This is so much the case that gender has been used as a proxy for an age (Meier, Mastracci, & Wilson, 2006) . Research also identifies antecedents to emotional labor, including organizational demands on the types of emotional behavior required of the individual worker (Morris & Feldman, 1996) . Finally, a significant body of research also identifies personal characteristics impacting the street contexts create pathways to either job satisfaction or burnout. There are moderating variables that impact the performance of emotional labor. For

PAGE 12

12 may enhance a sense of control and indicate higher job satisfaction (Wharton, 1996) . Personal efficacy at performing other focused emotional labor coupled with enjoyment in working with people can also influence the degree to which burnout occurs . Overall job control, feedback, and opportunities to learn are related to lower incidence of burnout as related to emotional labor (Guy et al., 2008) . Workers who report self efficacy at emotional labor also have positively associated service outcomes for clients (Hsieh & Guy, 2008) . Deleterious effects of emotional labor occur most promin ently when workers must fake their feelings, or constantly self (Hsieh et al., 2011; Sloan, 2014) . This is particularly so for women (Johnson & Spector, 2007) . In situations of low autonomy or high constraints for j ob tasks, workers were also more likely to suffer from burnout and its antecedents: emotional exhaustion and job stress (Johnson & Spector, 2007) . This makes sense given the relation of job satisfaction with self efficacy, and is an initial glimpse into the potential for organizational cul ture to impact emotional labor performance. Street Level Bureaucracy Theory Street Level Bureaucracy (1980/2010) provided a theoretical framework through which government agents may be viewed or understood. Metaphorically, Street Level Bureaucra cy Theory connected the vast body of scholarship in public administration with the realities of front line labor, work routines, and interactions in much the same way that street level bureaucrats connect government with citizens. The primary characteristi cs of this theory can be grouped into five areas: 1) front line status, 2) people processing, 3) inherent discretion, 4) irreducible autonomy, and 5) ultimate policy making (Maynard Moody & Portillo, 2010) . All five areas of this theory have implications for the research questions under examination.

PAGE 13

13 Due to the inherent discretion and irreducible autonomy afforded to the street level bureaucra t, scholars and managers seek to describe and attain accountability. This takes shape in practice as performance management. Effective performance management often focuses on outputs, seeking to quantify the routines and structures at the street level (Behn, 2014) . To tighten t he control over this workforce, bureaucratic and organizational constraints are applied (Wilson, 2000) . Emotional labor, on the other hand, is the least measured or acknowledged element of job performance (Meier et al., 2006) . When sho uld the street level bureaucrat focus the street level bureaucrat choose to mask frustration or irritation with their work by simply smiling their way through? In the name of accountability to agency and citizen, can constraints render the effective and sustainable practice of emotional labor impossible, or otherwise so challenging that direct practitioners either turnover or languish in their positions long aft er their effectiveness has been tapped? Given high levels of turnover and challenges in these policy areas, could this be a mechanism to address? Impact of Organizational Culture on Street level Bureaucrats Industrial psychology and the study of bureaucra tic behavior have common roots . At the turn of the twentieth century, with the industrial revolution in full swing , the American government had undergone significant changes since the drafting of the constitution. So too, the bureaucracy took on new charac ter and federal, state, and local levels. Using knowledge gained from factories and shops, scholars sought to apply these principles in public administration. Lamenting the national losses from a lack of productivity and efficiency in the public sector, Fr ederick Taylor (1911) drew connections between successes in current industry and the practice of scientific management. Taylor recommended that administrative processes be examined for

PAGE 14

14 ineffective components; perhaps, those rule of thumb actions passed down from gener ations before were unnecessary in the current context. Using the example of a bricklayer, Taylor suggested the introduction of simple apparatus might further eliminate superfluous actions and increase efficiency (1911). Taylor gained a great deal of gover nment attention for these administration , the seemingly practical roots of which remain to the current day. While the bureaucracy blossomed , practitioners urged standard ization of municipal management. Citing the standardization of typewriter keyboards, threading on water pipes, and admission requirements to institutions of higher education, Dr. Jesse Burks, then director of Municipal Research in Philadelphia, suggested t hat similar standardization could be applied to government processes and budgeting with successes ranging from saving mone y to saving life and limb (Burks, 1912) . Similar bureaus of research emerge d across the country, with the purpose of increasing efficiency of local government through scientific method s . With caution not often seen in latter day reform efforts, a New York Municipal Bureau of Research trustee commented, Municipal research is a method, not a panacea. It aims not to make over either the man in of fice or the men who vote, but to give men as they are better methods of watching and judging what their public servants do (Hopkins, 1912 , p. 244). The overt focus on efficiency in public administration is apparent in much of the other formative literature as well. Scholars like Frank Goodnow and William Willoughby called for budgetary and structural reform in government, which was a reflection of the need for efficient use of the pu blic trust and public monies. An underlying concern was that public servants would actively pursue pure self interest instead of the interests of the people they represented, as

PAGE 15

15 Goodnow had observed in the Italian system (1900) . Similarly, Willoughby is credited fo r moving the field toward cementing scientific management principles as a foundation for efficient and effective public administration (Sayre, 1958) . The first textbook on public administration further solidified scientific management as the dominant t heme (White, 1926) . These early scholars codified rationality as synonymous with efficiency in the practice of public administration (Sayre, 1958). Consequently, the public service was expected to be dispassionate and impartial to the changing winds of the political environment so as to maintain effectiv e and efficient performance. This is evident in emphasis on the oft debated politics administration dichotomy (Goodnow, 1900). Other scholars explored the application of scientific management principles as related to the structure of government, and, echoi ng Goodnow, emphasized that politics and administration could not be exercised simultaneously without producing inefficiency (Gulick & Urwick, 1937) . In addition to the P/A dichotomy, the theoretical demands of the new bureaucracy also weighed on the importance of neutrality. Weber or without emotion. However, at this same time, psychology experimenters engaged in an industrial study of the impact on illumination on work outputs at the Hawthorne Works of the Western Electric which h ad been an increased understanding the relation of the work environment (essentially the lighting in the production area) and the relationship to the generation of outputs by workers. Instead, the experiment showed that regardless of the physical work env ironment, workers were (Roethlisberger, 1941) . This new knowledge impacted the study of human resources in both industry and in government. The lead researcher commented on the implicati ons of this new

PAGE 16

16 and industrial in nature, and the producers need attent ion to their humanity, what can be expected of the impact of the work environment, or organizational context on the performance of emotional labor? Theories of street level bureaucracy are focused on the individual as well as the irreducible autonomy and discretion afforded members of this group. There is general consensus that this group can never be fully controlled by their organizational context. Ultimately, argues Lipsky, street level bureaucrats are called to be accountable to clients rather than and sometimes in spite of their organization (p. 161, 1980/2010). This creates tension, then, between two theoretical frameworks of human behavior: that of the individual and that of the organization. The practical observation of emotional labor provides a pl atform for this juncture. The study of emotional labor readily crosses these barriers, recognizing but not full examining the importance (Mastracci, Newman, & Guy, 2006 , p.124 ) . Like emotional labor performance by the individual, the influence of organizational c ulture on client outcomes is well established in the field of human services, including in substance use treatment (Ghose, 2009) and child welfare (Glisson, Green, & Williams, 2012) . In the field of industrial psychology, the culture of an institution is linked to poor performance of practitioners, but experience of emotional labor goes uninspected. Some aspects of organizational culture include decided focus on bureaucratic constraints. Specifically, centrality, or the degree to which decisions are limited to higher ranking individual s in the hierarchy, as well as rigidity, the extent to which front line staff have discretion and flexibility over

PAGE 17

17 bureaucratic rules and routines, may serve to operationalize bureaucratic constraints (Glisson et al., 2012) . The research on organizational culture forms a significant basis for human resource approaches in public agencies and is focused on burnout reduction, gains in retention, and maintenance of job satisfaction for employees, but it ignores the literature on emotional labor as the primary work task. In the context of child protection, the impact of the work is also characterized as burnou t, treats the latter two terms as relatively inevitable, and does not fully explain organizations where practitioners persevere in the workforce for longer tenu re and organizational cultures where teams demonstrate higher levels of job satisfaction and burnout avoidance despite similarities in everyday crisis or client groups. Essentially, these are deleterious after effects of the work, but do not take into acco unt the job design itself. The research on emotional labor provides an interesting counterfactual for empirical study that is absent in these other constructs: while prolonged practice of high levels of emotional labor are related to burnout, emotional lab or may also produce high levels of job satisfaction (Guy, Mastracci, Newman, & Maynard Moody, 2010) . As stated earlier, job satisfaction typically occurs when the perform er feels self efficacy toward emotional labor, and incurs a sense that their work is meaningful. Any attempt at empirically examining relationships outside the well established pathways between the practice of emotional labor and burnout resulting from it must include the organizational culture of the performer. Institutional Context for Practical Emotion Work The policy context for this inquiry is child protective services and specifically, the portion of that system relying primarily on encounters with front line caseworkers for the

PAGE 18

18 realization or maintenance of child safety. While much of the media, and also a good portion of many child welfare agencies, focus on the removal, placement, and permanency of children or on those children who die from maltre atment, this dissertation narrows to the front end of the child protection system, or the part most prevalently encountered by the public. Nationally, 3.2 million children and their families received an investigation or assessment for services in 2014. Of those children, approximately 1.3 million children received some sort of family preservation or safety related service following the assessment or investigation. During the same time, 241,919 children (less than 8% of total children encountered) were re mov ed for foster care services (US Department of Health and Human Services, Administration for Children Youth and Families, & . Due to limitations in reporting and definition, it is hard to pinpoint the number of children who die of maltreatmen t following system intervention, but the national report Child Maltreatment, 2014 has a count of 1580, making this occurrence a generally rare event. The focus on removal and foster care of very vulnerable and traumatized children is not unwarranted, nor i s the goal for better understanding situations for children who die as a result of child maltreatment following system involvement. However, it is important to note in cases of foster care and the expansive formal system of courts, that foster caregivers, guardians ad litem and court appointed special advocates play important roles in the delivery of services alongside the individual child welfare caseworker. Similarly, the wide variety of circumstances, social contexts, and involved entities for children w ho die from maltreatment make it difficult to generalize a fully preventative response (Commission to Eliminate Child Abuse and Neglect Fatalities, 2016) . While the community based system of care for children at risk or experiencing low to moderate levels of child maltreatment in their home is also expansive, the public child

PAGE 19

19 protection caseworker is often the only formal entity involved in these cases, and most often on a voluntary basis once a relatively low threshold for child safety has been determined. Typically, in home services are focused on tertiary prevention of child maltreatment and success is often mea sured by rates of recurrence. However, these rates are not often tied to the practices designed to reduce the rates, but are rather pared with more tangible outputs such as time to initial response and thoroughness of assessment and investigative practices . This performance oriented reality ignores the emotive components of child welfare work by claiming effectiveness can be measured quantitatively. However, consideration of the basic logic model of child welfare services indicates effectiveness is achieved when families alter their behavior to provide on going safety for their children. This change does not occur as a result of a well timed, coordinated process but is instead fostered by the emotion work on the street level. It is unclear whether it is desp ite or due to the aforementioned lopsided focus of the child welfare system, but broad sweeping child welfare policy changes in the wake of child fatalities and/or scandals are often focused on the front end portion of the system (G ainsborough, 2010) . One might postulate that these reactive policies might constrain front line caseworkers in the hope to further improve the system and subsequent outcomes for children and families. These constraints come in the form of performance regi mes that elevate the importance of policy and procedure (e.g. PerformanceStat) in importance for outcomes such as trust, transparency, and responsiveness (Behn, 2014) . Similarly, organizational leadership may apply constraints and limits on street level discretion, favoring ins tead tools like actuarial risk assessment or the algorithms of predictive analytics to determine workflow and emphasis. Emotional labor, on the other hand, including the display rules of the technology or the logic of the street level bureaucrat as change agent are subsequently ignored. How does the pursuit of accountability for

PAGE 20

20 relatively rare systemic events impact the emotional labor desired on the front line for the majority of children and families impacted by the child protection system? For these str eet level bureaucrats, emotional labor is their dominant work product, but do their jobs ever become highly difficult or untenable because of this quest for accountability? Research Questions In sum, the goal of this research is to further enliven the exp erience of street level bureaucrats engaged in emotional labor within a defined organizational culture that includes bureaucratic constraints. This dissertation queries these questions: How is the context of bureaucratic constraint in organizational cultur e related to the performance of emotional labo r by street level bureaucrats? How i s the performance of other focused emotion management limited or expanded in situations of high bureaucratic constraint? Similarly , is self focused emotion management simply survival for the street level bureaucrat and a disguise for self preservation ? How can bureaucratic constraints contribute to the creation of difficult jobs for street level bureaucrats, thus leading to high levels of emotional exhaustion, burnout and cost ly turnover?

PAGE 21

21 II. LITERATURE REVIEW The foundation for this research takes root from an enduring public administration examination of street level behavior. Debates about the limits of practical rationality of these actors as distinct from econom understand the implications or outcomes of each act of discretion (Simon, 1997) . This literature review explores street level behaviors with this underlying assumption. Further, it highlights the institutional contex t of street level bureaucrats that further bounds behavior. In particular, the review focuses on technologies that require greater degrees of emotion work. The choice to emote, to self regulate, or to create conditions under which others may emote, is an act of discretion. As such, the literature sets the stage to explore the impact of bureaucratic constraints in this area. Finally, this review explores the policy context of child welfare, as current reforms cause work technologies to shift from people pro cessing to people changing. Emotional Labor Emotional labor as a construct was first observed and described empirically by Hochschild in 1983. When observing flight attendants, she began to uncover the unique characteristics of this service industry, wh ere as a condition of employ ment , individuals were charged with tending to the emotional environment of an airplane as their primary job function. Air travel is stressful, and airlines found that the smiling, unflappable, servile but confident , attendant ( observations, which took a sociological bent, birthed an extensive body of research that spanned numerous industries, private and public. Further, this research also took p lace in a variety of disciplines, from sociology to industrial psychology, to public administration. The focus of this inquiry comes from the area of public administration, though as a construct, emotional labor, as

PAGE 22

22 well as its antecedents and consequences , exhibits a great deal of stability across contexts, fields of study, and specific professions and job responsibilities (Hülsheger & Schewe, 2011) . For the public servant, emotional labor is a characteristic of the job, and it occurs in pursuit of a policy goal ( Guy et al., 2008; Mastracci, Guy, & Newman, 2011) . Similar to emotional environment intended to elicit trust, satisfaction, and confidence from the public. Emotional labor takes place especially in those jobs which are crisis driven and emotion laden, such as 911 call takers, first responders, crisis response workers including domestic violence advocates, and human services employees like child welfa re caseworkers and mental health providers. Typically, emotional labor per se takes on two distinct forms, surface acting and deep schchild turned to the theater, where a paradigmatic shift had occurred in dramaturgical methods. In early days, actors were instructed to manipulate their actions, inflections and emotive performances to reflect the intention of the script and director. T the actor. On the other hand, Stanislovski, a 19 th century Russian actor and director, derived a exper ience the character as intended by the writer. In this manner, referred to as the Stanislovski System or Method Acting, the intended emotion is more genuinely felt by the actor and thus results in a more convincing, spontaneous, and authentic performance. incongruent emotion according to display rules of the organization (Grandey, 2003) . For

PAGE 23

23 example, the organization might require a cashier to smile at customers during a transaction (Rafaeli & Sutton, 1990) ( Guy et al., 2008) deep acting (Gelderen, 2016; Kruml & Geddes, 2000) . Conversely, deep acting involves a great deal of emotive effort, and occurs when the worker expresses emotion that s/he genuinely feels in an effort to bring about the type of situation or client experience desired by the organization. In deep acting, the worker then behaves as if they are truly feeling the emotion required by the job, and this results in a more authentic performance for the service recipient. Empirical evidence suggests that service recipients are adept at discerning the difference between surface acting and deep acting, and may even respond more positively to deep acting by the service employee (Zhan, Wang, & Shi, 2015) . The performance of surface acting and deep acting are both important in the delivery of human services. However, there are implications for the balance of th is work on burnout, job satisfaction, and retention of the employee. Often, research in this area utilizes Conservation of Resources (COR) theory to explore the manner in which emotional laborers select from a finite resource pool, and conserve efforts th roughout the day (Gelderen, 2016; Grandey, Foo, Groth, & Goodwin, 2012) . There exists a paradoxical relationship between the resources needed for different types of acting and the consequences of the performance. On one hand, m ore resources emotional labor, the deleterious impact of emotional la bor comes from prolonged surface acting, even though the emotional resources required are less. Time and again, research indicates that the most detrimental consequences of emotional labor result from prolonged surface acting

PAGE 24

24 (Chau, Dahling, Levy, & Diefendorff, 2009; Goodwin, Groth, & Frenkel, 2011; Hsieh et al., 2011; Roh, Moon, Yang, & Jung, 2016; Sloan, 2014) . Similarly, burnout in particula r has been identified to result from prolonged surface acting, and appears to be less about the amount but rather on the ability of the worker to manage the associated emotional dissonance (Brother idge & Grandey, 2002) . This decrease in the authenticity of worker emotions on the job has also been linked beyond just consequences for the job, including decreases overall in Public Service Motivation (PSM), specifically with the distinct motivator of c ompassion (Hsieh, Yang, & Fu, 2012) . Similarly, this inauthentic performance can have a negative impact on wellbeing, specifically de pressed mood and depression symptoms (Erickson & Wharton, 1997) . Increases in surface acting are also related to decreases in affective commitment to the job (Seery & Corrigall, 2009) . Some research proposes a connection between emotional labor and Secondary Tra umatic Stress (STS) in the field of child welfare (Caringi, Lawson, & Devlin, 2012) . efficacy regarding the performan ce. Self (Sloan, p. 277, 2014) . The positive impact of self efficacy is noted empirically in severa l studies of service workers (Guy et al., 20 08; Hsieh et al., 2011) . Self high emotional labor exists (Sloan, 2014) . characteristic of the overarching context. Typically, autonomy is measured by responses on a 1 10 scale to similarly versions of the question abou t the level of control they have over how job

PAGE 25

25 tasks are done (Wharton, 1993) . Autonomy on the job is related to lower levels of inauthenticity (Erickson & Wharton, 1997) . Sim ilarly, emotional labor as a whole is significantly related to positive consequences in situations where individuals have high levels of job autonomy (Wharton, 1993) . P ugliesi ( 1999) discussed the constructs of deep and surface acting within the confines of the audiences, or subjects, of the labor. This slightly different bent delinea tes the categories as labor in human services in that the client focus is important as a work technology, but the own emotions has implication for the closely behaviors. For example, Tu mmers, Bekkers, Vink, & Musheno ( 2015) , illustrate that a particular rules, o r offer personal resources. This construction completely lacks an understanding of the emotive components of work and incorrectly reduces these street level behaviors to a linear, cannot be ignored. This argument is supported by studies on the orientation of organizations, where client oriented firms see frontline workers elevated in status due to their distinct knowledge of clients and longer relationships as compared to others in the organization (Lively, 2002) . This client orienta tion as the emotion work of street level bureaucrats is less understood in the literature, and specifically with regard to the ever present constraints that form the institutional context where they are performed.

