...

The nature, causes and consequences of harm in emotionally-demanding occupations RR610

by user

on
Category: Documents
21

views

Report

Comments

Transcript

The nature, causes and consequences of harm in emotionally-demanding occupations RR610
Health and Safety
Executive
The nature, causes and consequences
of harm in emotionally-demanding
occupations
Prepared by Birkbeck College University of London
for the Health and Safety Executive 2008
RR610
Research Report
Health and Safety
Executive
The nature, causes and consequences
of harm in emotionally-demanding
occupations
Professor Rob B Briner BSc, MPhil, PhD
Dr Sarah Poppleton MA, MSc, PhD
Dr Sarah Owens BA, PhD
Dr Tina Kiefer lic phil, Dr phil
Department of Organisational Psychology
Birkbeck College
University of London
London WC1E 7HX
Traditional approaches to understanding psychosocial job characteristics and well-being have been quite general in that
they explore links between general job characteristics such as workload and control on workers in many different sorts of
occupations. One example of a more specific approach can be found in research into emotional labour - the requirement
to regulate both feelings and the expression of feelings for organizational goals. Early research into emotional labour
focused on customer service workers (CSW) but has more recently also considered human service workers (HSW) such
as nurses and social workers. A more specific approach to thinking about the outcomes of demanding psychosocial
job characteristics can be found in research on burnout which is thought to have three elements: emotional exhaustion,
depersonalization (also labelled cynicism), and (low) accomplishment (also called professional efficacy). Much recent
research has started to explore the links between emotional demands and burnout. The main aim of this project is
therefore to explore the nature of such links through undertaking three distinct tasks. The first is a literature review of
evidence and theory while the second two tasks comprise two empirical studies examining several key issues in burnout
research.
This report and the work it describes were funded by the Health and Safety Executive (HSE). Its contents, including any
opinions and/or conclusions expressed, are those of the authors alone and do not necessarily reflect HSE policy.
HSE Books
© Crown copyright 2008
First published 2008
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means (electronic, mechanical, photocopying, recording or otherwise) without the prior written permission of the copyright owner.
Applications for reproduction should be made in writing to:
Licensing Division, Her Majesty’s Stationery Office,
St Clements House, 2-16 Colegate, Norwich NR3 1BQ
or by e-mail to [email protected]
ii
ACKNOWLEDGEMENTS
The authors would like to thank all the organizations and participants who gave their time to share
their experiences of emotional labour with us. We are indebted to Gabriele Woelfle who prepared the
final version of this report. We would also like to thank Peter Kelly and Kevin Mantle from HSE for
their guidance and support throughout this project.
iii
iv
Contents
Executive summary .................................................................................................................................... ix
Background
ix
Literature review
ix
Study 1: Questionnaire survey
ix
Study 2: Daily diaries and interviews
x
Conclusions and recommendations
xi
1 Introduction ...........................................................................................................................................1
1.1
Rationale and background
1
1.1.1 Emotional labour .....................................................................................................................1
1.1.2 Emotions at work.....................................................................................................................2
1.1.3 Burnout.....................................................................................................................................2
1.2
Objectives and research questions
3
1.2.1 Literature review .....................................................................................................................3
1.2.2 Empirical work ........................................................................................................................3
2 Literature review...................................................................................................................................4
2.1
Introduction and overview
4
2.2
Correlates and causes of burnout
4
2.2.1 Demographics ..........................................................................................................................4
2.2.2 Occupation ...............................................................................................................................6
2.2.3 Personality................................................................................................................................7
2.2.4 Traditional job stressors ..........................................................................................................9
2.2.5 Summary ................................................................................................................................13
2.3
Burnout process
14
2.3.1 Introduction............................................................................................................................14
2.3.2 Sequential order of the three burnout dimensions...............................................................14
2.3.3 Other research into the processes underlying burnout ........................................................16
2.3.4 Summary and conclusion ......................................................................................................18
2.4
Emotional labour and burnout in customer service and human service work
19
2.4.1 Introduction............................................................................................................................19
2.4.2 Emotional labour ...................................................................................................................19
2.4.3 Emotional demands of customer service and human service work....................................21
2.4.4 Summary ................................................................................................................................23
2.5
Relationship between emotional demands and burnout in customer service workers and
human service workers
23
2.5.1 Introduction............................................................................................................................23
2.5.2 Customer contact and burnout ..............................................................................................23
2.5.3 Customer aggression .............................................................................................................24
2.5.4 Relationship between emotional labour and burnout ..........................................................25
2.5.5 Summary and conclusions ....................................................................................................29
2.6
Burnout interventions
29
2.6.1 Introduction............................................................................................................................29
2.6.2 Individual level burnout interventions .................................................................................30
2.6.3 Interventions at the individual-organizational interface .....................................................30
2.6.4 Organizational-level interventions .......................................................................................31
2.6.5 Burnout workshops................................................................................................................32
2.6.6 Concluding summary ............................................................................................................32
2.7
Summary and conclusions
33
3 Study 1: Questionnaire survey .........................................................................................................35
3.1
Purpose of study 1 and main research questions
35
3.2
Variables
35
3.2.1 Person characteristics ............................................................................................................35
3.2.2 Job characteristics..................................................................................................................36
3.2.3 Emotional demands ...............................................................................................................36
3.2.4 Outcome variables .................................................................................................................36
v
3.3
Describing the three human service work groups
38
3.3.1 Recruitment of the sample and the three main human service work groups .....................38
3.3.2 Person Factors: How can the main groups be characterized?.............................................38
3.3.3 How can their jobs be characterized?...................................................................................40
3.3.4 Summary ................................................................................................................................42
3.4
Research question 1: What is the level of burnout and well-being in the key human service
occupations in the UK?
43
3.4.1 RQ1a: How do these levels of burnout and well-being compare to other samples?.........43
3.4.2 RQ1b: How do the levels of burnout and well-being compare across different human
service occupations? ..........................................................................................................................44
3.4.3 Summary and conclusion ......................................................................................................44
3.5
Research question 2: Which key variables are associated with burnout and well-being? 45
3.5.1 RQ2a: Which key variables are associated with burnout and well-being?........................45
3.5.2 Brief summary .......................................................................................................................46
3.5.3 RQ2b: Are different key variables associated with burnout and well-being in different
human service occupations?..............................................................................................................47
3.5.4 Summary ................................................................................................................................49
3.6
Summary and conclusion
49
4 Study 2: Daily diaries and interviews...............................................................................................51
4.1
Introduction and overview
51
4.1.1 Method and design ................................................................................................................51
4.1.2 Data analysis..........................................................................................................................52
4.1.3 Brief overview of results.......................................................................................................52
4.2
Events engendering positive emotions
55
4.2.1 Enjoyable client contact ........................................................................................................55
4.2.2 Meeting or interactions with clients which had a positive outcome ..................................55
4.2.3 Client contact in which the client gave positive feedback ..................................................56
4.2.4 Other types of positive client interaction .............................................................................56
4.2.5 Other types of positive, work-related event.........................................................................56
4.2.6 Positive non-work events ......................................................................................................57
4.2.7 Concluding summary ............................................................................................................57
4.3
Events engendering negative emotions
57
4.3.1 An interaction with a difficult client ....................................................................................57
4.3.2 A client who is going through a negative experience .........................................................59
4.3.3 Verbal abuse ..........................................................................................................................61
4.3.4 Paperwork or “Admin” .........................................................................................................62
4.3.5 Problems with the working environment .............................................................................63
4.3.6 A colleague or subordinate’s handling of a situation with a client ....................................64
4.3.7 Difficult interactions with colleagues or third parties.........................................................66
4.3.8 Role of negative non-work events........................................................................................67
4.3.9 Concluding Summary............................................................................................................67
4.4
Emotional demands
68
4.4.1 Emotionally demanding situations .......................................................................................68
4.4.2 Emotion work in particular events .......................................................................................71
4.4.3 Concluding Summary............................................................................................................77
4.5
Stress and burnout
78
4.5.1 Diarists’ thoughts on burnout ...............................................................................................78
4.5.2 Stress and burnout-type emotions in human service work .................................................79
4.5.3 Concluding summary ............................................................................................................84
4.6
Additional findings
84
4.6.1 Detrimental consequences of human service work .............................................................84
4.6.2 Buffers and coping ................................................................................................................85
4.6.3 Concluding summary ............................................................................................................88
4.7
Summary of the qualitative results
89
5 Conclusions and recommendations ..................................................................................................91
5.1
Literature review
91
vi
5.2
5.3
5.4
Study 1: Questionnaire survey
93
Study 2: daily diaries and interviews
94
Recommendations
95
5.4.1 For research............................................................................................................................95
5.4.2 For practice ............................................................................................................................96
6 Appendices............................................................................................................................................97
6.1
Appendix 1: Measures used in the questionnaire survey
97
6.2
Appendix 2 : Daily diary
102
6.3
Appendix 3: Interview schedule topic guide
108
6.4
Appendix 4: The diary coding template
112
6.5
Appendix 5: Additional interview codes
120
7 References.......................................................................................................................................... 124
vii
viii
EXECUTIVE SUMMARY
BACKGROUND
Traditional approaches to understanding psychosocial job characteristics and well-being have been
quite general in that they explore links between general job characteristics such as workload and
control on workers in many different sorts of occupations. One example of a more specific approach
can be found in research into emotional labour – the requirement to regulate both feelings and the
expression of feelings for organizational goals. Early research into emotional labour focused on
customer service workers (CSW) but has more recently also considered human service workers
(HSW) such as nurses and social workers. A more specific approach to thinking about the outcomes
of demanding psychosocial job characteristics can be found in research on burnout which is thought to
have three elements: emotional exhaustion, depersonalization (also labelled cynicism), and (low)
accomplishment (also called professional efficacy). Much recent research has started to explore the
links between emotional demands and burnout. The main aim of this project is therefore to explore
the nature of such links through undertaking three distinct tasks. The first is a literature review of
evidence and theory while the second two tasks comprise two empirical studies examining several key
issues in burnout research.
LITERATURE REVIEW
The literature review examined the following six areas:
1.
2.
3.
4.
The main correlates and causes of burnout in human service work
Traditional work stressors that have been found to relate to burnout (and related outcomes)
The pathways through which these stressors lead to these end points
The extent to which and how the emotional demands of human service occupations are
different from emotional demands of other related and un-related occupations
5. The relationship between emotional demands and burnout (and related outcomes) in customer
service work and human service work
6. Interventions which may ameliorate the impacts of such emotional demands on outcomes
As the vast proportion of studies in this field is cross-sectional it was in nearly all cases only possible
to examine associations. Most demographic and occupational factors (area 1) correlate quite
inconsistently with burnout thought the few studies which examine personality do find consistent
relationships between negative affectivity and burnout. Some traditional work stressors (area 2), such
as role ambiguity, job demands, and social support have shown reasonably consistent associations
with burnout. While there is a general belief that burnout occurs in a sequence of stages (area 3), such
that its different elements appear in a particular order, the available evidence does not support any
specific sequence. Other approaches to burnout processes emphasize intra-personal processes such as
emotion regulation. The nature of emotional demands (area 4) does seem to be different in HSW and
CSW probably due to the different motivations behind and involvement in the emotional labour
required. While emotional demands such as the quantity of customer interaction (area 5) do not seem
to relate to burnout other demands, such as surface acting (or emotional dissonance), do seem to be
associated. While in some senses HSW have the most emotionally demanding jobs they also have the
lowest levels of emotional dissonance. There is very little research on burnout interventions (area 6)
but that which has been done suggests some of these interventions can be effective.
STUDY 1: QUESTIONNAIRE SURVEY
A survey of 398 participants (around 40% response rate) was undertaken. The sample had three sub
groups: CSW, mainly call centre staff (N=110); HSW-Other, mainly human service workers, such as
social and care workers, but excluding teachers (N=184); and, Teachers, mainly primary school heads
(M=104). The following research questions and sub-questions were addressed:
ix
1. What is the level of burnout and well-being in key human service occupations in the UK?
• RQ1a: How do these levels of burnout and well-being compare to other samples?
• RQ1b: How do the levels of burnout and well-being compare across different human service
occupations?
2. Which key variables are associated with burnout and well-being?
• RQ2a: Which key variables are associated with burnout and well-being?
• RQ2b Are different key variables associated with burnout and well-being in different human
service occupations?
In order to do this measures of person characteristics (e.g. demographics and personality), traditional
job characteristics (e.g. role conflict and autonomy), emotional demands (e.g. surface acting and
requirement to display positive emotions), and well-being (burnout and GHQ12).
The levels of burnout and GHQ12 scores for the sample as a whole were broadly similar to other
national and international norms. CSW show the highest levels of cynicism (referred to as
depersonalization in the literature review), the lowest levels of professional efficacy (referred to as
accomplishment in the literature review) but not the highest levels of exhaustion (which was found in
Teachers). There were no differences across the groups in GHQ12 scores.
After controlling for person characteristics and traditional job demands (which accounted for large
proportions of the variance), emotional demands did account for significant though small proportions
of variance in the three burnout dimensions but not GHQ12. The pattern of relationships between
specific emotional demands and burnout was different across the three groups. As the study was
cross-sectional no causal inferences can be made.
STUDY 2: DAILY DIARIES AND INTERVIEWS
This study addressed the following two research questions:
1. What kinds of events and situations at work are found to be particularly emotionally
demanding at work and in what ways?
2. What are the pathways and mechanisms through which specific emotional demands lead to
particular kinds of psychological health outcomes?
24 participants (including 10 head teachers, 4 social workers and assistants, and 3 care workers)
completed daily diaries about events at work and at home for two consecutive weeks and the majority
of participants took part in a follow-up face-to-face or telephone interview. The diaries contained a
checklist of emotions and a series of questions asking for descriptions of positive and negative events
and during the interviews participants were asked for further details about the events they had
described.
More events engendering positive than negative emotions were recorded in the diaries. The types of
positive event recorded included positive meetings with clients and those that had successful
outcomes. Types of negative events included dealing with difficult clients and being subject to verbal
abuse. However, there seemed to be particular features of negative events which help explain why
they were experienced as such, including not feeling listened to in one’s professional role and feeling
unable to resolve a negative situation.
Emotion work was often seen as something which could be both difficult and very rewarding where
the participant had felt able to turn a very negative situation into a more positive one and help resolve
clients’ problems. On the other hand certain types of emotional labour were viewed only as difficult
such as interactions where participants felt they had to suppress anger or were unable to voice their
opinion. Participants who appeared to be particularly distressed also reported emotionally demanding
situations at home as well as at work.
x
CONCLUSIONS AND RECOMMENDATIONS
Literature review: Most research about emotional labour and burnout is cross-sectional meaning that
is not possible to make causal inferences. A number of features of burnout, including its inconsistent
relationships with person and occupational factors and the mixed evidence about how it develops call
into question aspects of its validity and reliability as a construct. Researchers have different views
about how emotional labour is best construed and the nature of the relationship between emotional
labour and burnout appears to differ across HSW and CSW. There are so few studies of burnout
interventions it is difficult to draw any meaningful conclusions.
Study 1: Though scores on burnout and GHQ12 were not different from available norms for similar
groups the three groups differed in a number of important respects including the level of different
aspects of emotional labour reported and levels of different dimensions of burnout. Emotional
demands were found to be related to burnout (after controlling for person factors and traditional job
stressors) but the patterning of the relationships varied across the three groups again demonstrating the
importance of contextual factors in understanding emotional labour.
Study 2: The double-edged nature of emotional labour for participants provided important suggestions
for why it may be the case that HSW report higher levels of emotional labour but not necessarily
higher burnout. Some light was also shed on why and how negative interactions with clients may be
experienced as harmful. While all job characteristics depend to some extent on their context the
results suggest that emotional labour plays out very differently depending not only on context but the
specific nature of the events and episodes encountered by employees.
Recommendations for research and practice: Undertaking more longitudinal and less cross-sectional
research is essential for understanding links between emotional labour and burnout. Future research
should focus on theoretically unpacking both burnout and emotional labour in order to develop clearer
and more elaborate constructs and find ways of incorporating context more fully. Given the current
state of knowledge it is difficult to make sound practical suggestions. However, the available
evidence suggests that emotional labour cannot be managed in the same way as traditional ‘stressors’
and that practitioners should start to engage more with the idea of emotional labour when looking at
how work affects well-being.
xi
xii
1 INTRODUCTION
1.1
RATIONALE AND BACKGROUND
Significant advances in our understanding of the relationships between psychosocial work conditions
and well-being have been made through examining the effects of general job characteristics, such as
workload and control, on workers in a wide range of workplaces and occupations. However, it also
seems likely that there are important differences in workplaces and occupations which mean that the
nature, meaning, and effects of these general job conditions will depend on specific organizational and
job contexts. While it is useful therefore, up to a point, to look at these general job characteristics and
their general effects on workers it is also important to look at more specific job characteristics and the
specific effects they may have on worker well-being. For example, particular kinds of workload may
have particular effects on specific aspects of well-being.
One example of a more specific type of workload is emotional work demands. While work has always
demanded the use of physical and cognitive (or mental) resources it can also require the use of
emotional resources. This sort of work demand has been traditionally associated with those whose
jobs require them to deal with the pain and suffering and needs of others such as nurses, doctors, other
medical staff and social workers. Collectively, such work is sometimes referred to as human service
work. More recently, however, it has also been identified in customer service work – jobs in which
effective performance requires the display of particular emotions or the handling of the emotions in
those with whom they are paid to interact. Though, as will be mentioned later, this distinction is not
always a clear one.
HSE’s 2002 Strategic Research Outlook identified human service occupations, such as nursing,
teaching and social work, as an important area for further investigation in part because of the specific
nature and high level of emotional demands placed on those in such occupations. It is worth noting
that with the expansion of the service sector over the past few decades the number of people now
employed in jobs that have some element of emotional demands, such as those in call centres, has
increased dramatically. In the same document, it is also noted that: “HSE is particularly interested in
innovative thinking to take forward the identified issues rather than full technical solutions”
(paragraph 26, p.5).
The general purpose of this research is therefore to explore the nature and effects of emotional
demands at work. Before setting out the research objectives and research questions three key ideas
that form the background to this research will be discussed briefly.
1.1.1
Emotional labour
As indicated earlier, the physical and cognitive demands of work are not the only resources required in
some jobs and occupations. Hochschild (1983) was the first to note that some jobs require workers to
also use their emotional resources in displaying or withholding particular emotions. Her initial
research focused on air attendants but since that time emotional labour has been investigated in a wide
range of other occupations such as nurses, counsellors, supermarket cashiers, hairdressers, debt
collectors, social workers, and criminal interrogators. What all these jobs have in common is that
workers must display specific emotions at particular times in order to complete job tasks effectively.
A central mechanism though which emotional labour is considered to be harmful to well-being is
because of the possible discrepancy between displayed and felt emotions. A nurse, for example, may
be expected to display concern towards patients. If the nurse actually feels little concern but
nonetheless are still required to display such feelings this creates a discrepancy which, it is argued, is
very effortful to maintain and ultimately damaging to well-being. Clearly, this is more likely to occur
when a person has been in an occupation for some time. The disturbing nature of such emotional
1
discrepancies may be particularly marked in occupations where there is a strong vocational element,
such as teaching and nursing, as workers in these professions are more likely to believe it is important
to actually feel the emotions they should display and to not ‘fake it’. It is important to note that many
jobs which involve intense emotional labour also have this strong vocational draw.
As will be discussed later, the concept of emotional labour is central to developing an understanding
of how emotional demands may affect worker well-being.
1.1.2
Emotions at work
In parallel to the shift in thinking about demands of work from the physical and cognitive to the
emotional there has been a move towards considering the emotional aspects of workplaces,
organizations and jobs. The past decade has seen a very significant increase in research exploring the
causes and consequences of specific emotions at work (e.g. Briner, 1999; Briner & Totterdell, 2002).
In part this is because the emotional nature and demands of work have been largely ignored but also
because existing ideas such as job satisfaction and job stress are viewed by some as too general and
unfocused to now be helpful.
One lesson from the body of work on emotions in organizations for the current research is the
importance of specificity and considering the ways in which specific kinds of work events and work
demands are likely to lead to particular kinds of emotional reactions both in the shorter and longerterm. In other words, rather than thinking of stressful job conditions, in general, as having many
different kinds of possible outcomes it may be more productive to view these different outcomes as
being caused by more specific and different kinds of emotional states such as anger, fatigue, anxiety,
or sadness.
1.1.3
Burnout
One example of a specific type of outcome can be found in the ideas of burnout. As the term implies,
the notion is that workers in jobs with high levels of emotional demands eventually lose the ability or
energy to meet the demands of the job. For several decades, burnout research has examined its causes
and also more recently how it may be prevented (Maslach, Schaufeli & Leiter, 2001). Most
definitions of burnout suggest it has three components: Emotional exhaustion, lack of personal
accomplishment and depersonalization. Most of the literature reviewed in this report refers to these
three components. Note, however, that in Study 1 we had to use a scale by Maslach and colleagues
that uses a slightly different terminology, namely emotional exhaustion, cynicism (instead of
depersonalization) and professional efficacy (instead of accomplishment).
While there are debates concerning the precise nature of burnout and how it can be assessed there is
broad agreement that it is the emotional demands of human service work which may, over a number of
years or decades cause burnout or burnout-like outcomes. It is also considered that other outcomes
such as early retirement, absence, physical ill-health, and changing careers may be involved.
Although clearly related to human service occupations and to burnout there has thus far been relatively
little attempt to link these two areas of research together although initial work doing so has proved
fruitful (e.g. Brotheridge & Grandey, 2002).
The current research is therefore taking place in the context of a growing body of work into emotional
labour, emotions at work more generally, and the burnout as a specific consequence of emotional
demanding work.
2
1.2
OBJECTIVES AND RESEARCH QUESTIONS
As already stated, the main general purpose of the research is to examine the nature and effects of
emotional demands at work. In order to do this, the research has three different but complementary
elements, a literature review and two empirical studies, which are described below.
1.2.1
Literature review
Compared to research on traditional psychosocial job characteristics that are thought to be harmful
such as workload or control, work on emotional demands is relatively new though expanding rapidly.
For this reason the first element of the research aims to conduct a literature review in order to identify
existing thinking and evidence concerning the following five issues, some of which were adapted
directly from HSE’s 2002 Strategic Research Outlook:
1.
2.
3.
4.
The main correlates and causes of burnout in human service work
Traditional work stressors that have been found to relate to burnout (and related outcomes)
The pathways through which these stressors lead to these end points
The extent to which and how the emotional demands of human service occupations are
different from emotional demands of other related and un-related occupations
5. The relationship between emotional demands and burnout (and related outcomes) in customer
service work and human service work
6. Interventions which may ameliorate the impacts of such emotional demands on outcomes
1.2.2
Empirical work
Given the general purpose of the research it is important to collect data which allows some inference
to be made both about the presence and effects of emotional demands in human service workers and
also about the processes through which such demands may affect well-being. It is important to note
that because of the time-scales and budgets involved it has not been possible to conduct a longitudinal
study of the duration required to fully explore the effects of emotional demands on well-being which
are generally assumed to occur over years or even decades. However, the two studies conducted here
will in combination help to at least suggest some possible causal mechanisms. It should also be noted
that research in this area (which will be reviewed later), in common with other research examining the
impact of psychosocial job conditions on well-being, is characterized by a dearth of studies adopting
designs which permit reasonable inference of causality.
In order to explore the presence and effects of job demands a cross-sectional questionnaire study will
be conducted which aims to address two main research questions:
1. What is the level of burnout and well-being in key human service occupations in the UK?
2. Which key variables are associated with burnout and well-being?
In order to understand more about processes through which emotional demands may affect will being
a qualitative daily diary study and interviews will be conducted which address the following research
questions:
3. What kinds of events and situations at work are found to be particularly emotionally
demanding at work and in what ways?
4. What are the pathways and mechanisms through which specific emotional demands lead to
particular kinds of psychological health outcomes?
3
2 LITERATURE REVIEW
2.1
INTRODUCTION AND OVERVIEW
The main aims of this literature review are to identify existing evidence and thinking around the
following six issues:
1.
2.
3.
4.
The main correlates and causes of burnout in human service work
Traditional work stressors that have been found to relate to burnout (and related outcomes)
The pathways through which these stressors lead to these end points
The extent to which and how the emotional demands of human service occupations are
different from emotional demands of other related and un-related occupations
5. The relationship between emotional demands and burnout (and related outcomes) in customer
service work and human service work
6. Interventions which may ameliorate the impacts of such emotional demands on outcomes
The aim of the review is not, therefore, to generate research questions for the two empirical studies
reported later but, rather, to critically review the available theory and evidence around a number of key
issues within burnout research. However, the focus of the two empirical studies through to some
extent exploratory was also guided by the findings of the literature review.
While this will be discussed in more detail later, it is important to note again at this point that the term
human service work refers here to any worker whose job involves the day-to-day care of individuals in
need (e.g. teachers, nurses, social workers). Where relevant, this review will also consider customer
service work where employee emotional display and handling the emotions of customers and clients
are important parts of the job. However, as this review focuses largely on burnout as an outcome and
the processes through which burnout may occur, the vast majority of material in the review is drawn
from research conducted with human service workers. This is simply because burnout, as a
consequence of emotional labour, has been much more commonly studied in human service rather
than customer service workers as it assumed that human service occupations will experience the
greatest levels of emotional demands and hence the greatest levels of burnout.
2.2
CORRELATES AND CAUSES OF BURNOUT
The purpose of this section is to address the first and second literature review issues. Sections 2.2.1 to
2.2.3 examine the risk factors that are associated with burnout and Section 2.2.4 examines the
relationships between traditional job stressors and burnout. It should be noted that, in common with
other areas of research about the influence of personal and work factors on work-related well being,
almost all the research is cross-sectional and hence it is not possible in most cases to draw any causal
inferences. Where the study under discussion is longitudinal this will be mentioned.
Here we organize the correlates and causes of burnout under four main headings: Demographics,
occupation, personality, and traditional job stressors.
2.2.1
Demographics
Gender
The relationship between gender and burnout is not clear-cut. Overall, there do not seem to be reliable
differences in burnout levels between men and women, and several studies have found no gender
difference across any of the three burnout dimensions (e.g. Ben-Zur & Yagil, 2005; Dormann & Zapf,
2004; Kop, Euwema & Schaufeli, 1999; Rafferty, Friend & Landsbergis, 2001).
4
Where differences have been found, they tend to suggest that the genders are differentially susceptible
to different dimensions of burnout, specifically that women may be more prone to emotional
exhaustion and men more prone to depersonalization. This pattern was found in an early study by
Maslach and Jackson (1981), and has been supported by several other studies. For example, Johnson
(1991) found that female police officers scored relatively high on emotional exhaustion, whereas
males scored relatively high on depersonalization, and Lewig and Dollard (2003) found being female
was significantly associated with emotional exhaustion in a sample of call centre workers. Deery,
Iverson and Walsh (2002) also found that being female was associated with higher emotional
exhaustion in their bivariate analysis, although this relationship was not significant in their
multivariate structural equation model in which numerous other variables were taken into account. In
partial support of these gender differences, Dollard, Winefield, Winefield and de Jonge (2000) found
depersonalization was associated with being male, but that gender was unrelated to emotional
exhaustion and personal accomplishment.
However, the empirical data are mixed. A good number of studies have reported no gender difference
in emotional exhaustion (e.g. de Jonge & Schaufeli, 1998; de Rijk, LeBlanc, Schaufeli & de Jonge,
1998; Dickter & Sin, 2004; Grandey, Giebels & Janssen, 2005; Holman, Chissick & Totterdell, 2002;
Tummers, Landeweerd & van Merode, 2002; Van Vegchel, de Jonge, Bakker & Schaufeli, 2002) and,
as detailed above, some studies have found no gender differences in any of the burnout dimensions.
Occasionally reverse findings are reported, for example, Price and Spence (1994) found higher levels
of emotional exhaustion amongst men in their sample of drug and alcohol workers.
Where gender differences are found to exist, it is possible that this may reflect a confounding of
gender and occupation. It has been suggested that predominantly female jobs, such as nursing, suffer
more from burnout, and that predominantly male occupations, such as policing, are more prone to
depersonalization (Maslach & Jackson, 1981; Maslach, Schaufeli & Leiter, 2001).
Age
In many of the early studies, younger employees consistently reported higher levels of burnout (e.g.
Maslach & Jackson, 1981; 1985) and the demographic norms reported in the Maslach Burnout
Inventory Manual (Maslach, Jackson & Leiter, 1996) show a clear gradient of burnout decreasing with
age across all three dimensions. However, more recently there have been mixed findings, and several
studies have found no relationship between age and burnout.
Consistent with the early pattern of findings, Zohar (1997) and Dollard et al (2000) found that age was
significantly negatively related to all three burnout dimensions and Carson, Maal, Roche, Fagin, De
Villiers, O’Malley, Brown, Leary & Holloway (1999) found their low burnout group was significantly
older than their high burnout group in a sample of mental health nurses. Zellars and Perrewé (2001)
found age was negatively associated with depersonalization, but not significantly related to emotional
exhaustion or diminished personal accomplishment in nurses, whereas Cranswick (1997) reported that
emotional exhaustion, but neither of the other burnout dimensions, was significantly negatively
associated with age in a sample of rehabilitation workers. Complicating matters still further, Dormann
and Zapf (2004) found age was unrelated to emotional exhaustion, but negatively related to
depersonalization and positively related to personal accomplishment in a sample of customer service
workers, whereas Rafferty et al (2001) found that age was negatively associated with both emotional
exhaustion and depersonalization, but unrelated to personal accomplishment. Tang, Au, Schwarzer
and Schmitz (2001) reported two studies: in the first study age was significantly negatively associated
with emotional exhaustion in teachers, but in the second no relationship between age and burnout was
found.
A couple of other studies have also found no relationship between age and any of the three burnout
dimensions (e.g. Ben-Zur & Yagil, 2005; Payne, 2001) and several studies have focused only on the
emotional exhaustion dimension of burnout and have found this to be unrelated to age (e.g. De Jonge
& Schaufeli, 1998; Deery, Iverson & Walsh, 2002; Giebels & Janssen, 2005; Holman et al, 2002; Van
Vegchel et al, 2002; Tummers et al, 2002). De Rijk et al (1998) measured emotional exhaustion and
depersonalization and found no relationship with age for either variable.
5
As can be seen from this brief review, although age is sometimes cited as one of the variables most
consistently related to burnout (e.g. Maslach et al, 2001), there is actually very little consistency in the
empirical findings. Having said this, where significant relationships between age and burnout are
reported, they almost exclusively describe an inverse relationship. By implication, burnout is a
process that unfolds over time and hence it is curious that where relationships are found between age
and burnout they tend to show that burnout decreases rather than increases with age. However, it
should also be noted that as most of these studies are cross-sectional it is not possible to infer any
causal relationship. It may be, for example, that those who report most burnout actually leave their
jobs and hence those who remain will tend to have lower levels of burnout (see below for a discussion
of ‘survival bias’)
Tenure
A number of studies have investigated the relationship between job tenure and burnout and the
findings have been mixed, with positive associations, negative associations and a lack of association
all reported.
In studies carried out over the past ten years, the most common finding appears to be that burnout
increases with tenure. Zohar (1997) found tenure was significantly positively associated with all three
burnout dimensions, and emotional exhaustion has been found to be positively associated with tenure
by Deery et al (2002), Kop et al (1999) and Lewig and Dollard (2003). Alexander and Klein (2001)
found that sense of accomplishment diminished with increased length of service in ambulance
workers, but found no association between tenure and emotional exhaustion or depersonalization.
Few recent studies have found an inverse association between tenure and burnout. However, Holman
et al (2002) found that tenure was significantly negatively related to emotional exhaustion, and Tang et
al (2001) reported two studies, one of which found that emotional exhaustion was lower amongst more
experienced teachers, and the other found no association.
Fujiwara, Tsukishima, Tsutsumi, Kawakami and Kishi (2003) and Omdahl & O’Donnell (1999) both
found that number of years of experience were unrelated to all three burnout dimensions in home care
workers and nurses respectively, and Giebels and Janssen (2005) found no association between tenure
and emotional exhaustion in a sample of social workers. De Rijk et al (1998) also found tenure was
unrelated to emotional exhaustion or depersonalization in nurses.