PAGE 26

26 Street Level Bureaucracy Theory work to uncover the experiences of street level bureaucrats provides further foundation for this study ( 1980/ 2010) . At the street level of the bureaucratic hierarchy, citizens encounter the government with personal interactions and relationships with government agents. In many cases, these encounters are most tangible government related experience. Each encounter may serve, alongside media venues and popular culture, to shape public values such as trust and confidence in governme nt (Moore, 1995) . Further , street level bureaucrats are afforded discretion and autonomy to conduct their work, which is guided by accountability to both the agencies and clients they serve. In human services organizations, street context is shaped by the techno logies they deliver, whether people processing, people sustaining, people changing, or necessarily multi technology in design (Hasenfeld, 1983) . Their responses and behaviors to these varied contexts play an important role in public management literatures, specifically in personnel management str ategies and performance monitoring and management. Typologies emerge according to organizational situations, and efforts to understand these more nuanced contexts also form an important backdrop to this work. This section explores each of these facets of s treet level bureaucracy theory and applied research. The front line status of street level bureaucrats puts them face to face with citizens. Their focus may be in direct service delivery in law enforcement, social services, mental health, job counseling, or emergency services. They may serve as gatekeepers to citizens seeking access to services, as in entitlement benefits or medical insurance. In the last twenty years, citizens have been increasingly defined as consumers (Powell, Greener, Szmigin, Doheny, & Mills, 2010) or even co producers of services (Bovaird, 2007) . The front line, then, has rapidly become an arena where citizen satisfaction and engagement are tell tales of policy effectiveness .

PAGE 27

27 (Prottas, 1979) , activities at the front line take on additional weight. Personal, direct contacts characterize work designed to form both short and long term relationships with citizens (Maynard Moody & Portillo, 2010) . Ultimately, street level bureaucrats are called to be distinctly un bureaucratic in the traditional, efficiency focused, Taylor esque impression elicite illustrating their need for flexibility and responsiveness. A balance must be maintained between d uty to the agency and duty to the citizen. Instead of mechanic al repetition, then, individuals must exercise fluid discretion in everyday work at the street level. This inherent discretion further shapes the work and the desired outcomes. In addition to the adaptations and strategies identified by Lipsky (1980/2010), street level bureaucrats are known to go above and beyond in heroic services to some citizens, even at the expense of others or themselves (Maynard Moody & Musheno, 2003) . Others have purpose with program or p olicy . As they purs ue their understanding of justice, street level bureaucrats exercise discretion to make choices (Kelly, 2005) . The degree to which discretion should be limited or enabled at the street level is an ongoing debate in public administration scholarship. These limits on discretion come in the form of policies, mandates, performance goals, and outcomes measures (Radin, 2006; Wilson , 2000) . Limits on discretion only go so far in limiting the flexibility afforded the street level (Maynard Moody & Portillo, 2010) . It does not make sense that every decision, small, minor, or otherwise, would be observed and regulated by someone higher in the hierarchy, even in the name of

PAGE 28

28 accountability. Fur ther, the limited resources afforded to the street level worker demand hard to measure choices for and against citizens: whom to help and whom to ignore (Calabresi & Bobbitt, 1978) . Given that emotional labor can smooth interactions with citizens, it is important to understand the bounds, if any, of autonomy in the performance of emotion work, particularly e emotional resources. These organizational bounds are largely ignored in the literature, with the focus instead on bounds within the individual or according to the situation or client. Irreducible discretion and autonomy are also highlighted when conside ring the final characteristic of street level bureaucracy theory: ultimate policymaking. Full understanding of policy implementation remains a challenge in the field of public administration (Bardach, 1977; Pressman & Wildavsky, 1973; Robichau & Jr, 2009) . The street level bureaucrat is often the enforcer, gatekeeper, and/or vessel of service delive ry. As such, the discretion and autonomy afforded these individuals leads to the conclusion that they are the ultimate policy makers (Brodkin, 2011; Meyers & Vorsanger, 2007) . This commonly recognized idea of bottom up implementation is not without influence, of course, from the top down. Managers, jurisdictional characteristics, and politicians have all been found to have influence, albeit limited in some cases, on actions at the street level (May & Winter, 2009 ) . Beyond mere influence, how can accountability be ensured for these actors? Scholars and public managers alike shape responses for this question. In repeated new emergence of management eras, from New Public Management to Managerialism, researchers expl ore different ideas of accountability (Behn, 2001; Hood, 1991; Radin, 2006) . From these efforts emerge various strategies for performance management and ideas about administrative control (Lipsky, 2010) . In the 30 th anniversary edition of Street Level Bureaucracy , Lipsky tackl es this

PAGE 29

29 idea o The Assault on Human Services: Bureaucratic Control, Accountability, and the Fiscal Crisis for street ue to high levels of discretion and because accountability must simultaneously be to the agency and the client (2010, p. 159). While Lipsky acknowledges that accountability at heart is the relationship between people and groups, he does not explore the con sequences of this emotive dilemma in great detail. Specifically, street holds true in the emotional realm, I argue that street level burea u crats must weigh their emotion work on this balance between the organization and the client. Perhaps the impossibility of this task is the mechanism for the burnout Lipsky later observed: Thus, generations of thoughtful and potentially self sacrificing people are disarmed in t heir social purpose. They come to believe that it is impossible to find conditions conducive to good practice, and that the public agencies cannot be otherwise structured. Their choices appear to be to leave public service for other work or to resign thems elves to routine processing of clients. . . (2010, p.186). Impact of Organizational Culture on Street level Bureaucrats The impact of the organization as related to human behavior was also manifest in a formative, classic, parallel Friedrich/Finer debate . This debate turned largely on the understanding of human, specifically bureaucratic, behavior. The exchange began with responsibility entrusted to them to devel op and apply specialized knowledge in the interest of the public (Friedrich, 1968 /1940 ) . Finer responded by asserting that mutual professional responsibility was not a sustainable practice, and that performance of bureaucrats must be preserved through structures that ensured enduring responsiveness (Finer, 1941) . History reveals that neither of these scholars was entirely correct (Stewart, 1985) . However, echoes of both sides

PAGE 30

30 take shape in contemporary methods for performance management and a dministrative reform group processes of goal setting also emerged. A rguably, when seeking accountability for rules. This is particularly true in the child welfare field, where multidisciplinary child fatality review teams acro ss the nation review child deaths. Following reviews of child fatality cases across jurisdictions, these teams commonly recommend changes in the assessment process for child welfare, and specifically those surrounding assessment of safety and risk (Douglas & Cunningham, 2008) . level discretion would indeed resolve the proble ms at hand. This, however, is not the case for the former or the latter. In the context of child welfare and work that is dominated by emotional labor, perhaps was d managing to higher level human needs, similar to the hierarchy of human motivation, as was emerging in psychology at the time (Maslow, 1943) . The traditional Theory X, said McGregor, was inferior in that it focused on basic needs and utilized a carrot and stick approach to ( McGregor, 1957 ). The human side of enterprise is overtly relevant to human services organizations. The technologies employed by these organizations are primarily concerned with clients, patients, or citizens, as in the case of public administration. Hasen feld sets forth a taxonomy of technologies

PAGE 31

31 employed by human services that is useful for this inquiry. In short, three types of functional human service technologies are utilized in these organizations: 1) people processing, 2) people sustaining, and 3) pe ople changing. Each technology has a distinct purpose, a certain level of demand for emotional labor by the street level bureaucrat, and different implications for the organization and those receiving services from the organization. To further complicate t his taxonomy, the technologies employed by human services organizations are often intermingled, conflicting, and not based in current science (Hasenfeld, 1983; Sandfort, 2003) People processing was of initial concern in early street level bureaucracy literature (Prottas, 1979) and in sociology ( Hasenfeld, 1972) . The primary purpose of this technology is to Hasenfeld, 1983, p. 135). This technology is at play in many public administration functions such as licensing and inspection, eligibility determinations for social security, and assessment and investigation of abuse and neglect of vulnerable populations. Even the lowly speeding ticket may be considered the employment of people processing. The people processing technology has implications for the citizen, who is granted a label and all the accompanying social implications of such. The processing role and su bsequent citizen experience is explored in current day when considering (Epp, Maynard Moody, & Haider Markel, 2014) intended to enabl e or disable client behavior. From an organizational perspective, the locus of control for the agency is imparted by the boundaries it designs in the community or service population. Staff and client relationships are minimal, in that they are primarily f ocused on gathering applicable information and instruction

PAGE 32

32 related to the boundaries developed. Typically, these are short term encounters designed with efficiency as a priority to prevent backlog ( Hasenfeld, 1983) . Bureaucratic, administrative controls may be of the most assistance in this area, as the technology is conducive to manualized delivery. Managers in this technology are concerned with outputs and dispositions. It is not surprising, perhaps, that performance management strategies that focus primari ly on outputs emerged in agencies with substantial people processing roles [e.g. CompStat as implemented in policing and city management (Behn, 2003) ]. People sustaining technologies are concerned with those situations where relatively low capacity for change by the individual has been determined ( Hasenfeld, 1983) . Examples of these technologies are nursing homes, long term care facilities for the mentally ill, longer term residential tr eatment for youth, and prisons where individuals carry out long sentences. These technologies are concerned primarily with what Hasenfeld calls attribute stabilization (1983), or achieving and maintaining a particular desirable status (e.g. safe from onese lf or to others, fed, healthy, clothed, etc.). The most paternal of services, people sustaining technologies are concerned with custodial care, providing sustenance, and maintaining equitable allocation to similar individuals and situations. Similar to peo ple processing, people sustaining technologies use threats and promises to control clients, and bureaucratic, administrative controls to adequately ensure staff compliance with policies and procedures. Typically, there is great insolation within the instit ution from outside pressures, and the tightest of control over service recipients. Trends toward deinstitutionalization and its replacement, community based care have reduced the usage of these types of human services organizations (Bachrach, 1983) . Similarly, the The Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA), or

PAGE 33

33 welfare reform, took long term eligibility for financial benefits to alleviate poverty, which had onc e been considered people sustaining, and oriented it as a short term solution with a focus on employment (Blank, 2002; Hasenfeld, 1983) . Similar changes occurred in child welfare, which will be discussed in greater depth in a bit. Suffice to say, these p rimarily people sustaining organizations have been effectively thrust into a situation where the emphasis is on for people changing technologies, which require significantly different organizational and practitioner characteristics and skills. People chan 1983, p. 140). This orients the organization and the practitioner toward the objective of p artnering with clients to facilitate this change in the selected attribute(s). For example, in education, the focus is on the development of the student, and in mental health rehabilitation, the preparation of individuals to return to the community. In thi s context, managers must focus on the desired outcomes of planned change as result of intervention. Staff must be committed to the technology, according to Hasenfeld, and are most effective when they exhibit professional judgement, are recognized for their expertise, and are allotted legitimate autonomy (1983). The difficulty with this reality, however, echoes the dilemmas of accountability reflected in SLB literature. Ethnographic studies in the field reveal that the structuration of this technology takes on extensive variability, so much so that it renders the actual work of the organization ambiguous (Sandfort, 2003) . With that caveat, there exists little doubt that this effort to change attributes in cl ients takes on a considerably different tone than processing or sustaining. People changing requires a strong, dynamic, at times pro longed, relationship between the street level bureaucrat and the

PAGE 34

34 client. This is particularly true when the magnitude of t he change is great (Hasenfeld, 1983). I argue that to cultivate this relationship, emotional labor is required, and that people changing in general is primarily emotional labor as opposed to cognitive or physical labor. Hasenfeld acknowledges this shift as well when describing the evolution of human services delivery, workers and clients, we need to understand the environmental and organizational factors that sha (Hasenfeld, 1992) . Table 1.1. Hasenfeld's (1983) Typology of the Functions of Human Services Technologies People Processing People Sustaining People Changing Type of Product Alte red status Attribute stabilization Attribute change Core Activities Classification disposition Custodial care; sustentation Planned change Managerial Concern Disposition of product Acceptable allocation rules Demonstrative effectiveness Staff Client Relations Minimal Modest Extensive Client Control Threats and promises Threats and promises; reinforcement control Persuasion; Reinforcement control Staff Compliance Bureaucratic Bureaucratic Commitment Human service organizations rarely fit into only one column of technology delivery, and often, these technologies are muddled in practice, scope, and definition (Sandfort, 2003) . As noted earlier, policies and trends in governance may intermingle th ese technologies creating

PAGE 35

35 purpose in highlighting the management and organizational characteristics conducive to different types of human services work, and in further amplifying the impact of these characteristics on the SLB. Further, the emphasis in this typology on the types of relationships (and thus, the levels and var ieties of emotional labor) required for each technology will serve as a springboard for discussion of technologies in child welfare, the specific context for this dissertation. Recognizing the human as the primary vessel for work technologies also helps e xamine the possible impact of jobs demanding a balance of other focused and self focused emotional labor. Designing an organization where display rules are clear and also congruent with the general thoughts and feelings of the workforce in one area where r factory were to conduct their operation in a shop window, necessitating the public display of emotions that would build, say, trust in t heir brand or inspiration in the polity, the meaning each incongruous to the organizational display rules would not last long. In this manner, job satisfaction wo with the organizational intent. Public administration, or course, is not conducted behind glass. Metaphors aside, street level bureaucrat s must interact with citizens, be re sponsive to needs and varied contexts, and maintain nimble levels of relational focus on the emotions and behaviors of citizens. The relational elements of accountability that Lipsky (2010) identified may be ignored at the expense of all involved, as can t he requirement for relationship in the delivery of people changing technologies. However, the scales may be also be tipped in support of other focused emotion

PAGE 36

36 work and the direction of relational accountability. Consider an example from Sweden, where natio nal differences were identified from observations of safety inspectors in both the US and Sweden: American inspections are designed more as formal searches for violations of regulations; Swedish inspections are designed more as informal, personal missions to give advice and information, establish friendship ties between inspector and inspected, and promote local labor management cooperation. (Wilson, 2010, p. 296). Emotional labor is the job task in the Swedish example, where the focus is on establishing managing emotions and relationships to accomplish workplace safety. This then, has more flavor n contrast, the observations of American inspections seemed more concerned with people processing, or defining client rel ations (Hasenfeld, 1983), and thus, less emphasis on emotive requirements of the job. Discretion in this case is much more about determining compliance of the client, and controls are based on threats and rewards, or the carrots and sticks of Theory X to a chieve policy goals. The organizational context of people changing organizations is a common area of study in industrial psychology. In general, this literature examines the impact of organizational culture on the delivery of human services, as well as th e resulting client outcomes. While the differences and similarities of culture and climate as well as the merits of each are debated in the literature, this review considers organizational culture from a sociological perspective. Organizational culture is consequences for compliance and deviance (Schein, 2010; Verbeke, Volgering, & Hessels, 1998) . Culture has been constructed to define the level of constraints at hand in an organization,