Taken together, these results paint a somewhat confusing picture, and it would be difficult to say
anything definitive about the relationship between tenure and burnout. Some commentators have
described burnout as an early career problem (Maslach et al, 2001; Schaufeli & Enzmann, 1998), and
suggested that this may be due to an issue of ‘survival bias’, where those who burn out early are likely
to quit their jobs, leaving behind survivors with generally lower levels of burnout. However, this
characterisation of burnout is based largely on early findings of Maslach (1982) and Pines and
Aronson (1988), and does not sit so well with the more recent findings of a positive association
between tenure and burnout, detailed above. It seems likely that the relationship is a complicated one:
for instance, it is possible that, on the one hand, those most prone to experience burnout leave the
profession early and, on the other hand, those who remain in the profession become more emotionally
exhausted over time. What is really needed to needed to understand the relationship between tenure
and burnout is probably a greater understanding of the processes involved in the development of
burnout.
2.2.2
Occupation
The original concept of burnout was specific to human service workers and burnout was thought to be
a result of the unique demands, in terms of interpersonal interactions, specific to these kinds of jobs.
However, burnout research has been conducted in a wide range of occupational settings and the results
do not consistently support the idea that burnout is necessarily higher in human service workers.
6
Brotheridge and Grandey (2002) directly compared burnout levels in different types of occupation,
looking at human service workers, service/sales employees, managers, clerical employees and physical
labourers. This study did not find significant occupational differences in emotional exhaustion levels:
the absolute highest value was for service/sales jobs, but this did not differ significantly from the other
values. Mean levels of depersonalization were significantly higher for service/sales employees and
managers than for clerical employees and physical labourers. Human service workers reported
significantly lower levels of depersonalization than did service/sales employees, managers, and
physical labourers, which was contrary to the hypothesised relationship that human service workers
should experience higher levels of depersonalization. Similarly, personal accomplishment was highest
for human service workers, significantly higher than that for service/sales employees and physical
labourers. Taken together, the findings from this study did not support the hypothesis that those in
human service professions or prototypical emotional labour jobs experience higher levels of burnout
than managers, clerical employees, and physical labourers.
Police work is an example of an occupation which is widely assumed to be highly stressful and likely
to lead to burnout, but which has been empirically found not to lead to high levels of burnout in
members of the profession. Studies of police officers have generally found relatively low levels of
psychological strain and burnout, although some have found relatively high levels of depersonalization
(Schaufeli & Enzmann, 1998). Kop et al (1999) found the typical burnout profile of police officers
seemed to comprise relatively low levels of emotional exhaustion, average levels of depersonalization,
and levels of personal accomplishment slightly above average. Similarly, Hart, Wearing & Headey
(1995) found police officers had less psychological distress and greater well-being than the average
person, although this study did not measure burnout per se. A possible explanation for this pattern of
results may be the existence of a selection effect, such that police officers are especially selected for
their resistance to stress. In support of this hypothesis, Kop et al (1999) claim that personality
inventories are often used in the selection process, and high scores on neuroticism are generally a
reason to reject a candidate.
There have also been findings of surprisingly low levels of burnout in nursing and related healthcare
occupations, which are generally thought to be burnout prone professions. Payne (2001) found
burnout scores were generally low in hospice nurses, consistent with past studies of hospice nurses.
Similarly, Carson et al (1999) found that, contrary to popular belief, burnout was not a particularly big
problem in mental health nurses, with only a relatively low 5.7% falling into the high burnout group
(as defined by Maslach et al (1996) burnout inventory manual). Hayter (1999) reported remarkably
low levels of depersonalization in specialist HIV/AIDS community nurses with 97% of the sample
scoring low on depersonalization.
On the other hand, some findings have been in line with expectations regarding the kinds of work that
might lead to burnout. Reviewing a number of studies across various professions, Schaufeli and
Enzmann (1998) concluded that levels of emotional exhaustion were particularly high amongst
teachers. Johnson and Hunter (1997) also found high levels of emotional exhaustion in counsellors,
particularly those dealing with victims of sexual assault.
Overall, the question remains as to whether human service work and/or prototypical emotional labour
jobs do in fact cause high burnout, relative to other professions. Of course, levels of emotional
demands differ widely within professions, and emotional demands are present in many jobs, not just
those typically seen as causing burnout. Therefore, it may be more useful to look specifically at the
relationship between emotional demands and burnout, rather than carrying out analysis at the
occupational level.
2.2.3
Personality
The vast majority of burnout studies choose not to focus on personality variables (around 90% of
studies according to both Pines (2004) and our own literature search), but those studies which do
measure personality consistently find that it is a significant predictor of burnout, especially emotional
exhaustion. The reasons for the neglect of personality variables in burnout research are not obvious,
7
but appear to be more ideological than scientific, in that researchers wish to avoid appearing to ‘blame
the victim’. Burisch (2002, p.2) points out that Maslach and Leiter’s (1997) book The Truth About
Burnout puts the accountability for burnout ‘squarely on the shoulders of the organization’, and that
this is applauded by fellow burnout researcher Cary Cherniss, who agrees ‘that the causes - and
solutions - are to be found primarily in the organization, not the individual’. But, as Burisch (2002)
points out, “this question has been astonishingly little studied” (p.2).
Unsurprisingly, the bulk of research into the relationship between personality and burnout has looked
at trait affectivity, i.e. negative affectivity/neuroticism and (less commonly) positive affectivity.
Studies have unanimously found that negative affectivity is significantly associated with higher levels
of emotional exhaustion (Dollard et al, 2000; Grandey et al, 2004; Iverson, Olekalns & Erwin, 1998;
Van Vegchel et al, 2002; Wright & Cropanzano, 1998; Zellars & Perrewé, 2001; Zohar, 1997). This
association is generally fairly strong, with correlation coefficients as high as r=0.72 (Wright &
Cropanzano, 1998), but more commonly in the range 0.40-0.50. The association between negative
affectivity and the other two burnout dimensions is generally less strong but generally still significant.
Zellars and Perrewé (2001), Dollard et al (2000) and Iverson et al (1998) found negative affectivity
was significantly associated with all three burnout dimensions, but Zohar (1997) did not find a
significant association with depersonalization, where burnout measures were completed by the
employees’ partners.
In terms of positive affectivity, Deery et al (2002) focused on emotional exhaustion as their burnout
measure and found this to be significantly inversely associated with emotional exhaustion. Iverson et
al (1998) found positive affectivity to be inversely associated with all three burnout dimensions, and
Zellars and Perrewé (2001) obtained a similar result with extraversion, which is strongly related to
positive affectivity.
Although the majority of studies have focused on trait affectivity as the most pertinent personality
traits, a number of other personality traits have also been studied in relation to burnout. Self-efficacy
has been found to be negatively related to burnout in teachers (Tang et al, 2001) and in nurses
(Greenglass & Burke, 2002), and the related individual difference variable of internal locus of control
has also been shown to be inversely associated with burnout (Bond & Bunce, 2003; Westman, Etzion
& Danon, 2001). Type A personality has been reported as a positive predictor of burnout in teachers
(Burke & Greenglass, 1995; Jamal & Baba, 2001), whereas hardiness (Alexander & Klein, 2001) and
sense of coherence (Soderfeldt et al, 2000) have both been found to be negative predictors of burnout
in ambulance workers and social workers respectively.
As is generally the case in this field, there is little longitudinal research to draw on. Burish (2002)
investigated the relative contribution of personality (a battery of personality tests administered before
starting training) and environmental factors to the development of burnout in a sample of German
nurses. His results are difficult to interpret, as neither individual nor experiential variables could be
linked to intra-individual change in burnout scores, measured at 7 points in time; however, interindividual burnout scores could be explained about equally well by personality and environmental
factors. Deary, Watson and Hogston (2003) also carried out a longitudinal study of burnout in nursing
students, focusing on the ‘Big 5’ personality traits. They found that personality factors at course entry
contributed significantly to the prediction of burnout. In particular, neuroticism and openness
predicted emotional exhaustion, agreeableness negatively predicted depersonalization, and
conscientiousness negatively predicted depersonalization and positively predicted personal
accomplishment.
Overall then, there is good evidence that personality factors, especially trait affectivity, play an
important role in the burnout process. As in the wider stress field, however, the exact nature of this
role is potentially complex. Taking negative affectivity as an example, this may have a direct effect,
such that individuals high in negative affectivity are more prone to burnout irrespective of their job
experiences; it may act as a moderator, such that these individuals show stronger burnout reactions to
stressful job conditions; or it may act as a biasing factor in research, by inflating self-reports of both
stressful work conditions and burnout, thus producing artificially high correlations between the two.
8
In reality, it may be a combination of these options. Klein and Verbeke (1999) found that employees
with higher emotional reactivity (in terms of autonomic feedback) were more likely to score highly on
all three burnout measures, and that this relationship was even more pronounced under conditions of
high job stress, supporting the moderator hypothesis. In support of the biasing hypothesis, a number
of studies have found that trait affectivity is significantly associated with self-reports of stressors as
well as burnout scores (e.g. Iverson et al, 1998; Grandey et al, 2004) and Dollard et al (2000) found
associations between burnout and demand-control-support stressors were attenuated when trait anxiety
was controlled for.
It therefore seems important that future research includes measures of personality traits in order to
gain a deeper understanding of the role played by personality in the burnout process.
2.2.4
Traditional job stressors
Please note that this section deals only with traditional job stressors as causes and correlates of
burnout. The effects of emotional demands arising specifically from emotional labour will be
considered in a later section after a discussion of the nature of emotional labour.
Role conflict and role ambiguity
Role conflict occurs when conflicting demands are placed on employees, and role ambiguity describes
a situation where employees are unclear as to what they are expected to do in order to perform their
job well. Troyer, Mueller and Osinsky (2000) found that customer service jobs are particularly prone
to role conflict, where employees act as the ‘broker’ between the organization on one side and
customers on the other.
Empirical studies of the effects of role conflict and role ambiguity on burnout have generally found a
significant relationship. Both Zellars and Perrewé (2001) and Koniarek and Dudek (1996) found that
role conflict and role ambiguity were significantly positively correlated with all three dimensions of
burnout in nurses, and Klein and Verbeke (1999) reported the same pattern of results with travelling
salespeople. Greenglass, Burke and Koniarski (1997) used a combined measure of role conflict and
role ambiguity, which was also significantly positively associated with all three burnout scales in a
sample of teachers. Zohar (1997) used both traditional and ‘hassles based’ measures of role conflict
and ambiguity, and found that all four of these measures correlated significantly with all three
dimensions of burnout.
A few studies have focused specifically on the emotional exhaustion element of burnout. Thompson,
Kirk and Brown (2005) found role ambiguity was significantly associated with emotional exhaustion
in a sample of policewomen and Tummers et al (2002) found role conflict and role ambiguity were
significantly positively associated with emotional exhaustion and psychosomatic health complaints
amongst nurses. Barber and Iwai (1996) reported that role conflict but not role ambiguity was a
significant predictor of emotional exhaustion in a sample of nurses and social workers working with
elderly Alzheimer’s patients, whereas Stordeur, D’hoore and Vandenberghe (2001) found role
ambiguity but not role conflict predicted emotional exhaustion in nurses.
Overall then, a review of the empirical research reveals a fairly consistent association between role
stressors and burnout, despite the fact that specific associations failed to reach significance in a couple
of individual studies. The relationship tends to be weaker with personal accomplishment than with
emotional exhaustion or depersonalization: in a meta-analysis of early studies, Lee and Ashforth
(1996) reported the corrected mean correlations for role conflict as r = .53 (emotional exhaustion), r =
.37 (depersonalization) and r = -.21 (personal accomplishment), and for role ambiguity as r = .21
(emotional exhaustion), r = .34 (depersonalization) and r = .11 (personal accomplishment).
However, these studies reviewed above were all cross-sectional, so care must be taken in inferring a
causal relationship from stressors to burnout. A particular issue that needs to be considered is the
possibility that neuroticism is acting as a third variable, producing or inflating the association between
measures of stressors and strain. Few studies have measured trait affectivity, but those which have
9
done so have found neuroticism/negative affectivity to be significantly associated with perceived
levels of role conflict and ambiguity, as well as with burnout scores (Zohar, 1997; Zellars & Perrewé,
2001). Zohar (1997) controlled for negative affectivity and demographics in the first step of his
multiple regression analysis and found that this substantially reduced the variance accounted for by
role stressors; specifically, role conflict no longer explained a significant amount of variance in
personal accomplishment, and role ambiguity no longer accounted for a significant proportion of
variance in depersonalization or personal accomplishment. This would suggest that it is important for
future research to control for neuroticism in testing the effects of role stressors on burnout.
Demands
Studies have measured job demands in various ways, but these can be split broadly into those which
use simple measures of workload, sometimes referred to as role overload, and those which base their
research on Karasek’s (1979) demands-control theory. These latter studies tend to assess demands in
conjunction with job control, and usually use Karasek’s measures, e.g. the Job Content Questionnaire
(JCQ). These measures assess ‘psychological demands’, which are defined somewhat more broadly
than simple workload. These two broad categories of studies are reviewed separately here, although
the results are largely similar.
Workload/Role overload: In qualitative studies, excessive workload is often cited as the factor
contributing most to feelings of stress and low morale, particularly amongst human service workers
(e.g. Huby, Gerry, McKinstry, Porter, Shaw & Wrate, 2002; Muncer, 2001; Ramirez, Graham,
Richards, Cull & Gregory (1996).
Empirically, high workload or role overload is particularly associated with the emotional exhaustion
element of burnout, and a number of studies have focused on this relationship. These studies have
unanimously reported a significant association between workload and exhaustion (Deery et al, 2002;
Greenglass, Burke & Fiksenbaum, 2001; Houkes, Janssen, de Jonge, Nijhuis, 2001; Houkes, Janssen,
de Jonge, Bakker, 2003; Janssen, de Jonge & Bakker, 1999; Thompson et al, 2005; Tummers et al,
2002). These results are perhaps not surprising, particularly given that workload is almost assessed
through self-report measures.
Where studies have measured all three burnout dimensions, the findings have been more mixed,
although once again, all such studies report a positive relationship between workload and emotional
exhaustion. Iverson et al (1998) and Klein and Verbeke (1999) found that workload was significantly
associated with all three burnout dimensions, whereas Jenkins and Elliott (2004), Janssen, Schaufeli
and Houkes (1999) and Coffey and Coleman (2001) found workload was unrelated to
depersonalization or personal accomplishment. Other studies have found that work overload is
associated with emotional exhaustion and depersonalization, but not personal accomplishment
(Cordes, Dougherty & Blum, 1997; Cranswick, 1997; Zellars & Perrewe, 2001).
In a meta-analysis of early studies, Lee and Ashforth (1996) reported corrected mean correlation
coefficients for workload as r = .65 for emotional exhaustion, r = .34 for depersonalization and r = -.09
for personal accomplishment. This reflects the general pattern of results found above, that the
strongest relationship is to be found with emotional exhaustion and the weakest with personal
accomplishment.
However, as we have found throughout this literature review, the available research is almost
exclusively cross-sectional in nature, making it very difficult to draw conclusions regarding cause and
effect. Indeed, in a rare longitudinal study among healthcare professionals, Leiter and Durup (1996)
found that emotional exhaustion actually predicted work overload rather than the other way around,
although they found support for the effects of overload on depersonalization (no longitudinal effects
were found for personal accomplishment).
Psychological demands: As with workload, research has consistently reported a significant positive
association between psychological demands and emotional exhaustion, and has often, but not always,
reported a significant association with the other dimensions of burnout as well.
10
Holman et al (2002), de Jonge and Schaufeli (1998) and Lewig and Dollard (2003) measured only the
emotional exhaustion aspect of burnout, and found this was significantly correlated with job demands.
Heuven and Bakker (2003) and de Rijk et al (1998) measured emotional exhaustion and
depersonalization subscales of burnout: the former found both to be significantly related to job
demands, but the latter did not find a significant association with depersonalization. Considering
studies that measured all three dimensions of burnout, Janssen, Peeters, de Jonge, Houkes and
Tummers (2004) found demands were significantly correlated with all three subscales, whereas
Dollard et al (2000) and Rafferty et al (2001) reported that emotional exhaustion and depersonalization
but not personal accomplishment were significantly correlated with job demands. Schaufeli and
Bakker (2004) assessed quantitative demands separately from emotional demands and actually found
that emotional demands were less strongly associated with burnout. Specifically, quantitative
demands were significantly correlated with emotional exhaustion and cynicism, whereas emotional
demands were only significantly associated with emotional exhaustion; neither measure was
associated with a sense of accomplishment (or professional efficacy).
Control
Research carried out in the last decade demonstrates a fairly consistent association between low job
control/autonomy and burnout. In particular, a significant relationship between control/autonomy and
emotional exhaustion was unanimously reported (Dollard et al, 2000; de Jonge & Schaufeli, 1998; de
Rijk et al, 1998; Heuven & Bakker, 2003; Holman et al, 2002; Iverson et al, 1998; Janssen et al, 2004;
Lewig & Dollard, 2003; Rafferty et al, 2001; Tummers et al, 2002).
Fewer studies looked at the other dimensions of burnout, and these results were more mixed, although
still broadly positive. In terms of depersonalization, a number of studies found a significant inverse
association with control (Dollard et al, 2000; Heuven & Bakker, 2003; Iverson et al, 1998; Janssen et
al, 2004), although de Rijk et al (1998) did not find a significant association. Rafferty et al (2001)
measured job control as skill discretion and decision authority separately, and found a significant
correlation with the former, but not the latter. They also reported similar results with respect to
personal accomplishment, i.e. it was found to be significantly correlated with skill discretion but not
decision authority. Other studies supported a positive association between control and personal
accomplishment (Dollard et al, 2000; Iverson et al, 1998; Janssen et al, 2004).
The studies reviewed above seem to find a stronger relationship between control and burnout than the
older studies reviewed by Lee and Ashforth (1996). They review 11 studies reporting correlations
between autonomy and burnout, with mean corrected correlation coefficients of r=-0.15 for emotional
exhaustion, r=-0.13 for depersonalization and r=0.07 for personal accomplishment, for all three of
which the credibility interval included zero. However, Schaufeli and Enzmann (1998) cite six other
studies from this period showing higher correlations of autonomy and burnout, which were not
included in the meta-analysis. They suggest differences in results may be due to different
operationalisations of autonomy.
As with the other variables reviewed, there is a distinct lack of longitudinal data available to draw on.
Where such data is available, changes in job control do not seem to relate to changes in burnout
(Poulin & Walter, 1993; Wade, Cooley & Savicki, 1986), which suggests we should regard positive
findings from cross-sectional studies with a degree of caution. As with other variables, cross-sectional
studies run the risk that negative affectivity may artificially inflate associations between self-report
measures of autonomy and burnout. Few studies control for negative affectivity, but Dollard et al
(2000) found that the significant relationships found in the bivariate analysis between job control and
emotional exhaustion/depersonalization became non-significant when controlling for negative
affectivity in their LISREL analysis, supporting the possibility of a biasing effect.
Interaction between demands and control
Despite the fact that Karasek’s (1979) original theory postulated an interaction effect between job
demands and control in predicting job strain, few burnout studies have tested this. The few that have
done so have found no evidence in support of an interaction effect (e.g. de Rijk et al, 1998; Dollard et
11
al, 2000; Heuven & Bakker, 2003; Rafferty et al, 2001). This null result represents a common finding
in the wider stress literature.
Social support
A review of empirical studies investigating the relationship between social support and burnout
reveals fairly mixed findings, with many studies reporting an inverse relationship between support and
burnout, some reporting an association between social support and only some dimensions of burnout,
and others reporting no significant relationship between support and burnout.
Looking first at research which has measured all three burnout dimensions, a number of studies have
found a consistent association between high levels of support and low levels of burnout (e.g. Dormann
& Zapf, 2004; Iverson et al, 1998; Jamal & Baba, 2001; Janssen et al, 2004; Klein & Verbeke, 1999;
Schaufeli & Bakker, 2004). These studies include a range of measures of social support, including
supervisor support, co-worker support and general workplace social support. A couple of other studies
have found significant inverse associations between social support and all three burnout dimensions in
terms of zero-order correlations, but have found that some of these pathways became non-significant
in their structural equation modelling. For example, Greenglass et al (1997) found there was no
significant pathway from supervisor support to any burnout dimension in their LISREL model, and no
significant pathway from co-worker support to emotional exhaustion. Similarly, Dollard et al (2000)
found the relationship between social support and emotional exhaustion became non-significant in
their LISREL analysis, controlling for NA.
A number of other studies have found, even at the bivariate analysis stage, that social support is
associated with some aspects of burnout but not others. These studies have tended to measure both
supervisor and co-worker support, and have often found that these also predict different dimensions of
burnout. Generally, social support has been most often found to predict emotional exhaustion and
least often found to predict personal accomplishment (which also reflects the findings of earlier studies
summarised by Lee and Ashforth (1996) in their meta-analysis).
Several studies have found supervisor support to be associated with lower levels of emotional
exhaustion and depersonalization and unrelated to personal accomplishment (Janssen et al, 1999;
Cranswick, 1997; Rafferty et al, 2001), although Peeters and LeBlanc (2001) found supervisor support
to be unrelated to emotional exhaustion and Jenkins and Elliott (2004) found it was not significantly
associated with any dimension of burnout. Patterns of associations with co-worker support have been
even more variable, including findings that co-worker support is associated only with emotional
exhaustion (Jenkins & Elliott, 2004; Peeters & LeBlanc, 2001), only with depersonalization
(Cranswick, 1997), with emotional exhaustion and depersonalization, but not personal
accomplishment (Janssen et al, 1999) and with emotional exhaustion and personal accomplishment,
but not depersonalization (Rafferty et al, 2001).
Finally, a couple of studies have found that social support was unrelated to any of the three burnout
dimensions, including Fujiwara et al (2003), who used four measures of social support (from
supervisors, co-workers, family or friends) and Burke and Greenglass (1995). The latter was a rare
longitudinal study which found that social support did not significantly predict burnout a year later,
after the effects of demographics and work stressors were put into the regression equation (zero order
correlations were not reported).
Having reviewed studies that measured all three dimensions of burnout, we turn now to a number of
studies that measured only emotional exhaustion, as the core dimension. The vast majority of these
have reported a significant negative association between social support (measured in various ways)
and emotional exhaustion (e.g. Aiken & Sloane, 1997; Coffey & Coleman, 2001; de Jonge &
Schaufeli, 1998; Giebels & Janssen, 2005; Holman et al, 2002; Houkes, Janssen, de Jonge and Nijhuis,
2001; Houkes, Janssen, de Jonge & Bakker, 2003; Janssen et al, 1999; Lewig & Dollard, 2003;
Thompson, Kirk & Brown, 2005; Tummers et al, 2002). Only a couple of studies reported a nonsignificant relationship between social support and emotional exhaustion: Deery et al (2002) found
team leader support but not team member support was significantly associated with lower levels of
12
emotional exhaustion in call centre workers, and Barber and Iwai (1996) found that workplace support
was not a significant predictor of emotional exhaustion in a sample of nurses and social workers caring
for Alzheimer’s patients.
Overall then, the majority of studies have reported a significant inverse relationship between social
support and burnout, particularly emotional exhaustion, a pattern of results which is consistent with
Lee and Ashforth’s (1996) meta-analysis of early studies. However, as with other variables, it is
important to note that almost all of these studies are cross-sectional, and so it is dangerous to infer a
causal relationship. Where longitudinal data are available, they do not tend to support the effect of
social support on burnout (e.g. Burke & Greenglass, 1995; Lee & Ashforth, 1993). There is also the
issue of trait affectivity potentially acting as a third variable affecting both burnout and self-reports of
social support. Again, few studies have considered this, but where they have there is good evidence
that self-reports of workplace support from supervisors and co-workers are significantly associated
with positive and negative affectivity (e.g. Iverson et al, 1998). Again, this underlines the importance
in the future of undertaking longitudinal studies where possible, and of controlling for trait affectivity
in cross-sectional research.
Finally, a handful of studies (Stephens & Long, 2000; Zellars & Perrewé, 2001; Beehr , Jex, Stacy &
Murray, 2000) have looked in more depth at the nature of the communications involved in social
support at work, separating these out into positive communications about work, negative
communications about work and non-work-related communication. These studies have consistently
found that negative communications about work are positively associated with burnout and other
measures of psychological strain, whereas positive communications about work show the reverse
effect, i.e. are associated with lower levels of strain. Zellars and Perrewé (2001) and Beehr et al
(2000) found no relationship between non-job communications and strain, although Stephens and
Long (2000) found non-job communication was associated with lower PTSD and physical symptoms
in police officers. These are potentially very interesting results from the perspective of starting to
understand the processes involved in the relationship between social support and psychological strain.
However, it is difficult to know how such cross-sectional data should be interpreted, in other words,
does discussing and interpreting work in negative ways with ones colleagues lead to strain, or does
experiencing more negative events at work simply lead to more negative communications about work
as well as higher levels of strain?
Effort-Reward Imbalance
Siegrist’s (1996) effort-reward imbalance model has received relatively little attention in the burnout
literature, although three recent studies have looked at this: Bakker, Killmer, Siegrist and Schaufeli
(2000a) and Van Vegchel, de Jonge, Meier, Hamers (2001) and Van Vegchel et al (2002).
Bakker et al (2000a) found that nurses who experienced effort-reward imbalance (ERI) reported
higher levels of emotional exhaustion and depersonalization, but that ERI was unrelated to personal
accomplishment. Van Vegchel et al (2001, 2002) looked at ERI in ancillary healthcare workers, and
used the Utrecht Burnout Scale, rather than the more widely used Maslach Burnout Inventory.
Overall, there was no effect of high effort-low reward on exhaustion, but, perhaps surprisingly, there
was a relationship between high effort-high reward and exhaustion (Van Vegchel et al, 2002). The
analysis also looked at different types of effort separately, in terms of traditional psychological
demands, physical demands and emotional demands. In terms of traditional psychological demands,
as for the overall analysis, high effort-high reward, but not high effort-low reward, was associated with
exhaustion. For physical and emotional demands, ERI was not associated with exhaustion (Van
Vegchel et al, 2001). Overall, the pattern of results reported by Van Vegchel et al (2001) tend to
suggest that effort/demands have a direct effect on exhaustion, rather than supporting an effect of
effort-reward imbalance per se.
2.2.5
Summary
• The vast majority of burnout studies are cross-sectional and hence there are relatively few
studies which tell us anything about the possible causes of burnout and a relatively large
13
number of studies examining correlations between demographic factors, personality,
traditional job stressors and burnout.
• A number of demographic factors have been studied. Gender shows inconsistent and
sometimes no association with burnout. Even where relationships are found between gender
and burnout this may be confounding of gender and occupation as it maybe that jobs where
higher levels of burnout are found also happen to be predominantly female. Both age and
tenure, likewise, have been found to have mixed and no relationship with burnout. While we
would expect burnout to be higher in particular kinds of occupations, such as human service
occupations, the available evidence does not always support this and overall the relationship
between occupation and burnout remains unclear.
• While most, around 90%, of burnout studies do not examine personality as a cause or correlate
of burnout those that do find fairly consistent relationships. In particular, negative affectivity
has been shown to be fairly strongly correlated with burnout thought the mechanisms
underlying this relationship are unknown.
• The relationships between burnout and traditional job stressors have been examined in a
number of studies. Role conflict and role ambiguity have been found fairly consistently to be
associated with burnout. Workload, general psychological demands and control have all been
found to be associated with burnout and, in particular, emotional exhaustion. Social support
has been found to be inversely associated with burnout though this finding is not consistent
across studies with some finding relations with only some dimensions of burnout and others
no relationship at all.
2.3
2.3.1
BURNOUT PROCESS
Introduction
In contrast to the large number of studies into the correlates and causes of burnout, relatively few
studies examine burnout as a process. This chapter addresses the third literature review issue and
therefore examines theory and research about the process of burnout, with a view to yielding insights
into the pathways by which particular stressors lead to harm end points. First, it looks at those studies
that explore burnout as a sequence of stages. Second, it discusses those studies that examine
psychological and behavioural processes that mediate burnout. Throughout this section, the discussion
concentrates on longitudinal studies, because causal inferences cannot be made from cross-sectional
designs nor are they appropriate for studying processes. However, since there are very few
longitudinal studies of burnout in the literature, some cross sectional studies are also included where
relevant.
2.3.2
Sequential order of the three burnout dimensions
Most commonly, studies of the burnout process seek to identify the sequence in which the burnout
dimensions appear in order to also identify when the burnout process may be starting and hence the
appropriate point for early intervention.
There are essentially three different views about the order in which the three burnout dimensions occur
(see Figure 1). First, suggesting that emotional exhaustion is at the core of burnout, Leiter and
Maslach (1988) proposed that emotional exhaustion in human service work results from the emotional
overload associated with working with difficult and demanding recipients of their care. In order to
cope with this, workers distance themselves psychologically from these recipients (i.e.
depersonalization), so diminishing their capacity to deal with those in their care, which in turn
decreases their sense of personal accomplishment. Leiter (1993) later also proposed a variation on this
model, in which emotional exhaustion leads to depersonalization, and reduced personal
accomplishment develops separately.
14
Theorist
Sequence
Leiter and
Maslach
(1988)
Emotional exhaustion  depersonalization  reduced personal
accomplishment
Golembiewski
et al (1986)
Depersonalization  reduced personal accomplishment  emotional
exhaustion
Van
Dierendonck
et al (2001a)
Reduced personal accomplishment  depersonalization  emotional
exhaustion
Figure 1: Three models of the sequential ordering of burnout dimensions
A second view, proposed by Golembiewski, Munzenrider and Stevenson (1986), suggested that
depersonalization occurs first, as a maladaptive attempt to deal with the stresses of the job. This
reduces personal accomplishment (as above), ultimately resulting in emotional exhaustion. More
recently, Golembiewski, Boudreau, Munzenrider and Luo (1996) presented a phase model of burnout
in which the sequential ordering of the burnout dimensions is broadly the same, but in this model,
there are eight progressive phases of burnout (ranging from low MBI scores on all three dimensions in
Phase I, to high scores on all dimensions in Phase VIII). However, importantly, the phase model is not
describing a developmental sequence, but rather it describes a way of classifying individuals in
accordance with the severity of their burnout symptoms. Therefore, people can have different
pathways through the eight stages. This seems to allow for individual differences in the burnout
process, whilst still suggesting that depersonalization is typically first symptom of burnout.
Golembiewski et al (1996) also suggests that the order of importance of the dimensions is the reverse
of the sequential order, with emotional exhaustion being the most important and central feature of
burnout and depersonalization is the least.
More recently, van Dierendonck, Schaufeli and Buunk (2001a) hypothesised a third possibility: That
burnout begins with diminished personal accomplishment, which in turn causes depersonalization,
which in turn leads to emotional exhaustion.
Longitudinal studies testing two or more of these models present an inconclusive picture. For
example, a longitudinal study of child protection workers (Savicki & Cooley, 1994) comparing the
Leiter and Maslach (1988) and Golembiewski (1986) models, found that their results did not lend
great support to either model. However, Toppinen-Tanner et al’s (2002) cross sectional and
longitudinal data on Finnish white and blue-collar workers suggested the superiority of Leiter and
Maslach (1988) model. A cross sectional study using structural equation modelling by Cordes et al
(1997) also found greatest support for the Leiter and Maslach (1988) model, although, since the fit
with the data was only modest, its superiority compared to Golembiewski et al (1986) was pronounced
less than definitive.
Turning to the van Dierendonck et al (2001a) model, a cross sectional study by Miller, Birkholt, Scott
and Stage (1995) found that amongst workers providing services to the homeless, reduced personal
accomplishment preceded depersonalization, which in turn preceded emotional exhaustion. However
these authors also noted that this differed from the sequence found in their earlier study of hospital
employees (in which depersonalization and low personal accomplishment occurred simultaneously,
and these led to emotional exhaustion), prompting them to speculate that, “the causal order of burnout
15
dimensions does not appear to be robust across different service provision contexts” (p.141). Van
Dierendonck et al (2001a) carried out a secondary analysis of five comparable longitudinal data sets,
primarily based on samples of healthcare workers and teachers, and using fieldwork intervals of
between three months and one year. Of the three models, van Dierendonck et al’s (2001a) provided
the best fit across the five studies, because personal accomplishment influenced depersonalization,
which, in turn, influenced emotional exhaustion. Thus, reduced personal accomplishment emerged in
this secondary analysis as the start of the burnout process.
A rationale for the sequence proposed by van Dierendonck’s et al’s (2001a) emerged in a qualitative,
interview-based study by Ekstedt and Fagerberg (2005), which explored the, “lived experiences of the
time preceding burnout” (p.59) amongst a sample of nurses. The authors described the essence of
these experiences as, “being trapped, with stimulating challenges and self-nourishing drive on one side
and with responsibilities on the other” (p.59). Prior to burnout, respondents described feeling very
motivated and absorbed in their work, but increasingly concerned that they could not live up to the
emotional demands and responsibilities they held toward their families, driving them, in turn, to work
even harder. At this point, the self-image of “skilled professional” was threatened and they felt that
things were beginning to “slip between their fingers” (p.62) causing them to fear failure, both at work
and at home. This, perhaps, corresponds to the personal accomplishment dimension of burnout.
Attempting to defend their self-image, interviewees then neglected the signs of strain and became
increasingly detached from their own emotions, withdrawing socially but failing to respond to the
warning signs of burnout. Note that in these cases, de-personalization was directed towards the self
rather than others: The cynical withdrawal from recipients, as hypothesized by Maslach and Jackson
(1981), was not evident in this study. The final stages of burnout were bodily and psychological
manifestations of strain (e.g. headaches, insomnia, worry), followed by extreme fatigue or perhaps
emotional exhaustion until, ultimately, the burned out individual reached “the bottom line” (p.64) and
was forced, finally, into acceptance and confrontation of his or her predicament.
In sum, burnout researchers and theorists share an assumption of sequentiality, but the research is
inconclusive. There seems to be most support for emotional exhaustion as the first burnout symptom
(Schaufeli & Enzmann, 1998), although recent longitudinal research has lent support to the idea of
reduced personal accomplishment as the first symptom (van Dierendonck et al, 2001a). It is also quite
possible that, as some researchers have suggested, identifying the ‘correct’ causal sequence might be
an unsolvable conundrum: It may differ according to whether burnout is viewed as a short or long
term phenomenon, as reflected in the time intervals being tested (van Dierendonck, Schaufeli &
Buunk, 2001b), or according to individual differences (e.g. people’s causal attributions about their
emotional state; Moore, 2000). It may also vary according to the occupational group under scrutiny
(Miller et al, 1995), and, in this context, it may be worth noting that Golembiewski’s (1996) data
supporting depersonalization at the start of burnout comes mostly from commercial organizations
rather than the human services. Also, given that burnout is often viewed as a long-burn phenomenon,
it also seems plausible that the constellation of different symptoms in burnout wax and wane over
time.
2.3.3
Other research into the processes underlying burnout
A handful of other theory papers and longitudinal studies have attempted to illuminate the
psychological and behavioural processes involved in the burnout process. These are grouped and
discussed according to their key themes in the next sections.
Escalating effort
Büssing and Glaser (2000) presented a model of burnout based on German Action Regulation Theory,
in which work stressors are conceptualised as barriers or interruptions in the process of work
regulation. In nursing (the occupation studied) such hindrances might include telephone calls,
technical problems or unannounced doctors’ rounds. The authors hypothesise that such problems
impede progress, so generating ‘work stress’ in response, including extra effort (e.g. longer hours),
increased pace and risky action. This, in turn, causes ‘strain’, in the form of emotional exhaustion and
depersonalization. A cross-sectional study of 482 nurses supported major parts of the model,
16
particularly the mediating role of extra effort and risky action between work stressors and emotional
exhaustion.
This model takes a mechanistic approach to burnout, which says little about its emotional content. It
also has a somewhat unusual conception of work stressors (barriers and interruptions). However, it
does emphasize an association between additional effort and burnout that is underscored elsewhere.
For example, nurses in Ekstedt and Fagerberg’s (2005) qualitative study worked unreasonable
amounts of overtime and became overly absorbed in their work, as a result of the twin pressures of a
strong “inner incentive” (p.62) to work hard, combined with a fear of failing in their responsibilities
towards others. In this way, over-working was a precipitating factor in burnout.
Recovery
The variable of ‘effort’, in combination with insufficient recovery, features prominently in a study by
Demerouti, Bakker and Bulters (2004). These authors hypothesised and tested a burnout process
based on a model of effort-recovery. According to this model, if there is insufficient opportunity to
recover after a high workload day (experienced as work-home interference) then an individual begins
work again in a sub-optimal state, so increasing the general load and effort required to perform, so
making subsequent recovery more difficult. This is a cumulative, loss spiral process, which ultimately
results in exhaustion. A three-wave panel study of employment centre workers uncovered long term
reciprocal relations between the three variables of work pressure, work-home interference and
exhaustion. In other words, the variables were both causes and consequences of one another,
suggesting that the burnout process is a downward, negative spiral, involving escalating effort,
increasingly ineffective recovery, and increasing exhaustion.
Hypothesising a similar process, in which recovery is important to well-being, Cropley and Millward
Purvis (2003), conducted a diary study amongst 102 secondary school teachers, which explored
ruminative thinking after work. They found that ‘high strain’ teachers took longer to unwind than
‘low strain’ teachers, even when controlling for post-work activity. While there were no differences in
level of ruminative thinking between the two groups at 5 p.m., 6 p.m. and 7 p.m., there was a
significant difference at both 8 p.m. and 9 p.m. The high strain group also reported finding it difficult
to stop thinking about work, and thought more about future work tasks and recent events at work than
the low strain group. This highlights an association between job strain and inability to unwind that has
also been noted in other studies (e.g. Steptoe, Cropley & Joekes, 1999), although the direction of
causality is unclear - is the inability to unwind a symptom or a cause of strain?
Emotion and interpersonal processes
Finally, a handful of other studies have examined the emotional and inter-personal processes within
burnout. These are highly relevant in the current context for two reasons: First, because they centre
on emotion work (as opposed to more conventional stressors) and worker-customer or client
relationships (as opposed to burnout as a within-person process) and, second, they often focus on
human service workers (HSWs).
These studies focus on three key concepts, empathic communication, lack of reciprocity, and emotion
regulation in order to explain burnout. Each of these will be discussed in turn.
Miller et al (1995) tested an empathic communication model of burnout (Miller, Stiff & Ellis, 1988)
which proposes that HSWs often choose their jobs because they feel a high degree of empathy for
others. However, empathy takes two forms – empathic concern, in which one feels for another – and
emotional contagion, in which one feels with the other. It was hypothesised that the latter (and not the
former) is associated with burnout. Their cross-sectional study of 172 homelessness workers found
support for the model, in which emotional contagion correlated negatively with communicative
responsiveness and reduced personal accomplishment (placing this as the first of the three burnout
dimensions) and empathic concern correlated positively with communicative responsiveness, but was
unrelated to burnout.
17
The second concept used to explain burnout processes, lack of reciprocity, was proposed by Bakker,
Schaufeli, Sixma, Bosveld and Van Dierendonck, (2000b) and centres on lack of reciprocity in the
GP-patient relationship. This recalls the asymmetrical relationship between the service provider and
the client, which is a theme in the emotional labour literature. Bakker et al’s (2000b) study addressed
the central question of “what is it that makes demanding patients so stressful?” (p.428). Using data
from 407 GPs, collected in two waves with a five year interval, they tested the hypothesis that
demanding patients are indirectly related to burnout via lack of reciprocity, because contact with
demanding patient leads to a perceived lack of reciprocity in the doctor-patient relationship, which
depletes the GP of emotional resource and initiates burnout. The results lent support to the model,
because the model was confirmed at Time 2, even after controlling for Time 1 scores on each of the
model components. There was also evidence of reciprocal effects such that burnout leads to increased
perceptions of more demanding patients. The authors suggest that the results might generalise, because
similar associations between lack of reciprocity and emotional exhaustion have been found in studies
amongst other occupational groups such as nurses (Van Yperen, 1996) and teachers (Van Horn, 2001).
The third concept, emotion regulation, was explored by Côté and Morgan (2002) in a longitudinal
study of emotion regulation, job satisfaction and intention to quit conducted amongst 111 working
college students. If was found that suppression of unpleasant emotions at Time 1 decreased job
satisfaction at Time 2 (four weeks later), whereas the amplification or exaggeration of positive
emotions increased job satisfaction at Time 2. Suppressing negative emotions at Time 1 also increased
intention to quit at Time 2 and this was mediated by job satisfaction. The study also controlled for a
number of confounding variables, including negative affectivity, and found no evidence of reverse
causality. By way of explanation of these differential effects, the authors suggest that, “the suppression
of unpleasant emotions exerts its impact mainly through emotional dissonance” (p.959), which is a
“psychologically taxing” state, whereas, “the amplification of pleasant emotions exerts its impact
mainly through social interaction” (p.959), because displays of happiness produce a favourable
response in others, which in turn positively influences affective state.
Extending these ideas, Côté (2005) has recently proposed a social interaction model of the effects of
emotion regulation on work strain. In this paper, he criticises the emphasis on intra-personal (within
person) processes in explaining the fairly well supported causal link between emotion regulation and
strain, contending, instead, that interpersonal processes have a greater role in explaining this
association. This directly recalls some of the earliest work on emotional labour, which concerned itself
with the effects of emotional labour on the recipients of it such as customers and clients (e.g. Rafaeli
& Sutton, 1988, 1990).
Côté’s model suggests that the sender’s (employee’s) emotion regulation and ensuing display impacts
on the receiver’s (e.g. customer’s) response, which, in the manner of a feedback loop, impacts on the
sender’s strain. For example, as the sender’s surface acting increases, receivers’ responses become
more unfavourable (because of perceived inauthenticity), thus increasing the sender’s strain.
However, Côté (2005) suggests that consideration of inter-individual processes does not necessarily
preclude consideration of intra-individual processes – for example, both receivers’ responses and
senders’ emotional dissonance may partially mediate the relationship between surface acting and
strain.
Whilst this model is untested, it represents an interesting broadening out of ideas about how and why
emotional experience and strain (including burnout) might be linked. Consideration of inter-personal
processes also provides an explanation for why performing emotional labour seems to bring with it
both beneficial and harmful effects. It suggests that the inter-personal side of emotional labour should
not be overlooked in the context of understanding the causes of burnout in people work.
2.3.4
Summary and conclusion
• Whilst there is a general belief that burnout occurs in a sequence of stages, the available
evidence does not support one correct ordering of the burnout dimensions. In particular,
reduced personal accomplishment, generally considered the most peripheral dimension of
18
burnout, seems to have been found to be the initial dimension to emerge in several studies of
human service workers.
• Theorising and studies that have sought to unravel the intra-individual (within person)
processes within burnout have proposed a role for: (i) escalating effort and (ii) effort-recovery.
Other studies have examined burnout as an interpersonal process, the dynamics of which may
include: (i) lack of reciprocity; (ii) emotional contagion and (iii) recipients’ responses to
emotional regulation. As mentioned earlier, the strongest type of support for a particular
causal process comes from longitudinal studies. There is longitudinal data to support the
effort-recovery model (Demerouti et al, 2004), the role of lack of reciprocity in burnout
(Bakker et al, 2000b), and the role of emotion regulation in negative outcomes like reduced
job satisfaction (Côté & Morgan, 2002). A qualitative study of burnout amongst nurses
(Ekstedt & Fagerberg, 2005) also provides a descriptive (though retrospective) insight into the
burnout process.
• Overall, however, there is surprisingly little longitudinal data that examines how burnout
develops. Few conclusions can be drawn from this very limited amount of evidence.
2.4
2.4.1
EMOTIONAL LABOUR AND BURNOUT IN CUSTOMER SERVICE AND HUMAN
SERVICE WORK
Introduction
This section addresses the fourth literature review issue and so examines the extent to which the
emotional demands of human service occupations differ from, or accord with, the emotional demands
of other occupations, including customer service work in particular.
To this end, we first describe the theory of emotional labour, which is germane to almost all of the
studies into the emotional demands of paid work. The following section then examines the types of
occupations which require emotional labour, drawing attention to the similarities and differences in the
emotion work carried out within these jobs.
2.4.2
Emotional labour
There have been several approaches to understanding how emotional labour works. The first and most
influential of these was developed by Hochschild (1983) in her book The Managed Heart. This
approach and its limitations will be discussed first, followed by a discussion of other ways of thinking
about emotional labour.
The concept of emotional labour was first identified by Arlie Hochschild (1983). Her qualitative study
of flight attendants revealed that, “the emotional style of offering the service is part of the service
itself” (p.5) so that flight attendants were trained to, “go out there and really smile” (p.4), irrespective
of their underlying feelings. Thus, emotional labour, or “the process of regulating both feelings and
expressions for organizational goals” (Grandey, 2000, p.97), comprised a fundamental part of the
flight attendant’s job.
Hochschild (1983) saw emotional labour as a specific form of emotion management, and thus an
extension of the management of feelings which we all undertake in order to feel what we ‘ought’ to be
feeling at particular points in our everyday lives. However, in emotional labour, this emotion work
serves the organization’s ends, rather than the individual’s. That is why, according to Hochschild
(1983), it holds the potential to be exploitative and damaging for the worker.
In Hochschild’s (1983) view, service work is a performance in which the worker obeys corporately
managed emotional display rules in order to deliver the requisite level of sincere and caring service.
For example, the worker may have to amplify positive feelings, or suppress negative ones (Diefendorff
19
& Richard, 2003). This is achieved through either deep or surface acting. Deep acting parallels method
acting, in which spontaneous, felt emotions are worked on and changed (by, for example, ‘psyching’
oneself up, or evoking appropriate emotional memories). In surface acting, by contrast, the required
feelings are merely displayed (through bodily, facial and verbal communication) for the audience’s
benefit.
According to Hochschild (1983) both forms of acting bring costs: With surface acting, the emotional
cost is “a sense of being phoney or insincere” (p.21). By contrast, in deep acting, the distinction
between real and ‘worked on’ emotion blurs, and the worker risks losing the “signal function” (p.21)
of feeling, or, basically, an appreciation of what feelings tell us about ourselves. By extension, what is
and what isn’t my real self is obscured. This brings the risk that the worker will (i) over-extend herself
emotionally, (ii) over-identify with the job, and ultimately, (iii) burn out. According to Wharton
(1999), this is a particular danger in human service work: “In these jobs, then, the problem is not that
workers are expected to display a concern for clients’ well-being that they do not genuinely feel.
Instead, HSWs – or any workers that are too identified with their work role – are at risk precisely
because the feelings expressed at work are inseparable from the self. Over time this inability to
depersonalise and detach oneself increases these workers risk of burnout” (p.162/3).
The view that emotional labour is essentially harmful to workers has been criticised by Wouters
(1989) who suggested that Hochschild was overly preoccupied with the costs of emotion work. He
claimed that her “theoretical stance toward feeling makes her see costs where there are none, or hardly
any” (p.118) and also that her approach “hampers understanding of the joy the job can bring” (p.116).
Acknowledging the influence of Wouters, as well as Hochschild, emotional labour now seems to be
seen as something of a “double-edged sword” (Ashforth & Humphrey, 1993, p.96), having both
negative and positive personal consequences for workers. This complicates the straightforward link
between emotional labour and negative psychosocial outcomes, as originally hypothesised by
Hochschild (1983). Also, it is worth noting that Hochschild (1983) paid relatively little attention to
how and why emotional labour would lead to the negative consequences she hypothesised. Other
theorists have attempted to explain how such links may occur and these are discussed below.
There have been several other theoretical perspectives on emotional labour (e.g. Ashforth &
Humphrey, 1993; Grandey, 2000, Diefendorff & Gosserand, 2003; Morris & Feldman, 1996). This
section concentrates on Morris and Feldman (1996) and Grandey’s (2000) contributions, because these
are most directly concerned with the process of emotional labour and its consequences for the
individual.
Morris and Feldman (1996) made several important theoretical insights. First, they argued that even
when there is congruence between felt and displayed emotions, there will be some degree of emotional
labour involved in ensuring that the emotion is displayed in an organizationally appropriate way. This
proposition helps to explain the perceived ubiquity of emotional labour in working life. Second, they
elaborated the concept of emotional labour, arguing, in agreement with later theorists (e.g. Zapf, 2002)
that it is a multi-dimensional construct, varying according to dimensions such as the frequency, variety
and intensity of emotions displayed. Third, they proposed that variation along these dimensions might
explain the degree of emotional exhaustion felt by workers. So, for example, short, scripted
interactions performed by fast food workers will be more superficial, less laborious and therefore less
emotionally exhausting than longer, emotionally intense interactions between nurses and their patients.
Thus nursing is seen as more emotion-full and more emotionally exhausting than waiting on tables in a
restaurant.
Fourth, they highlighted the importance of emotional dissonance as a dimension of emotional labour.
This concept was first suggested by Hochschild (1983), to describe the state of discomfort or tension
caused by the inconsistency between real and required emotions. This is then resolved by either
working on the feelings or feigning them (Mann, 2004). Although Hochschild (1983) only mentioned
emotional dissonance briefly, it is now seen as the key source of work stress in emotional labour
(Bolton & Boyd, 2003), and, as such, it has since received much attention from researchers.
20
Grandey (2000), meanwhile, suggested a model of how emotional labour might lead to harm end
points. A study of daily events in customer service work tested and found support for this model
(Totterdell & Holman, 2003). In it, the antecedents of emotional labour are interaction expectations
(of the frequency, duration and variety of emotional display) and emotional events (positive and
negative). Thus, in contrast to Morris and Feldman (1996), who see frequency, duration, variety and
intensity of emotional display as dimensions of emotional labour, Grandey (2000) sees them as job
characteristics that are the antecedents or precursors to it (Zapf, 2002). Concurring with this, the
current review treats these aspects of emotional display as job characteristics, rather than dimensions
of emotional labour (see Section 1.1.2.2). In Grandey’s (2000) model, emotional labour is
operationalised according to how it is done, that is, through deep and surface acting. The outcomes are
considered to be individual and organizational measures of well-being. The emotional labour-outcome
relationship is moderated by individual factors (e.g. positive and negative affectivity) and
organizational factors (e.g. job autonomy). This model is comprehensive, and would seem to be
applicable in an event-based analysis of emotional labour. It omits emotional dissonance, however,
which is curious given the amount of evidence which links emotional dissonance to stress outcomes
(e.g. Holman et al, 2002; Zapf, 2002; see below). However, it can be argued that emotional dissonance
is a feature of acting, because otherwise there is a harmony between display rules and authentic
feelings. This is especially true of surface acting, because deep acting largely eliminates the
dissonance (Holman et al, 2002). This means that emotional dissonance can be subsumed within
surface acting. In support of this, many measures of emotional dissonance and surface acting within
the literature are identical, or near identical.
2.4.3
Emotional demands of customer service and human service work
Emotional labour is frequently seen as “a feature of contemporary work life” (Erickson & Ritter, 2001,
p.159; Steinberg & Figart, 1999). However, people in occupations where the focus of the job is
“interaction with other people” (Mann, 2004; p.205) are generally considered the main performers.
The term ‘people work’ subsumes both customer service and human service, or caring, jobs.
Brotheridge and Grandey (2002) classify these as prototypical emotional labour and prototypical
burnout occupations, respectively. Much of the remainder of this chapter will focus on customer
service workers and human service workers.
The distinction between these two types of people work is not absolutely clear cut, because, they have
much in common, as outlined below. Neither are the categories mutually exclusive, because it can be
argued that some jobs combine elements of both customer and human service work. Nevertheless, in
line with other research in this area, here we are defining human service work as work which,
“involves the day to day provision of care to individuals in need” (Miller et al, 1995, p.124) and
customer service work as jobs “that involve frequent customer contact and emotion displays controlled
by the organization” but where the primary task is not the provision of care (Brotheridge & Grandey,
2002, p.19). They have also been distinguished as “direct person-related jobs” in which “the primary
task is to modify the clients physically or psychologically” and “indirect person related jobs” in which
“it is only a secondary task to make the customer feel good” (Dollard, Dormann, Boyd, Winefield &
Winefield, 2003, p.84).
Thinking about the emotional demands of these two types of work, there are similarities and
differences. Firstly, both human service work and customer service work involve extensive contact
with clients or customers. However, there is likely to be a difference in the duration and intensity of
client contact, with sales encounters being briefer and less emotionally intense than human service
interactions. In a cross-occupational study of emotional labour, Brotheridge and Grandey (2002) found
that both types of people worker reported more frequent interactions with customers than did clerical
and physical labourers. However service/sales employees reported less variety of emotional
expression, less intensity of interactions and shorter duration of interactions than human service
workers. It was also the human service workers who were found, overall, to show the most positive
expression.
21
Secondly, if both types of work involve extensive customer contact it follows that they will each
attract their fair share of negative customer contact, such as encounters involving verbal and possibly
physical aggression. In line with this, exposure to verbal aggression has been identified as a significant
emotional demand for various types of customer and human service worker, including call centre
workers (Grandey et al, 2004), trolley car drivers (van Dierendonck & Mevissen, 2002), mental
healthcare workers (Lawoko, Soares & Nolan, 2004), hospital workers (Winstanley & Whittington,
2002), police officers (Le Blanc & Kelloway, 2002), nurses (Le Blanc & Kelloway, 2002; Walsh &
Clarke, 2003) and railway and airline workers (Boyd, 2002). The occurrence of physical violence, or
physical threats against the person, has been identified and measured in human service workers
(Lawoko et al, 2004; Walsh & Clarke, 2003; Winstanley & Whittington, 2002). A study of sexual
harassment in call centre work (Sczesny & Stahlberg, 2000) also found that 74% of female workers
reported receiving sexually harassing phone calls at work and 25% reported that this led to negative
emotional consequences.
Thirdly, as mentioned above, customer service workers and human service workers are widely seen as
the main performers of emotional labour. However, there is a marked discrepancy in the extent to
which emotional labour has been studied in these two types of work. Specifically, there is a long
history of studies into the emotional labour performed by customer service workers, usually in low
paid jobs that are highly governed by display rules. For example, early case study and ethnographic
studies (in the 1980s and early 1990s) established that emotional labour was carried out by an array of
customer service workers, including supermarket and store sales clerks (Sutton & Rafaeli, 1988;
Rafaeli & Sutton, 1990; Tolich, 1993), bill collectors (Rafaeli & Sutton, 1991; Sutton, 1991) criminal
interrogators (Rafaeli & Sutton, 1991) and Disneyland ride operators (Van Maanen & Kunda, 1989).
By contrast, the emotional labour literature has only recently turned its attention to human service
workers, and, with a few notable exceptions (e.g. Bolton, 2000; Brotheridge & Grandey, 2002) there
have been few empirical studies of emotional labour in jobs like nursing, teaching and social work.
However, in the literature on human service work, particularly the nursing literature, there is a great
deal of discussion around the nature and value of emotional labour (e.g. Staden, 1998; McQueen,
2004). This has emphasised the observation that emotional labour in human service work is
underpinned by the notion of caring. According to Himmelweit (1999), nurses perform “caring
labour” (p.27), which is a distinct type of emotional labour (although not all caring labour is emotional
labour, much of it being physical). Whilst most service jobs involve catering to someone else’s needs,
caring labour is distinctive because the motivation has an intrinsic importance to the activity: In other
words, caring labour involves caring for and about the caree, and in this way paid caring is more akin
to caring in a domestic setting than it is to other non-caring occupations (Himmelweit, 1999). For
example, a qualitative study by Bolton (2000) showed how gynaecology nurses offered “extra”
emotion work as a “gift” to bereaved and distressed patients (p. 580). In practical terms, this is likely
to mean that human service workers’ emotional labour is likely to be focused around sensitivity,
empathy and concern. This is likely to contrast with many types of customer service work, in which
the sensitivity requirements are relatively low (e.g. call centre work, Lewig & Dollard, 2003).
The relevance of this observation to an investigation of the relationships between emotional demands
and harm in human service work is twofold: First, it suggests that the caring labour performed by
many HSWs may be slightly different from classic emotional labour as identified by Hochschild
(1983). This is because it may be performed for intrinsic, caring motivations, rather than simply in the
service of the organization. The emotional labour performed by HSWs might therefore be better
described by the more general, umbrella term of ‘emotion work’. Second, emotion work may have
different consequences in human service work compared to customer service work because, in the
human service context, it is often not about acting. Himmelweit (1999) suggests that harm results from
the “full commodification of emotion” (p.37), when the worker-client relationship is superficial and
temporary and the worker’s underlying feelings frequently conflict with the emotions to be displayed.
In contrast, the caring labour performed by many human service workers may not alienate the worker
in the same way because it is incompletely commodified, so that it is underpinned by a more rewarding
form of emotional engagement with the client or patient. Nevertheless, HSWs may be at risk of
22
burning out because the feelings expressed at work are inseparable from the self and they are
insufficiently able to remain detached from their clients (Wharton, 1999).
Zapf, Isic, Bechtoldt and Blau (2003) carried out an interesting comparison of the job characteristics,
job stressors and emotional demands of 375 call centre workers, compared with three comparison
groups (blue collar and administrative workers, ordinary service workers and human service workers).
Human service worker’s jobs seemed to be the most demanding, scoring highest on measures such as
time pressure, organizational problems and, most notably, measures of emotional demands (e.g. the
requirement to express positive emotions and the need to be sensitive to the emotions of others). Call
centre workers ranked second of the four groups on these dimensions. However, the two positions
were reversed for emotional dissonance. This suggests that human service work is particularly
emotionally demanding and call centre work is also more emotionally demanding than other jobs.
However, the nature of human service work and the way in which it is done (e.g. with relative
autonomy, Zapf et al, 2003) may lessen the amount of dissonance experienced, which may, in turn,
exert a protective effect over well-being. The study also suggested that human service work is
characterised by high levels of other (not emotion-related) job and organizational demands, which may
help to account for burnout in this group.
2.4.4 Summary
• Emotional labour is “the process of regulating both feelings and expressions for organizational
goals” (Grandey, 2000, p.97). Whilst it is often seen as a ubiquitous feature of modern
working life, emotional labour is most often associated with different types of people work.
• Two important types of people work are human service work and people service work. Theory
and evidence suggests that both of these types of work are emotionally-demanding, compared
with other jobs. In particular, both involve high amounts of client/customer contact, and both
involve exposure to customer verbal aggression.
• However, the motivations for performing HSW are thought to be different from those
underpinning CSW, such that HSWs are expected to genuinely care about their clients and
patients, so that their emotional displays will be more authentic and heartfelt. This should
result in lower emotional dissonance in HSW, potentially reducing the harmful effects of
emotional labour. However, the picture is not clear cut: HSWs may remain at high risk of
burnout because they over-extend themselves emotionally, or over-identify with those whom
they care for, or because of the high emotional dissonance associated with having to appear
caring if you no longer do.
2.5
RELATIONSHIP BETWEEN EMOTIONAL DEMANDS AND BURNOUT IN
CUSTOMER SERVICE WORKERS AND HUMAN SERVICE WORKERS
2.5.1 Introduction
The purpose of this section is to address the fifth literature review issue and thus examines the extent
to which client interactions, and the attendant emotional demands these bring, are associated with
burnout in customer service and human service workers. First, customer contact per se as an
antecedent or correlate of burnout and other harm end points is examined. Next, customer aggression
is then examined in the same way. Last, relationships between emotional labour and burnout (and
other similar measures) is considered first in customer service workers and then in human service
workers.
2.5.2 Customer contact and burnout
Recent studies which have examined customer contact in detail have found little by way of an
association between customer contact per se and burnout (e.g. Erickson & Wharton, 1997; Deery et al,
23
2002). Rather, these studies have sometimes highlighted a potentially beneficial association between
customer interaction and feeling efficacious and accomplished in the job. For example, Cordes et al
(1997) looked at four aspects of the interpersonal interactions of human resources professionals:
Number of interactions, duration of interactions, emotional intensity of interactions, and percentage of
interactions that were face-to-face. The number of interactions was unrelated to emotional exhaustion
and depersonalization, but positively associated with personal accomplishment. Intensity of
interactions was positively associated with emotional exhaustion and depersonalization, but unrelated
to personal accomplishment. Duration of interactions was inversely related to depersonalization (i.e.
longer interactions were associated with lower depersonalization). Finally, percentage of face-to-face
interactions was related inversely to emotional exhaustion and depersonalization and positively to
personal accomplishment.
Similarly, Brotheridge and Grandey (2002) found, across occupations (including customer service
workers and human service workers), that frequency and duration of interactions and intensity and
variety of expression were not related to emotional exhaustion or depersonalization, but were all
positively related to personal accomplishment. They write, “Recent research in both burnout and
emotional labor literatures has found mixed or nonsupportive results for the frequency of interactions
and the display rules of the job (Cordes et al, 1997; Morris & Feldman, 1996; Rafaeli & Sutton, 1990;
Schaubroeck & Jones, 2000), and this was true for this study as well.” (p.32). They also suggested, in
common with other writers (e.g. Dormann & Zapf, 2004), that the findings relating to customer
interactions and personal accomplishment suggest that emotional labour can be rewarding, and
underscore, “the need to expand the traditional stress perspective of emotional labour” (Brotheridge &
Grandey, 2002, p. 32).
2.5.3
Customer aggression
As discussed earlier (Section 1.1.1.3) customer aggression has been found to be an occupational
hazard in both customer service and human service work. In general, recent studies have highlighted a
consistent association between customer aggression and emotional exhaustion, plus, in some cases, an
association between aggression and the other dimensions of burnout as well (e.g. Ben-Zur & Yagil,
2005; Dormann & Zapf, 2004; Grandey et al, 2004; van Dierendonck & Mevissen, 2002).
Customer service work
Thinking firstly about customer service work, Dormann and Zapf (2004) developed an instrument to
measure various forms of customer-related social stressors based on qualitative interviews. This
comprised four subscales: Disproportionate customer expectations, customer verbal aggression,
disliked customers and ambiguous customer expectations. Each of these subscales was found to be
significantly associated with all three burnout dimensions and the strongest associations were seen
with aggressive customers across all three measures.
Grandey et al (2004) studied the effects of customer verbal aggression on call centre workers.
Qualitative interviews carried out before the survey found that all interviews reported daily yelling and
threatening and insulting behaviour from callers. Indeed, half of the interviewees said that this was the
most stressful part of the job. In the survey, estimates of the typical frequency of aggression ranged
from 0 to 50 times per day, with a mean of 7. Emotional exhaustion was significantly associated with
frequency of customer aggression and stress appraisal of customer aggression. Interestingly this study
also found people who appraised customer aggression as more stressful were more likely to surface
act, whereas deep acting was associated with low stress appraisal. This raises the interesting and
possibly important question of causality: Does frequent surface acting lead to feelings of stress and
possibly burnout, as much of the literature suggests, and/or does an individual need to surface act
because, underneath, he or she is feeling stressed or burned out?
Lastly, a study by Deery et al (2002) analysed a number of aspects of call centre work as predictors of
emotional exhaustion. Using LISREL analysis they found that customers becoming more
abusive/demanding, having to speak in a scripted manner, management pressure on wrap-up time,
routinization, and management emphasis on quantity (not quality) were all significant predictors of
24
emotional exhaustion. Also, supporting the assertion that customer contact per se is not intrinsically
harmful, number of calls taken did not predict emotional exhaustion, and the average length of calls
was negatively associated with exhaustion.
Human service work
Research with human service workers has tended to focus on physical aggression and violence rather
than verbal aggression. Generally, these studies have supported the association between client
aggression and burnout and other harm end points. Walsh and Clarke (2003) studied the effects of
exposure to aggressive incidents on UK NHS workers, mainly nurses and nursing staff. For three
months, all staff reporting an aggressive incident at work (around 5% of staff over this period) were
asked to fill in a survey, including the Impact of Events Scale – Revised (IES-R: Weiss & Marmar,
1997), which has scales measuring the core elements of trauma symptomatology, i.e. intrusion,
avoidance and hyperarousal. Most incidents described were cases of minor physical aggression, and
the remainder were verbal. Perhaps surprisingly, cases of verbal aggression were associated with
significantly greater intrusive recollections, and significantly higher subjective ratings of overall
impact. However, only 40% of those reporting an incident completed the survey, which may limit the
representativeness of the findings.
Lawoko et al (2004) studied exposure to violence and aggression amongst mental healthcare workers
(nurses and psychiatrists) in the UK and Sweden, and found this was common, with around 75% of
respondents having been victims of violence or aggression during their career, and around 65% during
the last year. Victims perceived their psychological and physical health as poorer than non-victims.
A study by Winstanley and Whittington (2002) researched the effects of being exposed to violence
and aggression on anxiety and burnout amongst hospital staff. The survey required respondents to
report experiences of physical assault, threatening behaviour and verbal aggression from either
patients, or patients’ friends and family. State anxiety was significantly associated with threatening
behaviour by visitors and more frequent verbal aggression, but was not physical assault or threatening
behaviour by patients. Emotional exhaustion and depersonalization were significantly associated with
all four measures of aggression, but personal accomplishment was not significantly related to any of
these measures. The study also examined recency effects and found that emotional exhaustion and
depersonalization were higher for those who had been assaulted within the last three months,
compared with those who had been assaulted longer ago.