PAGE 37

37 (Glisson et al., 2008, 2012) . Rigidity reflects the degree to which individuals may exercise discretion, wher e they must ask permission from superiors, or the flexibility in following rules, policies, and procedures in the agency. rigidity of culture to resulting outcomes for children (Glisson et al., 2012) . This impact has been noted in acts of discretion as well, where organizational norms rather than needs of clients were found to be more influential in caseworker decisions about service and placement (Martin, Peters, & Glisson. , 1998) . Outside child services, discretion for emotion work was also found to be related to the organizational culture and climate of emergency room implementing a family centered intervention with a strong emotional support component. (Hemmelgarn, Glisson, & Dukes, 2001) . When organizations emphasized emotional support to families and patients as the pr imary mission (versus the more technical, life saving functionality of the emergency room), workers were more likely implement emotion work as a primary part of their job. Similarly, workers also reported less negative impact from their work on their perso nal well being in this type of culture (Hemmelgarn et al., 2001) . These studies provide support for the inquiry at hand, but do not spec ifically address the rigidity of organizational culture as connected to the performance of emotion work . Public Child Welfare as an Institutional Context for Practical Emotion Work This study will use public child welfare as the institutional context for street level workers engaged in emotion work. This setting is particularly relevant in the current era of constant reforms in child welfare systems, including debates about the organizational technologies best employed by child welfare agencies at the time of assessment and early intervention with families alleged to have maltreated their children. Historically, front line

PAGE 38

38 processing role. The primary work product for th processing technology, in that the work was to seek out the reported family and child, conduct an assessment or investigation of the family situation, and determine whether or not child maltreatment occurred, us ually with a preponderance of the evidence. Additionally, workers assess continuing danger to the child, and make the decision to remove the child weighed in light atment determine the course of action for agency intervention, including access to on going services. Further, a label was applied to the person found responsible for abuse or neglect, and that label is entered into a state central registry. In many states , potential employees can access c entral registries, with many states having rules about future employment in child or vulnerable adult serving agencies based on prior findings of child maltreatment. This emphasis on triage for the child welfare investiga paradoxically in response to a changing outcomes focus by the child welfare field. The onset of New Public Management (NPM) demanded a reexamination of the logic models and technologies of many human services agencies (Hasenfeld, 1992; Hood, 1995) . In child welfare, the subjectivity of findings and their relative impotence for promoting change in parental beh avior were identified as major difficulties (Kohl, Jonson Reid, & Drake, 2009) . A child welfare r eform adopted steadily in the United States since that time is Differential Response. Differential response in child welfare provides that in cases of low and moderate risk for child safet y, an alternative response may be used by the front end caseworke r. Rather than conducting an investigation to determine whether or not child maltreatment occurred, workers in an alternative response would instead assess child and family functioning to determine needs and strengths that might be reduced or augmented, re spectively. The philosophical shift was related to a practical

PAGE 39

39 belief that families might be more willing to engage with the child welfare worker in a less adversarial setting. Following initial assessment, the alternative response caseworker is instructe d to seek family engagement in making necessary changes to reduce risk factors for future child maltreatment. This engagement is voluntary on the part of the fam ily, according to model design (Merkel Holguin, Kaplan, & Kwak, 2006) . Differential r esponse as chi ld welfare reform, then, effectively changed the technologies to be used by the street level bureaucrat in an alternative response. Essentially, this change shifted the balance of the work from processing to changing , where the emphasis of the work is on r elationship with clients as the primary intervention of the agency. Training modules for caseworkers new to the reform focus on this change in technology, they are told You are the . Further, a mantra of th e reform is Safety through engagement , or the ability to achieve the desired outcome of child safety by developing family engagement and relationship in initial contacts . Reactions to this reform in the child welfare system vary from wildly enthusiastic to cautionary, and may be reflective of the fundamental change in the service delivery technology . Misunderstandings abound as to the intent of such reform, and often focus on the need to hold caseworkers and families accountable. Critics cite the volunta ry nature of the se services as a primary issue. A particularly vocal critic of differential response comes from a law professor, procedural fairness and court oriented interventions rather than relationship oriented approaches (Bartholet, 2014; Bertelli & Lynn, 2006) . Despi te criticism, advocates for Differential Response also abound, and agencies continue to expand and implement the practices. Further, in many

PAGE 40

40 jurisdictions, there exists renewed emphasis on family engagement, no matter the type of initial response or stage of service (Winokur, Ellis, Drury, & Ro gers, 2015) . Critics and advocates of child welfare reform often conflate the concept of family engagement with attitudes about family preservation and child safety. Indeed, caseworkers who express a proclivity for family preservation may exhibit differen ces in decision thresholds to take action in the course of casework (Dettlaff, Graham, Holzman, Baumann, & Fluke, 2015) . Since these individual decision making thresholds occur in the context of varying organization culture and climate, as wel l as related to case and family characteristics, this area of study is considered making (Baumann et al., 2011) . Child welfare in a time of reform was selected due to th e new emphasis on family engagement, or what I argue is a type of other focused emotional labor. Successful performance of other focused emotional labor results in increases in performance, positive service outcomes, and job satisfaction on the part of the performer. Arguably, these results might relate to increases in retention of child welfare services. Yet, for workers on the front end of the child welfare system, turnover remains an utmost concern. Nationally, the turnover rate for child protective serv ices workers is 22%, or roughly six times the national average (American Public Human Services Association, 2005) . The consequences of turnover are taxing on an already resource depleted system: retraining, abandoned caseloads, and significant impact on families served from discontinuity of care. Remaining colleagues must carry the bulk of this burden, which in turn may lead to more b urnout and subsequent turnover (DePanfilis & Zlotnik, 2008) . Conversely, self focused emotional la bor, or the masking of felt emotions and surface acting, is r elated to burnout, emotional exhaustion, and turnover. When the technology on the

PAGE 41

41 front end of the child welfare system changed from people processing to people changing with increased demands fo r emotion work, did the organizational contexts change with it? W hat is causing frontline workers to turn inward and burn out? Could this be a manifestation of self preservation and conservation of emotional resources? Are bureaucratic constraints imposed for process accountability limiting the ability of street level bureaucrats to perform emotion work? Based on the questions posed in this inquiry and this review of the literature, the next chap t er delineates five hypotheses for investigation, outlines th e operationalization of constructs, and describes the methods employed for testing the hypotheses.

PAGE 42

42 III. HYPOTHESES AND RESEARCH DESIGN This chapter provides a description of the research questions explored in the literature. The overarching question as well as the more specific sub questions, are addressed through five hypotheses that address the constructs of interest as operationalized and informed by the literature. As identified in chapter one, the overarching research question of this inquiry is : How is the context of bureaucratic constraint in organizational culture related to the performance of emotional labor by street level bureaucrats? This question is examined using three sub questions: (1) How is the performance of other focused emotion management limited or expanded in situations of high bureaucratic constraint? (2) How i s self focused emotion management survival for the street level bureaucrat and a disguise for self preservation ? and (3) How can bureaucratic constraints contribute to the creation of job stress for street level bureaucrats, thus leading to high levels of emotional exhaustion, burnout and costly turnover? The policy and practice area is child welfare, where street level bureaucrats work with families identified in the child welfare system who meet the legislative and jurisdictional cri terion for formal intervention, services, and out of home placement. Essentially , caseworker jobs are heavily weighted toward fostering non adversarial relationship s with families on their case load to assess child safety, family needs and strengths, and to aid in deliver ing or arranging services to address family concerns. There also may be variability i n emphasis on family engagement, family preservation , and child safety based on the stage of the case , the direction of the agency, or the needs of children, youth, and families . As noted in the review of literature, the success or performance of this group can be measured both federally and locally through the recurrence rates of families in the child pr otection system or of children in out of home placement due to safety concerns . Thus, this group also meets the one of the main aims of

PAGE 43

43 agencies delivering people changi ng human services technologies: the need to foster relationships in an effort to cultiv ate changes in human behavior in pursuit of a policy goal. Hypotheses The main research question for this dissertation examines the context of bureaucratic constraint in organizational culture and the relationship to the performance of emotional labor by street level bureaucrats. The first sub question, designed to elicit answers regarding this context is: How is the performance of other focused emotion management limited or expanded in situations of high bureaucratic constraint? When street level bureau crats experience depletion of resources, autonomy , and discretion, the literature notes that adaptive behavior is a possible may occur , or strategies and work routines that focus on the easier or more expedit ious routes of service delivery (Brodkin, 2011). There lies a key difference between other focused and self focused emotional management, in that other focused emotional labor demands deep acting, and thus time and resources to effectively engage with clie nts and citizens, either in seeking contact, or establishing relationship. Self focused emotional management, on the other hand, does not require relationship, and is a more efficient emotional choice for the street level bureaucrat. Further, it operates w ithin the relative vacuum of an individual psyche, thus conserving resources and effort. Consequently, focusing on self as compared to the other can be considered the emotional equivalent of selecting the easier or more expeditious behavior in the face of constraints. This line of thinking lends to the following hypothes e s: H1: Bureaucratic constraints will be negatively related to authentic expression of emotion. H2: Bureaucratic constraints will be negatively related to deep acting.

PAGE 44

44 Similarly , the second sub research question explores whether self focused emotion management is simply survival for the street level bureaucrat and a disguise for self preservation ? To further narrow street level behaviors within self focused emotional management, adapt ive behavior might point to false face or surface acting as a primary strategy in the context of organizational constraint. That is, deep acting may not be possible in the time afforded, and may not be valued by the agency as much as basic display rules. T hus, the following hypothesis: H3 for self focused emotion management. Finally, at the practical level, a sub research question is: How can bureaucratic constraints contribu te to the creation of job stress for street level bureaucrats, thus leading to high levels of emotional exhaustion, burnout and costly turnover? The final two hypotheses point to an important crux, where the coexistence of organizational constraints implem ented to shape street level behavior for other focused emotion work actually have the opposite impact . The consequence, then, is an untenable position for the street level bureaucrat . Self efficacy for the worker centers on a belief that s/he is successful at emotion work (Heuven & Bakker, 2006) . The following hypotheses highlight that autonomy and discretion might have finite lower bounds for street level bureaucrats, where emotion work is focused on self preservation and survival as opposed to service delivery. This work would then also be less satisfying, more burnout inducing, and might result in the intent ion to quit the work entirely. to both the client citizen and the street level bureaucrat. H 4 : Bureaucratic constrain ts will be negatively related to self efficacy in the performance of emotion management .

PAGE 45

45 H5: Bureaucratic co nstraints will be positively related to expression s by street level bureaucrat s that they intend to quit their position. Research Design This research utilized a data collection process that was part of a larger study of workforce needs and intervention s in the United States. Due to the magnitude of the larger effort, that project obtained all data from sites across the country under protocols approved by multiple institutional review boards, including those of project partner universities as well as the indiv idual state child welfare agencies involved . Under these protocols, participants in the overall study completed a survey containing the items of interest for this dissertation alongside a culture and climate measure discussed later in this chapter. T hese instruments were completed in local agencies during work hours, with the agreement of supervisors and agency leadership. For those individuals who were unavailable on the day of administration, the project offered an on line opportunity to participate . For the purpose of this research, staff of the larger project recoded the data to de identify individuals as well as their specific agency, team, and supervisor prior to sending the data to the author . Upon receipt of the data set, data were cleaned to exclude 8 units of non child welfare workers who participated in the culture and climate portion of the research in smaller agencies. Then, culture and climate data were merged with survey data on elements of interest for this study. Thirty eight cases wer e dropped when they failed to merge on their unique identifier. This failure was due to a flaw in survey collection, where unique identifiers of individuals who participated on line were not saved in the Qualtrix application. Finally, because of the neste d nature of the analysis in this dissertation, units with only one participating worker were dropped.

PAGE 46

46 The participants illustrated in Table 3 .1 reflect the participants whose responses are included in this study. Participants The sample of child welfare c aseworkers in this study (n=345) were employed at one of 17 agencies participating in a large workforce development initiative in one eastern state . These caseworkers were all part of child serving, case carrying units within the ir agency. At the time of s urvey administration, the larger project goal was to achieve as close to the population of those workers within the jurisdiction s as possible, with the intent to capture sufficient information related to organizational culture and climate . The 17 participa ting local agencies nest within a state supervised, locally administered child welfare system. The 1 7 agencies range in size from Table 3.1. Sample Agency # of units Total Casew orkers Original estimate Inclusion and Response Rate A 6 18 27 67% B 1 6 6 1 00% C 4 20 29 69% D 27 106 139 7 6 % E 2 8 10 80% F 4 21 30 70% G 1 3 13 23 % H 1 7 7 100% I 3 8 25 32 % J 9 37 53 70 % K 2 16 16 100% L 7 41 51 80 % M 1 8 8 100% N 2 13 13 100% O 1 3 4 75% P 2 17 17 100% Q 2 13 13 100 % Total 75 345 461 75 %

PAGE 47

47 fo ur to 139 child welfare workers, and serve a variety of populations, including rural, mid sized, and urban. Table 3 .1 illustrates particip ation by agency, which varied from 100% to 23% for an overall response rate of 7 5 % . Participants ranged in age from 22 70 years old, with a mean of 39 years. Eighty seven caseworkers stated their undergraduate degree was in social work, followed by 85 who reported their degree was in Psychology. Twenty eight caseworkers reported they had earned an advanced degree in soc ial work, and 50 reported advanced degrees ranging from criminal justice to psychology. Similar to national trends in child welfare casework, 87% of the sample reported identifying as women. Half of the caseworkers reported they were Caucasian/white, follo wed by Table 3.2 . Demographic Characteristics of Caseworkers Frequency Percent Gender (Missing=3) Female 298 87.1 Male 44 12.9 Race (Missing=0) African American/Black 129 37.4 Caucasian/White 188 54.5 Asian 5 1.4 Native American 1 0. 3 Biracial/Multiracial 20 5.8 Othe 1 r 2 0.6 Ethnicity (Missing=0) Hispanic/Latino Origin 29 8.4 Undergraduate Major (Missing=14) Counseling 5 1.4 Criminal Justice 18 5.2 Human Services 20 5.8 Other 61 17.7 Psychology 90 26.1 Social Wo rk 99 28.7 Sociology 38 11

PAGE 48

48 35% who identified as African American/Black. The rest reported race or ethnicity as Hispanic, biracial/multiracial, Native American, or Asian. A full report of demographics may be found in Table 3.2. Operationalization of Theo retical Constructs The performance of e motional labor is operationalized via two constructs: the emotive state (other focused). Indicators of self focused emotional la bor include masking of true Indicators of other focused emotional management are work to calm or deescalate clients and work to help others emote in a desirable manner according to desired outcomes (Guy et al., 2008) . Deep acting occurs when the laborer truly feels the emotion necessary for a certain situation (e.g. unconditional positive regard as in humanistic psych otherapy). These constructs are organized in Figure 3.1 an d correspond with individual measures. Figure 3.1 Operationalization of Emotional Management B ureaucratic constraints may be operationalized via three constructs: limited resources, limited aut onomy , and limited discretion (see figure 3.2 ) . Limitations to resources are illustrated by concern over liability in the work, which is an individually perceived constraint, as is the inability to do a complete job in the time allotted. Limitations to aut onomy are expressed in the Indicators Constructs Concept Emotional Labor Other focused Emotion management Deep acting: Helping others emote, relationship buidling Self focused Pretending expresssion ('false face') Inauthentic expression

PAGE 49

49 climate measures of stress in the OSC, with attention to role conflict and role overload, specifically. Finally, limited discretion is operationalized as organizational rigidity as indicated by formalization and centralization of decision making and authority. Figure 3.2. Operationalization of Bureaucratic Constraint Burnout, low self efficacy and expressed intent to quit may all be consequences of stress in the work environment, and have strong relationships to turnover in th e child welfare literature (Mor Barak, 2001) These concepts, constructs, and indicators are illustrated in Figure 3 .3. Figure 3.3. Operati onalization of Turnover Intent Measures The research design elicits understanding of the key constructs as manifested in the study sample. Given the primary research questions as related to the dependent variable of emotional l abor performance, the survey included selected items from the G uy, Newman, Mastracci (GNM) Emotional Labor Questionnair e (Guy et al., 2008) . Researchers have administered the Indicators Constructs Concept Bureaucratic Constraints Limited Autonomy Role conflict Role overload Limited Discretion Formalization Centralization Concerns for liability Indicators Constructs Concept Turnover of SLBs Employee expression Intent to quit Employeee experience Low self efficacy

PAGE 50

50 GNM internationally as a measure of emotional labor performance in public service. The survey included four constructs: Pretending Expression/False Face, Authentic Expression, Emotional Management, and Deep Acting. Constructs and known alpha coefficients were obtained from a larger pilot of these scales. Pretending Expression/False Face captures self focused emotional labor to mainta in assured original alpha coefficient for this measure is .746 (Guy et al., 2008) . For this study, the alpha coefficient was .61 9 . Authentic Expression addresses the manner in which workers are able to emote without when working with clients/customers original alpha coefficient for thi s measure is .638 (Guy et al., 2008) . For this study, the alpha coefficient was .647. Emotion Management addresses other focused emotion management, where wo rkers express the importance of a skill for assisting others in emoting according to the professional I am good at getting people and

PAGE 51

51 original alpha coefficient for this measure is .710 (Guy et al., 2008) . For this study, the alpha coefficient was . 81 4 . Deep Acting is measured by items that center around the effort necessary to sincerely feel items, I need to show to client at the time of administration for this construct. For this study, the alpha coefficient was .842. Self efficacy is related to performance of work tasks and people changing technology . The survey con tained a scale derived from a state training system, and contained child welfare casework related successes. Caseworkers responded to five items on a Likert scale of one through five, with one representing Never and five representing Always. The items for this scale n helping children and families educational well expressions of gratitude from children Following item reduction, Intent to quit was measured through five items items distributed throughout the survey instrument. Each were scaled on 1 to 7 scales that ranged in choi ces related to the specific question (e.g. not at all likely to very likely and terrible (chances) to excellent (chances)) and

PAGE 52

52 look for a new job in the next yea often used as a proxy in the absence of longitudinal data (Auerbach, Schudrich, Lawrence, Claiborne, & McGowan, 2014) . After item reduction, for all the items was .9 48 . Key organizational ele ments were measured through the Organizational Social Context (OSC) Questionnaire (Glisson et al., 20 12) . The OSC is a nationally normed instrument in both and climate constructs, but this research is concerned with those that exemplify bureaucratic constrain t. Rigidity and role stress were of primary concern. Rigidity within the unit, as measured by the subscales of formalization and centralization (seven items each, measured on a Likert scale ranging from strongly agree to strongly disagree), examines the c ulture where caseworkers conduct their work. Formalization measures the need for procedural compliance in everyday work, where creativity is restricted in favor of a step by every piece of Alpha coefficients for these scales are .6 1 for centralization and .69 for formalization (Glisson et al., 2012) . Stress within the unit, as me asured by the subscales of role conflict and role overload , captures the degree to which caseworkers understand their positions and roles and can operate freely within them. The OSC measures stress as the overall climate within a unit, as opposed to tradit ional measures of stress at the individual level. The seven items on role conflict, or being

PAGE 53

53 ith time and resource to prior administration in 81 child welfare a gencies in the United States. This produces a unit level T score for each overall concept and construct. The alpha scores for these items are .83 for role conflict and .82 for role overload (Glisson et al., 2012) . On an individual level, the survey assessed caseworker factors related to unique aspects of child welfare work using portions of a general st aff survey designed for caseworkers (Dettlaff et al., 2015) . Specifically, scales related to caseworker concerns for liability, time constraints related to workload, and an index of decision making thresholds were included in the survey. Concerns for liability are caseworker perceptions that he or she may be disciplined or fired if a child on one of their cases is harmed over the course of or after the agency intervention. It also examines worry over media attention and att ention to others that have been disciplined or fired in the course of their work. This real possibility is a constant backdrop to the high stakes and high accountability realm of child protection work. However, the scale also includes items that speak to t agency leadership if the worst were to occur, including the likelihood that the agency would conduct a thorough investigation of the situation prior to assigning blame to th e caseworker. After conducting item reduction, t 828 . The variables developed for this analysis are displayed in Table 3.3 , with their corresponding alpha coefficients.