Lastly, a US study by LeBlanc and Kelloway (2002) struck a note of dissent. They found, rather
surprisingly, that exposure to aggression and violence from the public (in occupations like nursing and
policing) did not affect emotional and physical well-being, including GHQ12 scores. However, coworker aggression was associated with reduced well-being. The authors suggest that certain
occupations may expect, and accommodate, aggression from the public as part of the job.
2.5.4
Relationship between emotional labour and burnout
This last part of the literature review examines the relationship between emotional labour and burnout,
plus other, similar harm end points. First, it gives a broad overview of the findings relating to
emotional labour and burnout across the literature. Then it reviews the literature on emotional labour
and burnout in customer service workers and human service workers.
It is interesting to note that despite the fact that emotional exhaustion is at the core of burnout, the
burnout literature contains relatively little research on the relationship between emotional labour and
burnout (Brotheridge & Grandey, 2002). However, a significant number of quantitative studies within
the emotional labour literature have established a consistent link between some aspects of emotion
work and burnout. It is important to note, however, that these studies are almost exclusively
correlational: It remains perfectly possible burnout generates emotion work, or both are in some way
reciprocally related.
25
There is consistent evidence that both emotional dissonance and the closely related concept of surface
acting are related to burnout (e.g. Dormann & Zapf, 2004; Grandey, 2003; Heuven & Bakker, 2003;
Holman et al, 2002; Lewig & Dollard, 2003; Zapf, Seifert, Schmutte, Mertini & Holz, 2001). Indeed,
Zapf (2002), reviewing the literature on the role of emotional labour in well-being, states that the
empirical findings on the relation between emotional dissonance and emotional exhaustion are
unequivocal, because, “significant positive correlations across twelve studies ranged from 0.20 to
0.48, with a sample size weighted mean correlation of r = 0.32” (p.256). A few studies have examined
de-personalization, and found a positive correlation with this and emotional dissonance (e.g. Heuven
& Bakker, 2003; Kruml & Geddes, 1998; Zapf et al, 1999). The relationship with personal
accomplishment is more problematic, because being able to respond to display rules can be seen as an
aspect of professionalism, contributing to a sense of personal accomplishment (Mann, 2004; Zapf,
2002). Equally, difficult client or patient interactions can be seen as a challenge, as opposed to being
purely stressful (Zapf, 2002). This reasoning supports the mixed findings for and against a negative
correlation between emotional dissonance and personal accomplishment.
Evidence suggests that deep acting is more weakly related to burnout than surface acting. For
example, Grandey’s (2003) study of university administrative assistants found that both surface acting
and deep acting were significantly associated with emotional exhaustion but, as predicted, the
relationship was much stronger for surface acting (r = 0.58 versus r = 0.33). Similarly, Holman et al
(2002) found that emotional dissonance and surface acting were strongly positively associated with
emotional exhaustion, job-related anxiety and depression. Deep acting, however, was more tentatively
related to emotional exhaustion and job related anxiety, and unrelated to depression. Lastly,
Brotheridge and Grandey (2002) found that deep acting was not related to emotional exhaustion or depersonalization, but did result in a heightened sense of personal accomplishment. Surface acting was
significantly related to emotional exhaustion, though in regression analysis it was not significant after
controlling for other variables. Surface acting was however, significantly related to de-personalization
and a diminished sense of personal accomplishment. The authors conclude that overall, these results
tend to undermine the argument that emotional labour and/or high customer contact is inherently
stressful: Only the faking of emotions resulted in negative consequences in this study.
Another strand of survey research has focused explicitly on the types of emotions being managed.
Pugliesi and Shook (1997) found that the necessity to inhibit emotional display was the most salient
for workers’ well being. Developing this, Erickson and Ritter (2001) studied emotional labour,
burnout and inauthenticity in a large sample of dual earner married parents. Emotional labour was
measured according to the frequency with which respondents experience twelve emotions at work, and
the frequency with which they had to cover them up. Agitated emotions (nervousness, anger and
irritation) were managed 73% of the time, and both the experience of agitation and having to hide
agitation were significantly related to both burnout and inauthenticity. This was not the case for the
other emotions. The authors suggest that the consequences of emotional labour might depend on the
specific emotion being managed.
Whilst the above studies have established that there is an association between emotional labour and
burnout, the importance of this (relative to the role of other more traditional stressors in the
development of burnout) is less clear. Schaufeli and Enzmann (1998) compared the results of sixteen
studies and found that, contrary to expectations, common job-related stressors like workload, time
pressure and role conflicts, correlated more highly with burnout than client and emotional labour
related variables. The concluded that, “on empirical grounds, the assertion that burnout is particularly
related to emotionally charged interactions with clients has to be refuted” (p.84).
More recently, Zapf et al (2001) examined emotion work, versus organizational and social variables as
predictors of burnout. Hypothesising a co-occurrence of high emotional demands and negative
organizational environment, the authors aimed to explore the unique contribution of emotion work to
burnout, after controlling for task and organizational variables, stressors and social support. The
sample (n = 1,241) consisted of both customer and human service workers. They found, as
hypothesised, a correlation between emotion work and the other predictor variables. Also as
hypothesised, emotion work was predictive of all three burnout dimensions: For example, emotion
26
work variables accounted for 12% of the variation in emotional exhaustion, and only 3% when the
effects of the other variables were partialled out.
Wharton (1993), found that workers in high emotional labour jobs (in the banking and hospital
industries) were no more likely than other workers to experience job related burnout. Another similar
study, using a different measure, yielded similar results (Wharton & Erickson, 1995). Burnout was, in
fact, better predicted by more conventional job characteristics (e.g. autonomy and number of hours
worked). This concurs with the conclusions of Schaufeli and Enzmann (1998), see above. Wharton
(1999) suggests that these types of finding do not necessarily refute Hochschild’s (1983) warnings
about the costs of emotional labour. However, they do, she suggests, imply that negative consequences
may only occur under limited conditions (e.g. Wharton’s 1993 study, emotional labour correlated with
increased emotional exhaustion among workers with low job autonomy, long hours and longer job
tenure).
In customer service work
Call centre workers are one of the most studied occupational groups in this literature because the
combination of high demands and low control found in such work is associated with high levels of
emotional labour and stress (Zapf et al, 2003). In particular, studies have shown that call centre work
involves adherence to restrictive display rules and is heavily monitored (Lewig & Dollard, 2003;
Holman et al, 2002; Zapf et al, 2003), emphasising the emotionally laborious nature of this type of
work.
A cross sectional study of 347 workers in two UK call centres (Holman et al, 2002) found that surface
acting and emotional dissonance were directly and negatively related to well-being (measured as
emotional exhaustion, job-related anxiety, job-related depression and reduced job satisfaction).
However, emotional dissonance was unrelated to deep acting, and deep acting did not show strong or
clear correlations with well-being.
Meanwhile, Lewig and Dollard (2003) examined the effects of emotional labour (measured as
frequency of interactions, duration of interactions and emotional dissonance) on emotional exhaustion
and job satisfaction, amongst 98 call centre agents. These workers experienced moderate levels of
emotional exhaustion (similar to police officers and probation officers, but less than other human
service workers). Of the emotional labour dimensions, only emotional dissonance was associated with
emotional exhaustion and reduced job satisfaction, concurring with other studies that suggest that
emotional dissonance mediates the emotional demands-emotional exhaustion relationship (e.g.
Abraham, 1998; Morris & Feldman, 1996). Furthermore, emotional exhaustion was largely explained
by high emotional dissonance, low rewards and low support. Interestingly, the inclusion of two openended questions found that employees rated providing good customer service as the most satisfying
aspect of the job, and dealing with angry and abusive customers as the most stressful aspect of the job.
This underscores the impact of verbal aggression, and also highlights the fact that emotional labour
can be rewarding, as well as stressful (see later).
Perhaps because of Hochschild’s (1983) pioneering work, flight attendants are another customer
service group who continue to receive researchers’ attention. For example, supporting the importance
of emotional dissonance, Heuven and Bakker’s (2003) study of flight attendants found that emotional
dissonance was a more important predictor of burnout than quantitative job demands and control.
Meanwhile, Williams (2003) asked nearly 3000 Australian flight attendants whether they found
emotional labour costly or satisfying. More respondents found it costly or stressful (44%), than a
source of satisfaction (34%), or neither (22%). She also found that certain factors affected how
respondents viewed their emotional labour. In particular, respondents who had been sexually harassed
were 51% more likely to view emotional labour as a cost. By contrast, those who felt valued by the
organization were six times more likely to see emotional labour as satisfying. Williams concludes that,
“the variables which emerge most prominently with regard to whether emotional labour is a cost or a
satisfaction (at the most powerful level of the multiple regression equation) are organizational ones,
sexual harassment and fatigue” (p.543).
27
Boyd (2002) reported data from a 1998 and 1999 questionnaire survey of railway and airline staff
which documented a sharp year on year rise in customer violence and verbal abuse. These authors
argued that the intensity of emotional labour correlated with escalating customer abuse. Customer
abuse can therefore be seen both a direct occupational health hazard, and an indirect one, via its effect
on emotional labour.
By contrast, a semi-structured questionnaire survey of 926 members of UK airline cabin crews (Bolton
& Boyd, 2003), gave a more optimistic and empowered view of emotional labour in this industry. This
study found little evidence of deep acting, and little evidence of an exploitative “invasion of self”
(p.303), amongst cabin crews. Rather, respondents were both highly aware and accepting of their
empty performances. The authors argue that the emotional labour literature has a deterministic feel,
implying the inevitability of harm. This undervalues the independent outlook which organizational
members bring to their work, and downplays the potential for “spaces for resistance and
misbehaviour” (Bolton & Boyd, 2003, p.294) in which real or unfaked emotions are expressed in
instances of humour, innuendo, imperfect service and so forth. An ethnographic study of airline cabin
crew and telephone sales assistants (Taylor & Tyler, 2000) confirmed this idea of spaces for resistance
to display rules. For example, airline telesales staff had learnt to detect when they were and weren’t
being monitored, and on numerous occasions were observed expressing deviant feelings and making
the interaction less one-sided, by, for example, disconnecting calls from rude or ignorant customers. In
contrast to much of the quantitative work in this area, a recent, qualitative study conducted amongst
tour representatives (Guerrier & Adib, 2003) also captured the positive sides of emotional labour,
portraying it as having few costs and being largely beneficial to these individuals.
In human service work
There are arguments both for and against the idea that burnout should be related to emotional labour in
human service work. Human service workers have emotionally demanding jobs (Zapf et al, 2003),
involving long and emotionally intense interactions (Brotheridge & Grandey, 2002), which may
involve the risk of over-identification and emotional over-extension (Hochschild, 1983; Wharton,
1999). Human service workers can be expected to give a great deal in terms of emotional support, so
there can be a lack of reciprocity in the client-carer relationship and lack of reciprocity, in turn, has
been associated with burnout (Kop et al, 1999). Also, human service work is likely to be particularly
emotionally exhausting in situations in which either caring has to be feigned because actual emotional
engagement is low, or strong emotions have to be suppressed in the interests of professional
detachment (Grandey, 2000; McQueen, 2004). On the other hand, the caring and vocational
motivations of human service workers should mean that their emotional displays are relatively
spontaneous and authentic, so that emotional dissonance is likely to be low (Himmelweit, 1999). Also,
human service workers often have high job autonomy and high work involvement, and are largely
unsupervised in their emotion work, all of which have been shown to mitigate the effects of emotional
labour on well being (e.g. Schaubroeck & Jones, 2000; Wharton, 1999).
Surprisingly, given these interesting arguments and counter-arguments, there has been little research
into the effects of emotional labour amongst human service workers. The handful of studies which
have examined this do not provide persuasive evidence that the caring labour performed by human
service workers leads to burnout, or other harm end points. For example, Fujiwara et al (2003) found
that the inter-personal conflict with clients was significantly associated with emotional exhaustion and
depersonalization in Japanese home care workers. However, this type of conflict did not correlate with
lack of personal accomplishment. Similarly, Jenkins and Elliott (2004) found that client-related
difficulties were significantly associated with emotional exhaustion and depersonalization amongst
nursing staff, but not personal accomplishment. Hayter’s (1999) qualitative study with specialist
AIDS/HIV community nurses showed that grief and loss associated with the death of patients was a
particularly stressful part of the role. However, this was also inextricably linked to the most rewarding
aspect of the role, i.e. the close, long term relationship with clients. Perhaps reflecting these client
relationships, depersonalization was remarkably low in this employee population.
Peeters and LeBlanc (2001) examined different types of job demands – quantitative, emotional and
organizational – as predictors of burnout in oncology healthcare workers. Surprisingly, emotional
28
demands were not significant predictors of either emotional exhaustion or depersonalization, whereas
quantitative demands were predictors of both and organizational demands predicted emotional
exhaustion. Ramirez et al (1996) researched stress and burnout in hospital consultants. Twenty-five
specific sources of stress were included in the survey, which aggregated into four main factors: Work
overload, feeling poorly resourced and managed, having managerial responsibilities, and dealing with
patients’ suffering. Of these, work overload made the greatest contribution to stress and dealing with
patients suffering, the least. Global reported stress was significantly related to psychiatric morbidity
(GHQ12) and burnout. Stordeur et al (2001) researched stress and burnout in nursing, using the
Nursing Stress Scale. This has three sub-scales: Physical environment/workload, social environment/
conflict and psychological environment (including dealing with death and dying). All three subscales
were individually significantly associated with emotional exhaustion, but the psychological
environment scale was not significant in the multiple regression analysis.
Finally, Brotheridge and Grandey (2002) give a comparative view of the effects of emotion work in
different types of job. The authors compared job-focused and employee-focused emotional labour as
predictors of burnout, controlling for the effects of negative affectivity, across five occupational
groups: Human service workers (n = 29), service/sales employees (n = 143), managers (n = 15),
clerical staff (n = 22) and physical labourers (n = 29). They found that despite the finding that the
different types of people work were more emotionally demanding than other types of occupation, there
were no significant difference in the levels of emotional exhaustion across groups. Moreover, human
service workers reported significantly lower levels of depersonalization, and higher levels of personal
accomplishment, compared with the other groups. The authors’ comments support the arguments
made by the writers in the nursing literature that human service workers are likely to truly care about
their patients and their work, and this will mitigate the stressful outcomes of their emotional labour.
2.5.5
Summary and conclusions
• From the large amount of evidence into customer service work, there is little evidence that
amount of customer contact per se is linked to burnout or other harm-type outcomes.
However, there is evidence in support of an association between emotional dissonance (or
surface acting) and burnout. This suggests that surface acting can contribute towards burnout,
although the consequences of deep acting are not clear. There is also good evidence that
suppressing negative emotions and managing agitation is harmful.
• Human service workers have the most emotionally demanding jobs, although emotional
dissonance may be lower for these workers compared with customer service workers. Whilst
there is much commentary in, for example, the nursing literature about the role and
importance of emotional labour in human service work, there is has been very little empirical
research into the consequences of emotional labour in human service occupations. The one
pertinent, methodologically strong study (Brotheridge & Grandey, 2002) suggests that high
emotional demands do not correlate with high burnout for human service work. This may be
due to the particular features of caring labour (or emotion work in a human service role) as
opposed to more conventionally understood emotional labour, with its emphasis on display
rules and acting. This confirms the value of examining the association between emotional
labour and harmful outcomes in human service work.
2.6
2.6.1
BURNOUT INTERVENTIONS
Introduction
This chapter addresses the sixth and final literature review issue and therefore examines the evidence
for and against the efficacy of various burnout interventions. These are organised by level, so that the
first section examines individual level interventions, the second examines interventions focused on the
individual-organization interface, the third section examines organizational level interventions and the
final section looks at burnout workshops.
29
2.6.2
Individual level burnout interventions
Examples of individual level interventions include: Self-monitoring (e.g. keeping a ‘stress diary’);
self-assessment (using the MBI as a diagnostic tool); didactic stress management (i.e. disseminating
information about stress and burnout); relaxation; and cognitive-behavioural techniques (i.e. therapies
aimed at changing appraisals in order to change people’s reactions to situations) (Schaufeli &
Enzmann, 1998).
Unfortunately, there is a paucity of evidence in the literature about the relative merits of different
types of burnout interventions. Schaufeli and Enzmann (1998), reviewing thirty different approaches
to preventing or combating burnout conclude: “Most interventions are rather general and not
specifically tailored to reduce burnout. Furthermore, there are only a few well–designed studies that
document the effectiveness of interventions” (p.182).
Of the well-designed studies that there are in the literature, several have examined the relative
effectiveness of relaxation-based therapies compared to cognitive-behavioural techniques. Schaufeli
and Enzmann (1998), for example, quote a study by Higgins (1986), which found that seven sessions
of both types of therapy were equally effective in reducing the levels of emotional exhaustion amongst
two samples of care-workers, compared with no-treatment controls.
Salmela-Aro, Naatanen and Nurmi (2004) carried out a similar evaluation of two types of therapeutic
intervention amongst 62 severely burned out workers, compared with 28 controls. Both interventions
used group therapy and focused on participants’ personal goals and projects. Both aimed to change
“the ways in which an individual perceives and deals with unfavourable work situations” (p.212).
However, one used active therapeutic interventions (e.g. muscle relaxation and drama) and the other
analytic group therapy (i.e. free association in the group). Both interventions consisted of 16 one-day
sessions over a ten-month period. The study found that burnout symptoms decreased from pre to post
measurement significantly more in the intervention groups than the controls, and that both
interventions were equally effective.
Turning to cognitive-behavioural interventions, Schaufeli and Enzmann (1998) cite a study by
Malkinson, Kushnir and Weisberg (1997), which examined the efficacy of rational emotive therapy (a
form of cognitive-behavioural training) in reducing burnout amongst a sample of blue collar workers.
The experimental group exhibited lower burnout than a non-treated control group, both immediately
after training and at the one-year follow up.
Van Dierendonck, Schaufeli and Buunk (1998) evaluated a five week, group-based burnout
intervention programme carried amongst care professionals. The intervention worked on the premise
that burnout results, to an important extent, from perceptions of inequity. It therefore involved
cognitive re-structuring exercises which aimed to reduce perceived inequity between the worker on the
one hand, and the recipients of his or her care, plus the employing organization, on the other. The
study found that levels of emotional exhaustion were reduced in the experimental group compared to
controls, and that this effect persisted after one year. The other burnout dimensions were unaffected
however, and the authors conclude that this confirms ‘previous findings’ that emotional exhaustion is
the easiest burnout dimension to influence using cognitive and relaxation-based therapies.
In short, it is difficult to form any meaningful conclusions from the above studies. There is some
suggestion of an equivalence paradox found also in psychotherapy in which different forms of therapy
appear to have broadly similar outcomes. There is also some suggestion that emotional exhaustion
may be the burnout dimension most amenable to change through individual-level intervention.
2.6.3
Interventions at the individual-organizational interface
Stress and burnout interventions at this level still focus on the individual, but in the context of his or
her organization – that is, they do focus on the individual but rather than simply dealing with reducing
burnout symptoms they aim to enhance work and organizational skills which may help reduce
30
burnout. Examples of such interventions include time management training, interpersonal skills
training (e.g. with aims like anger management or assertiveness training), job expectancy training (to
promote a realistic image of the job), career planning and peer support. In general, these interventions
seem targeted towards specific skills rather than on stress and burnout. As such, they are usually
effective in addressing these specific outcomes (like anger reduction) and sometimes effective in
influencing more general outcomes (like mental health or burnout) when these are measured.
For example, a study by Heaney, Price and Rafferty (1995) measured the effects of a worksite coping
intervention aimed at increasing the social support given to human service workers. The intervention
took a ‘train the trainer’ approach in which several people from each of a number of care-homes
attended, and then they, in turn, trained other staff. This yielded a large sample size (in the region of
1,000 individuals in the experimental group and the same in the control group). Five weeks after the
series of six training sessions the intervention positively influenced some aspects of individual and
work team functioning. Although in general it had little effect on mental health, it did reduce
depressive symptoms and somatization in those most at risk of leaving their jobs.
Gerzina and Drummond (2000) evaluated a multimodal cognitive-behavioural intervention aimed at
anger reduction in small sample of police officers. The course of six, weekly group-based sessions
(with homework assignments in between) involved training in relaxation skills, cognitive re-appraisal
and problem solving. It was found to be effective in teaching the police officers how to relax, and in
reducing their anger, both immediately post-intervention and at eight week follow up. The beneficial
effects of the treatment also seemed to generalise, as shown by a decrease in trait anxiety during the
study and at follow up.
Shimizu, Mizoue, Kubota, Mishima and Nagata (2003) evaluated a programme of communication
skills training (aimed principally at assertiveness) amongst a small sample of Japanese nurses.
Between pre-training and five month follow up they found an improvement in communication skills
and in personal accomplishment, but not in emotional exhaustion or de-personalization. Finally,
Ewers, Bradshaw, McGovern and Ewers (2002) evaluated the effects training a group of forensic
mental health nurses in psychosocial interventions. The trained nurses showed a significant decrease in
burnout, post training, whilst a control group showed a small increase. They conclude that equipping
the nurses with “a better understanding of serious mental illness” and “training them in a broader
range of interventions” helped them to “experience less negative effects of stress resulting from the
caring role” (Ewers et al, 2002, p.470).
2.6.4
Organizational-level interventions
Few, if any, organizational-level interventions seem to be aimed specifically at reducing burnout.
Rather, burnout seems to be an outcome occasionally measured in studies evaluating more general,
organizational-level stress interventions. Such interventions focus on surveillance (e.g. stress audits),
reducing stressors (like work overload), improving the fit between the employee and the organization
(e.g. career management and re-training) and the institutionalisation of projects and services (e.g.
employee assistance programmes and organizational development) (Schaufeli & Enzmann, 1998).
In the recent literature, there are just one or two examples of studies which examine the effects of an
organizational intervention on burnout. Schaufeli and Enzmann (1998) discuss an organizational
development programme introduced in four Dutch community mental health centres (Van Gorp &
Schaufeli, 1996). The intervention consisted of identifying problems within and across the centres
(e.g. work overload, and tensions between professional staff and support staff) and introducing sitespecific measures to tackle these issues. At the 18 month follow up, levels of burnout had not
decreased markedly, although employees were satisfied with how the problems had been tackled.
There were also significant improvements in other measures across one or more sites (e.g. absenteeism
and psychosomatic complaints).
Büssing and Glaser (1999) carried out an interesting longitudinal evaluation of an organizational
change programme that aimed to redirect patient care from a functional to a holistic model of nursing
31
in a German hospital. The essence of the change was to make the nursing process more individual
patient-oriented. The study compared the levels of stressors and levels of burnout amongst a group of
nurses in the experimental (or model) hospital, with a control group of nurses in two matched
hospitals, both at T1 (before the intervention) and T2 (immediately after the intervention, two years
later). The quantitative results showed that work stressors were substantially reduced in the model
hospital between T1 and T2. Interestingly and unexpectedly, however, emotional exhaustion and
depersonalization increased during this time. Qualitative research suggested that this apparent increase
in burnout was attributable to the greater social and emotional demands that were placed on nurses
enacting the new, holistic model. Supporting the putative link between emotional labour and harm end
points, it seemed that when nurses became less task oriented and more patient oriented, they felt more
burned out.
Lastly, writing about organizational level stress interventions, Briner and Reynolds (1999) suggest that
writers’ enthusiasm for these types of intervention is unmatched by empirical evidence to suggest that
they work. The lack of effects is unsurprising because the causes of stress in an organization are
unlikely to be easily identifiable and easily controllable, and interventions are equally unlikely to be
clean and controllable manipulations with uniformly positive effects. Rather, the effects of such
interventions are usually mixed: A wide range of outcomes are measured (e.g. mental health, job
satisfaction, performance, absence levels, burnout and stress) and typically, organizational-level
interventions affect some of these measures positively, but have no effect (or even an adverse effect)
on other measures (e.g. Reynolds, 1997; Schaubroeck, Ganster, Sime & Ditman, 1993; Van Gorp &
Schaufeli, 1996; Workman & Bommer, 2004). For interventions to work, they argue, decisions about
which intervention to use should be based on a careful, preferably longitudinal assessment of the
causes of negative employee states within an organization. Organizational-level interventions are only
appropriate if common, job-centred causes of stress are established throughout the organization, or at
least amongst identifiable groups. This is likely to be equally the case for burnout, suggesting that
organizational-level interventions will improve burnout if (and only if) organization-wide causes can
be identified and addressed.
2.6.5
Burnout workshops
The final intervention to be discussed, burnout workshops, combine some of the methods discussed in
the previous sections into one intervention. Such workshops aim to (i) increase participants’ awareness
of their work related problems and (ii) increase their coping resources by cognitive and behavioural
skills training and establishing support networks. Workshops last several days and include features
like self assessment, relaxation, inter-personal skills training and time management. Schaufeli and
Enzmann (1998) write that although burnout workshops are “booming business” (p.179), evidence
relating to their effectiveness is, once again, scarce. However, these authors cite three relatively recent
evaluative studies of burnout workshops, all carried out with human service workers, which showed a
significant beneficial effect on emotional exhaustion. They conclude, ‘Taken together, it seems that
multifaceted burnout workshops are effective in reducing levels of emotional exhaustion, even across
relatively long periods of time (up to one year). Other burnout dimensions are usually not affected
though.” (p.182). This concurs with evaluations of individual-level strategies, which tend also to
impact on emotional exhaustion but not the other dimensions (e.g. Van Dierendonck et al, 1998).
Emotional exhaustion thus appears to be the facet of burnout which is most easy to influence,
particularly through the cognitive-behavioural and relaxation-type therapies which dominate burnout
interventions.
2.6.6
Concluding summary
• There is a paucity of evidence for the effectiveness or otherwise of burnout interventions
hence any conclusions drawn can only be tentative.
• Less than a handful of studies have examined quite different forms of individual level burnout
interventions (e.g. cognitive-behavioural therapies). The results suggest some effect across
most interventions with emotional exhaustion being most amenable to change.
32
• The few interventions at the level of the individual-organization interface usually target
specific skills (e.g. anger reduction or communication skills training) and few address burnout
as such. The evidence suggests some effects on some dimensions of burnout and other aspects
of well-being.
• The very small quantity of evidence about burnout workshops suggests that they may have a
positive influence on emotional exhaustion.
• Overall, it seems from the small amount of available evidence that specific, targeted
interventions may be effective in reducing burnout and having other positive consequences.
However, there is insufficient evidence to say whether or not organization wide interventions
can reliably reduce burnout.
2.7
SUMMARY AND CONCLUSIONS
The purpose of this literature review was to address the following six issues.
1.
2.
3.
4.
The main correlates and causes of burnout in human service work
Traditional work stressors that have been found to relate to burnout (and related outcomes)
The pathways through which these stressors lead to these end points
The extent to which and how the emotional demands of human service occupations are
different from emotional demands of other related and un-related occupations
5. The relationship between emotional demands and burnout (and related outcomes) in customer
service work and human service work
6. Interventions which may ameliorate the impacts of such emotional demands on outcomes
It is important to note that while there is a reasonable quantity of empirical work about burnout the
quality of this work is severely constrained by the designs and methods adopted by almost all research
in the field. In particular, most studies are cross-sectional which means it is not possible to clearly
identify causes of burnout. Also, although burnout is assumed to be a process which unfolds over
time, there is little research which specifically examines this process. In spite of the popularity of the
burnout concept many very basic questions about its causes, the processes around it, and indeed its
status as a specific syndrome or psychological condition remain unanswered.
The main summary points are as follows.
• Most demographic and occupational factors correlate quite inconsistently with burnout
thought the few studies which examine personality do find consistent relationships between
negative affectivity and burnout.
• Some traditional work stressors such as role ambiguity, job demands, and social support have
shown reasonably consistent associations with burnout.
• While there is a general belief that burnout occurs in a sequence of stages different researchers
have proposed that different elements appear in a particular order or sequence. However, the
available evidence does not provide support for the existence of any specific sequence. Other
approaches to burnout processes emphasize intra-personal processes such as emotion
regulation.
• The nature of emotional demands does seem to be different in HSW and CSW probably due to
the different motivations behind and involvement in the emotional labour required. While
many jobs involve emotional demands, contextual differences across such jobs may make
mean that such demands while superficially similar may be qualitatively very different.
33
• While emotional demands such as the quantity of customer interaction do not seem to relate to
burnout other demands, such as surface acting (or emotional dissonance), do seem to be
associated. While in some senses HSW have the most emotionally demanding jobs they also
have the lowest levels of emotional dissonance.
• There is very little research on burnout interventions but that which has been done suggests
some of these interventions can be effective. However, given the limited of understanding of
the nature of burnout it is not always clear how such interventions work or whether they are
intervening specifically in burnout as such or other related processes.
34
3 STUDY 1: QUESTIONNAIRE SURVEY
3.1
PURPOSE OF STUDY 1 AND MAIN RESEARCH QUESTIONS
This study had two main purposes. The first was to explore the level of burnout within key UK human
service professions. This is done here by establishing levels of burnout and well-being in different
kinds of human service professions and comparing these levels to those found in other countries. The
second aim was to take a novel approach to identifying key variables associated with burnout. While
most studies of links between emotional work demands and burnout only explore the emotional
demands of the job there is evidence that many traditional ‘stressors’ may also be related to burnout.
In order therefore to explore whether emotional demands are associated with burnout after controlling
for traditional stressors, traditional dimensions of stressors, work demands and other job
characteristics were taken along with other measures of emotional demands developed specifically in
the emotional labour literature. A second category of relevant variables concerns personal dispositions
and attitudes such as negative affectivity and professional commitment. While these individual
differences are not directly relevant to the broader HSE objectives, examining and where necessary
controlling for their influence helps clarify the nature of relationships between job characteristics and
burnout. A third type of variable, which is likely to be important, are those concerned with
characteristics such as job tenure, gender and age. Here too, the findings from the literature review
were used to inform the choice of measures included in the survey.
Two main research questions each with two sub-questions were addressed in the quantitative study:
Research Question 1: What is the level of burnout and well-being in key human service
occupations in the UK?
• RQ1a: How do these levels of burnout and well-being compare to other samples?
• RQ1b: How do the levels of burnout and well-being compare across different human service
occupations?
Research Question 2: Which key variables are associated with burnout and well-being?
• RQ2a: Which key variables are associated with burnout and well-being?
• RQ2b Are different key variables associated with burnout and well-being in different human
service occupations?
3.2
VARIABLES
The variables relevant to understanding factors related to burnout were identified on the bases of the
literature review. Wherever possible, standardized and validated measures were chosen. All measures
showed reasonable to good internal consistency (reliability).
Table 1 shows the inter-correlations and reliability coefficients (Cronbach Alpha) of these variables.
Cronback Alpha is a measure of internal consistency of a scale and its value ranges from zero to one.
Values around or over .70 are considered acceptable and all the scales used are within this range. The
variables have been grouped into the three main categories, person characteristics, job characteristics
and emotional job demands. Appendix 1 contains full details of the measures.
3.2.1
Person characteristics
This group of variables consists of demographic factors (age, gender, marital status, school-aged
children), job related person factors (occupational tenure, job tenure) as well as positive affectivity
35
(PA) and negative affectivity (NA). PA and NA refer to trait affectivity which describes the
disposition to experience positive or negative affect or moods and emotions.
3.2.2
Job characteristics
Job characteristics identified as relevant were role overload, role conflict (conflicting work roles and
tasks), control and autonomy (control over tasks and the way they are performed) and social support
from colleagues and supervisor.
3.2.3
Emotional demands
We measured three types of emotional demands: Verbal aggression refers to the amount of verbal
aggression shown towards the participant by a client or customer; emotional labour measures the
degree to which surface acting (superficially displaying emotions) and deep acting (trying to feel the
emotions one has to show) are performed as part of the job; and, emotional dissonance is the
frequency with which positive emotions need to be displayed even if not felt and the degree to which
negative emotions need to be hidden at work.
3.2.4
Outcome variables
While we also measured variables such as job satisfaction or intention to quit, the main purpose of this
research was to examine the effects of emotionally demanding work on well-being hence this analysis
focuses mostly on health and well-being outcome variables. These include general psychological wellbeing and burnout, consisting of exhaustion, cynicism (referred to as depersonalization in the literature
review), professional efficacy (referred to as accomplishment in the literature review).