PAGE 54

54 Table 3.3 . Variables Used in the Analyses 1 Cronbach's Alpha False Face 0.619 I hide my true feelings so as to appear pleasant at work In my job I act confiden t and self assured regardless o f how I actually feel I wear a "mask" in order to deal with families and children i n an appropriate way Authentic Expression 0.647 I let my true feelings show when working with families and children It is easier for me to show my true feelings than to pretend I am good at expressing how I feel. Emotion Management 0.814 I am goo d at getting people to calm down Dealing with emotionally charged issues is an important part of my job In my job, I am good at dealing with Emotional Issues * Deep Acting 0.842 I try to actually experience the emotions that I must show to families and children I work hard to actually feel the emotions that I need to show to families and children I work at developing the feelings in side of me that I need to show families and children Concerns for Liability 0.828 I believe I have had adequat e training to help me make the right decision about the safety and well being of my clients* I am worried that one of my cases may draw media attention.* I have known caseworkers that have been disciplined or fired because of real or perceived mistakes .* If a child in one of my cases is harmed, I believe the agency will conduct a thorough investigation into what happened before assigning blame I know my supervisor will be supportive of me and the decisions I made if a child is harmed in one of my ca ses I know my leadership will be supportive of me and the decision I made if a child is harmed in one of my cases Success/Self Efficacy 0.832 H ow often do you feel you are successful in helping families make positive changes in their lives

PAGE 55

55 Table 3.3 cont d How o ften do you feel you are successful in helping children and families achieve permanency How often do you feel you are successful in helping children and families remain safe How often do you feel your efforts have contributed to the physical, mental, and educational well being of children and families How often do you receive thanks or expressions of gratitude from children and families you have worked with?* Intent to Quit I will probably look for a new job in the next year * N/A How often do you think about quitting your job? * What are the chances that you will leave your current job for a job OUTSIDE the agency sometime in the next six months ? * What are the chances that you will leave your current job for a job OUTSIDE the agency sometime in the next year ? *Asterisks designate items excluded from analysis following dimension reduction. 1 The OSC items are proprietary , but a full description including reliability and validity analysis of factors is available in Glisson et al., 2012 . Factor Analysis In SPSS, construct validity was establish ed using Principal Components Analysis (PCA) with Varima x rotation with Kaiser normalization. This approach was appropriate for this data set given the ordinal nature of the items, the size of the sample, and established suitability for data reduction of a ll items . Each set of items was non There were also no significant outliers in this data set. Emotional labor factors loaded on originally intended constructs, similar to past analysis using these items. Th on self efficacy at .558, which was lower than all other items in that factor (.885, .752, and .669,

PAGE 56

56 respectively) . This item loaded lightly on authentic expression, at .286. Conseq uently, this item was dropped from analysis. Concerns for liability loaded into two separate factors: worry and support. For the worry support for worker discre tion and fairness of determining accountability in light of a child being factor were dropped from analysis. Items in the success/ self efficacy scale did not divid e into factors. However, the item How often do you receive thanks or expressions of gratitude from children and families you have work had a considerably lower value after rotation than the other items in the scale. Consequently, this item was dr opped from the analysis. Finally, the items for Intent to Quit were analyzed. The questions were quite similar and all items loaded on one factor. After rotation, items ranged from .684 and .914. After dropping the two lowest items, the remaining items ex that you will leave your current job for a job OUTSIDE the agency sometime in the next year This variable was somewhat skewed, with most respondents stating they did not intend to quit their position so . This variable was transformed into a binomial distribution, with 1 4 being 0 and 5 7 representing at least some desire to quit ones job in the subsequent six months. Analy tic Technique Each of the five hypotheses were examined using either Hie rarchical Linear Modeling (HLM) or Ordinary Least Squares (OLS) regression . Because survey participants were clustered

PAGE 57

57 in units, and the v ariables of rigidity, role conflict, and role overload were measured at the unit level, data are clustered in such a manner that regression i ntercepts may also be clustered. This clustering violates assumptions of linear regression and can lead to erroneou s conclusions should OLS be utilized in those instances (Bryk & Raudenbush, 1992) . Consequently, each dependent variable was firs t test ed using a null hypothesis to see if its interclass correlation coefficient (ICC), or variance within units, was sufficient to warrant the use of HLM. If the data were sufficiently clustered, hypotheses were tested using a homogen e ous and heterogeneo us full random coefficients model, where both level one and level two contain independent variables [ e.g. rigidity at the unit (level 2) and liability (level 1) as potential independent variables related to false face acting] . For those dependent variables without sufficient variance by unit , OLS was used to test each hypothesis. To address these hypotheses, thirteen variables were developed using data from an organizational culture and climate survey and a survey containing the variables of interest for th is inquiry. See Figure 3.4 for an illustration of the analysis plan for these data, with level one and level two independent variables (bureaucratic constraints) and the dependent variables in each hypothesis . Fi gure 3 .4 Variables for Analysis Variable Na me Description Uses by Analysis Unit This is the de identified unit number for each unit. Grouping variable for HLM Social Work Major Indicates whether or not the participant has an undergraduate major in social work. Control variable Gender Indicates w hether participant identifies as a woman. Control variable Support Identifies the level to which participant anticipates they might receive support from their agency if a child were harmed on their caseload. Independent variable in all hypotheses

PAGE 58

58 Ta ble 3.4 c on t d Rigidit y (centralization & formalization) A t perception of centralization and formalization in their agency. Independent variable in all hypotheses Role Overload (Stress subdomain) A t perception of time constraints and overload in the work. Independent variable in all hypotheses Role Conflict (Stress subdomain) A t concern with incongruent pressures in their work. Independent variable in all hypotheses Authentic Expre ssion Mean d egree to which participant is authentic with emotions at work. Dependent variable in H1. Deep Acting Mean d egree to which participant engages in efforts to feel emotions required by job. Dependent variable in H2. False Face Mean d egree to wh ich participant uses false face acting Dependent variable in H3. Emotion Management Participants mean of perception that they are efficacious at other focused emotional management. Dependent variable in H4. Success Mean d egree to which participant per ceived they are successful in achieving people changing policy goals in child welfare. Dependent variable in H4. Intent to Quit Binomial variable that i ndicates whether participant gives any indication of seeking new employment within the subsequent six months. Dependent variable in H5.

PAGE 59

59 IV. ANALYSIS AND FINDINGS This chapter report s results of hypothesis testing for the five hypotheses in this study. The five hypotheses are designed to address the overarching research question : How is the context of bure aucratic constraint in organizational culture related to the performance of emotional labor by street level bureaucrats? Addressing this research question was accomplished by examining three sub questions: (1) How is the performance of other focused emotio n management limited or expanded in situations of high bureaucratic constraint? (2) Similarly , is self focused emotion management simply survival for the street level bureaucrat and a disguise for self preservation ? and, (3) How can bureaucratic constrain ts contribute to the creation of job stress for street level bureaucrats, thus leading to emotional exhaustion, burnout and turnover? Table 4 . 0 1 reports each hypothesis alongside its corresponding research question. Table 4. 0 1 . Research question and corre sponding hypotheses Primary Research Question: How is the context of bureaucratic constraint in organizational culture related to the performance of emotional labor by street level bureaucrats? Sub research question: How is the performance of other focu sed emotion management limited or expanded in situations of high bureaucratic constraint? H1: Bureaucratic constraints will be negatively related to authentic expression of emotion. H2: Bureaucratic constraints will be negatively related to deep acting. Sub research question: Is self focused emotion management simply survival for the street level bureaucrat and a disguise for self preservation ? H 3 : Bureaucratic constraints will be positively for self focused emotion ma nagement. Sub research question: How can bureaucratic constraints contribute to the creation of job stress for street level bureaucrats, thus leading to high levels of emotional exhaustion, burnout and costly turnover? H4: Bureaucratic constraints will b e negatively related to self efficacy in the performance of emotion management. H5: Bureaucratic constraints will be positively related to expression by the street level bureaucrat that they intend to quit their position.

PAGE 60

60 Missing Data Prior to running ad ditional analysis, a missing value analysis in SPSS was conducted to look for missing data and to understand whether to proceed as if the missing data occurred randomly in this sample. Forty at Random (MCAR) test indicated a non significant finding ( p =0.861), allowing the analysis to proceed under the assumption that these data might have been randomly omitted. T he remaining analyses were conducting using list wise deletion of cases with miss ing values. Table 4. 0 2 presents missing data in this analysis. List wise deletion resulted in a reduction in units from 75 to 73. Table 4. 0 2 . Missing Data Pattern Analysis Variable False Face Emotion Management Support SW Major Intent to Quit Missing 7 10 5 6 5 *Patterns with less than 1% cases (3 or fewer) are not shown. Results The variables related to the performance of emotional labor were authentic expression, deep acting, false face, and emotion management. Univariat e statistics are displayed in Table 4.03. Of note is the low variability on the emotion management factor. No one in the sample deviation is small, at .866. The skewed nature of this dependent variable and lack of variability makes it unsuitable for hypothesis testing using parametric tests. Because there are two variable s

PAGE 61

61 that address self efficacy, the success variable , or the extent to which participants indicated they experience success in their jobs, was used instead. The success at efficacy for people changing technologies variable was also skewed, indicating that all respondents averaged at least a little bit of success in people changing with regard to child safety, permanency, and well being. The skew was less than emot ion management scores however , so the variable w as retained in the analysis. Table 4 . 0 3 . Univariate statistics , Dependent Variables Min M a x Mean Std. Deviation Skewness Kurtosis False Face 1.00 7.00 3.8600 1.22301 .110 .309 Authentic Expressio n 1.00 7.00 4.0633 1.26873 .046 .542 Emotion Management 3.00 7.00 5.6200 .86643 .577 .078 Deep Acting 1.00 7.00 3.9422 1.17938 .099 .066 Success 2.00 5.00 3.7533 .56846 .035 .080 The final dependent variable in the analysis was related to the intention of respondents to quit their position, a proxy for turnover behavior in the workforce. After transforming to a binomial distribut ion, the data indicate that 37% percent or 111 caseworkers in the sample f quitting their job for one outside the agency within one year. Four var iables served to operationalize bureaucratic constraints: concern for liability, rigidity (including centralization and formalization), role conflict, and role overload. The latter t hree are t scores as related to normed data from 81 child welfare agencies nationally. As such, the t scores center near the 50% percentile , with the units in this analysis trending slightly higher on rigidity and slightly lower on role overload tha n the n orm . Concern for liability to the caseworker in the event of harm to a child ranged from a lot (1) to very little (7) (these items were reverse coded), with the mean (3.31) slightly favoring confidence in terms of support by the

PAGE 62

62 agency, leadership, and sup ervisors . Univariate statistics for these independent variables may be found in Table 4. 0 4 . Table 4. 0 4 . Univariate statistics for independent variables Min M a x Mean Std. Deviation Skewness Kurtosis Rigidity 32.61 84.81 57.6163 1.22301 11.49965 .399 R ole Conflict 29.94 72.36 50.7639 1.26873 10.21078 .134 Role Overload 13.04 79.56 43.9378 .86643 13.33511 .426 Liability 1.00 7.00 3.3178 1.17938 1.30279 .435 o explore bivariate re lationships. This matrix is reported in table 4. 0 5 . The coefficients were all lower than r = .329, with the exception of role conflict and role overload, which was significant ( r = .510, p < .01). This is not surprising given that t hey exist in the same do main (stress) of the culture/climate instrument , though this does flag some concern related to discriminant validity for these two constructs . The choice was made, however, to exclude exhaustion under the stress domain given that it more closely reflects i ssues related to burnout than bureaucratic constraint. The hope is that teasing out the unique contributions of each construct might occur through regression analysis. Using a two tailed test of significance, social work majors and women (control variables ) were negatively correlated to Deep Acting ( r = .142, p < .05 and r = .144, p < .05, respectively) . Women were also correlated negatively to authentic expression ( r = .115, p < .05) . Success at people changing behavior was negatively correlated with fa lse face ( r = .173, p < .01) , but positively correlated with authentic expression ( r = .175, p < .01) , emotion management ( r = .282, p < .01) . Conversely, concern for liability was positively related to false face acting ( r = .114, p < .05) , but negativel y related to authentic expression ( r = .237, p < .01) ,

PAGE 63

Table 4. 0 5 . Correlation Matrix 1 . 2 . 3 . 4 . 5 . 6 . 7 . 8 . 9 . 10 . 11 . 12 . SW Gen FF Auth EmoM DeepA Quit Succ Liab Rigid Rcon Rover 1. Social Work Major 1.00 2. Gender 0.09 1.00 3. False Face 0.06 0.06 1.00 4. Authentic Expression 0.04 .115 * .236 ** 1.00 5. Emotion M anagement 0.02 0.04 0.02 .222 ** 1.00 6. Deep Acting .142 * .144 * 0.09 .132 * .212 ** 1.00 7. Intent to Quit 0.01 0.02 .252 ** 0.08 0.03 0.00 1.00 8. Success 0.05 0.09 .173 ** .175 ** .282 ** 0.09 0.09 1.00 9. Liability 0.02 0.10 .114 * .237 ** .235 ** .158 ** .329 ** .123 * 1.00 10. Rigidity 0.04 0.00 0.05 0.04 0.03 0.04 .150 ** 0.05 .174 ** 1.00 11. Role Confli ct 0.01 0.06 .207 ** 0.06 0.04 0.06 .239 ** 0.10 .213 ** .157 ** 1.00 12. Role Overload 0.01 0.02 .186 ** 0.03 0.05 0.10 0.08 .148 * 0.05 .134 * .510 ** 1.00 *. Correlation is significant at the 0.05 level (2 tailed). **. Correlation is signifi cant at the 0.01 level (2 tailed).