36
Table 1: Correlations and scale reliability coefficients in the diagonal
Control
Role
Role
&
overload
conflict
autonomy
Role overload
Control and
autonomy
Role conflict
Social support
Surface acting
Deep acting
Display
positive
emotions
Hide negative
emotions
Exhaustion
burnout
Professional
efficacy
burnout
Cynicism
burnout
Negative trait
affectivity
Positive trait
affectivity
Psychological
well-being
Job
satisfaction
Intention to
quit
Social
support
Surface
acting
Deep
acting
Display
positive
emotion
Hide
ExhauNegative
stion
emotion
Professional
efficacy
Negative Positive
General
Cynicism trait
trait
Health
affectivity affectivity
Job
Satisfaction
.85
-.00
.75
.49**
.37**
.37**
.03
-.25**
-.15**
-.25**
.05
.61
.37**
.42**
.07
.80
-.29**
.01
.67
.09
.69
.09
.11*
.19**
-.01
.06
.23**
.71
.18**
.03
.25**
-.03
.24**
.21**
.50**
.81
.49**
-.30**
.40**
-.32**
.43**
.02
.01
.21**
.88
-.15**
.33**
-.20**
.13*
-.17**
.18**
.22**
.15**
-.18**
.84
.26**
-.41**
.34**
-.27**
.32**
-.05
-.11*
.06
.64**
-.31**
.76
.26**
-.22**
.24**
-.20**
.37**
-.02
.05
.17**
.46**
-.20**
.36**
.85
-.13*
.41**
-.27**
.17**
-.24**
.15**
.14**
-.02
-.42**
.53**
-.56**
-.28**
.92
-.41**
.30**
-.36**
.32**
-.39**
.01
-.04
-.14**
-.54**
.26**
-.50**
-.54**
.50**
.91
-.23**
.53**
-.36**
.27**
-.31**
.11*
.15**
-.03
-.57**
.47**
-.75**
-.32**
.69**
.49**
.94
-.01
-.41**
.19**
-.18**
.13*
-.06
-.16**
-.01
.34**
-.27**
.49**
.14**
-.47**
-.28**
-.61**
** Correlation is significant at the 0.01 level (2-tailed); * Correlation is significant at the 0.05 level (2-tailed).
37
3.3
DESCRIBING THE THREE HUMAN SERVICE WORK GROUPS
In order to better understand each occupational group we will first describe the sample and each
occupational group in relation to emotional demands and outcome variables.
3.3.1
Recruitment of the sample and the three main human service work groups
The sample was recruited from a list of organizational contacts derived from commercial lists and
publicly available directories (e.g. Local Authority Directory, 2006), as well as the personal contacts
of the researchers. Each organizational contact was emailed a letter giving details of the research (e.g.
aims, methods, confidentiality and how to participate).
Eight organizations agreed to participate (e.g. insurance companies, social service departments of local
authorities). Additionally, a number of smaller organizations (e.g. doctor’s and dental surgeries,
schools) were recruited via personal contacts. Participating organizations were sent or given a batch of
questionnaires, with reply paid envelopes, to distribute to staff. The only criteria for participation were
that staff should have direct contact with the public, and be engaged in either customer service or
human service work. Staff were informed that participation was entirely voluntary, that the
questionnaires were anonymous, and were asked to return the questionnaire to the researchers in the
reply paid envelope provided. Additionally, respondents were asked to return a reply paid postcard,
giving their contact details, if they were interested in participating in the qualitative study.
1,000 questionnaires were distributed and 398 usable questionnaires were returned (i.e. a response rate
of almost 40%). The participants can be divided into three main groups:1. Human service workers (without teachers) – mainly social workers, care workers and housing
workers, plus a small number of medical staff (e.g. GPs) (N = 184), labelled HSW-Other.
2. Customer service workers – mainly call centre employees (N = 110), labelled CSW
3. Teachers – mainly primary school head teachers (N = 104)
The following analyses will explore the data for these three main groups of human service workers.
3.3.2
Person Factors: How can the main groups be characterized?
Table 2 shows that the majority of the sample is female, married or living with a partner, without
children in school age or younger. The vast majority of the sample is aged 50 or younger.
38
Table 2: Demographics
Gender
Age
Marital status
School age children at home?
Male
Female
Under 20
21-30
31-40
41-50
51-60
61+
Single
Married/with
partner
Divorced /separated
Widowed
Yes
No
HSWOther
(%)
44.5
47.0
0
27.2
51.1
64.4
47.2
66.7
25.8
CSW
(%)
Teachers
(%)
Total
(%)
36.4
24.2
100
57.9
23.9
9.9
2.8
8.3
50.8
19.1
28.8
0
14.9
25.0
25.7
50.0
25.0
23.3
28.1
71.9
1.8
28.9
22.3
25.6
18.3
3.0
30.5
55.3
16.5
28.3
60.2
54.3
100
58.6
41.5
22.9
0
17.1
31.3
22.9
0
24.3
27.1
8.9
.5
28.1
71.9
39
3.3.3
How can their jobs be characterized?
In this section we describe the characteristics of three groups with respect to job characteristics,
emotional demands and outcome variables. These analyses are conducted using one-way
Anovas with post-hoc Bonferroni tests to determine significant differences between the three
groups. In doing so, we will be better able to understand the nature of the three occupations with
respect to their emotional demands and job characteristics.
Emotional demands
On the whole, emotional labour (deep or surface acting) is on average only required rarely to
sometimes. Participants report the need to sometimes hide negative emotions and often display
positive emotions (no matter what they feel). Participants were further asked how often they
experienced verbal aggression against them in their job. Most participants answered “rarely” or
“sometimes” (see Table 3).
The three groups vary significantly with respect to reporting of emotional demands, except for
deep acting. Compared to CSW and Teachers, HSW-Other report significantly more verbal
aggression against them at work and a higher need to manage their emotions (both in relation to
showing positive emotions and hiding negative emotions). Teachers report significantly more
surface acting than the other two groups. CSW overall have the lowest emotional demands at
work.
Table 3: Emotional demands at work across the three occupational groups
HSW-Other
Mean
SD
Surface acting
Deep acting
Display positive
emotions (emotion
management)
Hide negative
emotions (emotion
management)
Verbal aggression
CSW
Mean
Occupational groups
Teachers
SD
Mean
SD
Total
Mean
SD
bc
2.58
.74
2.52
.78
2.87
.70
2.64***
.75
2.88
.82
2.84
.75
2.89
.70
2.87
.77
4.23
.56
3.67
.77
3.97
.55
4.01***abc
.67
3.28***ab
.83
3.51
3.03
3.15
3.01
1.19
2.55
1.08
2.24
.92
2.68***ab 1.14
a
b
Note. Anchors range from 1 to 5. = significant difference between HSW-Other and CSW, = significant difference
between HSW-Other and Teachers, c= significant difference between Teachers and CSW. *** Mean difference is
significant at the 0.01 level. ** Mean difference is significant at the 0.05 level. * Mean difference is significant at
the 0.10 level.
Job characteristics
Participants experience medium levels of role overload, conflict, clarity, control and ambiguity.
Also, participants report frequent social support by employees and supervisors. Apart from role
clarity and lack of ambiguity, the three groups differ with respect to their job characteristics (see
Table 4).
Teachers report higher role overload than the other two groups, more control and autonomy than
CSW, and less social support than both other groups. CSW report the least role overload, less
control and autonomy than both other groups, more social support than Teachers and less than
HSW-Other. HSW-Other report significant less role overload than Teachers, but more than
40
CSW, and less control and autonomy than Teachers, but more than CSW. HSW-Other also
report the highest levels of social support. In summary this means that Teachers report the
highest levels of role overload, and at the same time report the highest levels of control and the
lowest levels of social support.
Table 4: Job characteristics across the three occupational groups
HSW-Other
Mean
SD
CSW
Mean
Occupational groups
Teachers
SD
Mean
SD
Role overload
Role conflict
Control and
autonomy
2.91
.90
2.45
.77
3.83
.87
Total
Mean
SD
abc
3.03***
1.00
2.94
.80
2.80
.66
2.99
.83
2.91
.77
3.37
.76
2.73
.68
3.55
.70
3.24***ac
.79
Social support
3.58
.49
3.41
.58
3.19
.67
3.43***abc
.59
Note. Anchors range from 1 (strongly disagree) to 5 (strongly agree). For social support from 1 (never) to 4 (often).
a
= significant difference between HSW-Other and CSW, b= significant difference between HSW-Other and Teachers,
c
= significant difference between Teachers and CSW. *** Mean difference is significant at the 0.01 level. ** Mean
difference is significant at the 0.05 level. * Mean difference is significant at the 0.10 level.
Outcome variables
Do the three main job groups differ with respect to burnout, well-being, job satisfaction,
intention to quit? Again, providing this information gives us a better sense of the nature of each
of the three occupational groups, which provides a basis for discussing and understanding
findings that relate to the research questions.
Table 5 shows that Teachers report significantly higher levels of exhaustion than HSW-Other
and less professional efficacy than CSW. Further, Teachers report lower intention to quit and
higher job satisfaction than CSW. HSW-Other report significantly less exhaustion than
Teachers, more professional efficacy than CSW and less cynicism than CSW. Further, HSWOther report lower intention to quit than CSW. CSW report less professional efficacy than
HSW-Other, and more cynicism, higher intention to quit and less job satisfaction than both
other groups. In summary, Teachers report the highest levels of exhaustion but also the lowest
intention to quit and the highest job satisfaction.
41
Table 5: Outcome variables across the three occupational groups
HSW-Other
Mean
SD
CSW
Mean
Occupational groups
Teachers
SD
Mean
SD
Total
Mean
SD
Exhaustion
(burnout)
2.43
1.52
2.77
1.51
3.08
1.33
2.69***b
1.49
Cynicism
(burnout)
1.65
1.39
2.48
1.61
1.70
1.30
1.89**ac
1.47
Professional
efficacy
(burnout)
4.63
1.02
4.34
1.09
4.64
.88
4.56**ac
1.01
Psychological
well-being
2.97
.47
3.02
.52
2.89
.50
2.96
.49
Intention to quit
1.92
1.08
2.81
1.34
1.78
1.20
2.13***ac
1.26
Job satisfaction
3.19
1.01
2.60
1.07
3.45
.95
3.09***ac
1.06
Note: For intention to quit 1 (almost certainly stay) to 5 (almost certainly quit); Burnout 0 (never) to 6 (every day);
health 1 (worse than usual) to 4 (better than usual); job satisfaction 1 (little) to 5 (high).
a
= significant difference between HSW-Other and CSW, b= significant difference between HSW-Other and Teachers,
c
= significant difference between Teachers and CSW*** Mean difference is significant at the 0.01 level. ** Mean
difference is significant at the 0.05 level. * Mean difference is significant at the 0.10 level.
3.3.4
Summary
So, overall, how can the three groups be characterized? The three groups differ with respect to
job demands and outcomes as follows:
• HSW-Other reported the highest requirements for emotion management (hiding
negative emotions and displaying positive emotions), experience the highest verbal
aggression and feel the highest role overload, but also report the most social support.
With respect to burnout, well-being, intention to quit and job satisfaction HSW-Other
report average levels.
• CSW reported relatively low emotional demands (in particular the lowest level of the
need to display positive emotions), the lowest role overload and conflict while also the
lowest control & autonomy; they showed the highest levels of cynicism and the lowest
levels of professional efficacy. They also reported the highest level of intention to quit.
• Teachers have the highest levels of role overload and report the highest level of surface
acting, but only medium levels of emotion management. They report the lowest level of
social support and the highest level of exhaustion, but relatively low levels of cynicism
and relatively high professional efficacy. At the same time they also show relatively
high levels of job satisfaction and low levels of intention to quit.
In summary, the professional groups differ significantly with respect to reported job demands
and health-related outcomes. We would therefore also expect that the nature of the relationships
between emotional demands and burnout will vary across these groups. In the next section we
will try to explain the relationships between these variables.
42
3.4
RESEARCH QUESTION 1: WHAT IS THE LEVEL OF BURNOUT AND WELLBEING IN THE KEY HUMAN SERVICE OCCUPATIONS IN THE UK?
The results discussed in the previous section and Table 5 also give a partial answer to the first
research question: What is the level of burnout in key human service occupations in the UK first
compared to other national norms and, second, compared across the three occupational groups?
3.4.1 RQ1a: How do these levels of burnout and well-being compare to other
samples?
The authors of the Maslach Burnout Inventory (Maslach et al, 1996) advise that each of the
three dimensions of burnout should be treated separately. For this reason we will discuss the
three dimensions separately. Unfortunately Maslach et al (1996) do not provide clear norm
tables or a large range of samples for comparison purposes. We will therefore have to compare
the results found here to the Dutch, Finnish and Canadian samples (total N=5259) provided in
by Maslach et al (1996).
Exhaustion
The average level of exhaustion for our participants was M=2.69 (SD = 1.49). This means that
those items in the questionnaire measuring, in this case exhaustion, were reported on average as
being experienced by participants between “once a month or less” (scored 2) and “a few times a
month” (scored 3). According to the authors of the scale, Maslach et al (1996) this is an average
level for the Canadian sample (N=3727). It is considerably lower than the levels of exhaustion
found, for example, in Dutch civil servants or Dutch rural workers, however, it is comparable to
other human service workers from a Canadian sample (nurses, psychiatric workers, clerical).
Cynicism
The level of cynicism in the overall sample is M=1.89 (SD= 1.47), which means between “a few
times a month” (2) and “once a year” (1). According to the manual this again is an average level
rate in the Canadian sample and comparable to other human service work samples, however
significantly higher than Finnish participants in a computer job (N=289) or the Dutch civil
servants (N=956) and rural workers (N=761).
Professional efficacy
The mean for efficacy is 4.56 (SD=1.01), which is between “once a week” and “a few times a
week”. This is similar to the average level found in the Canadian sample. It is higher, however,
than the average human service worker in Canada, and more comparable to military, clerical
and technical workers in Canada. It is higher than the Dutch civil servants and lower than the
Dutch rural workers.
GHQ12 – levels of minor psychiatric disorder
The GHQ12 is a self-administered screening instrument that was developed to: A) detect an
inability to carry out one’s normal “healthy” functions; and, b), to detect the recent appearance
of new experiences of a distressing nature (Goldberg, 1972). There are several ways to score
this measure. The GHQ12 can be used to assess psychological well-being (see above and
Appendix 1) and to detect minor psychiatric disorders. So far in this report we have used the
measure as an indicator for well-being. As described above we scored the GHQ12 as a Likertscale measure and re-coded it, so that higher values indicate higher well-being (see Hardy,
Shapiro & Borril, 1997).
43
In order to compare our participants to other UK samples, we used the GHQ12 to detect the
level of psychiatric disorder. To do this so, we followed the case-scoring method (see e.g. Hardy
et al 1997). Participants who responded indicating that their health or symptoms were worse
than usual in relation to four or more of the GHQ12 items were labelled, using the same
terminology, as ‘cases’ (potential cases or likelihood of psychiatric disorder) and 26.9% of the
sample fitted into this category. All other participants were defined as ‘non-cases’. For this
comparison we draw on the case-rates for samples from NHS trust and samples from the BHPS
(British Household Panel Survey N=501), provided by Wall, Bolden, Borrill, Carter, Golya and
Hardy (1997, p. 520). Wall et al’s (1997) study revealed that overall 26.8% of respondents were
classified as probable “cases” with minor psychiatric disorder, which is virtually identical to the
participants in our study. However, this case rate is significantly higher than that of the BHPS
sample of 17.8% (χ2 = 153.4, df1, p<.01). In Wall et al’s (1997) sample women managers
(41%) and women doctors (35%) had the highest rates, followed by male medicine staff and
male and female nurses (30%/39%).
Overall, we can draw the conclusion that while our sample has a higher case rate than the
general population (as assessed by the BHPS), the case rate is broadly comparable to other
human service workers (NHS employees) in the UK.
3.4.2 RQ1b: How do the levels of burnout and well-being compare across
different human service occupations?
In this section we will compare the levels of burnout in the three occupational groups to each
other and to Maslach Burnout Inventory (Maslach et al, 1996) and GHQ norms. Table 5 shows
the relevant data. Turning first to exhaustion, Teachers report significantly higher levels of
exhaustion than HSW-Other, but not significantly higher than CSW. Also, the difference
between HSW-Other and CSW is not statistically significant. According to the norms, the
exhaustion levels of our three groups fall within the normal range.
With respect to cynicism, CSW report the highest rates of cynicism, which would also be
considered high in the norm table provided by Maslach et al (1996). However, HSW-Other and
Teachers do not significantly differ from each other (or from the norm). CSW not only report
the highest levels of cynicism, but also the lowest levels of professional efficacy. However, this
level is still within the norms provided. The efficacy levels of HSW-Other and teaches do not
significantly differ from one another.
We also compared the three occupational groups with respect to cases of minor psychiatric
disorders (according to the procedure described in the above section). As shown in Table 5,
there are no significant differences between the three occupational groups. A χ2 - analysis with
the case-rates does not show any significant differences either (χ2 = 1.40, df2, p>.10)
3.4.3 Summary and conclusion
•
•
•
•
•
While there are clearly differences between the levels of burnout amongst the three occupational groups, the overall level of burnout of this UK sample is well within
international averages
Only one score can, in comparison to a large North American sample be considered
outside the norm, and that is the cynicism level of CSW
CSW show the highest level of cynicism and the lowest level of professional efficacy.
Teachers experience higher exhaustion rates than HSW-Other, but not than CSW
Overall our participants appear to have a higher case rate of minor psychiatric disorder in
comparison to a representative sample of the UK population, however, they show similar
44
rates to NHS employees. The three occupational groups do not significantly differ from
one another with respect to their case rate of minor psychiatric disorder
3.5
RESEARCH QUESTION 2: WHICH KEY VARIABLES ARE ASSOCIATED
WITH BURNOUT AND WELL-BEING?
In this section the results relevant to research question 2 are discussed. Guided by the literature
review, we examined three groups of variables related to burnout: Person factors, job
characteristics and emotional demands. In the first step we analyze how emotional demands
explain burnout and well-being over and above person factors and job characteristics as these
are already known to affect burnout and well-being. In the second step we explore the statistical
predictors of burnout for each occupational group separately. As this is a cross-sectional study,
it should be noted that it is not possible to make any claims about causality though we can
analyze what statistically contributes to burnout and well-being.
3.5.1
RQ2a: Which key variables are associated with burnout and well-being?
One important question this study seeks to address is which factors are associated with burnout
and well-being. So, what is associated burnout and well-being? While it is known, in general,
that trait affectivity and job characteristics are related these outcomes, the main purpose of these
analyses is to explore if the emotional demands of the job also affect burnout and well-being
once the effects of trait affectivity and job characteristics have been taken into account. In other
words, do emotional demands affect these outcomes over and above the effects of trait
affectivity and traditional job characteristics?
In order to address this question, we conducted hierarchical regressions, entering the variables
in three steps: First, person characteristics (PA, NA, gender, age, occupational tenure, job
tenure), secondly, job characteristics (role overload, control & autonomy, role conflict and
social support) and thirdly, emotional demands (verbal aggression, surface and deep acting,
having to display positive and negative emotions). Table 6 summarizes the results of the third
step. Note that the first and second steps were omitted from the tables for the sake of
readability. We present the results for the three burnout dimensions, exhaustion, cynicism and
professional efficacy and well-being (GHQ12).
Explaining Exhaustion
Table 6 shows that experiencing verbal aggression in the job and surface acting are significantly
related to exhaustion after controlling for the effects person factors and traditional job
characteristics.
Explaining Cynicism
Verbal aggression and demands to display positive emotions are significantly related to
cynicism beyond job characteristics and trait affectivity. In other words, the more aggression
and the lower the requirement to display positive emotion, the more cynical participants report
to be.
Explaining professional efficacy
For professional efficacy, positive trait affectivity and job characteristics explain the majority of
variance. However, two of the emotional demands do play a role beyond these variables. The
more individuals feel they have to display positive emotions at work and the more they perform
deep acting, the higher their professional efficacy levels.
45
Explaining well-being
Well-being is explained by trait affectivity, control and social support, but none of the measures
of emotional demands were found to relate to well-being once these other variables had been
taken into account.
Table 6: Results of hierarchical regression analysis (full sample)
Exhaustion
Model
3
(Constant)
Negative trait affectivity
Positive trait affectivity
Gender
Age
Children at home
Tenure occupation
Tenure job
Role overload
Control and autonomy
Role conflict
Social support
Verbal aggression
Surface acting
Deep acting
Display positive
emotions
Hide negative emotions
F
R2
Model 2/3 ΔR2
Cynicism
Beta
Sig.
Beta
Sig.
.184
-.191
.010
-.037
-.004
-.034
.065
.353
-.156
.018
-.041
.126
.106
.032
***
***
.131
-.356
.133
-.105
-.021
-.036
.155
.139
-.187
.088
-.041
.098
.033
.021
***
***
***
**
-.069
-.109
**
.076
-.009
22.77
.48
.03
***
***
***
**
***
***
22.87
.47
.02
***
***
***
*
**
Professional
Efficacy
Beta
-.025
.415
.073
.108
.016
.005
-.073
-.089
.098
-.067
-.008
.074
-.032
.094
.112
Sig.
**
-.335
.315
.026
.000
.010
.061
-.071
-.236
.083
-.001
.096
-.029
-.045
-.026
**
-.040
***
**
*
*
.054
***
**
12.67
.33
.03
Wellbeing
Beta
***
***
***
**
.009
***
***
24.03
.50
.00
Note. *** Significant at the 0.01 level. ** Significant at the 0.05 level. * Significant at the 0.10 level. Model 2/3
∆R2 refers to the additional variance accounted for by including the five emotional demands variables once person
and traditional job characteristics have been taken into account.
3.5.2
Sig.
Brief summary
The results are presented visually in Figure 2. Generally speaking, after controlling for trait
affectivity and traditional job characteristics, emotional demands related quite weakly (though
statistically significantly) to burnout and were unrelated to well-being (as assessed by the
GHQ12). Exhaustion is related to only two emotional demands – verbal aggression and surface
acting – such that more aggression and more surface acting is associated with higher levels of
exhaustion. Cynicism is significantly positively related to role overload, role conflict, verbal
aggression and NA. However, control and autonomy and PA are negatively related to cynicism
such that higher levels of control and autonomy and PA are related to lower levels of cynicism.
Professional efficacy is positively related to two emotional demands - deep acting and having to
display positive emotions at work.
46
***
ns
Positive correlation +
Negative correlation -
Role
overload
Exhaustion
Verbal Aggression
Role
conflict
NA
PA
1. Step (NA/PA)
Cynicism
Control &
autonomy
Prof.
Efficacy
Social
support
Wellbeing
Surface Acting
Display positive
emotion
Deep Acting
2. Step (job characteristic)
3. Step (emotional demands)
controlling for NA/PA
controlling for NA/PA & job characteristics
Figure 2: Summary of results from research question 1a
3.5.3 RQ2b: Are different key variables associated with burnout and well-being
in different human service occupations?
Given the differences in the nature of the three job groups described above, it seems likely that
there are differences between the three occupational groups in the ways in which person factors,
job characteristics, and emotional demands relate to burnout and well-being. The following
analyses are again hierarchical regressions for each of the three occupational groups separately.
Table 7 summarizes the significant results.
Human service workers
For this group, three emotional demands measured significantly related to the outcome variables
beyond person factors and job characteristics:
• Surface acting and hiding negative emotions statistically predict exhaustion
• Displaying positive emotions statistically predicts professional efficacy
47
Table 7: Hierarchical regressions for burnout and GHQ12 split by occupational group, controlling for person factors and job characteristics
Exhaustion
HSWCSW
Other
Beta
Beta
Model
3
(Constant)
Negative trait affectivity
Positive trait affectivity
Gender (1 male, 0 female)
Age
Children at home
Tenure occupation
Tenure job
Role overload
Control and autonomy
Role conflict
Social support
Verbal aggression
Surface acting
Deep acting
Display positive emotions
Hide negative emotions
F
R2
Model 2/3 ΔR2
.181***
-.059
.053
.041
.035
-.052
.034
.371***
-.057
.077
-.012
.083
.143**
.029
-.093
.156**
Teacher
Beta
Cynicism
HSWCSW
Other
Beta
Beta
Teacher
Beta
.211**
.273*** .201***
.120
.136*
-.323*** -.243*** -.187*** -.468*** -.392***
-.054
-.065
.194***
.073
.107
.149*
-.167*
-.077
.087
-.032
-.008
-.066
-.018
.006
-.023
.068
.006
.019
.099
-.115
-.057
.156
.202***
.011
.106
.104
.322*** .269***
.019
.028
-.242***
-.136
-.075
-.210***
-.180
.012
.050
.090
.035
.185*
-.106
.017
-.001
-.170**
.142
.267***
.149*
.059
.141*
.133
.019
.109
.081
-.032
.187*
.065
-.034
.045
-.083
-.020
.003
.057
-.082
-.038
-.043
.008
.055
-.015
.001
.050
Professional efficacy
HSWCSW
Teacher
Other
Beta
Beta
Beta
Psychological well-being
HSWCSW
Teacher
Other
Beta
Beta
Beta
.066
.433***
.124*
.178**
-.009
.012
-.062
-.167**
.036
-.125
.054
.029
-.007
.062
.146*
.136
-.066
.424***
.083
.068
.133
.082
-.104
.040
.054
.139
-.057
.076
-.077
.155
.103
.038
-.180**
.419***
.097
-.070
.031
-.032
-.035
-.113
.249***
.071
.048
.243***
.011
-.017
.119
-.159
-.437*** -.256*** -.303***
.284*** .492***
.158*
.011
-.039
.044
.008
-.016
.180*
.055
-.119
.023
-.011
.004
.247**
.036
.049
-.372***
-.135**
-.089
-.376***
.135**
.000
.187**
-.029
-.076
.027
.018
.165**
.022
.023
-.151**
-.035
-.044
.018
-.068
-.042
.000
-.083
.004
-.136
.120
.008
-.028
-.015
11.33***
6.80***
7.67***
10.19***
6.6***
7.00***
7.00***
2.94
4.93***
14.11***
9.12***
5.50***
.49
.05***
.48
.53
.46
.50
.40
.43
.01
.02
.02
.03
.55
.00
.56
.02
.36
.05***
.23
.05**
.46
.00
.03*
.00
Note that these are the results of a three-step hierarchical regression (see Table), where the data for model 1 and 2 has been omitted for the sake of readability.
*** Significant at the 0.01 level. ** Significant at the 0.05 level. * Significant at the 0.10 level
48
Customer service workers
For CSW, verbal aggression is the only emotional demand measured that statistically predicts
any variance over and above person factors and job characteristics. Specifically, verbal
aggression relates to exhaustion, cynicism and well-being. In other words, the more verbal
aggression participants reported, the higher their levels of exhaustion and cynicism and the
lower their level of well-being.
Teachers
For this group, verbal aggression and surface acting explain variance in outcome measures over
and above person factors and job characteristics. Specifically,
• Verbal aggression positively relates to exhaustion as well as positively to professional
efficacy. In other words, while more verbal aggression is related to higher levels of
exhaustion, it is at the same time also related to higher levels of professional efficacy.
• Surface acting positively relates to cynicism, the more surface acting, the higher the
level of cynicism.
3.5.4
Summary
The three occupational groups differ somewhat with respect to what is associated with burnout
and well-being beyond person factors and job characteristics. Exhaustion is significantly
associated with surface acting and having to hide negative emotions (HSW-Other), and with
verbal aggression (CWS and Teachers). Cynicism is associated with verbal aggression (for
CSW) and surface acting (Teachers). Professional efficacy is associated with having to display
positive emotions (HSW-Other) and verbal aggression (Teachers). The only emotional demand
associated with well-being over and above person factors and job characteristics is verbal
aggression (CSW).
3.6
SUMMARY AND CONCLUSION
• The first set of analyses compared scores on emotional demands, job characteristics and
outcomes across the three groups HSW-Other, CSW and Teachers.
• The three groups vary significantly in the emotional demands they report with the
exception of deep acting where there is no difference across the groups. For surface
acting, Teachers report higher levels than both HSW-Other and CSW but there is no
difference between HSW-Other and CSW. In terms of the requirement to display
positive emotion, all the groups were different from each other with HSW-Other
reporting the highest levels, followed by Teachers, and then CSW. HSW-Other also
reported higher levels of the requirement to hide negative emotions than both CSW and
Teachers with was no difference between these latter two groups. Again, HSW-Other
reported higher levels of verbal aggression than CSW and Teachers with no difference
between these groups. This suggests that the nature or quality of emotional labour may
vary in important ways across different occupational groups and contexts.
• The three groups also varied significantly in relation to scores on measures of job
characteristics with the exception of role conflict. In relation to role overload all the
groups were significantly different from each other with Teachers reporting the highest
levels followed by HSW-Other then CSW. Teachers and HSW-Other reported a similar
level of control and autonomy and both reported higher levels than CSW. Differences
in social support were found between all the groups with HSW-Other reporting the
highest levels followed by CSW and then Teachers.
49
• Addressing RQ1b, comparing levels of outcome variables across the three groups there
were significant differences on all the measures with the exception of psychological
well-being (GHQ12). In relation to exhaustion, Teachers reported a higher level than
HSW-Other but otherwise there were no differences between the groups. For cynicism,
CSW reported higher levels than both HSW-Other and Teachers (with no differences
between these two groups). In relation to professional efficacy, HSW-Other and
Teachers reported a similar level and both more than CSW. Intention to quit was higher
for CSW than for both HSW-Other and Teachers (with no differences between these
two groups). Similarly, job satisfaction was lower for CSW than for both HSW-Other
and Teachers (with no differences between these two groups). The difference in
outcome variables across the three groups again emphasizes the different emotional
experiences in different occupations engaging in emotional labour.
• The second set of analyses addressing RQ1a compared scores on the burnout measure
and GHQ12 with the available norms. For exhaustion, the average level reported by
participants was similar to the average found in a large Canadian sample but lower than
the levels found in samples of Dutch civil servants and rural workers. The average level
of cynicism in the sample was similar to averages found in the large Canadian sample
but was significantly higher than Finnish computer workers or Dutch civil servants.
Professional efficacy was broadly similar to the average found in the large Canadian
sample, but higher than human service workers in Canada and Dutch civil servants but
lower than Dutch rural workers. Using case scoring, the percentage found in our
sample was higher to that found in the general population (as assessed by the British
Household Panel Survey) but very similar to the percentages found in samples of NHS
staff. Generally, however, the average levels of GHQ12 and burnout scores in the
sample are comparable with norms found elsewhere.
• The third set of analyses addressing RQ2a, examined the extent to which emotional
demands were related to the outcome variables once person characteristics and
traditional job characteristics had been taken into account. In each case, with the
exception of well-being (GHQ12), emotional demands were significantly though not
strongly associated with outcome measures. Experiencing verbal aggression and the
requirement to engage in surface acting were associated with higher levels of
exhaustion. Higher levels of verbal aggression and the lower requirements to display
positive emotions were associated with higher levels of cynicism. The requirement to
display positive emotions and engage in deep acting were associated with higher levels
of professional efficacy. In general, these results suggest two things. First, that
measures of well-being are not equally sensitive to emotional demands and, second, that
personality and traditional ‘stressors’ or job characteristics may play a much more
important role in burnout than emotional demands.
• The fourth set of analyses, addressing RQ2b, explored whether there were differences in
the three groups in relation to which emotional demands were associated with
outcomes. For HSW-Other, surface acting and hiding negative emotions were related to
exhaustion while displaying positive emotions was related to profession efficacy. For
CSW, verbal aggression was related to higher levels of exhaustion, cynicism, and lower
levels of well-being. For Teachers, verbal aggression was related to both exhaustion
and professional efficacy while surface acting was related to cynicism. These specific
patterns of relationships strongly suggest that it is not possible to generalize across
contexts about how the emotional demands associated with emotional labour impact on
well-being.
50
4 STUDY 2: DAILY DIARIES AND INTERVIEWS
4.1
INTRODUCTION AND OVERVIEW
Section 4 reports the results of Study 2, which was a qualitative diary and interview-based study
of events at work. This aimed to:
• Provide detailed descriptions of the types of situation that may ultimately cause harm in
emotionally demanding occupations
• Explore which types of event are experienced as emotionally demanding, and in what
ways
• Provide insights into the processes involved in burnout
Section 4.1 provides the background to the results of Study 2. Firstly, it details the methods
used, explaining the data collection, the sample and the data analysis. Next, it gives a very
broad overview of the findings.
4.1.1
Method and design
The above aims were addressed using a qualitative methodology, in which participants
completed a diary of events at work and home for a period of two consecutive weeks. The
findings from the diary study were augmented by follow up telephone and face-to-face
interviews. It should be noted that while the earliest research on emotional labour did use
qualitative methods, more recently quantitative methods have dominated. This study is also
unusual in that it uses diary studies and an event-based approach to explore the experience of
emotional demands and how these may lead to burnout. The diaries, the follow-up interviews
and the sample will now be discussed.
The diaries
Qualitative diaries were chosen because they are an effective way of capturing detailed
descriptions of events soon after they happen (Symon, 1998). They have also been found to be
an effective way of collecting data on emotions (e.g. Macdonald & Morley, 2001) and of
providing accounts of phenomena as they unfold over time (Symon, 1998).
The diary contained a checklist of emotions and asked participants to describe the following
type of events at work:
• An interaction with a client that made you feel positively
• An interaction with a client that made you feel negatively
• A work-related event that made you feel positively that did not involve a client
interaction
• A work-related event that made you feel negatively that did not involve a client
interaction
• A home-related event that made you feel either better or worse, overall
The diaries were either completed via email or hard copy according to preference.
Appendix 2 for full diary content.
51
See
Follow-up interviews
The findings from the diary study were augmented by follow-up telephone and face-to-face
interviews with the majority of the diarists. The purpose of these interviews was to gather more
detailed information on some of the events described in the diaries and to probe, in more detail,
the role of emotion work in these events. A sample interview guide is included in Appendix 3.
Sample
The sample of diarists was recruited from respondents to the quantitative survey. Respondents
who were interested in participating in the diary study were asked to return a reply-paid
postcard, giving their contact details, to the researchers. They were then contacted by the
researchers. Approximately 100 postcards were received, of which 38 people agreed to
participate. 12 people either dropped out of the study or completed one week (or less) diary
entries. Also, two individuals were excluded as their diaries contained very few events and very
limited detail so did not add useful information.
The findings are based on the diaries of 24 research participants, and interviews with 18 of
these. Three quarters of the sample (18) of the diarists were female, and six were male. A
breakdown of the sample by occupation is given in Table 8.
Table 8: Breakdown of the diarists/interviewees by occupation
Occupation
Number of Diarists
Head teachers
Social workers and social
work assistants
Care workers
Housing workers
Miscellaneous other e.g. home
tutor, community substance
misuse worker, police doctor
4.1.2
10
Number
telephone
interviewed
4
Number
interviewed face
to face
4
4
3
3
1
0
1
2
2
2
4
2
0
Data analysis
The diary data were analysed according to the guidelines for template analysis (King, 1998).
The events were coded according to both (i) descriptive codes, which described the type of
event (e.g. a positive meeting with a client); and (ii) more interpretive codes, which described
the meaning that diarists gave to the event and the reasons behind their emotional reaction (e.g.
turning a difficult situation around). The diary coding template can be found in Appendix 4.
The interviews were structured around the diary entries. Because of this, the interview data were
coded according to the same coding template as the diaries (Appendix 4). However, a number of
new codes, which were specific to the interviews, were also added (see Appendix 5).
The diary and interview data were then summarised in a matrix format (Miles and Huberman,
1994). These matrices were then used as the basis for the narrative accounts.
4.1.3
Brief overview of results
The data were quantified to give a very broad overview of the patterns (see Table 9). The
numbers of events which were categorised as positive overall, versus the number that were
52
categorised as negative overall, within each category (i.e. client interaction versus other work
event) were calculated. There were more positive work events than negative work events (352:
311), and nearly twice as many positive home events as negative (183:94). There were also
more positive client interactions than negative (182:134) and ‘positive client interactions’ was
the most popular category of work event (182, compared with 170, 134 and 177). This suggests
that client contact is one of the most positive and rewarding aspects of the service worker’s role.
However, with 134 negative client interactions and 177 other negative non-work events, there
were clearly plenty of stresses and strains in human service work, and these are the main focus
of the qualitative findings.
Turning now the qualitative findings, the first section describes the content of the positive
events in the diaries. This places the causes of harm and the nature of emotionally-demanding
situations (i.e. the negative work events) into a wider context, which takes account of the
diarist’s entire (positive and negative) experience of work and non-work. Thus it tries to
provide a more complete, rather than partial, view of these data.
Next a more detailed examination of the negative events is provided in order to yield an
understanding of (i) which types of event were seen as difficult or potentially harmful and (ii)
which types of event were experienced as emotionally demanding and why. The data reported
here are largely taken from the diaries.
The third section examines the emotional demands of human service work, drawing mostly on
the interview data. Firstly, it discusses emotionally demanding situations which were associated
with stress and burnout-type emotions. Secondly, it examines how people performed ‘emotion
work’ within particular events, and how they experienced different types of emotion work as
emotionally costly and/or rewarding.
Some insights into stress and burnout in human service work that were revealed by the
interview and diary data are discussed in the fourth section. In particular, it examines some
factors associated with stress and burnout-type emotions. These were mainly (but not
exclusively) derived from on a sub-group of diarists who are here termed the ‘distressed group’.
These particular diarists seemed to experience a relatively high frequency of stress and burnouttype emotions across the diary period. Their data revealed insights into the potential causes of
stress and burnout and the processes underlying these harm end points. However, it is important
to note that the diarists were not classified according to any scientific measures of stress or
burnout (e.g. their score on the MBI), so that it would be inappropriate to label them either of
these. They are called the ‘distressed group’ for ease of reference, and also because we have
chosen to interpret sections of their data according to what they seem to reveal about harm end
points like stress and burnout.
The final section of the qualitative results examines two sets of additional findings from Study
2. These findings are more indirectly linked to the aims of the study and are included because of
their interest value. First, the chapter describes diarists’ perceptions of the harmful
consequences of stress and emotional load at work. Secondly, it explores some potential buffers
to stress and burnout, and some insights into how workers coped with the emotional
components of their jobs.
53
Table 9: Overview of positive and negative events, by diarist
Diarist
1
2
3
4
5
8
11
13
16
19
20
21
22
23
24
25
28
29
31
32
33
36
37
38
Total
Occupation
Head teacher
Head teacher
Head teacher
Manager - Education
Head teacher
Head teacher
Head teacher
Head teacher
Head teacher
Housing worker
Housing worker
Head teacher
Social worker
Social worker
Social worker
Housing worker
Residential care worker
Home tutor
Manager in residential care
home
Head teacher
Manager in residential care
home
Social worker
Community Substance
Misuse worker
Consultant and Police
Doctor
Number of
positive client
interactions
across diaries
11
11
6
8
9
1
7
15
8
6
4
14
9
4
11
4
5
8
7
Number of other
positive work
events across
diaries
7
15
2
6
11
11
6
9
8
4
8
8
2
2
6
1
3
7
6
Number of
positive non-work
events across
diaries
11
12
6
8
7
9
5
12
6
5
9
14
3
0
13
4
6
10
3
6
8
13
7
9
11
6
3
8
12
2
8
7
7
7
10
5
5
3
7
6
11
10
3
6
11
10
8
8
2
Total = 182
Total = 170
Total = 183
Total = 134
Total = 177
Total = 94
54
Number of
negative client
interactions
across diaries
7
7
2
7
5
8
8
10
3
8
4
8
2
4
7
3
2
5
7
Number of other
negative work
events across
diaries
10
9
4
6
6
6
11
10
4
7
4
17
3
4
5
0
7
6
13
Number of
negative nonwork events
across diaries
4
4
1
0
1
3
0
8
5
11
3
2
9
3
1
0
2
5
7
4.2
EVENTS ENGENDERING POSITIVE EMOTIONS
It is noteworthy that in this study, the majority of reported events were regarded as positive.
Generally speaking, according to this brief snapshot of the content of human service work, the
rewards of most jobs outweighed the hassles and difficulties. This chapter gives a brief
overview of the rewarding aspects of human service work, as reported in the diaries. This is
done in order to place the ‘causes of harm’ into a broader context, which takes account of both
the positive and negative features of the job.
4.2.1
Enjoyable client contact
Interestingly, the most frequently reported type of event (positive or negative) in the diaries
centred on enjoyable interactions with clients. In these events, diarists simply took pleasure in
spending time with the people in their care. For example, a worker in a residential children’s
home wrote:
“I was in the office, the door was wide open. Key child walked in and kissed me and
said ‘Hi mum’. This made me laugh. He was just being cheeky. I needed to relax a bit.”
Reported instances of enjoyable client contact were most prevalent amongst those working with
children (i.e. head teachers and care home workers) and the elderly (i.e. social workers),
although it also featured in the diaries of an individual working with the mentally ill and another
working with drug addicts.
Sometimes, a client interaction was seen as enjoyable because the diarist felt that he or she was
handling it successfully or well. These types of event were also associated with the satisfaction
of being in a helping role. For the head teaches in particular, enjoyable client contact was
frequently referred to as the most important part of the job, or the part of the job that gave it
meaning and personal satisfaction.
4.2.2
Meeting or interactions with clients which had a positive outcome
Another commonly-reported type of event was a more purposeful interaction with a client (often
a meeting) that had gone well. In some of these cases, the positive feelings arose from a sense
that the diarist, and/or the wider organization, was managing a client’s situation effectively.
Also, diarists were often pleased or satisfied because the interaction had run more smoothly, or
been more successful, than they had anticipated.
Some particularly interesting client interactions centred on how a diarist was able to turn a
difficult or challenging situation around, by, for example, calming a client down or allaying
their fears. In such circumstances, diarists sometimes alluded to using specialist skills, focused
on a client’s emotions, to bring about a change in attitude and ensure a positive outcome. The
two examples below were written by members of the management team in two care homes:
Example One: “A child was very angry and upset. She was kicking, biting, screaming
as she wanted to see her mum. Myself and another colleague restrained her but it
ended positively as we encouraged her to get her anger out in a more positive way –
hitting an inflatable toy, stamping feet and kicking a football really hard. She then
calmed down and had a cuddle and was really tearful but I felt positive as she has now
started to express herself and get her anger out.”
55
Example Two: “Met with child to discuss his visit to a potential placement. Initially he
tried to control the conversation/situation but I was able to turn this around and used
many different skills in order to get him to agree to a ‘trial’ at the placement. I had
gone into the meeting feeling that it would be difficult and confrontational and was
pleased with the outcome.”
In both of these events, the diarist was required to deal with a client’s extreme emotional
reaction and the result, in each case, was successful. The diarists emphasised how they were
using their professional skills effectively to achieve the positive outcome. These examples
therefore serve to show how emotionally charged situations at work can be experienced as both
challenging and personally rewarding. This point is revisited later.
4.2.3
Client contact in which the client gave positive feedback
Diarists frequently reported instances in which they received positive, personal feedback from
their clients, in the form of gratitude, appreciation, praise or affection. For example, a careworker reported a timely hug from a child, whilst a housing worker reported a rare verbal ‘thank
you’ from a tenant. Sometimes, diarists imbued such interactions with greater meaning by
explicitly linking them to the demands of the job, inferring that events like these made the job
worthwhile in spite of the many challenges they faced.
4.2.4
Other types of positive client interaction
Other types of rewarding client interaction included instances in which clients were observed to
be making progress towards a goal, as in the case of a teacher reporting a students’ learning, or a
care worker referring to a child’s emotionally mature behaviour. Sometimes in these instances, a
diarist reported a vicarious pleasure in the client’s own enjoyment of his or her achievement. An
example of this was given by a diarist who worked in the community with drug addicts:
“My session with my first client made me feel very proud - I managed to keep her coming
even when things were very negative. I enabled her to get help for some of the
underlying problems and then encouraged her to go on a creative writing course. I am
very excited and optimistic for her. I was happy and relieved that she had a good week
and when she said that life was good - that was my reward for the work I have done with
her - so for a short while I was buzzing.”
Some positive events were occupation-specific. For example, head teachers sometimes wrote
about observing the successes of their staff, like a teacher’s effective lesson and the beneficial
effect of this on the children. Likewise, social workers and housing workers enjoyed resolving
clients’ problems, often by securing positive action from a third party on behalf of a client.
4.2.5
Other types of positive, work-related event
There were fewer positive events in the diaries that did not involve client contact compared to
those that did. In particular, head teachers often wrote about pleasing interactions with third
parties or peers (e.g. school governors, local authority personnel and other heads). Such
interactions were viewed positively either because they involved helping to develop the school,
or they involved feelings of status, privilege and co-operation in the context of other schools or
the wider community. As such, these events were associated with the status of head teachers and
their leadership role.
56
Another group of events was themed around progress on work tasks, most frequently
administrative tasks. For head teachers, this seemed to be the positive corollary of having so
much paperwork to deal with in the first place. Because of the volume of paperwork and the
head teachers’ attitude towards it (see later), progress with administrative tasks seemed to be a
relatively minor satisfaction rather than an actively rewarding aspect of the role.
Social and care home workers seemed to value their formal, managerial supervision and this
was sometimes cited as a positive non-client interaction which provided valued social support.
Diarists across occupational groups also included different types of colleague interaction (e.g.
meetings and lunches) as positive events.
4.2.6
Positive non-work events
The diaries also contained many instances of positive non-work (or home/family centred)
events. Often these were not described in great detail, as exemplified by the most popular of
these events, which simply centred on pleasure and enjoyment in being at home with a partner
or family. Above all, home life was seen as a buffer, or antidote, to the stresses and strains of a
difficult job.
4.2.7
Concluding summary
• More positive or rewarding events were reported in the diaries than negative, or
potentially harmful, events.
• Most of the diarists enjoyed spending time with their clients and this was a very
rewarding aspect of their role.
• Another frequently mentioned type of event was a meeting or interaction with a client
that went well and had a successful outcome. These types of event sometimes involved
dealing with a client’s negative emotions and, because the diarist put her skills to good
use to turn the situation around, they were rewarding.
• Positive events involving client contact outweighed other types of positive work event.
Broadly speaking, this suggests that client contact is the most rewarding aspect of
human service work.
• Home or family life was usually perceived as buffering the negative effects of work on
the individual. That is to say, it was an antidote to the stresses and strains of the job.
4.3
EVENTS ENGENDERING NEGATIVE EMOTIONS
There were fewer negative events in the diaries than positive events, but the negative events
were reported in insightful detail. This chapter has two purposes: First, to help to elucidate the
causes of harm in emotionally demanding occupations by examining events at the day to day
level; and secondly, to answer the research question, ‘What kinds of events and situations at
work are found to be particularly emotionally demanding and in what ways?’ To fulfil these
both aims, this section describes the negative events described in the diaries.
4.3.1
An interaction with a difficult client
The most frequently cited type of negative event involved an interaction with a client who was
difficult to deal with. However, it was not usually the difficult behaviour of the client in itself
57
that presented a problem. Rather, these events were experienced as emotionally demanding for a
handful of more subtle reasons, which included: The diarist feeling that her actions in her
professional role were being questioned; feeling that she was not being listened to in her
professional role; a client making a complaint (or threatening to make a complaint) against the
diarist, often to someone at a higher level; and a diarist feeling relatively powerless to help a
difficult client. Four relevant examples are given below.
Example 1: A head teacher
This head teacher was dealing with a long running dispute between a small group of boys in one
class. On Day One, she wrote:
“Two letters from parents who feel I am ineffective when dealing with a child’s
behaviour in their children’s class. All the evidence points to it being 50-50 between
the one child and the others but these parents will not accept this and want to see the
governors and me about my inability to do anything of use (in their opinion. [ ]1 I got
both letters in the school office then got angry which made me cry in anger and
frustration that they don’t seem to understand the whole picture or their child’s role in
it. [ ] I then called the chair of governors who was very reassuring and made me feel a
lot better. But I found that the letters and my feelings about them coloured my day. I
was short tempered and was cross with a child for not a very good reason which I then
immediately felt sorry about. [ ] The negative feelings totally colour everything and
even though I’d had that lovely exchange with the first parent, it is the other issues that
stay with me through my working day.”
In this example, the head teacher’s actions were being questioned, in writing, by two sets of
parents. She wrote candidly about how this affected her feelings and actions during the course
of the day (she also circled ‘emotionally drained’ on the emotion checklist in this day’s diary).
The criticism felt personal, and this was a feature of some of the emotionally demanding events
described by the ‘distressed group’ (see 4.5.2). Her anxiety about the situation persisted over
several days and left her feeling a measure of self-doubt. For example, three days later, whilst
attending a course, she wrote:
“I’m worrying about these children whose parents have written complaints. It feels like
a courtroom drama file where you know which character should win but you don’t find
out until the last scene whether they do or not. One of the parents said that this is going
to get big. I’m scared of how big they want to make it and I’m starting to doubt myself
and my actions. Then I feel angry that their actions make me do that!”
Here, there was also an anxiety that the situation might persist and escalate (although in point of
fact, it never did). This fear of escalation was common to several, similar events recounted by
other diarists. For example, a housing worker wrote of a difficult tenant, who had placed a
corporate complaint, “Expecting some trouble”.
Example 2: A care worker
One of the management team in a care home wrote about a difficult interaction with a child who
was about to leave their care:
“At child’s review the child was very negative about the placement. His behaviour was
appalling and I felt both disappointed and embarrassed. Although the professionals did
not believe the allegations he was making, it did not make me feel good to hear him say
1
Square brackets within quotes indicate that portions of speech have been omitted
58
them. On return decisions I had made were questioned by the team. When they heard
my reasoning they agreed with me, but I felt it would have been more appropriate to let
me explain the discussions that were held rather than jump straight in. I know I am
capable of making the right decisions but constant questioning chips away at
confidence.”
In this example, the child’s difficult behaviour had a direct negative effect on the care worker (it
disappointed and embarrassed her). Her actions were also questioned, and this had the perhaps
more personal effect of undermining her confidence.
Example 3: A housing worker
A diarist who dealt with anti-social behaviour in the community gave another interesting
account of an interaction with a difficult client:
“Call from a client living near to a drug dealing anti-social family with a disabled
child at risk. Action is being taken to remedy this problem but much of the information
is confidential. I therefore cannot divulge what precise action is being taken but, the
client would not believe anything was being done, I was a liar, incompetent, and
anything else you care to throw in. The gentleman was belligerent and totally
unconvinced. Therefore arranged to call tomorrow with a colleague as a witness as to
what is being said as this gentleman is well known for altering the meaning of a
conversation and then making a complaint. Bingo…full house today…another client
calling about the same job, in the same manner. Must have been put up to it by client
No 1.”
This diarist regularly endured serious verbal abuse as he ran the gauntlet past trouble-making
clients on his ‘beat’. Although this event included verbal abuse, the most frustrating part of this
particular event seemed to be the client’s failure to listen. However, in common with most of his
other diary entries, the diarist’s tone of voice was dispassionate and cynical (particularly
compared to the tone of the head teacher and care worker in the previous examples). Indeed, the
diarist circled both ‘cynical’ and ‘emotionally-drained’ in that day’s diary. The diarists’
differing tones of voice seemed significant in the context of burnout, and this is discussed later.
Example 4: A police doctor
Lastly, a police doctor wrote:
“2 clients so far in the police station. I have to admit that I rarely feel positively
following these interactions. They’re mostly helpless, frustrated and angry and very
often guilty of the crime for which they’ve been arrested and it’s very difficult to have
any sort of positive impact on them.”
In some ways this type of interaction was the opposite of the events themed around enjoyable
and positive client contact (see previous section), in which the diarist often felt an easy empathy
with a client and also a pleasure in being able to help him or her. Rather, here, the diarist felt
relatively unsympathetic and helpless in the face of the client’s negativity. There were only a
few of these type of events, but they were definitely experienced as stressful, with diarists
circling the more extreme emotions (burned out, helpless, emotionally drained).
4.3.2
A client who is going through a negative experience
The second most commonly reported group of negative client interactions centred on clients
who were going through a negative experience. In most of these instances, the diarists felt
59
emotions like worry and sadness because they were concerned about the client’s welfare and/or
they felt empathy with the client’s predicament.
A simple example of this was given by a social worker, who wrote: “Client talked about losing
her husband this year, made me think about the loss of my mother and sister”. It is interesting
to note that the social workers, in particular, were often dealing with the sort of commonplace
problems that they themselves faced in their own lives (e.g. bereavement or the declining health
of a parent). This quote suggested that the client’s problem had a personal resonance for the
diarist, which in turn prompted an emotional reaction.
Another, hospital-based social worker wrote:
“Have a client with dementia who has been living with his son who was sole carer. He
had a lot of bruises and sores on admission which could be attributed to falls but ward
have raised the possibility that he may have been abused by son and is thereby a
vulnerable adult. Due to client's level of confusion it would be impossible to obtain any
information from him. I have met with son and we are looking at a placement as son
feels that due to the deterioration in his Dad's physical condition he would not be able
to cope with him back home. As the client is being removed from the possible risk I feel
it would be opening a can of worms unnecessarily to instigate vulnerable adult
proceedings but feel that I need to take advice as it’s not my place to make this
decision.”
When interviewed about this event later, the diarist explained her sympathy for the son, whom
she felt had lacked social services’ support in caring for his father over a long period of time.
However, she was also worried on a more personal level about making a morally correct and
politically defensible decision over the vulnerable adult proceedings. This obligation to act
correctly was similar to the fear of ‘failing’ a client (see below). It was discussed by several
diarists, at interview, as being a burdensome part of the caring role.
The worker who helped drug addicts in the community wrote (of one client):
“My home visit was very depressing. As much as the client it was his home
environment - not just the poverty and its shabbiness but also the grubbiness and
untidiness. I am not particularly tidy or house-proud myself so find it difficult to
understand why it affects me this way. I think it is because I have better, more
interesting things to do than housework - whereas my client lacks many skills and is
depressed or out of it a lot of the time and it is the grubbiness and lack of anything in
his life that his surroundings represent. He was brought up in a series of children's
homes and has had very little love in his life and it is that emptiness that gets to me. I
am ashamed to say that I actually spent less time there than I should have done
because of my distaste. Yet many of my clients have similar homes and it doesn't
always have such a dramatic effect. I think it is the emptiness of his life that makes the
difference.”
In this example, the diarist’s personal circumstances had a “dramatic effect” on the diarist. The
diary entry had a personal tone, and there was a sense that she had to work hard to overcome her
personal distaste, and she felt ashamed of this.
A final example of compassion for a client’s predicament was given by a head teacher:
“Y3 child returned from exclusion, feels rather hopeless, mother after school seemed
desperate and incapable. Feel frustrated that work of social services appears to be
60
having no impact on improving home or parent child relationship. [ ] A year has gone
by and no positive change, child is more problematic, less motivated than ever. Can't
raise any support from Pupil Referral Unit as they are fully stretched. Need to look at
how I can stretch school resources to put in more support otherwise it will be one
exclusion after another. [ ] Frustrated by social services, the play is so cautious and I
see child's time going by, they only get one chance. If teacher gives a poor lesson it is
manageable if a one off, many poor lessons means life chances are lost. Get frustrated
when I see life chances drifting by and little to counter it.”
In this event, the head teacher clearly felt desperately sorry for the child and his/her mother.
However, perhaps more significantly, she was very concerned that the child would not get the
help and support he or she needed, because resources were stretched and social services did not
seem to be sufficiently proactive. Indeed, this concern that the system might ‘fail’ the client was
another theme that often featured alongside empathy or sympathy with a client’s predicament.
For example, the diarist who dealt with anti-social behaviour wrote: “Informing clients that
their case would be closed through lack of evidence available for a prosecution, knowing that it
is going on but do not have the resource in place to gather the evidence”. One social worker in
the sample felt particularly strongly about this point, writing several times about her anger over
assessments (of clients’ needs), which were rejected by a cost-conscious funding panel. At
interview, she said that she had often commented that the funding panel would not reject social
workers’ assessments so readily if it was they (rather than she and her colleagues) that had
responsibility for relaying the bad news to the client.
4.3.3
Verbal abuse
There were twelve instances of verbal abuse in the diaries. However, interestingly, the tone of
these diary entries tended to be rather resigned, accepting and sanguine. For example, a care
worker, who was eloquent about her emotions elsewhere in the diaries and at interview, wrote
sparingly about being called a “tight c**t” by her designated key-child, who, elsewhere in the
diaries, had given her pleasure by hugging her and by jokingly calling her “mum”. She seemed
more upset by a mild instance of what she deemed to be racist language, used by a co-worker,
than by verbal abuse by the children in her care. This may have been because the children’s ages
and their problems rendered their behaviour excusable, and also because the children’s
behaviour was a source of satisfaction as well as stress.
Likewise, the housing worker who dealt with anti-social behaviour wrote:
“Called to a complaint of neighbour nuisance that has been going on for a long time.
Have injuncted the neighbour in the past. Had to walk past this person to get to my
destination and was subjected to a tirade of verbal abuse. I also know that there will
be another complaint against me as every time he sees me he makes an official
complaint to the council that I am spying on him. C'est la vie!!!”
In contrast to the previous example, here it seemed to be the diarist’s air of ‘having seen it all
before’ and his prevailing mood of cynicism (see 4.5.2.3) which lessened the effects of the
verbal abuse.
Also, in similar situations, it was often not the verbal abuse in itself that bothered the diarists.
Rather, it was another feature of the interaction (such as not being listened to in one’s
professional role, or feeling unable to turn the situation around to bring about a positive
outcome), which rendered it emotionally demanding. This was similar to the reasoning
underlying the diarist’s feelings about difficult client interactions (see above).
61
For example, a member of the management team in a care home wrote of how a child was
“angry” and “hostile” and how most of this anger and hostility was directed at her. She wrote,
“a lot of property was damaged and I was threatened a lot”. Although she circled ‘scared’ on
the emotion checklist in her diary that day, the diary entry focused on her feelings of frustration
at being unable to remedy the situation by pacifying the child. She also noted feeling frustrated
at the actions of other people who were “buying into” the child’s arguments (unhelpfully, in her
view), and also isolated by her colleagues’ failure to step in and challenge these threats.
A head teacher wrote about a negative interaction with a difficult and abusive parent:
“This morning's interview with Ms Y. She is known to be a VERY difficult and
dangerous parent. She arrived with her father in tow. She did stay calm and didn't
shout at me. She wasn't happy that her son was being reprimanded for his recent poor
behaviour i.e. kicking a ball inside the building at a child on crutches, who was
knocked over. She also maintains that the Deputy Head Teacher speaks to her uncivilly
- this is a mother who screams down the phone on occasions and whose father is
shouting abuse in the background. She didn't want the DHT who is also the SENCO
(Special Education Needs Co-ordinator) to have anything to do with her son at all.
When I stated that this just wasn't possible she began to get irate. She told me she
wanted to make a complaint. When I explained that this was done through the Chair of
Governors she became crosser. She told me that I had no right to ever be away from
school, my job was to be there dealing with the children. I explained why this wasn't
going to happen and suggested to her that she had lost faith in the school and asked if
she'd like some help finding another school for her son (as she allegedly can't read or
write). She stormed out, shouting abuse. All this before 9 o'clock.”
In this diary entry, the tone was pragmatic and a little ironic. The verbal abuse did not seem to
upset the head teacher, perhaps because she did not take it personally, given that the mother was
known to be a difficult client who had created problems on other occasions and in other schools.
However, the next day the head teacher wrote more emotionally, writing about her sadness that
the child was being moved and her concern, “that the child won't be so well supported at any of
the other local schools.” Therefore in this case, as per the previous examples, the verbal abuse
was less emotionally difficult than other aspects of the situation.
Finally, one diarist/interviewee talked about how he found verbal abuse more emotionally
demanding than usual at times when he was feeling tired. He wrote, “Tenant complaining that
work not carried out on time wanted it doing sooner or would go to his Councillor. I said we
would not do the work any sooner due to resource shortages. Tenant verbally abused me which
agitated me and put some stress on me especially as it was my first day back and because I felt
tired. This is normal occurrence each day but when tired it seems to bother me more.” This
concurs with the observation that diarists/interviewees found emotion work to be more effortful
at times when they felt tired or less mentally robust than usual (see 4.4.2.9).
4.3.4
Paperwork or “Admin”
Turning to the negative events, which did not involve a client interaction, paperwork or admin
emerged as a significant stressor. Indeed, in terms quantity, “admin/paperwork” was the most
important theme of the diaries, being mentioned 47 times. However, although most of the
sample wrote about admin, it was the head teachers who experienced it as a real stressor of the
job.
62
For example, a social worker wrote about her frustration about not completing an application for
Court of Protection, which had been on her agenda of things to do for several weeks. However,
she saw this as a problem of her own making, writing, “This is more to do with time
management than anything else. [ ] I know it is my own fault for not managing my time”. Other
types of worker wrote similarly about paperwork, seeing it as something which they
procrastinated over and disliked. Importantly, however, they did not seem to feel over-burdened
with it.
By contrast, the head teachers in this sample conjured up the modern stereotype of teaching as a
profession that is increasingly stifled by red-tape. They made frequent reference to needing to
author a diverse range of reports and plans, in accordance with official requirement and
deadlines. For the majority, this metaphorical mountain of paperwork came across as an
unrewarding, burdensome and frequently overwhelming feature of the job. For example, one
head teacher wrote of “throwing away” her briefcase “years ago” and substituting it with “a
supermarket crate”, in order to accommodate all the paperwork which she needed to take home.
Another wrote of how a particular report was “hanging over” her like the “Sword of
Damacles”. A third noted a “fear” of the paperwork she was going to need to complete in
response to meetings that day. And a fourth head teacher wrote of his annoyance at waking up
early on a Sunday morning in order to work on the School Development Plan, and his worry
about needing to “get (the) plan done soon”. In an otherwise fairly uneventful diary set, this
diarist also circled ‘burned out’ and ‘emotionally drained’ in the emotion checklist on a day in
which the two negative events consisted of, “A teacher who was impatient with a parent” and
“Amount of post and e-mails from local authority”.
A caveat to this observation about the importance of admin was that the diaries were completed
at the very end of the school year (July), when deadlines were perhaps more pressing than usual.
Nevertheless, the way in which paperwork featured so ubiquitously and heavily in the head
teachers’ diaries, and the way in which it was referred to as a pretty constant demand, suggested
that this was perceived as an ongoing stressor of the job. Moreover, paperwork also seemed to
be seen as disproportionately time-consuming and stress-inducing compared to other, more
important aspects of the head teacher’s role. One diarist made the following comparison:
“Spoke to child today who said he hated himself and that he wanted to commit suicide
- he is only eight. Seriously worried and have spoken to parent and teacher. Arranged
to meet them tomorrow. This is the important stuff - not the piles of paper on my desk.
But the paper and files make me feel more and more stressed as I hate not being in
control of my work space.”
In this quote, the child’s crisis was experienced as emotionally demanding, but this was
acceptable and perhaps even rewarding to deal with because it was important and meaningful in
the context of her key responsibility (i.e. safeguarding the well-being of the children in her
care). By contrast, the “piles of paper” were not emotionally demanding, but their very lack of
importance and meaning seemed to render them arguably more stressful than the more
emotionally taxing event which they were contrasted with. Paperwork therefore came across as
background stressor which was not emotionally demanding but which did have emotional
consequences, not least because of its questionable value. When this interpretation was checked
with this diarist (at interview) she agreed that it, “hit the nail on the head”.
4.3.5
Problems with the working environment
Problems with the environment (including computers and IT in general) were another
background irritant in these diaries. For some of the head teachers, incidents to do with the
63
school premises were difficult because first, they felt inexpert in dealing with
property/accommodation issues and/or, second, they had ultimate responsibility for the health
and safety of many children. On the first point, one diarist talked at interview about her feelings
of stress and frustration that were associated with having to assume responsibility for diverse
areas in which she had little expertise. She wrote:
“Trying to sort out the Fire Log books etc. We have a notice from the Fire Service
(with many other schools) and I am finding the expectations very stressful as I do not
feel confident.”
On the second point, another diarist wrote several times about his discomfort with being
ultimately accountable for the children’s welfare and safety. To illustrate this at interview, he
explained that his first question to a child who had fallen over in the playground that day was
not about how she was feeling, but rather, “Was it a plant or the fence?”, because if the fall had
been caused by the fence it might have been a health and safety issue. His first anxiety was
therefore about whether the fall was preventable and ultimately his ‘fault’.
The other occupational groups in the study were, by and large, not responsible for their working
environments. Their environment-related problems centred on technology and were seemingly
more straightforward than those of the head teachers. For one hospital based home tutor (a local
authority employed home tutor who deals with children who are temporarily out of the
education system because they’re ill or excluded from mainstream school), the inefficiency and
decrepitude of the computers and photocopiers in her department was treated as a sort of joke
that ran through the diaries. She wrote, “Trying to leave at end of afternoon – needed to
photocopy next week’s timetable – guess what?! Couldn’t find a photocopier working …”.