PAGE 64

emotion management ( r = .235, p < .01), and deep acting ( r = .158, p < .05). Intent to quit was positively related to liability ( r = .329, p < .01), rigidity ( r = .150, p < .01), and role conflict ( r = .239, p < .01). Hypothesis Testing Hypothesis One . The first hypothesis tested was H1: Bureaucratic constraints will be negatively related to authentic expression of emotion. The dependent variable for this test is Authentic Expression. Because the variable is the mean of a combination of scores, it is continuous and close to normally distributed. Using HLM Version 7.03, a null model was utilized to test whether there is a unit (level 2) effect on the dependent variable (level 1) intercept. This test guides the suitability of HLM as a regression model for this hypothesis. The test employed is Full Information Maximum Likelihood (FIML) estimators given that unit sizes vary considerably in this sample. The estimation of variance components in the model was not sig nificant ( p = .254) meaning that after other variables in the null model were controlled, there was no residual between groups variance from the random effects of the Unit o n the dependent variable Authentic Expression (level 1). HLM was not the appropriat e estimation method for this hypothesis. Consequently, OLS regression was conducted using IB M SPSS Statistics, version 25. The model s, including controls, are illustrated in Table 4. 0 6 . The first model, including just the control variables of social work major and gender, indicated a statistically significant relationship for gender ( p < .05), wherein those individuals identifying as women were less likely to exhibit authentic expression. Adding liability in Model 2 reduced that relationship and it was no longer significant. However, liability, or the indication of low confidence in agency, leadership, and supervisor to fairly assess blame for child harm was significant at p = < .001. Finally, model three, including all the variables for bureaucratic

PAGE 65

65 constr aint, also exhibited significance for liability ( p = < .01). The models did not identify significant relationships with the other independent variables. Hypothesis one was partially accepted. Table 4. 0 6 . Regression Results of Bureaucratic Constraints on Au thentic Expression Model 1 Model 2 Model 3 Social Work Major 0.051 0.780 0.042 Gender .120 * .097 0.099 Liability .227 *** 0.234 ** Rigidity 0.016 Role Conflict 0.033 Role Overload 0.065 R Squared 0.016 0.067 0.069 Adjusted R Squa red 0.009 0.057 0.050 F 2.398 7.044 *** 3.646 ** a. Dependent Variable: Authentic Expression b. Standardized coefficients are presented in the table c. *Significant at p < .05 **Significant at p < .01 ***Significant at p < .001 Hypothesis Two : Nex t, testing was conducted for H2: Bureaucratic constraints will be negatively related to deep acting. The dependent variable for this test is Deep Acting. Because this variable is the mean of a combination of scores, it is continuous, and close to normally distributed. Using HLM Version 7.03, a null model was utilized to test whether there is a unit (level 2) effect on the dependent variable (level 1) intercept. This test guides the suitability of HLM as a regression model for this hypothesis. The test empl oyed was Full Information Maximum Likelihood (FIML) estimators given that unit sizes vary considerably in this sample. The estimation of variance components in the model were not significant ( p = .473) meaning that after other variables in the null model a re controlled, there is no residual between groups variance from the random effects of the Unit on the dependent variable Deep Acting (level 1). HLM is not the appropriate estimation method for this hypothesis. Consequently, OLS regression

PAGE 66

66 was conducted us ing IBM SPSS Statistics, version 25. These models, including controls, are illustrated in Table 4. 0 7 . The first model, which included just controls of social work major and women, exhibited a significant relationship to both variables, with both social w ork majors and women negatively related to the performance of deep acting ( p = <.05). This continued throughout the other two models. Model two, which including just liability in addition to the controls, indicated a significant negative relationship ( p = <.001) to the performance of deep acting. Finally, model three, including all the variables for bureaucratic constraint, also exhibited significance for liability ( p = < .01). The models did not identify significant relationships with the other independent variables. Hypothesis two was partially accepted. Table 4. 0 7 . Bureaucratic Constraints on Deep Acting Model 1 Model 2 Model 3 Social Work Major .129* .134 * .135 * Gender .131* .116 * .119 * Liability .149 ** .161 ** Rigidity .023 Role Conflic t .088 Role Overload .041 R Squared .037 .059 .0 72 Adjusted R Squared .031 .049 .053 F 5.735** 6.088*** 3. 8 46 ** * a. Dependent Variable: Deep Acting b. Standardized coefficients are presented in the table c. *Significant at p < .05 **Sig nificant at p < .01 ***Significant at p < .001 Hypothesis Three . Next, H3: alse for self focused emotion management , was tested . The dependent variable for this test was False Fac e. Because this variable is the mean of a combination of scores, it is continuous, and close to normally distributed. Using HLM Version 7.03, a null model was

PAGE 67

67 utilized to test whether there is a unit (level 2) effect on the dependent variable (level 1) int ercept. This test guides the suitability of HLM as a regression model for this hypothesis. The test employed was Full Information Maximum Likelihood (FIML) estimators given that unit sizes vary considerably in this sample. In the case of this null hypothes is, the final estimation of variance components was mildly significant at p = .022, which is less than p = .05. Consequently, HLM was employed to estimate coefficients for this hypothesis. Results of this testing are in Table 4. 0 8 . Table 4. 0 8 . HLM Analysi s of False Face Model Variable Coefficient SE t ratio df p Value Random effects only Constant 3.849** 0.08 47.822 73 <.001 Unit Variance .123 Residual Variance 1.36 Chi square 99.231* 73 .022 ICC .082 Unit Level and Caseworker level fixed effects Unit Constant 2.39 .471 5.079 70 <.001 Rigidity .003 .007 .422 70 .674 Role Conflict .015 .008 1.902 70 .061 Role Overload .006 * .006 2.067 70 .042 Caseworker Social Work Major .013 .157 .82 223 .413 Gender . 259 .189 1.373 223 .171 Liability .091 .054 1.679 223 .094 Unit Variance .05 Residual Variance 1.34 Chi Square 84.309 70 .117 *p<.05 **p<.001 The full model including both unit and caseworker independent variables and controls indicated a statistically significant relationship with role overload on the performance of false face or surface acting. Role overload captures the degree to which participants indicated their

PAGE 68

68 strain in doing all the parts of the job demanded of them. Non e of the other independent variables were significant. The model comparison test was significant ( p < .01) with a Chi Square of 18.57. Hypothesis three was partially accepted. Hypothesis Four . Next to be tested was H4: Bureaucratic constraints will be neg atively related to self efficacy in the performance of emotion management. The dependent variable for this hypothesis is Success. Because this variable is the mean of a combination of scores, it is continuous, and close to normally distributed. Using HLM V ersion 7.03, a null model was utilized to test whether there is a unit (level 2) effect on the dependent variable (level 1) intercept. This test guides the suitability of HLM as a regression model for this hypothesis. The test employed was Full Information Maximum Likelihood (FIML) estimators given that unit sizes vary considerably in this sample. The estimation of variance components in the model were not significant ( p = >.500) meaning that after other variables in the null model are controlled, there is no residual between groups variance from the random effects of the Unit on the dependent variable Success (level 1). HLM was not the appropriate estimation method for this hypothesis. Consequently, OLS regression was conducted using IBM SPSS Statistics, ve rsion 25. The se model s, including controls, are illustrated in Table 4. 0 9 . Model one, which included just the control variables, did not exhibit statistical significance. Model two, which included just liability in addition to the controls, indicated a s ignificant negative relationship ( p = <.05) to self efficacy in people changing technologies. Finally, model three, including all the variables for bureaucratic constraint, also exhibited significance for liability ( p = < .05). The models did not identify significant relationships with the other independent variables. Hypothesis four was partially accepted.

PAGE 69

69 Table 4. 0 9 . Bureaucratic constraints on self efficacy Model 1 Model 2 Model 3 Social Work Major .039 .036 .036 Gender .087 .101 0.1 Liability . 132 * .147 * Rigidity .064 Role Conflict .014 Role Overload .136 R Squared .010 .027 .054 Adjusted R Squared .003 .017 .034 F 1.461 2.749* 2.770* a. Dependent Variable: Self Efficacy b. Standardized coefficients are presented in the table c. *Significant at p < .05 **Significant at p < .01 ***Significant at p < .001 Hypothesis Five. Finally, testing was conducted of H5: Bureaucratic constraints will be positively related to expression by the street level bureaucrat that they inte nd to quit their position. The dependent variable for this hypothesis is Intention to Quit. Because this variable is has 0/1 Bernoulli distribution, estimation will be conducted by using a logit link function to transform the variable. Using HLM Version 7. 03, a null model was utilized to test whether there is a unit (level 2) effect on the dependent variable (level 1) intercept. This test guides the suitability of HLM as a regression model for this hypothesis. The test employed is Full Information Maximum L ikelihood (FIML) estimators given that unit sizes vary considerably in this sample. In the case of this null hypothesis, the final estimation of variance components is significant at p = .006, which is less than p = .01. Consequently, HLM was employed to e stimate coefficients for this hypothesis. The final estimation of fixed effects, with a population average model with robust standard errors is illustrated in Table 4.11 . The full model including both unit and caseworker independent variables and controls indicated a statistically significant relationship with concerns for liability positively impacting a

PAGE 70

70 independent variables were significant. Hypothesis five was partially accepted. Table 4. 1 1 . HLM Analysis of Intent to Quit Model Variable Coefficient SE t ratio df p Value Random effects only Constant .517** * .148 3.490 73 <.001 Unit Variance .468 * .255 73 .006 Chi square 107.147 Unit Level and Caseworke r level fixed effects Unit Constant 5.144 *** 1.18 4.35 70 <.001 Rigidity .016 .012 1.329 70 .188 Role Conflict .035 .018 1.938 70 .057 Role Overload .005 .013 .447 70 .656 Caseworker Social Work Major .069 .311 .223 223 .824 Gender .102 .446 .229 223 .819 Liability .502*** .110 4.559 223 <.001 Unit Variance .170 Chi Square 87.823 *p<.05 * **p<.001 Results of Hypotheses Testing All five hypotheses were partially supported by this analysis , with at least one in dependent variable significantly modeling an influence on each dependent variable in each hypothesis . The results of this hypotheses testing are illustrated in 4.12 . The implications of these results and their relation to the research questions of this stu dy are contained in Chapter Five.

PAGE 71

71 Table 4.12 Summary of Hypothesis Testing Hypothesis Results H1 Bureaucratic constraints will be negatively related to authentic expression of emotion. Partially supported H2 Bureaucratic constraints will be negati vely related to deep acting. Partially supported H3 Bureaucratic constraints will be strategies for self focused emotion management. Partially supported H4 Bureaucratic constraints will be negatively related to self efficacy in the performance of emotion management. Partially supported H5 Bureaucratic constraints will be positively related to expression by the street level bureaucrat that they intend to quit their position. Partially supported

PAGE 72

72 V . DISCUSSION This chapter summarizes the results section in relation to the overarching research question, the sub research questions, and the hypotheses proposed in Chapter 3. Each dependent variable for the five hypotheses (authentic expression, deep acting, false face, s elf efficacy, and intent to quit) are reviewed alongside the extant literature to compare and contrast the results of this study. Where these results deviate from theoretical positions in the literature, this chapter reviews differences and develop s meanin g as applicable. Where practical implications exist for public administration, each will be delineated as takeaways or areas of attention in the realm of practic e. Limitations of the study are considered, including threats to internal and external validity , study design, sampling, and other factors that impact the generalizability of these results. Finally, the chapter concludes with suggestions for further study and outlines a research agenda related to this inquiry, with attention to establishing the impo rtance of this line of research to both practice and theory. Summary of Findings Table 5.1 contains a summary of the overarching research question, the sub research question, and hypotheses, alongside significant findings. For the first sub research ques tion, which queried how the performance of other focused emotion management limited or expanded in situations of high bureaucratic constraint , two hypotheses were tested. The first, that b ureaucratic constraints are negatively related to authentic expressi on of emotion, was partially accepted. In this sample, increases in worry by caseworkers that they will not be supported by their agency in the event of a child being harmed on their caseload (liability) are negatively related to authentic expression of em otion. The second hypothesis, that b ureaucratic constraints are neg atively related to deep acting was also partially supported. Having majored in social

PAGE 73

73 work, identifying as a woman, and increases in worry by caseworkers that they will not be supported by their agency in the event of a child being harmed on their caseload (liability) , are negatively related to the performance and effort related to deep acting. The second research question of how self focused emotion management might be simply survival for t he street level bureaucrat and a disguise for self preservation, was answered using one hypothesis. This hypothesis, that bureaucratic constraints are strategies for self focused emotion management was partially supported by the analysis. Increases in expressions of role overload such as feeling constrained by time and resources are positively related to false face strategies for self focused emotion management. The third research question queried as to how bureaucratic co nstraints contribute to the creation of job stress for street level bureaucrats, thus leading to high levels of emotional exhaustion, burnout and costly turnover? This research question was addressed by two hypotheses. First, it was hypothesized that b urea ucratic constraints are negatively related to self efficacy in the performance of emotion management. This hypothesis was partially supported. Analysis indicated that increases in worry by caseworkers that they will not be supported by their agency in the event of a child being harmed on their caseload (liability) are negatively related to their experience of self efficacy, or ability to successfully assist families and children in achieving necessary changes to promote safety, permanency, and well being. F inally, it was hypothesized that b ureaucratic constraints will be positively related to expression by the street level bureaucrat that they intend to quit their position. This hypothesis was partially supported. Increases in worry by caseworkers that they will not be supported by their agency in the event of a child being harmed on their caseload (liability) increased the odds that they express an intention to seek work outside their agency in the subsequent year.

PAGE 74

Table 5.1 . Summary of Findings Primary Research Question: How is the context of bureaucratic constraint in organizational culture related to the performance of emotional labor by street level bureaucrats? Sub Research Question Hypothesis Findings How is the performance of other focused emot ion management limited or expanded in situations of high bureaucratic constraint? H1 Bureaucratic constraints will be negatively related to authentic expression of emotion. Increases in worry by caseworkers that they will not be supported by their agency in the event of a child being harmed on their caseload (liability) are negatively related to authentic expression of emotion. H2 Bureaucratic constraints will be negatively related to deep acting. Having majored in social work, identifying as a woman, and increases in worry by caseworkers that they will not be supported by their agency in the event of a child being harmed on their caseload (liability) are negatively related to deep acting. How is self focused emotion management simply survival for the street level bureaucrat and a disguise for self preservation ? H3 Bureaucratic constraints will be strategies for self focused emotion management. Increases in expressions of role overload such as feeling constrained by time and resources are positively related to false face strategies for self focused emotion management. How can bureaucratic constraints contribute to the creation of job stress for street level bureaucrats, thus leading to high levels of emotional exha ustion, burnout and costly turnover? H4 Bureaucratic constraints will be negatively related to self efficacy in the performance of emotion management. Increases in worry by caseworkers that they will be not be supported by their agency in the event of a child being harmed on their caseload (liability) are negatively related to their experience of self efficacy, or ability to successfully assist families and children in achieving necessary changes to promote safety, permanency, and well being. H5 Bureauc ratic constraints will be positively related to expression by the street level bureaucrat that they intend to quit their position. Increases in worry by caseworkers that they will not be supported by their agency in the event of a child being harmed on the ir caseload (liability) increase the odds that they will express an intention to seek work outside their agency in the subsequent year.

PAGE 75

Authentic Expression Authenticity of emotional display is empirically linked to worker well being, including depressi on, depressed mood, and secondary traumatic stress (Caringi et al., 201 2; Erickson & Wharton, 1997; Grandey et al., 2012) leadership, or supervisor in the event of harm to a child under her purview is related to decreases in feelings of authenticity in the current stu dy. This is consistent with multiple studies that consistently report th inauthenticity and subsequent mental health outcomes as expressed by workers in the service and manufacturing industri es (Erickson & Wharton, 1997; Hochschild, 2003) . While the operationalization of autonom y was more nuanced in this study, to account for the unique facets of the bureaucracy within which child welfare work is performed, the implications are similar. It is logical to assume that if one expresses considerable worry about outcomes that are for t he most part largely not under her control, this can be construed as decreased autonomy of the worker. As discussed in the literature review , caseload following agency intervention is nationally a relatively ra re event as compared to the daily work tasks of child welfare caseworkers. Similarly, direct connections of worker behavior to child harm often oversimplify very complex family, community, and social dynamics. Nonetheless, worry that her agency will not pr ovide a fair and balanced investigation of the facts surrounding such cases may also be considered a constraint on worker discretion, as worker and supervisor decision making throughout the case is the focus of these types of internal investigations. While accountability under such circumstances is certainly justified and is a key component of the responsibilities assumed by child welfare workers, the relation of these constraints to authenticity of emotional displays is an important finding.