Despite their resigned and amused tone, these diaries conveyed an undercurrent of
dissatisfaction about having to cope with this particular inconvenience in addition to all the
important, client-based stresses of the job. In this way, the tutor’s problem was reminiscent of
the head teachers’ feelings about their admin.
4.3.6
A colleague or subordinate’s handling of a situation with a client
For those diarists in managerial positions within institutional environments, negative emotions
sometimes arose from either observing, or having to deal with, another staff member’s poor
practice. Such occasions were difficult for two reasons: Firstly, because the staff member’s
actions were often deemed to be have an adverse effect on the client, so that there was an
overlapping concern for a client’s welfare; and secondly, because the diarists often found
staffing issues personally delicate and difficult to deal with.
One head teacher wrote about an event that conveyed both of these points effectively:
“Early morning interaction with teacher who arrived late and wanted to leave to go to
the dentist. Her practice is poor and I do not feel that she meets the needs of children,
especially those with special needs. Her act this morning demonstrated a lack of care
for colleagues and children that I felt was unacceptable. I have undertaken several
observations recently where it has been important to feedback negatively as children
are not getting a good deal. She has had a good deal of professional support and I now
need to move towards a professional procedure.
On a personal level this is unpleasant and having had to do it several times over the
years I know it to be stressful for the client and for me, it also can impact other
colleagues. Keeping the needs of the children in mind makes it tolerable to undertake
64
because I cannot let them continue to have poor emotional, social and educational
experiences.”
Another event, written about by the same teacher on a different day, effectively demonstrated
the point about how concern for a client’s welfare was embedded in issues to do with poor staff
practice:
"I am concerned about Liam's well being. [ ] She is coping with a class with many SEN pupils ,
many of whom are needy and several of whom present very challenging behaviour. She
was fed up with Liam today, he had been off task, scribbling graffiti on his work. Her
practice is limited, too often she does not meet the learning needs and they deserve
better. I have started a performance related procedure with her, she has additional
support in the class. Liam had not been listening in class, and when I spoke to him he
had no recollection of the lesson so when it came to writing about it he was at a loss.
Liam’s dad died three months ago, he has become more troubled, he was before, he
appears to have lost weight and is disinterested in most things. [ ] I understand that
the teacher is frustrated with him, he can be disruptive in class, and his recent
bereavement has meant that he is dealing with even more complexities than previously.
School is very important to him.”
The head teacher also wrote about how she, “Spoke with her (the teacher) about this, trying to
get her on board to empathise with need”. Thus she highlighted how a perceived lack of
empathy, or the member of staff’s failure to care enough, was another potentially emotionallychallenging feature of these sorts of situations. In a similar vein, another head teacher wrote
about a school trip which one disadvantaged little boy was likely to miss because of his parents’
behaviour. She wrote:
“Not much effort had been made on behalf of the child [ ]. None of the staff seemed to
care if the child went on the trip. They seemed more concerned about the rudeness of
the parent. This father is inappropriate but not directly rude, for instance he calls me
‘love’. I would imagine he says the same to other staff but they won’t make
allowances”.
The same head teacher had an ongoing issue with staff performance. She wrote of how this
teacher did not know what a CRB (Criminal Records Bureau) check was (“give me strength []
!!”) and how, on another occasion, she failed to complete “one of the very unchallenging tasks I
had set her”. However, she was also candid about feeling “rather disinclined to deal with”
someone who had been working at the school for many years, and had been in a position of
some responsibility since before the diarist had joined as the new head. Her exasperated sign
off was “Help!”. This issue was therefore so sensitive and, by implication, so stressful, that she
was procrastinating about tackling it.
The last example was given by a member of the management team in a care home. In one diary
entry she had reviewed the ‘incident recordings’ from the previous night and wrote:
“One of the children had a difficult night and I was annoyed that if the staff on duty
had followed the management plans the situation could have been dealt with better. I
feel why do I bother! I feel as though I’m constantly repeating myself.”
She also wrote about observing a similar event:
“He (a child) was being exceptionally rude and I was left feeling frustrated and irritated
with the staff member who was dealing with him. Despite numerous conversations about
65
how to work with this child, the staff member was giving the child too many choices and
chances which escalated the behaviour.”
Like the other examples, the latter two events centred on a subordinate’s poor performance, and
included, by implication, concern about the effect on the clients. Interestingly, however, the
diarist’s emotional response (i.e. her frustration and irritation) was rooted in her feeling that she
was not being listened to in her professional role. This exemplifies how this theme of not being
listened to, either by client’s or co-workers, cut across different types of events to emerge as one
of the key reasons why, and how, particular events at work led to negative feelings.
4.3.7
Difficult interactions with colleagues or third parties
The final theme of the negative work events, which will be discussed here, revolved around
difficult interactions either with colleagues or third parties (outside the organization). The
majority of the diarists mentioned one or two instances of this type of event and, as discussed
later (see 4.5.2.4), these events seemed to be quite significant in the context of burnout.
In some of the more prosaic examples, the causal factor underpinning the difficulty seemed to
be a tension over roles and responsibilities. For example, a care manager wrote about how she
felt it increasingly incumbent on her to take over tasks that were not in her “job description”,
because the social workers she worked with seemed to be “incredibly stressed and put upon”
and felt the need to “offload this anxiety onto me”. She gave an example of this, illustrating
how she ended up taking on “additional tasks and roles so that the clients do not suffer”.
Quite often, the “difficult interaction” was a meeting involving a client and the representatives
of two or more agencies/organizations. In these cases, the diarist usually felt that another party
mishandled the meeting or acted insensitively towards the client. The diarist was left feeling
frustrated. Two examples of this are given below (the first from a head teacher, the second from
the substance misuse worker):
Example 1 (a head teacher): “Really upset with a SENCO from another school who
was horrid, insensitive and lacked respect to one of our children who is transferring to
her school in September. He has been really happy and settled in our school.”
Example 2 (the substance misuse worker, referring to “probably the most difficult
meeting I have ever been to”, in which her client was a young schizophrenic): “The
meeting with the client made me feel very frustrated and confused. It felt the mental
health services were treating some-one who was clearly ill as a rational person. Also
they asked him repeatedly what he wanted but it seemed to me we only explained why
he couldn't have it. Also the attitude to the illegal drugs problem seems to be more
punitive and judgemental than practical. My heart went out to the client who is having
people running his life. I am also cross and frustrated because resources are so split
into boxes and each box excludes people through their own rules so people like this
boy fall through the net. I ended up eating all the sweets I shouldn't have.”
A final example was a one-to-one interaction between a head teacher and another party:
“The centre manager of a referral unit behaved in an unprofessional manner both in
her actions (she didn’t inform the school about a situation) and also her reaction to
me. She did not acknowledge wrong-doing and reached aggressively towards me. I did
rise above this and the tone of the language/content of the language was calm but I felt
very very angry and could have cheerfully decked her!”
66
Despite the upbeat tone of this latter event, it was clear that the head teacher had to curb her
anger during this interaction. It is a good example of how ‘emotion work’ featured
spontaneously in some of the diary entries, as discussed later (see 4.4.2.6).
4.3.8
Role of negative non-work events
There were a fairly large number of negative non-work or home events in the diaries and they
included themes like a worry about a child or a worry about one’s health or that of a family
member. Such events seemed to be particularly significant to the ‘distressed’ group (see
4.5.2.3).
There were some events in which work could be pinpointed as having a negative effect on the
diarists’ personal lives. For example, one head teacher felt very strongly and very negatively
about the amount of out of hours work she was expected to do. A social worker wrote about
how she tried to either ring or see her sister, who had cancer, every day, but had forgotten to do
so that day because she was busy with work. A care manager wrote of how she was not “getting
on well” with her husband at that time because both she and he were working long hours at that
time. One diarist spoke, at interview, about negative health behaviours (like drinking a little too
much), which she attributed to the stresses and strains of the job. In sum, just as non-work (or
home life) was believed to buffer or lessen the stress of work, work was thought to exert a
negative effect at home in an inverse way. However, this should not be overstated: Looking
across the diaries, work was not usually portrayed as particularly harmful to home life. Rather,
for most people, home life gave them the strength, emotional security and necessary respite in
order to cope with difficult and demanding jobs.
4.3.9
Concluding Summary
• The diaries, supplemented by the interviews, revealed a number of types of event which
engendered negative emotions.
• Thinking first about client interactions, such types of event included: Interactions with
difficult clients; dealing with clients who were going through a negative experience; and
being subject to verbal abuse from clients. Thinking secondly about other types of
events, the key themes were: Paperwork (for the head teachers); problems with the
working environment; a colleague or subordinate’s poor handling of a situation with a
client; and difficult interactions with colleagues or third parties.
• All of these different types of events could be seen as potential ‘causes of harm’ in
human service work. However, a careful analysis of these situations also revealed that
certain over-arching themes were common to different types of events. These recurrent
themes often revealed important reasons why particular situations were experienced as
difficult or emotionally demanding. They included: Not feeling listened to in your
professional role; having your actions questions in your professional role; feeling
unable to turn a situation around in order to bring about a positive outcome; and
concerns about ‘failing’ a client.
• From these data, a distinction might be drawn between events which might be ‘harmful’
simply because they generated negative emotions for the worker, and events which
might be harmful because they were emotionally demanding. Paperwork and problems
with technology, for example, were not emotionally demanding. However, paperwork
was seen as stressful, partly because it lacked the meaning, value and importance of
67
other challenges of the job. Difficult interactions with clients and co-workers were
emotionally demanding, by contrast, and were experienced as difficult for the types of
reasons listed above (e.g. not being listened to in your professional role).
• Furthermore, it is also important to note that emotionally demanding situations were not
synonymous with stress and harm. This is because dealing with a client’s negative
emotions sometimes felt rewarding (see previous section), or sometimes felt
simultaneously rewarding and emotionally-draining (see next section).
• Finally, the role of home life seemed to be significant in the context of burnout (see
next section). Work also exerted a negative effect on home life, but the importance of
this should not be overstated: On balance, home life’s positive impact on work (in terms
of inoculating people against the stresses and strains of the job) was greater than work’s
negative impact on home life.
4.4
EMOTIONAL DEMANDS
This chapter examines the emotional demands of human service work, drawing mostly on the
interview data. First, it discusses emotionally demanding situations, which were general
descriptions of work situations which workers found personally difficult to deal with, and which
they associated with stress and burnout-type emotions. Secondly, it examines how people
performed emotion work within particular events, and how they experienced different types of
emotion work as emotionally costly and/or rewarding.
4.4.1
Emotionally demanding situations
Within the interviews, research participants commonly talked about what may be termed
‘emotionally demanding situations’. These tended to be seen as habitual stressors of the job,
and were described in general terms as difficult situations that workers were frequently
confronted with in their particular job role. They are discussed here because they had emotional
consequences for the worker and were explicitly or implicitly associated with stress and
burnout-type outcomes.
Emergencies, which ‘hijacked’ the working day
Head teachers, in particular, spoke of difficult situations to do with the children that required
immediate attention and urgent action. Relevant examples were sudden child protection issues
and serious incidents of unacceptable behaviour (e.g. those resulting in a child’s exclusion from
school). One head teacher labelled these “drop everything” situations, saying, “I hate it when
that happens”.
One of the most difficult aspects of these events seemed to be their complexity: Decisions had
to be made quickly; third parties had to co-operate; staff, parents and children required
emotional support; and the burden of responsibility for the child’s safety and/or a good outcome
lay with the head teacher. Moreover, these events happened unexpectedly, and were overlaid
onto all the normal demands of the working day. For example, one head teacher attributed
feeling ‘burned out’ on one diary day (a Friday) to the demands of a “very busy week” which
involved dealing with two emotionally-charged pupil situations (a serious bullying incident and
a child who felt suicidal), in addition to a busy week of meetings that, “I couldn’t actually
avoid”. Of the two situations with the children, she said, “I think they do take it out of you” and
“it is emotional and you do worry about it, not just in school, and think about what to do”.
68
However, it also worth noting that head teachers gave the impression that they felt and acted
very calmly and competently in these situations. For example, in one interview the researcher
commented on how pragmatically a diarist had reported a particular incident in which a child
had absconded. The interviewee replied that although she thought it should have felt alarming
(because they were just about to call out the dog unit):
“I didn’t see it as dramatic [ ] I think that whilst all these events are damaging and
I’m on an emotional roller coaster and for the most part I’m acting, there are certain
situations and when I’m in those situations I’m very focused and calm and I always
think things are going to be fine.”
It is therefore likely that these types of event were self-validating as well as stressful for the
head teachers.
Responsibility for the client’s welfare
Allied to the above situations, the interviewees found their responsibility for their client’s
welfare burdensome at times. For example, a head teacher wrote of a child protection issue in
which he had difficulty generating a response from social services. At interview he summed up
his feelings of frustration with the words, “I’m saying these children could die tonight and
they’re saying ‘we haven’t received your fax!’” He also talked about his great relief when the
situation was finally dealt with and the children were safe. A second head teacher talked about
how she had uncharacteristically lost her temper with the children when the fire alarm went off
unexpectedly one day. She said:
“I mean, the adrenalin kick when the fire alarm goes off is off the wall. You know, the
very minute it kicks in if you’re not expecting it, ‘Is there a fire? What am I responsible
for? Can I ensure the safety of everyone on this site?”’
Although the head teachers seemed to feel these responsibilities most acutely, other types of
workers also referred to the strain of responsibility. For example, a child-care worker spoke of
taking the children on a day trip: “And then again, God forbid something happens and the
paperwork isn’t filled in and that’s it!”
Questioning one’s own actions in a situation
Many of the human service workers in this sample were encouraged to be highly self-aware and
reflective about their work. This process sometimes generated an anxiety about the correctness
of their actions in particular situations. For example, a social worker spoke about placing a lady
with mild dementia in a residential care-home where most of the residents were significantly
more disturbed than she was. She said:
“Yeah, and it still really bothers me, actually, because I just felt, when we did the final
assessment [ ] she did not have anybody round her that would lift and encourage her.
They were all worse than her and I felt unhappy about that.”
Another social worker used an interesting map metaphor to describe her thoughts about the
rightness or wrongness of how she handled cases. For example, she talked about “going in the
right direction” with cases, and about “very bad days when I feel lost, very lost”. She talked in
particular about an ongoing case that she found “difficult”, because, “I’m not sure where I’m
going with them”.
Interestingly, in both of these examples, the social workers felt that their dispassionate decisionmaking was compromised by the opinions of the clients’ adult children. Social workers
69
sometimes found it difficult to reconcile a client’s needs with their families’ wishes, and this
was sometimes a significant factor in their more troubling cases.
Having to be an expert in many different areas
Most of the head teachers felt that they were required to be an expert in many different areas
beyond teaching, including finance, premises, grounds, personnel, health and safety. This was
allied to the large amount of paperwork in all of these different areas (see 4.3.4). One head
teacher described it as “being the engine that drives everything”. Another referred to “dreading
Monday” because she was required to attend eight meetings on the types of areas listed above.
She said:
“It’s just exhausting and draining. It’s always about something different and that
person always expects you to know what you are talking about as well.”
This requirement was experienced as emotionally demanding because the head teachers did not
feel competent in most of these areas and they did not find it rewarding. As such, it contrasted
sharply with teaching and the other aspects of the job, which played to their strength in the area
of inter-personal skills.
Staffing issues
Many of the research participants in this sample had managerial responsibilities and one of the
most difficult aspects of their job was dealing with problematic staffing issues. The head
teachers often spoke having to overcome their own inner reluctance to initiate disciplinary and
performance-related discussions with teachers. For example, one head teacher spoke of, “times
when you think I’ve got to have that conversation and I really don’t want to”.
Another head teacher had had to make a number of redundancies when she took over the post.
At interview, she linked the demands of this period with a number of stress outcomes (R =
researcher, P = participant):
P: “I mean, I was trying to deal with my feelings, deal with the process and deal with
their (the staff’s) feelings as well. It’s like everything coming onto me. I mean, I didn’t
really know what I was doing. I was having to ask for help all the time cos it’s the first
time, I mean I’ve never done anything like redundancies before. So I was constantly
having to be guided by somebody else and a lot of people became very accusatory,
which I mean they are bound to it in that kind of scenario. It did eventually get
sorted.”
R: “Accusatory about each other?”
P: “Yeah, and also me and what my plans were and ‘you’ve decided that’, ‘you’ve
done this’ and ‘you’ve already decided that I’m going’”.
R: “That must have been really difficult.”
P: “Yes, it was incredible. The stress was unbelievable. I mean came out in rashes all
over my legs, couldn’t relax, I didn’t sleep for weeks, didn’t eat for weeks. It was a
horrible, horrible time.”
The emotional component of this situation was also interesting. For example, she spoke about
having to deal with her own feelings about the situation, as well as those of her staff. She also
spontaneously talked about the “act” she put on at this time, saying, “I came out with rashes
70
and everything but in front of the children and the rest of the staff you wouldn’t have known it.”
Thus these were complex situations, which necessitated ‘emotion management’ on a number of
different levels.
Having described some of the ongoing and/or recurrent situations which these human service
workers found stressful, the second half of the chapter now focuses on the emotion work that
the diarists/interviewees performed in their work roles.
4.4.2
Emotion work in particular events
The interviews and diaries revealed much about emotion work in particular events. The phrase
‘emotion work’ is used here to describe the conscious or semi-conscious management of one’s
own and/or other’s emotions. It is a broader term than Hochschild’s (1983) ‘emotional labour’,
which is defined as emotion management in the service of the employing organization, and
which, as such, is quite specific to customer service roles.
Interestingly, many interviewees spoke about emotion work spontaneously, before the theme
was further probed by the researcher. Words like ‘perform’ and ‘act’ were also used
spontaneously, suggesting that this group were highly conscious of working on their feelings
and maintaining a particular persona in their work role. This supports the idea that emotion
work is a fundamental aspect of human service work.
More tentatively, the fact that people spontaneously discussed emotion work in the context of
talking about the negative aspects of the job suggests that it does contribute to ‘harm end
points’. However, this study also suggested that the association between emotion work and
stress is far from clear-cut, because emotion work was also experienced as rewarding. Section
4.2.2 discussed diary events in which emotion work generated positive feelings in themselves,
as well as their clients. The next section elaborates on this, explaining how people sometimes
found emotion work to be simultaneously difficult and rewarding. The remaining sections of the
chapter look at other aspects of emotion work, which were variously experienced as both more,
and less, emotionally demanding.
Emotionally demanding interactions which were both difficult and rewarding
Some client interactions in the diaries were both rewarding and difficult. The example below
was written by a home tutor, attached to a hospital:
“First lesson with a new student (off school through long term illness due to an
accident). Not sure if she’d engage at all as she’d been feeling tired/ill/depressed and
hadn’t engaged with the previous tutor. Managed to keep her involved for almost 2
hours. At times she was quite positive about the work. Her mother was surprised that
she’d responded positively.
BUT
The session described in 3a was very exhausting. I came home feeling worn out and
quite ratty. The student is angry about being ill and quite sure she won’t be able to do
the work I’m supposed to be doing with her (GSCE English Coursework). I had to
provide all the energy for the lesson to keep her motivated and convince her the task
was not insurmountable.”
This event serves to illustrate the dual nature of some emotionally taxing client situations, which
were neither wholly positive, nor wholly negative. The interaction was pleasing because it went
71
better than expected and the client made progress. However, it also incurred costs, in terms of
depleting the tutor of energy and goodwill. Talking about this at interview, she made sense of
this event as positive, “To be honest, I did actually come away feeling positive. I was exhausted,
but we’d got somewhere! We’d made a start, she’d done much more than I thought she would.”
One of the head teachers wrote in a similar way about an interaction with two staff members.
Unusually for her, she had circled ‘emotionally-drained’ and other negative emotions on that
day. She wrote:
“Good meeting with two teachers at 5pm. [ ] Was able to make some helpful suggestions
which they are going away to work on. Positive mostly because they seemed lifted at the
end of the meeting, they had begun expressing annoyance and dejection. In lifting them
of course able to lift myself after a difficult day.”
At interview this diarist explained how, in a job where you “don’t get any bouquets”, events
like these helped her to “create my own successes”. She therefore implied that it was helpful to
see events, retrospectively, in a positive light. This suggests, perhaps, that human service
workers looked for the positives in difficult and demanding situations, which is partly why,
perhaps, the rewards of the job seemed to outweigh the perceived costs.
Role modelling
In their jobs, many of the diarists/interviewees had to demonstrate best practice behaviour to
other staff, as well as their clients. For example, many of the head teachers believed that ‘an
ethos of calm’ was fundamental to good behaviour in schools. With this in mind, they rarely
raised their voices to pupils. One said, “I try to demonstrate that all the time to my staff”.
Role modelling was not usually seen as particularly difficult or stressful, although it did seem to
require effort. For example, a member of the management team in a care home said:
P: “Because of my role it can be a bit difficult because you need to present this image
to the staff team that we can manage and discuss things in a way that is positive,
rather sometimes you just want to go out there and say ‘I’m bloody pissed off’ but you
can’t … So, you need to be able to encourage them to talk about feelings and you need
to share feelings yourself so that they are able to do that, but at the same time because
of the role you need to keep that level of professionalism that maybe they wouldn’t
keep, if that makes sense.”
R: “Do you ever feel like you are acting?”
P: “Not acting, but at times it can feel (that you’re) constantly being watched and
observed….and that you have to be on your best behaviour performing. It’s role
modelling with staff and the children.”
R: “And is that draining, is that difficult?”
P: “Depends what you are doing at the time. If you’re feeling particularly stressed,
particularly tired, it can be.”
It is interesting that this interviewee did not equate role modelling with acting, perhaps because
it felt more spontaneous and authentic than acting, although she did then talk about
“performing”. Also, she made an observation about the conditions under which it became
difficult (i.e. when she felt stressed or tired), at which point, she implied, role modelling
72
required more effortful emotion work than usual. This theme is discussed again later, in the
context of doing ‘extra’ emotion work (see 4.4.2.9).
Surface acting
In the interviews in particular, references were made to emotion work which conformed to
Hochschild’s (1983) idea of surface acting. This is a form of relatively superficial emotion
work, in which the actor is conscious of seeming to be a certain way for the benefit of his or her
audience. For example, a head teacher talked about acting ‘cross’ when a misbehaving child was
sent to report to her. Asked if this was stressful, she replied, “No. That’s fine. That’s part of
getting that child through whatever it is they’ve done”. Another head teacher described a
tactical piece of “play acting”, which helped her to successfully “manage” a dispute between
two members of staff.
Surface acting was also performed tactically in meetings with difficult parents. One head
teacher spoke of a parent of a child with special needs who would not accept his diagnosis,
preferring to label his condition differently. She described review meetings with this parent
where, “She doesn’t listen [] I don’t know why we bother really, because nothing goes in. You
just sit there and you find you’re saying things for the sake of it, to record it and…”. The
teacher said that she felt very “frustrated” because of the superficiality of these meetings, but
also (asked if it was stressful) she replied, “the meeting itself’s not stressful, you’re just going
through the motions”. Interestingly, however, she did find the underlying situation difficult,
“just because you know you’re not going anywhere. And it’s stressful because poor old Daniel
is still sitting there suffering and we’re just coping.”
A housing worker also highlighted how surface acting became stressful under certain
conditions. Like the teachers, he did not see tactical pieces of surface acting as difficult or
demanding. Rather, he enjoyed having “a bit of banter” with his clients and prided himself on
his ability to “work the audience”, with a view to placating the clients and bringing them
around to his way of thinking. However, he also said:
“It doesn’t happen every time. I mean sometimes you have really stubborn people and
they’re never going to listen. And it becomes stressful when you know it’s going the
wrong way. You know, it can be from time to time you get tenants who will just not
listen to what you are going to say and they will say ‘right, I’m going to write to my
MP’ [ ] And you get frustrated but it’s just one of those things. It’s part of the job.”
Looking across these examples, it seemed that surface acting for short-lived, tactical purposes
was not found to be particularly difficult or emotionally demanding. However, a meeting could
sometimes become emotionally demanding when the audience was unresponsive (as per the
example immediately above), and an underlying situation, or the context in which the surface
acting took place, could be worrying and frustrating too (as per the previous example with the
head teacher). Moreover, surface acting was also sometimes a feature of situations in which a
strong negative emotion, like anger or dislike, was being suppressed. These situations of
emotional dissonance (see 4.4.2.6) were undoubtedly stressful.
Deep acting
By its very nature deep acting is less easy to access than surface acting, because the emotion
management is less conscious and the discrepancy between felt and displayed emotions is not
obvious. Perhaps unsurprisingly then, there were few instances of deep acting in either the
diaries or the interviews. Nevertheless, there was some evidence that deep acting took place,
and that it was important in facilitating empathy.
73
Several teachers, for example, spoke of how they viewed some difficult parents (e.g. those who
were very young and lacking in education) as almost children themselves. This clearly helped
them to generate feelings of empathy which, in turn, allowed them to deal more sympathetically
and sincerely with these individuals.
One social worker, in particular, seemed to be talking about deep acting in the context of how
she related to clients. This social worker lacked confidence, and she used deep acting to build
her confidence in client interactions. Talking about a slightly awkward exchange with a client’s
very well educated daughter, she said, “I noticed a change. More so in my voice. I talked
politely (said in a mock ‘posh’ voice). In a way, you don’t intentionally mimic. But you face up
to that situation”. This social worker also talked about how she tended to put herself “on the
other side of the fence” and tried to think, “Well what if it was my grandma or my mum or
whatever, what would I want to know?”. It seemed that this technique of placing herself,
mentally, in the client’s position, helped her to be empathetic in the same way that the teachers
felt empathetic towards unruly, adult ‘children’.
Not having voice/lack of reciprocity
One type of emotion work that did emerge as unequivocally negative centred on ‘not having
voice’ in interactions. In these events, the client (or another party) conveyed their opinions and
feelings (often their negativity), but the human service worker was unable to do likewise. This
left the worker feeling stonewalled and unequal in the relationship.
For example, a social worker told the story of how a client’s son had pressed for his mother to
be taken into residential accommodation, and this was somewhat contrary to the social worker’s
assessment of the client’s wants and needs. For example, he said things like, “I am telling you
this is what she needs at the moment and I would basically blame you if you took a different
course of action”. Because of this, she found the case personally very troubling. However,
interestingly, she also told of how she engineered a situation in which her opinions were
indirectly aired, because at the final review she asked leading questions of the elderly client in
front of the son and, through these, the client voiced her unhappiness at being “dumped” in the
home and largely ignored by him. The social worker said, “I was there for two hours and he
wanted to leave. He wanted to leave because he was going on holiday and I wasn’t gonna let
him go.”
A tutor gave a second very good example of feeling stonewalled. In her diaries, she wrote about
having to tutor an extremely unresponsive Spanish girl. At interview, she talked about how this
was her most difficult client and how, “Certainly on the day when I teach her in afternoon, so I
come straight home after teaching her, I would be feeling a bit wound up at that point.” The
researcher then asked her whether in that situation she had to work on keeping up the
enthusiasm and energy to teach her. The tutor replied, “I was at that point, when I wrote the
diary, that’s how I was behaving. But the last couple of sessions I’ve had a family member in to
translate how I’m feeling about her.” Asked if this made a difference, she replied:
“No (laughing). But, has it made…it made me feel better. [ ] Um saying stuff like ‘you
know your family has moved to this country and you really must try and learn some
English so that you can make some friends and go out and have a good time.’ Then I
said that...er…‘we have a waiting list of children that want home tuition and it makes
me feel disappointed that she has lessons at home but she isn’t letting herself join in
and learn anything.’ That sort of thing really.” The tutor then explained how this was
a “relief” because, “she’s not giving any indication that she’s understanding what I’m
saying. Now I know she is understanding some of what I say but she’s not going to
74
respond, so at least if someone can say that to her in Spanish at least I know that she is
understanding it.”
Suppressed anger
Suppression of anger was the second type of emotion work that was universally portrayed as
difficult and potentially damaging. This was a form of emotional dissonance (in which there is a
tangible difference between the emotion being displayed and the emotion felt). However it
seemed to be more damaging than other forms of emotional dissonance (for example, the
difference between real anger and faked anger, as exhibited by the teacher in 4.4.2.3) because of
the fact that a florid, negative emotion was being suppressed. Anger needed to be suppressed
because it was seen as a particularly negative and damaging emotion, particularly if shown
towards vulnerable people, like children. For example, a head teacher said (of a member of her
staff), “I know things will go wrong very quickly if she continues to shout at the children, to use
language which is slightly aggressive, then it’s very easy to damage”.
In the diaries, the community substance misuse worker wrote about a young schizophrenic who
“has recently been introduced to hard drugs by his father”. She wrote:
“My session with the young man gave me a lot of negative feelings. Anger at the
stupidity of the father has been quite difficult to deal with as I cannot share this with
the client as he feels he has a good relationship with his father. I also feel bewildered
and helpless as I feel he sees positive benefit in taking drugs and that therefore his
heart is not in doing anything to stop it. I feel it is the waste of a life and that he will
be involved in drug taking for the rest of his life. The prognosis for drug-taking
schizophrenics is not good as many commit suicide.”
A head teacher spoke similarly about the damage another father was inflicting on his child. At
interview, she spoke about how she had to overcome her anger in an altercation with this parent
(here she used the word ‘acting’ spontaneously):
“I found myself feeling very resentful of him because I know he is at the root of the
child’s problems. There’s an awful lot of acting, I find that I’m acting constantly, that
‘I’m OK’, that I’m on an even keel, ‘I’m happy, I don’t care that you’re abusing that
child’. My instinct is to grab them by the lapels and tell them what you really
think…but you don’t….”
She then talked more generally about how she had learnt over the years as a teacher to rarely
lose her temper, although she was naturally “quite fiery”. However, she felt that recently her
anger had become more difficult to control. By way of illustration, she made a “confession”
about a recent incident in which she had “overstepped the mark” and become “hotter than I
should”, which, she felt, “shows it’s really getting to me”. In this incident, she brought a child
into her office, calmly, and then when he didn’t respond to what she said “I really yelled!”.
This shocked the office staff and when she went to pull her office door closed she held the
handle with such force that she bent the metal and pulled the two bottom screws out. She saw
this as an indication of how furious she was inside, and said, “I didn’t notice at the time I’d
done it [ ] I didn’t like that, I was cross with myself. It was just too out of control”.
Interestingly, the head teacher then went on to link this with stress and burnout-type outcomes
saying, “I’m kind of like a born again calm person but it doesn’t work completely, because of
the pressure I’m under. It’s considerable and it’s bound to have a damaging effect and I think it
already is”. Asked what these outcomes were, the interviewee listed drinking too much and “a
pent up anger that as a deputy I didn’t have”. She also linked this chronic anger with
75
“outbursts of anger” at home and with “a whole roller coaster of emotions that you would have
when you’re going over the edge”. The interview therefore became a candid and eloquent
account of her perception that there was a link between her emotion work, centring on her
anger, and her feelings of stress and burnout.
A care worker within the ‘distressed group’ gave another good example of suppressed anger.
Her story is discussed in detail in 4.5.2.4.
Suppressed dislike and lack of empathy
In the diaries, there were one or two instances of events in which the worker found it very
difficult to be empathetic, and had to suppress her dislike, or distaste, or disapproval of the
client. In these cases, the client was blamed for his own misfortune. As in the case of suppressed
anger, these were very negative events (the diarists circled emotions like ‘helpless’,
‘emotionally-drained’ and ‘burned out’ in the emotion checklist).
In the first example (from the substance misuse worker) the diarist was referring to a client who
had been about to go into ‘rehab’, but had failed to attend:
“My session with the client who had relapsed made me feel very frustrated and
discouraged. I felt very helpless in knowing how to help him best and I found it very
difficult to be empathic - not only because he had used drugs just prior to the session
but also because he showed little sense of responsibility for what he had done. There
was little insight into why he had started using again - he just blamed his
homelessness.”
The second example was written by the police doctor. It gave a very clear indication of her
effortful emotion work as she struggled to be “objective” and overcome her distaste of the
offender and his behaviour:
“Interview and medical examination of person arrested as a result of allegation of
rape of his wife on 7 occasions. I tried to be objective on the basis that he’s not guilty
just because he’s in a police station. Nonetheless, I could not. I found this interview
particularly distasteful and went out of my way to decline a repeat examination he
requested 3-4 hours later, on the grounds that it was not medically indicated, as I was
aware that he wanted to leave his cell for a five minute break.”