PAGE 76

76 It is inte are good at emotional management and efficacious in its delivery. No one in the sample indicated disagreement that emotional management is a key component of their job in child welfare casework, or that they felt competent at calming others and maintaining agency display rules. The sample also demonstrated prior affinity for interacting with and dealing with people ther social work, psychology, or other related fields such as human services, criminal justice, or sociology. Even so, in this study, and the everyday realities of their work. Deep Acting Deep acting is the core of other focused emotion work, and as I argue in this dissertation, it is integral to people changing technology, or family engagement, as in the child welfare literature. Deep acting is other focused, and facilitates the change process by allowing relationships to develop (Hasenfeld, 1983; Pugliesi, 1999) . Change occurs, then, through the be consider ed instrumental to child protection work that seeks to change parent and caregiver behaviors in the interest of child safety, permanency, and well being. Client oriented firms in business, for example, recognize this dynamic and elevate the position of fro ntline workers to leverage their intricate knowledge of the client experience and their relationships (Lively, 2002) . Given the primacy of deep acting to the work of child welfare, it is concerning to observe the impact found in this study by bureaucratic constraint on performance of deep acting by agency, leadership, or superv isor in the event of child harm was negatively related to the

PAGE 77

77 related to outcomes for the workforce, including job satisfaction, mediation of burnout, and well be ing (Brotheridge & Grandey, 2002; Grandey, 2000) . Also of concern, however, are possible impacts on outcomes for children, youth and families who receive services from the child welfare system through the conduit of the caseworker. In customer service situations, customers differentially assess their satisfaction with the interaction according to whether the service worker was surface (false face) or deep acting. This was compounded when the service recipient accurately detected which strategy (false face or deep acting) was being utilized, which occurred with some frequency (Groth, Hennig Thurau, & Walsh, 2009) . Similarly, concerns related to emotional contagion may point to impacts on service recipients in emotionally intense situations where intervening caseworkers do not work to feel the emo tions they are to express in carrying out child welfare responsibilities and working toward facilitating change in detrimental behaviors, though these have not been fully investigated in child welfare work (Hatfield, Cacioppo, & Rapson, 1992) . In this case, we again see worry for accounta bility impinging on the emotion work of the job. With regard to deep acting, two surprising factors emerged as influential to caseworker performance of deep acting. Having an undergraduate major in social work and identifying as a woman both exhibited neg ative impacts on performance of emotional labor. This is interesting, because some studies utilize the proportion of women employed in an organization as a proxy for the degree of emotional labor involved ( Meier et al., 2006) . Similar, given the emphasis in undergraduate social work curricula on the employment of people changing technologies and the importance of relationship, it is puzzling that in the context of the agencies in this sample, both social wo rkers and women employ less deep acting in their work. The setting, of course, exhibits

PAGE 78

78 a large proportion of women (83% in this sample) with a high number of colleagues who have entered the workforce with the expectation that people changing technologies may be employed, based on their undergraduate education. While further analysis is warranted and necessary regarding this finding, one may speculate that those with strong proclivities toward emotion work might exhibit repressive behaviors in favor of cons erving valuable emotive resources, as we might find in studies based on conservation of resources (COR) theory (Gelderen, 2016; Grandey et al., 2012) . False Face False face, or surface acting, is the operationalization of self focused emotion work. False face strategies co ilies and children within the unit. Role overload is a characteristic of bureaucratic constraint, where time and resources are limited. The example item provi construct makes up the climate in the agency, and the variable is a normed t score from across a sample of 81 child welfare agencies nationwide. This means false face occurs more to individuals who work in agencies that are above the norm in terms of their constraints on time and resources (Calabresi & Bobbitt, 1978) . Role overload points to lack of autonomy in the job, with limits to time and resources. This is congruent with the literature, in that false face or surface acting is about doing the job while

PAGE 79

79 different place due to outside circumstances (Gelderen, 2016; Guy et al., 2008; Hochschild, 2003; Kruml & Geddes, 2000) . the service recipient, and false face has been observed empirically to determine response of the customer or client to the service worker (Zhan et al., 2015) . The implications in the applied field of child welfare, then, become clear. William Schwartz, a prominent social work educator and author of numerous foundational textbooks writes abo ut the role of the social worker, client. Thus the worker should ask [herself]; what kind of incident will I (Lee , 2001 , p. 186 ) . Indeed, what incident might be represented by the intervention of a child welfare caseworker who is preoccupied with self focused emotional labor due to constraints that impose enormous pressure and burden? This emotional restraint is de leterious to relationship building, and people changing technologies in human services require extensive relationship and emotional engagement for their efficacy (Hasenfeld, 1983) . Of note, false face acting behaviors in this study varied significantly by unit, to the point where statistical ana lysis not grouped at the unit level might have falsely reported trends in , and influences on , performance. Similarly, role overload is not an individual perception, but rather the mean of perceptions of all the caseworkers in a unit. Organizational climate , then, is the unit takes little effort to imagine a frenetic scene, wherein individuals in a unit are scrambling to accomplish the important work of child welfare, unable to assist team members, and unable to slow down for reflection and critical thinking. The consequences of this unit level reality in the

PAGE 80

80 office are reflected in restricted emotion in the field by caseworkers interacting with families and c hildren. Practically speaking, it seems the mission of child welfare might be at cross purposes with this scenario, and as noted above, this is most likely not lost on those citizens receiving interventions and services from the agency. Self Efficacy The caseworkers in this sample emphasized the importance of emotion management in , I used a compilation of items representing caseworker reflections of success in the people changing technology of child welfare (child safety, permanency and well being). For this measure of self efficacy, participants reflected on their success for impacting child ren and families according to the policy goals of child welfare as set forth in Federal law. These goals also represent categories where agencies measure caseworker performance (e.g. recurrence of maltreatment or time to permanency for children in foster c are). Arguably, these categories are of primary importance to the work of child welfare. In this sample of caseworkers, however, participants expressed that when they experienced worry that if a child is harmed on their caseload they will not be supported by their agency, leadership, or supervisor, their experiences of self efficacy in their job decreased. This observation mirrors the impact on authentic expression and deep acting. In this finding, two areas of the work are again at cross purposes. Casework ers who feel enormous accountability for the safety of children on their caseload, and who worry about their agency supporting their discretion and autonomy if that were to happen, at the same time report they feel less effective at meeting overall goals i n child welfare. There are two ways to consider this observation in light of the literature: first, the bureaucratic burden must be lightened to free

PAGE 81

81 up autonomy to practice people changing technologies including emotional labor to pursue policy goals, or that the measures of achievement of goals and mission for these caseworkers are out of balance. The first reflects the debate in the field related to how child welfare work should be conducted, and the disproportionate attention to relatively rare events, particularly in light of child welfare scandals and ad hoc policy development (Gainsborough, 2010) . The second possibility drives at issues related to performance management. Child safety, child permanency, and child well being are the expressed goals of child welfare agencies in the United States, but the mechanisms for achieving them are varied and the metrics for their measurement can be distorted in pursuit of management strategies. Often performance management metrics reflect a linear line of cause and effect, or inputs (adherence to policy) and outputs (child safety, permanency, and well being) without including the unpredictable (and in most cases uncontrollable) dynamics of human behavior (Radin, 1998) . Thus, street level bureaucrats must make a complicated assessment of their achievement of goals and performance, within the reality that their agencies grant them a deg ree of discretion and autonomy in their work, but their impact is ultimately determined by the choices of individual caregivers and families, over which they have little perceived or actual control. This is decidedly not a linear situation, and the bureauc ratic constraint in this equation contains within it the fear that agencies, leadership, and supervisors may judge casework in a linear manner despite this reality. Intent to Quit Turnover is a prevalent issue in child welfare organizations. The most rece nt national study, which is now quite dated, estimated that turnover rates for child welfare agencies is 22% per year (American Public Human Services Association, 2005) . Local, state, and national jurisd ictions spend considerable time, effort and money to address this issue. In this study, the

PAGE 82

82 odds of preliminary thinking by caseworkers about quitting their position increased, once again, in relationship to concerns that they might not be supported by the ir agency, leadership or supervisor in the event of a child being harmed on their caseload. This finding is unique in the literature related to turnover in child welfare systems specifically, which typically targets burnout, job dissatisfaction, low commit ment to the work and the organization, stress, and lack of social support as antecedents to turnover (Lizano & Mor Barak, 2012; Mor Barak, 2001) . While this phenomenon does mirror social support, it specifically operationalizes restraints on autonomy and discretion, and encompasses a portion of the accountability child welfare workers feel for children and families on their caseloads. It is important to note that self efficacy is tied in the literature to job satisfaction, which is subsequ ently tied to intent to quit, so more research on these linkages, mediators, and moderators is warranted in light of this finding. Similarly, high frequency of false face acting is also empirically tied to burnout and job satisfaction, specifically in publ ic services like child welfare (Hsieh et al., 2011) . Limitations and Recommendations for Future Research As with all inquiries, there are limitations to this study that impede interpretation and generalizability of findings. Because of the criti cal nature of emotion work in public service and specifically in child welfare, these limitations provide outline of areas for future research and inquiry. Recommendations for future research directly follow limitations in this section. The limitations of this study may be described in the following areas: sample strategy, measurement strategies including self report via survey, cross sectional design, and solely quantitative analysis. Each limitation is outlined in this section with attention to its descri ption, explanation, potential impacts on findings and conclusions, and suggestions for on going attention to the research questions.

PAGE 83

83 The first limitation is the sample strategy. The sample for this inquiry was one of convenience, as part of a larger proje ct related to workforce development in child welfare. While care was taken to access large portions of the population of child serving caseworkers in each of the seventeen jurisdictions, the jurisdictions themselves were selected based on a variety of fact ors outside the purview of this study. These factors included: interest in the larger workforce project, lack of participation in other initiatives within the state, availability and willingness of leadership to participate in the larger initiative and ove rall perception that there were workforce needs to address within the agency. These factors could introduce bias at least in terms of the population sampled. It is possible that the agencies that self selected into the study already had some indication in the jurisdiction that there were difficulties with retention and turnover, or that there were concerns related to organizational culture and climate. One possible study mitigation of this potential bias was the normalization of variables related to organiz ational culture and climate. Bias at the unit level may have been reduced by basing these scores on norms across child welfare and child serving agencies. On the other hand, however, there was no such normalization with the individually based variables wit hin the study, which were based on self reported perceptions of individuals within the agency. Norms do not exist in the same manner for these variables. Similarly, a second limitation of this study is use of survey methodology to understand the experience s of the sample. Self report certainly captures important aspects of experiences of street level bureaucrats, but such methodologies may introduce bias. For example, study participants might have over or under estimated their experiences upon hearing abo ut the workforce initiative designed to improve their work life. While participants were assured of their confidentiality through informed consent, it is difficult to ascertain sincerity of response or the

PAGE 84

84 match of response with the reality of the work. Th e survey, attached in Appendix 1, was developed for a national research study on building evidence for effective interventions in the child welfare workforce. To assist in this aim, the attached survey, which is quite lengthy, was administered alongside th e Organizational Social Context (OSC), which contains 105 items. Participants filled out the survey and the OSC (not necessarily in that order) typically in one sitting, introducing the possibility of survey fatigue with items that occur later in either of the study. The design of this study is cross sectional, whereas the organizational life and the experience of child welfare caseworkers is dynamic. The choice for the cross sectional nature of this study was based on time constraints as well as logistic feasibility. There is not analysis of contextual informat ion pertaining to study participants such as recent or significant events that impacted the workforce. Thus, survey responses and perceptions of factors might be based on events immediately preceding the administration of the survey and OSC rather than on the general experience of the caseworker. Finally, analysis and conclusions are based solely on quantitative analysis of survey items. While useful, these items only capture a portion of the lived experience of street level bureaucrats. Several conclusion s in the prior sections point specifically to this limitation, including that related to lower levels of performance of deep acting by women and social work majors, and the last finding related to caseworker intention to quit her/his job. In particular, th e variable related to concern for liability and support by agency, leadership, and supervisors cannot fully be understood through statistical analysis. For example, how does this bureaucratic

PAGE 85

85 constraint play out in the agency? Are caseworker worries that t hey will lack support during difficult times based on watching this play out with peers? Are they based on some part of the culture and climate that was not explored through this study? While still significant in finding, much more could be understood thro ugh qualitative methods. The preceding limitations, in addition to the conclusions and findings in this study contribute to the development of a research agenda that continues to address this area of public administration. This research agenda contains th ree parts: (1) continued analysis of this data set, (2) replication of this design paired with a mixed methods approach, and (3) longitudinal designs that measure the dynamics in this study over time and in the context of dynamically changing child welfare agencies. The first approach, continued analysis of this data set, draws on the rich data still unexplored in this extensive survey design. Attention to social desirability factors (discussed earlier) and analysis of skip patterns or answer patterns if e vident might further inform the limitation related to survey fatigue from the length of the instrument. Further, the survey contains several cross examinations of related concepts that might enliven the analysis conducted here. These concepts and construct s include personality through a version of the Big Five Inventory (BFI) (John & Srivastava, 1999) as well as measurements of burnout through the Maslach Burnout Inventory (MBI) (Maslach, Schaufeli, & Leiter, 2001) . Further, scales are included that measure relationships with co workers, supervi sory and social support, empathic orientation, coping and resilience characteristics, time constraints for completion of paperwork and family engagement, characteristics of workload including caseload size, and overall orientation toward family preservatio n as compared to child safety (Dettlaff et al., 2015; Graham, De ttlaff, Baumann, & Fluke, 2015) .

PAGE 86

86 The next opportunity for continued research on these questions is the replication of the design of this study outside the context of the larger workforce study, with attention to triangulating qualitative data with the qua ntitative results. Specifically, more detailed explanation of deep acting by the caseworkers who score themselves on either side of the spectrum (high disagree or highly agree) might assist in characterization that sheds light on how deep acting plays out in practice with children and families, and assist in understanding how those dynamics might impact people changing technologies. Further, semi structured interviews might create space to examine organizational dynamics that lend to the worry described in the main findings of this study. How do caseworkers receive signals about support for incidents that have likely not yet occurred? How overt is the messaging in their unit or organization, and for those that perceive less worry in this area, how are they b eing treated that reassures them? What types of of awareness to workers have about resulting responses from children and families for either deep acting or f alse face (surface acting)? This level of triangulation also introduces approaches to concerns related to performance management, with the ability to examine subsequent administrative data and client level outcomes, as well as target analysis that includes families and children who experience these workers in the field. Longitudinal research designs, especially when paired with a mixed methods approach, also present opportunities for continued inquiry. Specifically of interest based on findings in this stu dy is the degree to which certain individuals might enter the field with proclivity toward the performance of emotional labor and specifically deep acting, but who later may resort to surface acting for self preservation and/or in the race of constraints o n time and resources related to their

PAGE 87

87 focusing on the child welfare workforce, where there are high and frequent rates of turnover for such large portions, might shorten the time needed for such inquiry (Mor Barak, 2001) . Conclusion This chapter contained a summary of the f indings in this dissertation, alongside an examination of the extant literature in relation to these findings. It then concluded with limitations of the overall study as well as recommendations for further research. In sum, some bureaucratic constraints we re related to impacts on street level performance of emotional labor, perceptions of self efficacy in achievement of policy outcomes, and overall intention of street nomy and discretion may result in restrained emotion, which could impact the deployment of people changing technologies and overall outcomes of the agency. More research is needed to understand the organizational and individual characteristics that mediate and moderate this dynamic, and theory must consider the degree to which autonomy and discretion may be restricted in organizational contexts.

PAGE 88

88 REFERENCES American Public Human Services Association. (2005). R eport from the child welfare workforce study . Washington, DC. Retrieved from http://www.theprofessionalmatrix.com/docs/WorkforceReport2005.pdf Auerbach, C., Schudrich, W. Z., Lawrence, C. K., Claiborne, N., & McGowan, B. G. (2014). Predicting Turnover. Res earch on Social Work Practice , 24 (3), 349 355. http://doi.org/10.1177/1049731513494021 Bachrach, L. L. (1983). An overview of deinstitutionalization. New Directions for Mental Health Services , 1983 (17), 5 14. http://doi.org/10.1002/yd.23319831703 Bardach, E. (1977). The implementation game: What happens after a bill becomes a law . Cambridge, MA: MIT Press. Bartholet, E. (2014). Differential response: A dangerous experiment in child welfare. Florida State University Law Review , 42 , 573. Baumann, D., Dalgleis h, L., Fluke, J., & Kern, H. (2011). The decision making ecology . Behn, R. (2001). Rethinking democratic accountability . Washington, DC: Brookings Institution Press. Behn, R. (2003). Why measure performance? Different purposes require different measures. P ublic Administration Review , 63 (5), 586 606. http://doi.org/10.1111/1540 6210.00322 Behn, R. (2014). Brookings Institution Press. Bertelli, A. M., & Lynn, L. E. (2006). constitution . Baltimore, MD: Johns Hopkins University Press. Blan k, R. M. (2002). Evaluating Welfare Reform in the United States. Journal of Economic Literature , 40 (4), 1105 1166. http://doi.org/10.1257/002205102762203576 Bovaird, T. (2007). Beyond engagement and participation: User and community coproduction of public services. Public Administration Review , 67 (5), 846 860. http://doi.org/10.1111/j.1540 6210.2007.00773.x Bozeman, B. (2000). Bureaucracy and red tape . Upper Saddle River, NJ: Prentice Hall. Brodkin, E. Z. (2011). Policy work: Street level organizations unde r new managerialism. Journal of Public Administration Research and Theory , 21 (Supplement 2), i253 i277. http://doi.org/10.1093/jopart/muq093 Brotheridge, C., & Grandey, A. (2002). Emotional labor and burnout: Comparing two Jo urnal of Vocational Behavior , 60 (1), 17 39. http://doi.org/10.1006/jvbe.2001.1815 Bryk, A., & Raudenbush, S. (1992). Hierarchical Linear Models: Analysis and Data Analysis Methods . London, UK: SAGE Publications.

PAGE 89

89 Burks, J. (1912). Efficiency standards in mu nicipal management. National Municipal Review , 1 (3), 364 371. Calabresi, G., & Bobbitt, P. (1978). Tragic choices . New York, NY: WW Norton and Company. Caringi, J., Lawson, H., & Devlin, M. (2012). Planning for emotional labor and secondary traumatic stres s Google Scholar. Journal of Family Strengths , 12 (1), Article 11. Retrieved from https://0 scholar google com.skyline.ucdenver.edu/scholar?q=planning+for+emotional+labor+and+secondary+trau matic+stress&btnG=&hl=en&as_sdt=0%2C6 Chau, S. L., Dahling, J. J., Levy, P. E., & Diefendorff, J. M. (2009). A predictive study of emotional labor and turnover. Journal of Organizational Behavior , 30 (8), 1151 1163. http://doi.org/10.1002/job.617 Commission to Eliminate Child Abuse and Neglect Fatalities. (2016). Within o ur reach: A national strategy to eliminate child abuse and neglect fatalities . Retrieved from http://www.acf.hhs.gov/programs/cb/resource/cecanf final report DePanfilis, D., & Zlotnik, J. (2008). Retention of front line staff in child welfare: A systematic review of research. Children and Youth Services Review . Retrieved from http://0 www.sciencedirect.com.skyline.ucdenver.edu/science/article/pii/S0190740908000091 Dettlaff, A. J., Christopher Graham, J., Holzman, J., Baumann, D. J., & Fluke, J. D. (2015). D evelopment of an instrument to understand the child protective services decision making process, with a focus on placement decisions. Child Abuse and Neglect . http://doi.org/10.1016/j.chiabu.2015.04.007 Douglas, E. M., & Cunningham, J. M. (2008). Recommend ations from child fatality review teams: results of a US nationwide exploratory study concerning maltreatment fatalities and social service delivery. Child Abuse Review , 17 (5), 331 351. http://doi.org/10.1002/car.1044 Epp, C., Maynard Moody, S., & Haider M arkel, D. (2014). Pulled over: How police stops define race and citizenship . Erickson, R. J., & Wharton, A. (1997). Inauthenticity and Depression: Assessing the Consequences of Interactive Service Work. Work and Occupations , 24 (2), 188 213. http://doi.org/ 10.1177/0730888497024002004 Finer, H. (1941). Administrative responsibility in democratic government. Public Administration Review , 1 (4), 335 350. Friedrich, C. (1968). Public policy and the nature of administrative responsibility. In F. Rourke (Ed.), Bure aucratic power in national politics (3rd ed., pp. 316 26). Boston: Little Brown. Gainsborough, J. (2010). Scandalous politics: Child welfare policy in the states Google Scholar . Washington, DC: Georgetown University Press. Gelderen, B. van. (2016). Emoti onal labor among police officers: a diary study relating strain, emotional labor, and service performance. The International Journal of Human Resource Management , 1 28.

PAGE 90

90 Ghose, T. (2009). Linking Organizational Factors to Substance Abuse Treatment Outcomes: Multilevel Correlates of Treatment Effectivenes. In Y. Hasenfeld (Ed.), Human Services as Complex Organizations (2nd ed., pp. 429 451). Los Angeles, CA: University of California Press. Glisson, C., Green, P., & Williams, N. J. (2012). Assessing the Organi zational Social Context (OSC) of child welfare systems: implications for research and practice. Child Abuse & Neglect , 36 (9), 621 32. http://doi.org/10.1016/j.chiabu.2012.06.002 Glisson, C., Landsverk, J., Schoenwald, S., Kelleher, K., Hoagwood, K. E., May Health, T. R. N. on Y. M. (2008). Assessing the Organizational Social Context (OSC) of Mental Health Services: Implications for Research and Practice. Administration and Policy in Mental Health and Mental Health Services Research , 35 (1 2), 98 1 13. http://doi.org/10.1007/s10488 007 0148 5 Goodnow, F. (1900). Politics and administration: A study in government . New York, NY: Macmillan. Goodwin, R., Groth, M., & Frenkel, S. (2011). Relationships between emotional labor, job performance, and turnover . Journal of Vocational Behavior , 79 (2), 538 -548. Graham, J. C., Dettlaff, A., Baumann, D., & Fluke, J. (2015). The Decision Making Ecology of placing a child into foster care: A structural equation model. Child Abuse & Neglect . http://doi.org/10.1016/j.c hiabu.2015.02.020 Grandey, A. (2000). Emotional regulation in the workplace: A new way to conceptualize emotional labor. Journal of Occupational Health Psychology , 5 (1), 95. Grandey, A. (2003). When the show must go: Surface acting and deep acting as deter minents of emotional exhaustian and peer rated service delivery. Academy of Management Journal , 46 (1), 86 96. http://doi.org/10.2307/30040678 Grandey, A., Foo, S. C., Groth, M., & Goodwin, R. (2012). Free to be you and me: a climate of authenticity allevia tes burnout from emotional labor. Journal of Occupational Health Psychology , 17 (1), 1 14. Groth, M., Hennig Thurau, T., & Walsh, G. (2009). Customer reactions to emotional labor: The roles of employee acting strategies and customer detection accuracy. The Academy of Management Journal , 52 (5), 958 974. Gulick, L., & Urwick, L. (1937). Papers in the science of administration . New York, NY: Institute of Public Administration. Guy, M., Mastracci, S., Newman, M., & Maynard Moody, S. (2010). Emotional Labor in th e Human Service Organization. In Y. Hasenfeld (Ed.), Human Services as Complex Organizations (pp. 291 309). Thousand Oaks, CA. Labor. Public Administration Review , 64 (3), 289 298. http://doi.org/10.1111/j.1540 6210.2004.00373.x

PAGE 91

91 Guy, M., Newman, M., & Mastracci, S. (2008). Emotional labor: Putting the service in public service . New York, NY: M.E. Sharpe. Hasenfeld, Y. (1972). People processing organizations: An exchange app roach. Sociological Review , 37 , 256 263. Hasenfeld, Y. (1983). Human Service Organizations . Englewood Cliffs, NJ: Prentice Hall. Hasenfeld, Y. (1992). Human services as complex organizations . Newbury Park, CA: Sage. Hatfield, E., Cacioppo, J. T., & Rapson, R. L. (1992). Primitive emotional contagion. In M. Clark (Ed.), Review of Personality and Social Psychology: Emotion and Social Behavior (pp. 151 177). Newbury Park, CA: Sage. Hemmelgarn, A. L., Glisson, C., & Dukes, D. (2001). Emergency Room Culture and the Emotional Support Component of Family Centered Care. , 30 (2), 93 110. http://doi.org/10.1207/S15326888CHC3002_2 Heuven, E., & Bakker, A. (2006). The role of self efficacy in performing emotion work. Journal of Vocational Behavior , 69 (2), 222 235. Hochschild, A. (2003). The managed heart: Commercialization of human feeling . Retrieved from https://books.google.com/books?hl=en&lr=&id=UiSsLXeERF4C&oi=fnd&pg=PP2&dq=th e+managed+heart&ots=NVf2WGhlJF&sig=jJ5MA2qgxhvCfDrxWTx1zRU2kic Hood, C . (1991). A public management for all seasons? Public Administration , 69 (1), 3 19. Accounting, Organizations and Society , 20 (2), 93 109. Hopkins, G. (1912). The New York bure au of municipal research. Annals of the American Academy of Political and Social Science , 41 (1912), 235 244. Hsieh, C. W., & Guy, M. (2008). Performance outcomes: The relationship between managing Review of Pub lic Personnel Administration , 29 (1), 41 57. Hsieh, C. W., Jin, M., & Guy, M. (2011). Consequences of work related emotions: Analysis of a cross section of public service workers. The American Review of Public Administration . Retrieved from http://0 arp.sag epub.com.skyline.ucdenver.edu/content/early/2011/07/27/0275074010396078.abstra ct Hsieh, C. W., Yang, K., & Fu, K. J. (2012). Motivational Bases and Emotional Labor: Assessing the Impact of Public Service Motivation. Public Administration Review , 72 (2), 241 251. http://doi.org/10.1111/j.1540 6210.2011.02499.x Hülsheger, U. R., & Schewe, A. F. (2011). On the costs and benefits of emotional labor: A meta analysis of three decades of research. Journal of Occupational Health Psychology , 16 (3), 361 389. http://do i.org/10.1037/a0022876

PAGE 92

92 John, O., & Srivastava, S. (1999). The big five trait taxonomy: History, measurement, and theoretical perspectives. In L. Pervin & O. John (Eds.), Handbook of personality: Theory and research, Volume 2 (pp. 102 138). New York, NY: Gu ilford Press. Johnson, H., & Spector, P. (2007). Service with a smile: Do emotional intelligence, gender, and autonomy moderate the emotional labor process? Journal of Occupational Health Psychology , 12 (4), 319. Kelly, J. M. (2005). The dilemma of the unsa tisfied customer in a market model of public administration. Public Administration Review , 65 (1), 76 84. http://doi.org/10.1111/j.1540 6210.2005.00432.x Kohl, P. L., Jonson Reid, M., & Drake, B. (2009). Time to Leave Substantiation Behind: Findings From A National Probability Study. Child Maltreatment , 14 (1), 17 26. http://doi.org/10.1177/1077559508326030 Kruml, S. M., & Geddes, D. (2000). Exploring the dimensions of emotional labor. Management Communication Quarterly , 14 (1), 8 49. Lee, J. A. B. (2001). The empowerment approach to social work practice: Building the beloved community . New York, NY: Columbia University Press. Lipsky, M. (2010). Street level bureaucracy (30th Anniv). New York, NY: Russell Sage Foundation. Lively, K. J. (2002). Client Contact an d Emotional Labor: Upsetting the Balance and Evening the Field. Work and Occupations , 29 (2), 198 225. http://doi.org/10.1177/0730888402029002004 Lizano, E. L., & Mor Barak, M. E. (2012). Workplace demands and resources as antecedents of job burnout among p ublic child welfare workers: A longitudinal study. Children and Youth Services Review , 34 (9), 1769 1776. Martin, L. M., Peters, C. L., & Glisson., C. (1998). Factors Affecting Case Management Recommendations for Children Entering State Custody. Social Serv ice Review , 72 (4), 521 544. http://doi.org/10.1086/515777 Maslach, C., Schaufeli, W., & Leiter, M. (2001). Job burnout. Annual Review of Psychology , 52 , 397 422. Maslow, A. H. (1943). A theory of human motivation. Psychological Review , 50 (4), 370 396. http ://doi.org/10.1037/h0054346 Mastracci, S., Guy, M., & Newman, M. (2011). Emotional labor and crisis response: Working . Armonk, NY: ME Sharpe. Mastracci, S., Newman, M., & Guy, M. (2006). Appraising emotion work: Determining whether emot ional labor Is valued in government jobs. The American Review of Public Administration , 36 (2), 123 138. Retrieved from http://0 arp.sagepub.com.skyline.ucdenver.edu/content/36/2/123.short

PAGE 93

93 May, P. J., & Winter, S. C. (2009). Politicians, Managers, and Stree t Level Bureaucrats: Influences on Policy Implementation. Journal of Public Administration Research and Theory , 19 (3), 453 476. http://doi.org/10.1093/jopart/mum030 Maynard Moody, S., & Musheno, M. (2003). Cops, teachers, counselors: Stories from the front lines of public service . Ann Arbor, MI: University of Michigan. Maynard Moody, S., & Portillo, S. (2010). Street level bureaucracy theory. In R. Durant (Ed.), The Oxford handbook of American bureaucracy (pp. 252 277). New York, NY: Oxford University Press . McGregor, D. (1957). The human side of enterprise. Management Review , 46 (11), 22 34. Meier, K., Mastracci, S., & Wilson, K. (2006). Gender and emotional labor in public organizations: An empirical examination of the link to performance. Public Administra tion Review , 66 (6), 899 909. Retrieved from http://0 onlinelibrary.wiley.com.skyline.ucdenver.edu/doi/10.1111/j.1540 6210.2006.00657.x/full Merkel Holguin, L., Kaplan, C., & Kwak, A. (2006). National study on differential response in child welfare . Washing ton, DC. Retrieved from https://0 scholar google com.skyline.ucdenver.edu/scholar?cluster=1498212247039603311&hl=en&as_sdt=0,6 Meyers, M., & Vorsanger, S. (2007). Street level bureaucrats and the implemenation of public policy. In B. Peters & J. Pierre (Ed s.), Handbook of public admininstration . Moore, M. (1995). Creating public value: Strategic management in government . Boston, MA: Harvard University Press. Mor Barak, M. E. (2001). Antecedents to Retention and Turnover among Child Welfare, Social Work, and Other Human Service Employees: What Can We Learn from Past Research? A Review and Metanalysis. The Social Service Review , 75 (4), 625 661. Morris, J. A., & Feldman, D. C. (1996). The dimensions, antecedents, and consequences of emotional labor. Academy of Management Review , 21 (4), 986 1010. http://doi.org/10.5465/AMR.1996.9704071861 The ethics of dissent: Managing guerrilla government . Washington, DC: CQ Press. Powell, M., Greener, I., Szmigin, I., Doheny, S., & Mills, N. (2010). Broaden ing the Focus of Public Service Consumerism. Public Management Review , 12 (3), 323 339. http://doi.org/10.1080/14719030903286615 Pressman, J., & Wildavsky, A. (1973). Implementation: How great expectations in Washington mazing that federal programs work at all, this being a saga of the Economic Development Administration as told by two sympathetic observers who seek to build morals . Berkeley, CA: University of California Press. Prottas, J. (1979). People processing: The street level bureaucrat in public service bureaucracies . Lexington, MA: Lexington Books. Pugliesi, K. (1999). The consequences of emotional labor: Effects on work stress, job satisfaction, and well being. Motivation and Emotion , 23 (2), 125 154.

PAGE 94

94 Radin, B. ( 1998). The Government Performance and Results Act (GPRA): Hydra headed monster or flexible management tool? Public Administration Review , 58 (4), 307 316. Radin, B. (2006). Challenging the performance movement: Accountability, complexity, and democratic val ues . Washington, DC: Georgetown University Press. Rafaeli, A., & Sutton, R. I. (1990). Busy stores and demanding customers: How do they effect the display of positive emotion? Academy of Management Journal , 33 (3), 623 637. http://doi.org/10.2307/256584 Ric cucci, N. (2009). The Pursuit of Social Equity in the Federal Government: A Road Less Traveled? Public Administration Review , 69 (3), 373 382. http://doi.org/10.1111/j.1540 6210.2009.01984.x Robichau, R., & Jr, L. L. (2009). The implementation of public pol icy: Still the missing link. Policy Studies Journal . Retrieved from http://0 onlinelibrary.wiley.com.skyline.ucdenver.edu/doi/10.1111/j.1541 0072.2008.00293.x/full Roethlisberger, F. (1941). Management and morale . Cambridge, MA: Harvard University Press. R oh, C. Y., Moon, M. J., Yang, S. B., & Jung, K. (2016). Linking Emotional Labor, Public Service Motivation, and Job Satisfaction: Social Workers in Health Care Settings. Social Work in Public Health , 31 (2), 43 57. http://doi.org/10.1080/19371918.2015.10879 04 Sandfort, J. R. (2003). Exploring The Structuration Of Technology Within Human Service Organizations. Administration & Society , 34 (6), 605 631. http://doi.org/10.1177/0095399702239167 Sayre, W. (1958). Premises of public administration: Past and emergin g. Public Administration Review , 18 (2), 102 105. Schein, E. (2010). Organizational culture and leadership (4th ed.). San Francisco, CA: Jossey Bass. Seery, B. L., & Corrigall, E. A. (2009). Emotional labor: links to work attitudes and emotional exhaustion. Journal of Managerial Psychology , 24 (8), 797 813. http://doi.org/10.1108/02683940910996806 Simon, H. (1997). Administrative behavior: A study of decision making processes in administrative organizations (4th ed.). New York, NY: The Free Press. Sloan, M. ( 2014). The consequences of emotional labor for public sector workers and the mitigating role of self efficacy. The American Review of Public Administration . Retrieved from http://0 arp.sagepub.com.skyline.ucdenver.edu/content/44/3/274.short Stewart, D. (19 85). Professionalism vs. democracy: Friedrich vs. Finer revisited. Public Administration Quarterly , 9 (1), 13 25. Taylor, F. (1911). . New York and London: Harper Brothers.

PAGE 95

95 Tummers, L., Bekkers, V., Vink, E., & Musheno, M. (2015). Coping during public service delivery: A conceptualization and systematic review of the literature. Journal of Public Ad ministration Research and Theory , 25 (4), 1099 1126. http://doi.org/10.1093/jopart/muu056 US Department of Health and Human Services, Administration for Children Youth and Families, Child Maltreatment 2014 . Verbeke, W., Volgerin g, M., & Hessels, M. (1998). Exploring the Conceptual Expansion within the Field of Organizational Behaviour: Organizational Climate and Organizational Culture. Journal of Management Studies , 35 (3), 303 329. http://doi.org/10.1111/1467 6486.00095 Wharton, A. (1993). The Affective Consequences of Service Work: Managing Emotions on the Job. Work and Occupations , 20 (2), 205 232. http://doi.org/10.1177/0730888493020002004 Wharton, A. (1996). Service with a smile: Understanding the consequences of emotional labo r. In C. MacDonald & C. Sirianni (Eds.), Working in the service society (pp. 91 112). Philadelphia, PA: Temple University Press. Whitaker, G. (1980). Coproduction: Citizen participation in service delivery. Public Administration Review , 40 (3), 240 246. Whi te, L. (1926). Introduction to the study of public administration . New York: Macmillan. Wilson, J. (2000). Bureaucracy: What government agencies do and why they do it . New York, NY: Basic Books. Winokur, M., Ellis, R., Drury, I., & Rogers, J. (2015). Answe ring the big questions about differential response in Colorado: Safety and cost outcomes from a randomized control trial. Child Abuse and Neglect , 39 (98). Zhan, Y., Wang, M., & Shi, J. (2015). Interpersonal Process of Emotional Labor: The Role of Negative and Positive Customer Treatment. Personnel Psychology , 69 (3), 525 557. http://doi.org/10.1111/peps.12114

PAGE 96

96 APPENDIX ID _______ ___ _______ Indicate your undergraduate degree major _________________ _____ Year Graduated _____ From what institution did you receive your undergraduate degree? ________________________ Are you a graduate of a Title IV E Child Welfare Stipend Program (i.e., a program which paid for college in return for work in the field of child welfare which also included coursework on child maltreatment and a field placement in child welfare)? Yes ___ No If YES, what year________________ Do you have an advanced degree? If so what type of advanced degree do you hold? Check a ll that apply MA/MS _______________(in what discipline) MSW or MSSW MEd. MPH M.Div. Masters in Nursing JD Ed. D. Ph.D. _______________ (in what discipline) Other ______________________(please specify) Did you complete an MSW Stipend Program through VDSS o r in another state? Yes ___________________ From where? ____ If yes, what year? ____ No Are you of Hispanic/Latina/o Origin? Yes ___ No What is your self identified race? African American/Black Caucasian/White Asian Native American Biracia l/Multiracial Other (please specify): ___________________________ What year were you born?: _______________

PAGE 97

97 Gender ___ Male ___ Female ___ Transgender ___ Gender nonconforming Over the course of your career working in child welfare, how much experience have you had serving as a caseworker managing the following types of cases? (Please do not count experience supervising others.) 1(None at all) 2(A little) 3(A moderate amount)) 4(A lot) 5(A great deal) Intake 1 2 3 4 5 Investigations 1 2 3 4 5 Family Assessments 1 2 3 4 5 CPS Ongoing 1 2 3 4 5 Foster Care 1 2 3 4 5 Adoption 1 2 3 4 5 Prevention 1 2 3 4 5 Which types of cases are on your current caseload? (Please check all that apply.) __ Intake (hotline) __ Investigations __ Family Assessments __ CPS Ongoing __ Foster Care __ Adoption __ Prevention __Other ______________________________________ Thinking about your current caseload, how many of your current cases reflect the following case types? (Please enter 0 for those case types you do not currently serve.) Type of case Number of cases Intake (hotline) ___ Investigations ___ Family Assessments ___ CPS Ongoing ___ Foster Care ___ Adoption ___ Prevention ___ Total number of cases ___ Average hours you have worked on your job per week in the past 6 months (not just what were paid for but actual hours it takes you to do your job each week) : ____39 hours or less ___ 40 49 ___ 50 59 ___ 60 69 ___ 70+

PAGE 98

98 Circle the number that corresponds to your level of agreement with each statement 1(Strongly Disagree) 2(Disagree) 3(Neutral) 4(Agree) 5(Strongly Agree) Things I learn are useful 1 2 3 4 5 It is easy for me to use what I know in new situations. 1 2 3 4 5 It does not matter if I do well or poorly, I improve after finding out 1 2 3 4 5 People I care about think I should try to know more or do things better 1 2 3 4 5 My friends will disapprove if I do not strive to improve myself 1 2 3 4 5 People that mean a lot to me expect me to learn. 1 2 3 4 5 When I have time to spare, I try to do something better 1 2 3 4 5 When I am bored, I improve my skills. 1 2 3 4 5 Through learning I am more able to do my job 1 2 3 4 5 Learning helps me to do better in what I do. 1 2 3 4 5 Using the transcription service will improve my job performance 1 2 3 4 5 Using the transcription service will increase my pr oductivity. 1 2 3 4 5 Using the transcription service will enhance my effectiveness 1 2 3 4 5 I think the transcription service will be useful. 1 2 3 4 5 The amount of time I spend on paperwork will decrease after I have mastered the transcriptio n service. 1 2 3 4 5 I will have more time for contact with families and children after I have mastered the transcription service. 1 2 3 4 5 Learning to use the transcription service will be easy for me 1 2 3 4 5 I will find it easy to get the tr anscription service to do what I desire. 1 2 3 4 5 I will find the transcription service easy to use 1 2 3 4 5 I intend to use the transcription service frequently. 1 2 3 4 5 I believe I will have the ability to transcribe notes after client visi ts. 1 2 3 4 5 I believe the transcribed notes will be accurate. 1 2 3 4 5 I believe I will have the ability to upload the transcribed notes into the case management system. 1 2 3 4 5 I believe I will have the ability to use the transcribed notes t o complete assessments, plans, case updates on progress, etc. 1 2 3 4 5 I am satisfied with my job. 1 2 3 4 5 My supervisor genuinely cares about me. 1 2 3 4 5 My supervisor gives me help when I need it. 1 2 3 4 5 My supervisor supports me i n difficult case situations. 1 2 3 4 5 My supervisor helps me learn and improve. 1 2 3 4 5 My supervisor values and seriously considers my opinions in case decision making. 1 2 3 4 5 My supervisor helps me prevent and address burnout. 1 2 3 4 5 I have tender, concerned feelings for people less fortunate than me. 1 2 3 4 5 1 2 3 4 5 When I see someone being taken advantage of, I feel kind of protective towards them. 1 2 3 4 5 1 2 3 4 5 pity for them. 1 2 3 4 5 I am quite touched by things I see happen. 1 2 3 4 5 I would describ e myself as pretty soft hearted person. 1 2 3 4 5 Co workers in my unit professionally share and learn from one another 1 2 3 4 5 Co workers in my unit share work experiences with each other to improve the effectiveness of client services. 1 2 3 4 5 Co workers in my unit encourage each other to exercise professional judgment when making decisions. 1 2 3 4 5

PAGE 99

99 Circle the number that corresponds to your level of agreement with each statement 1(Strongly Disagree) 2(Disagree) 3(Neutral) 4(Agree) 5(Stro ngly Agree) Co workers in my unit are willing to provide support and assist each other when problems arise. 1 2 3 4 5 Co workers in my unit accept support from their colleagues. 1 2 3 4 5 My supervisor assists me in setting and assessing long term c ase goals. 1 2 3 4 5 My supervisor encourages creative solutions. 1 2 3 4 5 My supervisor is knowledgeable about effective ways to work with children and families. 1 2 3 4 5 My supervisor demonstrates leadership. 1 2 3 4 5 My supervisor teaches me the skills I need in this job. 1 2 3 4 5 My supervisor requires that I use standards (i.e., criteria) to address case decisions. 1 2 3 4 5 In uncertain times, I usually expect the best. 1 2 3 4 5 If something can go wrong for me, it will. 1 2 3 4 5 1 2 3 4 5 I hardly ever expect things to go my way. 1 2 3 4 5 I rarely count on good things to happen to me. 1 2 3 4 5 Overall, I expect more good things to happen to me than bad. 1 2 3 4 5 I see myself as extraverted, enthusiastic. 1 2 3 4 5 I see myself as critical, quarrelsome. 1 2 3 4 5 I see myself as dependable, self disciplined. 1 2 3 4 5 I see myself as anxious, easily upset. 1 2 3 4 5 I see myself as open to new experiences, complex. 1 2 3 4 5 I see myself as reserved, quiet. 1 2 3 4 5 I see myself as sympathetic, warm. 1 2 3 4 5 I see myself as disorganized, careless. 1 2 3 4 5 I see myself as calm, emotional ly stable 1 2 3 4 5 I see myself as conventional, uncreative. 1 2 3 4 5 I often think about quitting my job 1 2 3 4 5 The primary reason I stay in this job is the families and children I work with 1 2 3 4 5 The primary reason I stay in t his job is that I enjoy protection and safety work 1 2 3 4 5 The primary reason I stay in this job is that I like to work for this agency (my employer) 1 2 3 4 5 Indicate how you have felt over the past month I tend to bounce back quickly afte r hard times 1 2 3 4 5 I have a hard time making it through stressful events. 1 2 3 4 5 I does not take me long to recover from a stressful event . 1 2 3 4 5 It is hard for me to snap back when something bad happens. 1 2 3 4 5 I usually come thr ough difficult times with little trouble. 1 2 3 4 5 I tend to take a long time to get over set backs in my life. 1 2 3 4 5

PAGE 100

100 Circle the number that corresponds to your level of agreement with each statement 1 (Strongly disagree) 2 (Disagree) 3(Some what disagree) 4(Neither agree nor disagree) 5 (Somewhat agree) 6 (Agree) 7 (Strongly agree) Even working overtime, I cannot finish all of my work 1 2 3 4 5 6 7 My caseload is too high 1 2 3 4 5 6 7 I have too many cases to do a good job, yet I am expected to do so. 1 2 3 4 5 6 7 I cannot spend enough time with my clients 1 2 3 4 5 6 7 It is difficult for me to keep up with agency policies and guidelines 1 2 3 4 5 6 7 I will probably look for a new job in the next year. 1 2 3 4 5 6 7 I believe I have had adequate training to help me make the right decision about the safety and well being of my clients 1 2 3 4 5 6 7 I am worried that one of my cases may draw media attenti on 1 2 3 4 5 6 7 I have known caseworkers that have been disciplined or fired because of real or perceived mistakes 1 2 3 4 5 6 7 If a child in one of my cases is harmed, I worry that I will be disciplined or fire d 1 2 3 4 5 6 7 If a child in one of my cases is harmed, I believe the agency will conduct a thorough investigation into what a happened before assigning blame 1 2 3 4 5 6 7 I know my supervisor will be supportive of me and the decisions I made if a child is harmed in one of my cases 1 2 3 4 5 6 7 I know my leadership will be supportive of me and the decision I made if a child is harme d in one of my cases 1 2 3 4 5 6 7 I hide my true feelings so as to appear pleasant at work. 1 2 3 4 5 6 7 In my job I act confident and self assured regardless of how I actually feel. 1 2 3 4 5 6 7 in an appropriate way. 1 2 3 4 5 6 7 I let my true feelings show when working with families and children. 1 2 3 4 5 6 7 It is easier for me to show my true feelings than to pre tend. 1 2 3 4 5 6 7 I am good at expressing how I feel. 1 2 3 4 5 6 7 I am good at getting people to calm down. 1 2 3 4 5 6 7 Dealing with emotionally charged issues is an important part of my job. 1 2 3 4 5 6 7 In my job I am good at dealing with emotional issues. 1 2 3 4 5 6 7 I try to actually experience the emotions that I must show to families and children 1 2 3 4 5 6 7 I work hard to actually feel the emotions that I need to show to families and children 1 2 3 4 5 6 7 I work at developing the feelings inside of me that I need to show to families and children 1 2 3 4 5 6 7

PAGE 101

101 How often do you have any of the following experiences as a result of your job? 1(Never) 2 (once in a while) 3 (rarely) 4 (Sometimes) 5 (Often) 6 (Usually) 7 (Always) Being tired 1 2 3 4 5 6 7 Feeling depressed 1 2 3 4 5 6 7 Having a good day 1 2 3 4 5 6 7 Being physically exhausted 1 2 3 4 5 6 7 Being emotionally exhausted 1 2 3 4 5 6 7 Being happy 1 2 3 4 5 6 7 Being wiped out 1 2 3 4 5 6 7 1 2 3 4 5 6 7 Being unhappy 1 2 3 4 5 6 7 Feeling run down 1 2 3 4 5 6 7 Feeling trapped 1 2 3 4 5 6 7 Feeling worthless 1 2 3 4 5 6 7 Being weary 1 2 3 4 5 6 7 Being troubled 1 2 3 4 5 6 7 Feeling disillusioned and resentful 1 2 3 4 5 6 7 Being weak and susceptible to illness 1 2 3 4 5 6 7 Feeling hopeless 1 2 3 4 5 6 7 Feeling rejected 1 2 3 4 5 6 7 Feeling optimistic 1 2 3 4 5 6 7 Feeling energetic 1 2 3 4 5 6 7 Feeling anxious 1 2 3 4 5 6 7 How likely will you be to do the following? 1 (Not at all likely) 3 (Somewhat likely) 5 (Qu ite likely) 7 (Extremely likely) How likely is it that you will actively look for a new job in the next year 1 2 3 4 5 6 7 1 ( Terrible) 2 (Bad) 3 (Not so good) 4 (So so) 5 (Good) 6 (Very good) 7 ( Excellent) What are the chances that you will leave your current job for another job in the agency sometime in the next six months 1 2 3 4 5 6 7 What are the chances that you will leave your current job for another job in the agency sometime in the next year 1 2 3 4 5 6 7 What are the chances that you will leave your current job for another job outside the agency sometime in the next six months 1 2 3 4 5 6 7 What are the chances that you will leave your current job for another job outside the agency sometime in the next year 1 2 3 4 5 6 7 1 (Never) 2 (Seldom) 3 (Sometimes) 4 (Usually) 5 (Always) How often do you feel you are successful in helping families make positive changes in their lives 1 2 3 4 5 How often do you feel you are successful in helping children and families achieve permanency 1 2 3 4 5 How often do you feel you are successful in helping children and families remain safe 1 2 3 4 5 How often do you feel your efforts have co ntributed to the physical, mental, and educational well being of children and families 1 2 3 4 5 How often do you receive thanks or expressions of gratitude from children or families you have worked with 1 2 3 4 5 How often do you think about quitting your job? 1 2 3 4 5

PAGE 102

102 For the next 6 items, think of your WORK SITUATION when responding to each question. In the last three months, how often have you: 1 (Never) 2 (Almost Never) 3 (Sometimes) 4 (Fairly Often) 5 (Very Often) been upset because of something that happened unexpectedly at work? 1 2 3 4 5 1 2 3 4 5 found that you could not cope with all the things you had to do? 1 2 3 4 5 been angered because of things that happened that were outside your control? 1 2 3 4 5 found yourself thinking about things that you had to accomplish? 1 2 3 4 5 felt difficulties were piling up so high that you could not overcome them? 1 2 3 4 5 Below you will find a list of statements. Please read each statement carefully and decide if that statement describes you or not. If it describes you, circle the word "true"; if not, circle the word "false". I sometimes litter. True False I always admit my mistakes openly and face the potential negative consequences. True False In traffic I am always polite and considerate of others. True False True False I take out my bad moods on others now and then. True False There has been an occasion when I took advantage of someone else. True False In conversations I always listen attentively and let others finish their sentences. True False I never hesitate to help someone in case of emergency. True False When I have made a promise, I keep it no ifs, ands, or buts. True False I occasionally speak badly of others behind their back. True False xpense. True False I always stay friendly and courteous with other people, even when I am stressed out. True False During arguments I always stay objective and matter of fact. True False There has been at least one occasion when I failed to retu rn an item that I borrowed. True False I always eat a healthy diet. True False Sometimes I only help because I expect something in return. True False

PAGE 103

103 In the following six items you will be presented with two statements. Please choose between them. We understand that you might agree with both statements, but please choose only one statement, the one that best reflects YOUR general work focus and beliefs. You will see a statement more than once, but each pairing is differen t. There are no right or wrong answers. After selecting the statement that best reflects your general work focus and beliefs, please rate the strength of your preference for that statement, over the other, on the scale of Very Weak to Very Strong. For ex ample, if you strongly prefer Statement A OVER Statement B, then you will first select "Statement A" and then select "Very Strong." Or, if you barely prefer Statement A over Statement B, then you would first select "Statement A" and then select "Very Weak. " Item 1 Please select the statement, A or B that best reflects your general work focus and beliefs. Then, rate the strength of your preference for that statement. (A) Work should be focused on keeping the family together... OR... (B) Child welfare w orkers should be willing to advocate for the child. ____ A ____ B Very weak Weak Moderate Strong Very Strong Preference for A or B chosen above 1 2 3 4 5 Item 2 Please select the statement, A or B that best reflec ts your general work focus and beliefs. Then, rate the strength of your preference for that statement. (A) Families are the best place for children to achieve their full potential... OR... (B) There is a need to ensure the physical and emotional well bei ng of all children. ____ A ____ B Very weak Weak Moderate Strong Very Strong Preference for A or B chosen above 1 2 3 4 5 Item 3 Please select the statement, A or B that best reflects your general work focus and bel iefs. Then, rate the strength of your preference for that statement. (A) The child is the client and all other work is secondary... OR... (B) Work should be focused on keeping the family together. ____ A ____ B Very weak Weak Moderate Stro ng Very Strong Preference for A or B chosen above 1 2 3 4 5 Item 4 Please select the statement, A or B that best reflects your general work focus and beliefs. Then, rate the strength of your preference for that statement. (A) C hildren's rights should be safeguarded so they can achieve their full potential... OR... (B) The family's right to guide the development of their children should be safeguarded. ____ A ____ B Very weak Weak Moderate Strong Very Strong Prefere nce for A or B chosen above 1 2 3 4 5

PAGE 104

104 Item 5 Please select the statement, A or B that best reflects your general work focus and beliefs. Then, rate the strength of your preference for that statement. (A) Families are the best place for children to achieve their full potential... OR... (B) The state has a responsibility to protect children . ____ A ____ B Very weak Weak Moderate Strong Very Strong Preference for A or B chosen above 1 2 3 4 5 Item 6 Please select the statement, A or B that best reflects your general work focus and beliefs. Then, rate the strength of your preference for that statement. (A) Families are the best place for children to achieve the ir full potential... OR... (B) Children's rights should be safeguarded so they achieve their full potential. ____ A ____ B Very weak Weak Moderate Strong Very Strong Preference for A or B chosen above 1 2 3 4 5