Other instances of emotion work
Several other examples of emotion work emerged in the interviews. A hospital based social
worker referred to the difficulties inherent in dealing with a client’s negativity and feelings of
depression. Talking about a visit to a particularly difficult client (cited in the diaries), she said,
“When you meet people in very difficult circumstances they can be so negative about everything
that it’s very difficult to take anything good out of it.” During the interaction, she portrayed
herself as rather lost for words, and struggling “to think what to say next, to find a suggestion
that they might engage with or be positive about”. Also, afterwards, when she reflected on the
case she felt similarly at a loss, thinking, “How am I going to handle this one?” And “What can
we offer?” This social worker also talked about how being in a helping role was personally
important to her. Given this, it seemed that the difficulty with theses cases was that she
struggled to interpret them positively and integrate them into her image of her caring and
helping role.
A second, different example of emotion work was given by a head teacher, who disliked the
large amount of out or hours work she was required to do. She spoke of an outburst of petulance
76
with the local vicar at the end of a church service, which she had felt obliged to attend and
contribute to in her head teacher role. The researcher asked if it felt like a relief to be that honest
about her resentment at having to give her time to this, and she replied, tellingly, “No, actually I
felt bad afterwards. I felt that I should be saying this at school, I should be saying this to the
parents, I should be saying this to the governors, I should be saying this to the staff, I shouldn’t
be saying this to the vicar!” There was therefore emotion work inherent in saying ‘yes’ to all
these different people when she really wanted to say ‘no’, and her reluctance and/or resentment
(a little like the other head teacher’s anger) had been suppressed and displaced in another,
arguably less appropriate, context.
Having to do ‘extra’ emotion work
Finally, there were hints in both the diaries and the interviews that workers felt more sensitive
and vulnerable at times when they felt tired, or less mentally robust than usual. In these
situations, coping with verbal abuse or poor behaviour required more effortful emotion work
than usual. It was as if human service workers had a particular tolerance level for emotionally
demanding situations, but when their reserves of patience or energy were low, emotion work
became particularly difficult and effortful.
In the diaries, a housing worker wrote:
“Tenant complaining that work not carried out on time wanted it doing sooner or
would go to his Councillor. I said we would not do the work any sooner due to
resource shortages. Tenant verbally abused me which agitated me and put some stress
on me especially as it was my first day back and because I felt tired. This is normal
occurrence each day but when tired it seems to bother me more.”
A head teacher wrote similarly (under ‘client interactions that made you feel negatively’):
“Several today unfortunately. By shortly after 9am several children had been given
red cards - must have been supporting France last night. Calming them down was
more difficult than it often is, possibly because my frame of mind was not as positive.”
At interview, the housing worker spoke about this in general terms, saying that sometimes he
just felt “I’m tired”, and “that’s why you’re more sensitive to criticism, more sensitive to what
people say, more sensitive to how you’re feeling…”. Emotion work thus became more effortful
when people felt vulnerable, including when they felt stressed and/or burned out, and this is
discussed in the next chapter.
4.4.3
Concluding Summary
• Certain types of emotionally demanding situations were seen as important stressors by
the interviewees. Sometimes these were occupations specific. They included having to
deal with emotionally-charged ‘emergencies’; having overall responsibility for the
client’s welfare; having to be an expert in many different areas; questioning your own
actions in situation; dealing with conflicting pressures from within a client’s family; and
dealing with difficult staffing issues.
• Interviewees talked spontaneously about emotion work in their jobs and spontaneously
used words like performing and acting. This underlines the importance of this aspect of
caring labour. Given the theme of the diaries and interviews, it also highlights its
probable importance in the context of stress and burnout.
77
• Emotion work was frequently seen as both emotionally challenging and rewarding,
particularly when it resulted in desirable outcomes. It seemed to be adaptive and
functional for workers to view this aspect of the job positively.
• Deep acting was talked about as a useful aid to generating empathy, modelling
behaviour for the benefit of others, and creating an atmosphere of calm.
• Tactical pieces of surface acting seemed to be easy to perform. However, the underlying
situation was sometimes stressful, and surface acting also became more difficult when
the interaction was not going the worker’s way. Surface acting was also sometimes a
feature of situations of high emotional dissonance, involving the suppression of strong
negative emotions (see below).
• Certain types of emotion work were seen as unequivocally difficult and damaging.
Examples of this included interactions involving the suppression of anger or other
strong negative emotions, like dislike and distaste. Situations in which the worker was
stonewalled and/or was unable to voice his feelings and opinions also involved effortful
emotion work. In all of these situations there was a marked discrepancy between how
the diarist felt and how she tried to act (emotional dissonance).
• There was also evidence, from both the diaries and interviews, that workers had an
implicit tolerance level for emotion work. When they felt tired or otherwise vulnerable
their tolerance dipped, and emotion work (be it in the context of role modelling good
practice to staff or dealing with verbal abuse from clients) became more difficult and
personally taxing.
4.5
STRESS AND BURNOUT
This chapter describes some insights into stress and burnout in human service work, as revealed
by the interview and diary data. The chapter begins by examining how people used the term
‘burned out’. It then examines some factors associated with stress and burnout-type emotions.
These were mainly (but not exclusively) derived from on a sub-group of diarists who are here
termed the ‘distressed group’ (as explained in 4.4.2).
4.5.1
Diarists’ thoughts on burnout
Before discussing stress and burnout, it is interesting to note that the people who circled ‘burned
out’ in the diary emotion checklist, often meant ‘extremely tired’. This was ascertained at
interview: For example, one housing worker in a residential home circled ‘burned out’ the day
after he was woken at 4 a.m. by a difficult client and then had to work a full day shift. When he
explained this, it was clear that ‘burned out’ was a comment on his temporary, physical and
mental state, rather than anything to do with his attitude towards the job.
However, some interviewees also held beliefs about burnout in its psychological sense (these
were expressed spontaneously, as this was not a prompted discussion point). For example, a
care manager talked of the high burnout rate in residential care, because “the shifts, the training,
and the emotional (side of the job) can be very difficult for people and they can leave quite
quickly. It tends to be about two years people manage”.
A head teacher (of two years standing) said,:
78
“I mean, I think if I hadn’t taken time out and leave when I had to go through all the
redundancy and everything, I can see myself having a breakdown of some sort. And I
really can’t see how somebody can do this job for twenty-five or thirty years, for long,
it’s just exhausting, it’s exhausting, it’s a completely draining job”.
In this quote, she seemed to suggest that: (i) she might have burned out early in the job; and, (ii)
that head teaching was untenable as a long term job (for her at least), because of the high
likelihood of burnout. Therefore the care manager and the head teacher both seemed to see
burnout as a phenomenon that happens early on in a job/career. However, in the context of job
tenure, it may also be worth noting in passing that the two diarists who came across as most
classically burned out in this study had both been in highly demanding, human service work for
many years. They had an air of being exhausted and depleted by it. As one of these diarists said
(using the word ‘burnout’ spontaneously):
“It’s just, it’s come to, I tell you I’ve reached a point now where I, you call it burn out,
burn out or whatever you wanna call it. But it’s just the thought, you know, having to
put yourself into frame of mind, you know, actually going into work, and it’s like ‘oh
god here we go again’, you know, I don’t think I can cope with all the problems,
certain members of staff and…. I’ve just reached a point where I had enough, you
know. I want to leave, I would love, I would like to leave, but not only is my age
against me I feel when you do this type of work it really deskills you it really does. I
feel, I’m totally deskilled. [ ] I’ve got my age against me, I’ve got my health against
me, (I’m) totally deskilled. So there is a lot of frustration. It’s like I’ve stayed to long, I
stayed in the job too long, so now I’m stuck. [ ] But I can’t see any way out and I think
that’s what’s really, really frustrating.”
4.5.2
Stress and burnout-type emotions in human service work
As discussed in the previous chapter, it was clear from these data that emotionally demanding
situations (e.g. ‘emergencies’ in schools) and certain aspects of emotion work (e.g. suppressing
anger) gave rise to the type of emotions associated with stress and burnout. Additionally, a few
of the diarists seemed to be experiencing a relatively high number of stress and burnout-type
feelings (like ‘stressed’ and ‘emotionally drained’, as circled in the emotion checklist) and a
relatively low number of positive emotions (like ‘happy’ and ‘cheerful’) at the time of the
diaries. This group will sometimes hereafter be referred to as ‘the distressed group’. The diaries
and interviews of the distressed group contained some common features that seemed to be
associated with stress and burnout, and these are outlined below.
However, it is important to note that it was often difficult to tell whether these factors were
causal or symptomatic: For example, does the propensity to ‘take things personally’ eventually
lead to burnout, or do people take things more personally because they are feeling stressed
and/or burned out? Therefore, it is important to note that any inferences that are made about the
processes involved in burnout, or about causality, are made only tentatively.
A loss of professional detachment, or ‘taking things personally’
Most of the sample, most of the time, wrote and spoke in a tone of professional detachment.
This seemed to be underwritten by a belief that detachment was an important defence against
absorbing the clients’ stresses, and against becoming overly emotionally affected by their
problems. For example, a housing worker who dealt with mentally ill residents of a care home
said (of a particularly difficult client), “No, we can’t (take it personally). Well if you take it
personally then you get stressed out, then she (the difficult client) can do it some more. You
have to just take it and play her game. You don’t want to put too much stress on yourself”.
79
A social worker said similarly:
“You have to detach yourself slightly from people, which is a bit of a contradiction.
But if you can slightly detach yourself from it, it gives you a clearer view of exactly
what’s going on and not to taking on people’s stress because that’s what they want to
give you, they want to give you their stress and their anxiety. But actually by taking it,
it doesn’t help because you cannot think clearly yourself and they’re still just as
anxious as they always are.”
In contrast to this, the tone of the diary entries in many of the most upsetting events was more
personal (see 4.3.1.1). Furthermore the problem or, more usually, the criticism, was taken
personally. Sometimes, this was simply because the criticism was personal. For example, a head
teacher wrote about an incident in which someone snidely implied that she had lost her
“cultural heritage”. At interview, she said, “I was quite surprised that I was so cross”. In this
instance, it seemed that the very personal criticism had penetrated her armour of professional
detachment.
However, ‘taking things personally’ also seemed to be a mode of interpreting events, because
some of those in the distressed group appeared to generally take things more personally than
many of the other diarists. One example of this was the diary set of the head teacher cited earlier
(4.3.1.1) who was involved in a very difficult dispute with, and between, a group of parents. She
referred to this dispute in quite personal terms, writing variously about her ‘fear’, ‘worry’,
‘dread’, ‘anger’ and ‘frustration’ at the situation. Given that the situation was dramatic and
personal, this was understandable. However, she also reported other incidents in quite a
personal tone. For example, she wrote: “On another note, I received a reference request for a
teacher who is still working here, who hasn’t even hinted at her intention to relocate [ ] this
made me feel like a fool for not knowing what is going on in my own school”. In her words, the
conflict seemed to “colour” her view of other events, such that she had become personally
sensitive and apt to construe negative events as a reflection of her and her abilities.
‘Taking things personally’ in the context of stress and burnout
Thinking in more detail about the theme of taking things personally, it also appeared that this
was seen as playing a role in stress-type outcomes. This was suggested by comments by two
interviewees. The first of these, a head teacher who was a model of calm and professional
detachment, was asked about the personal consequences of dealing with a child’s aggression.
She replied, “I don’t find it difficult and it doesn’t cost me a lot emotionally. I don’t take it
personally”. She then referred to a time earlier in her career when she was unable to remain
detached, saying, “I think when I didn’t do it, when I couldn’t do it, it cost a lot.” She thus
implied that professional detachment was an acquired skill, which acted as an important defence
against the negative emotional consequences of the job.
The second interviewee, a housing worker, talked about a recent period of depression in which
work was a contributory factor. At the time, he felt that he was on the receiving end of
substantial pressure from senior managers to perform. Moreover, he took this departmental
responsibility to perform personally. In his words, “I suppose I felt…because I take it
personally: This is my department, this is my section. And I want it to work.” He then explained
how he had recovered from his depression with the help of counselling:
“Just the counsellor and I think it was just realising that I got to just stop whittling and
worrying and people aren’t out to get me. My boss and workers aren’t here just to
bring me down. And I think I had a long chat with my boss [ ] and I said, ‘Look I feel
80
really crap, I said I am struggling, you know, I said I feel as if you are on my case all
the time. And other people, criticism, it’s there, it sticks, it’s not bouncing off, it’s
sticking whereas constructive criticism that’s fine. But I feel that everything (I’m
doing) I think I’m doing wrong’.”
In this quotation, the interviewee spoke of how his egocentric view of events at work seemed to
be a factor in his illness and how the realisation that “people aren’t out to get me” was
significant to his recovery. This again re-iterated the possibility of a link between stress-type
outcomes and taking things personally (or, indeed, the loss of professional detachment).
However, there is a caveat here, in the context of burnout, because one of the key symptoms of
burnout is a sense of depersonalization, or cynicism, which would seem to be quite the opposite
of ‘taking things personally.’ This means one of two things: Either that erosion of professional
detachment could be a pre-stage that happens before burnout develops, or that ‘taking things
personally’ is associated with phenomena like stress and depression (as in the above quote), but
is not associated with burnout.
Additional emotional demands, usually home and family-related
Most of the distressed group had very demanding home lives at the time of the diaries. This
placed an extra emotional burden on them, in addition to their work. Moreover, because these
home-related factors were tangible problems, it seemed that they were causal i.e. they
contributed to the negative feelings, rather than being part and parcel of an overall jaundiced
view of life. Looking across the diary sets, it seemed that having to perform a great deal of
emotion work both at work and at home was a potent trigger for stress and burnout-type
feelings.
Several people suggested that occupation also had a causal effect over this, because friends,
relatives and colleagues held particularly high expectations of the amount of support they would
receive from a professional carer. For example, a manager in a care home said, “I’m the
lynchpin of the family, the one that keeps everybody together and obviously because of the job I
do, if anyone’s got a problem then it’s me they offload it onto”. Thus the total emotional load
might be very high, because of high expectations of empathy and caring at home, as well as at
work. The same interviewee also said, “It’s important (in this job) that you can just switch off
and relax. It makes it more difficult when you’ve got issues going on at home because you might
switch off from work but then switch on to something else. So you’re emotionally constantly
going.” This notion of being “emotionally constantly going” neatly articulated the problem of
facing high emotional demands in both (home and work) spheres of life.
A housing worker provided a particularly good example of this. He was injured at the time of
the diaries and was in pretty constant pain. Additionally, his mother had very recently been
diagnosed as terminally ill. He had significant caring responsibilities for her, and his diaries
contained several instances of his emotion work in this context (e.g. on Day One, “trying to stay
optimistic for her benefit but she gets weaker each day” and on Day Four, “trying to keep a
happy atmosphere going for her”). This diarist received a high number of complaints and a
great deal of verbal abuse from his clients in his job, and he believed that his personal
circumstances were now impacting on his job, such that, “I’m less patient with fools. Certainly
now”. This was also evident in his diaries, (e.g. on Day Two, “fed up of listening to moaning
people”). He also went on to describe how he had to “try not to show” this impatience with his
clients.
Equally, when he arrived home from work to find his children fighting (Day 11) he wrote, “It’s
bad enough dealing with nasty people all day without coming home to conflict.” In this context,
he explained how he felt that “it makes my family suffer in the long run” because he was
81
becoming “less patient, more grumpy, intolerant” with them as a result of the “roller coaster
ride” which he was “on” at work. Work and non-work therefore seemed to be locked into a
reciprocal relationship, in which high emotional demands in one part of life heightened the
requirement to do effortful emotion work in the other part, resulting in feelings of stress and
emotional exhaustion.
With reference to burnout, it is interesting that this diarist was the only individual in the study
who came across as cynical. This was evident at interview, and also in one or two of his diary
entries. For example, on Day Five he wrote:
“Big argument with manager about being positive with a job where known offenders
have moved to another place and started committing the same act of ASB (anti-social
behaviour). I wish to go back to court straight away, she wants to sit back and see
what develops just in case they are innocent??? Feel very sorry for the innocent
people that are suffering at the hands of these people.”
It was not possible to tell whether or not his cynicism had increased recently. Nor was it
possible to verify that his cynicism was, as he believed, almost a character trait, acquired a
number of years previously during his time in the police force. Nevertheless, it was clear that
this diarist displayed some of the signs of burnout at that particular time, and that this situation
seemed to be connected with first, prolonged exposure to difficult clients and emotionally
demanding situations at work and, second, concurrent problems in his personal life, which
substantially increased his total emotional load.
Non-client relationships at work
For several diarists in the distressed group, negative feelings seemed to be primarily associated
with internal (non-client) relationships at work. As discussed previously in 4.2, client contact
brought many rewards for the workers in this study, as well as difficulties, and it is interesting to
think about the relative importance of worker-client relationships, versus worker-manager (or
co-worker) relationships, in the development of stress and burnout.
An illustrative example was provided by the care worker in a children’s home cited earlier
(Section One), who used the term ‘burnout’ in relation to her current feelings about work. Her
diaries contained many testimonies to her unhappiness at work at that time. For example, she
wrote:
“Dread the thought of going into work. Hate that damned place” (Day Four), “WORK
JUST PISS(es) ME OFF ESPECIALLY ONE OF THE MANAGERS AND MY (names
manager’s position)” (Day Four), “My whole work experience is making me feel
NEGATIVE. Just being here is like a form of torture” (Day Eleven) and “JUST HATE
THAT PLACE” (Day Twelve).
Several factors contributed to her distress. In particular, she had a medical condition which
caused her to take significant amounts of sick leave and this created tensions between herself
and her managers. This was referred to several times in her diaries, when she was ill, took time
off, and was criticised for it afterwards. Moreover, there were factors to do with the culture (e.g.
her belief that there was institutional racism at the home) and to do with her own career (e.g. her
belief that she was now de-skilled, too old to change occupations, and had been overlooked for
promotion, in spite of her long job tenure and good qualifications). Thinking about the
processes involved in burnout and emotion work, it seemed, in this case, that her emotion work
was extremely effortful because she felt so negative and disliked being at work so much. This
82
was reminiscent of how other diarists described doing more emotion work, and more effortful
emotion work, when they were tired or feeling vulnerable.
One of the most interesting aspects of her diaries was the contrast between her relationship with
the children and her relationship with her managers. The children were often verbally abusive,
but she saw this as usually “not meant”, and simply a function of being, “the closest person the
child has got in their life and they will say things to you, they will try and hurt you.” She
elaborated:
“Sometimes, you know, you might have a young person, come and sort of call you a
c**t or something a bit more, go and screw your kids or whatever, you get all the
abuse, it’s unmentionable and then they’ll come back and they’ll say, ‘look I’m sorry I
shouldn’t have (inaudible word) like that’ and then they talk, yes, ‘but this is how you
make me feel’, and then they make you a cup of tea and then they’ll still talk to you.
And that in itself it doesn’t sound like much but the very fact that a young person took
on board that they upset you and they came back and apologised to you and tried to
make amend, it might not seem much but it’s quite a lot. [ ] you don’t look for
gratitude, you never look, well, I don’t look for anything, but it’s just little gestures like
that. They come and give you a hug. [ ] It’s little things like that, that’s the real
pleasure of it, this is the reward.”
However, in the next sentence she discussed two diary entries in which she had been criticised
twice by her managers. In one diary entry she wrote, “Really pissed off. Told by my (manager)
that I can no longer shift lead [gives medical reason]. Not happy, but have to put the clients and
their needs first. I am a childcare worker.” In the interview, she explained:
P: “Ehm, yes, what happened was, I think, the previous night when he told me, again,
I was so, so angry and so stressed out. My colleague kept asking if I was OK. Ehm, my
colleague kept asking, ‘are you alright?’ because, you know, she could see from my
facial expression that something was not right. My instinct was, ‘I can’t talk, I need to
get out of here, I need to go home.’ [ ] I’m a quite easy going person but I had so much
pain and anger. I was in such a fowl and acrid mood that I wasn’t thinking straight at
all. But I had to think if I go home, there was only two of us on shift that night, what
would happen to the children? There is only two of us. So regardless of how I was
feeling I had to be there for the kids because that’s my job. It doesn’t matter how what
resentment or frustration, you know, I had to put all that aside and get on with my
work because that’s what I was there for [ ].”
R: “You’ve got all that going on inside…”
P: “And you have to suppress it because I can’t show my emotions to the kids, you
know, I can’t. I have to be, it’s like a clown, you know, you’ve got a smile painted on
your face but on the inside you’re crying. Because you can’t give in, you know, your
full emotions to the kids, you can’t. You have to be smiling and you have to be joking
but inside you’re in a total and complete mess.”
The description of how she suppressed her anger and felt like a “clown” in front of the children
was an excellent example of surface acting to hide anger. She went on to say how this anger was
played out at home, when it stopped her sleeping. However, the point here is that her anger was
related to her interactions with management. It was her relationship with her managers, not with
the very demanding children in her care, which was making her feel thoroughly “pissed off”
with her job. This was the most dramatic example of how, for some people, working
relationships seemed to be relatively more consequential than client relationships in causing
83
distress. This seemed to be particularly the case when the clients were children, whose
behaviour was readily understood and rarely taken personally.
4.5.3
Concluding summary
• The term ‘burned out’ was often taken to mean extremely tired. A couple of the diarists
suggested that burnout happens early in careers, although the seemingly most ‘burned
out’ of the diarists had relatively long careers in human service work.
• Amongst the distressed group of diarists, certain factors seemed to be associated with
their feelings of stress and burnout. These were (i) the propensity to take things
personally and (ii) relationships with other parties (not clients) at work. Additionally,
non-work factors seemed to play an important role: Some of the seemingly most
stressed (and even burned out) of the diarists were in the grip of difficult, emotionallychallenging situations at home.
• Emotion work seemed to become more effortful when people felt stressed and burned
out. Because of this, emotion work seemed to be as much a symptom of stress and
burnout as a cause.
4.6
ADDITIONAL FINDINGS
This, the final chapter of the qualitative results, examines two sets of additional findings from
the qualitative study which are included here as they provide some additional insights. First, it
looks at diarists’ perceptions of the harmful consequences of stress and emotional load at work.
Second, it examines some potential buffers to stress and burnout, and some insights into how
workers coped with the emotional component of their jobs.
4.6.1
Detrimental consequences of human service work
The sample (particularly those in the distressed group) suggested that the demands of their jobs
at the time had a number of negative effects. These are now briefly discussed.
Negative health behaviours
In the diaries and interviews, there were several instances of negative health behaviours. For
example, one interviewee felt that she drank too much as a result of the pressures of her job (see
4.4.2.6). Another head teacher made healthy and unhealthy choices in accordance with her
feelings about the events of the day. She discussed this at interview (talking about a particularly
difficult event in the diaries):
“That just makes me want to eat loads. On the night that I felt really down about it – it
was the Monday night –I went home and felt really awful about it. I was going to go to
yoga and didn’t bother. I went home and I didn’t want to do anything except sit of the
sofa and eat. [ ] And I think I did. I think I got a take-away on the way home.”
Another diarist wrote about how she “ended up eating all the sweets I shouldn’t have”,
following, “probably the most difficult meeting I have ever been to”. In this way, she implicitly
associated her anxiety with comfort eating.
The emotional roller coaster
Two interviews used the metaphor of a roller coaster to describe the emotional effects of their
job. A head teacher (see 4.4.2.6) summed her current feelings up as, “a whole roller-coaster of
84
emotions that you would have when you’re going over the edge”. Similarly, a housing worker
spoke of how the diary, “made me look at the roller coaster ride that I’m on. And obviously that
can’t be doing me any good.”
The metaphor is interesting because it implies large vacillations in mood and also a sense of
being somewhat out of control and at the mercy of unpredictable events. There was something
quite dramatic about these interviewees’ accounts of how they felt, as if they believed that they
might ‘crash and burn’ if things carried on this way, or if they didn’t metaphorically climb off
the roller coaster and quit. The metaphor seemed to be a powerful description of a process
(perhaps burnout), which directly linked work-related feelings to negative personal
consequences.
Anger at home
Such consequences included the expression of anger at home, which were mentioned by both
interviewees in the context of the roller coaster. The head teacher, in particular, referred to a sort
of chronic, suppressed anger that originated at work (“a pent up anger that I didn’t have as a
deputy”) and was vented in “outbursts of anger” at home.
A home tutor in the sample also spoke about the overspill of work stress into home life at
interview. Her working day was divided into ‘sessions’ with different pupils, and difficult
sessions would leave her feeling a range of negative emotions (e.g. “frustrated”, “wound up”
and being driven “round the bend”). She felt that, “I probably come home feeling most stressed
if my last teaching session of the day hasn’t been stunning”. She also said that after a difficult
day with her students, “I’m rattier and more likely to fly off the handle”. In this way, as per the
other examples, suppressed negativity at work seemed to be vented in the more relaxed and
permissive sphere of home.
Other negative effects
One of the most distressed diarists had another view of how suppressed emotions at work
affected her at home. This was the childcare worker who talked about feeling like a “clown” in
front of the children, smiling and joking whilst inside “you’re in a total and complete mess”
(see 4.5.2.4). In her words:
“You have to just get on with it. And then…when you go to bed, in my case I can’t
sleep. I’m up all night because I’m so angry. [ ] I just can’t sleep. Feelings are going
round and round and round in my head because if you’re feeling victimised, you’re
feeling completely on your own.”
4.6.2
Buffers and coping
More optimistically, the data also revealed several factors that seemed to act as buffers,
protecting the worker against the harmful effects of handling emotionally demanding situations.
One buffer, already discussed in previous chapters, was a smooth-running, relatively
unproblematic home life. Similarly, the most obvious buffer at work was social support from
co-workers and managers. This was particularly strong amongst social workers and some careworkers, who received formal supervision in the form of sessions with their managers, in which
they discussed their cases and received support and guidance. In addition to this, the team
structure was particularly supportive. Head teachers, too, drew on a network of support which
included school governors, other heads, and counselling-trained learning mentors. The latter
sometimes played a valuable role in helping to plan and co-handle difficult situations (like
meetings with aggressive parents). All of these sources of support seemed to be of great value,
85
particularly in personal, work-related crises (e.g. when a social worker was threatened with
court action and when a head teacher’s received a low performance rating in an Ofsted [school
inspection] report).
A less obvious buffer centred on the division of the working day. For example, head teachers
enjoyed planned time with the children, such as practising for a play or, for one, working in the
attached nursery for children with special needs. This head saw this as her “time out”, which
meant, “I come out feeling a lot happier and ready to tackle the paperwork”. In other
occupations, the division of the working day into discrete visits and the handling of multiple
clients also seemed to act as a buffer. This was because, by the law of averages, if one case is
going badly then another is likely to be going well, so that the individual diaries often had a neat
balance between positive and negative events. When this idea was suggested by the researcher
to the tutor in the sample, she agreed, saying:
“Yes. Like one day of the week I’ve been working with silent teenager for two and a
half hours and then the nine year old boy who started off by not responding. But he’s
now settling down so, surprisingly, I find myself breathing a sigh of relief when I leave
the silent girl, knowing that I’m going to spend an hour with this boy who will
probably… his concentration skills are very poor…but at least I know that we will get
something done.”
In addition to these buffers, workers in the sample mentioned consciously using several coping
mechanisms to help them in emotionally demanding situations. This final section of the results
will briefly describe these.
Self talk in emotionally charged situations
A number of interviewees used self-talk to help them manage difficult situations. For example,
the housing worker (see 4.5.2.3) talked about how he used self-esteem bolstering thinking to
help to feel superior, and thus cope, when he was being verbally attacked. Here he was asked
about whether or not verbal abuse “bothered” him:
“It depends what mood I’m in. Sometimes, but to be honest, I know it’s a sad thing to
say but I’ve just got used to it over the years. You think, ‘I’m better than that. You’re
still on benefits, mate, you’ll not work towards your life. I’m a bit better than you’.
Because I think this is the only way to cope with it.”
Several of the head teachers also used self-talk in interactions with argumentative and
potentially aggressive parents. One recounted a recent incident with a father at the school:
“The last major run-in I can think of, we had father come in and he was objecting to
the fact we were asking him to pay his children’s dinner money. [ ] And he put his fist
up to me. And at that point, my other self kicks in and I said (calm voice) ‘breathe, put
your arms by your side…’, I had take myself out of any confrontational body language
I could be giving off, but deep inside I am petrified. [ ] And I just said really calmly,
‘Please put your hand down, if you hit me, the police will be called, because it is
assault. Please put your hand down.’ And on the third time he put his hand down.”
Thinking about emotion work, the self-talk in these situations can be seen as an aid to surface
acting in situations of high emotional dissonance. For example, in the above example, a calm
demeanour was needed in order to defuse the situation, but, for the head teacher, emotional
dissonance was high because she felt very frightened, and anything but calm, underneath. In her
words, “And I’ll do it and get rid of him, but my secretaries know at that point I will go grey,
86
I’ll cry, I need a cup of tea with sugar.” This shows how suppressed fear had a powerful,
negative impact on the actor.
Being able to voice feelings, and give emotions ‘back’
This coping mechanism was the corollary to feeling stonewalled, or not having a voice, in
interactions. It involved being able to voice feelings and, in this way, give responsibility for the
negative emotions back to the person who caused them. In this way, the inequality between the
carer and the cared for was often reduced, and the relationship became more reciprocal. The
coping mechanism cropped up earlier in the discussion: For example, the social worker in 4.4.2
was able to voice her discomfort with her elderly client’s placement by asking leading questions
in front of the eager-to-leave son; similarly, the tutor in 4.4.2 was able to politely convey some
of her resentment to her sullen Spanish student, via an interpreter.
A head teacher also talked about giving emotions back in the context of a problem with a very
disadvantaged and impoverished family, whom, she felt, social services were failing to help. In
the diary she wrote about her frustration at seeing the children’s time “dribbling away” without
anyone making any impact in improving the quality of their lives. Asked if she was able to act
on this frustration, she replied:
“The action you can make is to communicate with Social Services that I’m frustrated
at the lack of action. I did communicte it. [ ]They came in [ ]. Because it’s no good not
putting it to them. It (the frustration) has to go where it belongs. Otherwise I end up
taking it home and it doesn’t belong in my head, it belongs with social services. Put the
problem where it belongs.”
Here the interviewee deftly described the function and value of voicing the frustration and
passing it back to its source: Like a piece of unwanted emotional baggage, it was removed from
her mind, leaving her less troubled, less encumbered and presumably, with renewed hope of
action.
Pausing and separating/Compartmentalization
The coping technique of pausing and separating, or compartmentalization, involved creating and
maintaining a mental boundary between events, or between spheres of life like ‘work’ and
‘home’. It seemed to be an antidote to emotional carry-over from one event to the next. In these
situations, an individual’s emotions in one context affected how they felt in a different situation.
One head teacher described it like this, “But if something else happens that’s either really nice
or a really horrible thing I just feel like it tips me the wrong way, it tips me over the edge and I
just want to cry”. Such emotional carry-over generally seemed to leave workers more
emotionally drained and stressed than when events were more clearly separated.
The head teacher discussed in the previous section was very adept at pausing and separating.
Her school was in a deprived area and there were many problems. One way in which she coped
with a barrage of emotionally demanding situations was to take a small amount of time after
each event to pause, reflect and clear her mind. This helped her to minimise the emotional carryover between events, so that each new problem or challenge could be tackled in a calm and
purposeful way, unimpeded by worries, frustration or anger about what had happened
previously.
The technique was referred to in her diaries, when after a very difficult day (“had to take over a
class for maths, teacher distressed sent her home”, and “more of the same, end of afternoon,
two parents rowing with each other just outside the office, abusive and aggressive [ ] I removed
them from the premises and split them up” ) she wrote: “Have a meeting for new parents for
87
reception this evening [ ]. Don’t feel well disposed at the moment, will need to find a quiet
space for ½ hour”.
At interview, she described it like this:
P: “One of the things that I’ve learned to do is very simple it’s to pause, to literally
pause and stop. Because sometimes, in the past, I’ve gone from being with a child
who’s been angry, distressed or upset straight into perhaps talking to a member of
staff or a parent and if I don’t separate them out then I can carry all the baggage that
comes from that into the next thing. What I do is very simple, I just give myself a few
seconds, just to pause and stop, to let go of ‘that’ before I start ‘this’.”
R: “What do you do in those few seconds?”
P: “Focus on the breathing. I try to clear my mind. Because I meditate a couple of
times a day – it’s a sort of mini-meditation. It’s just letting go of ‘stuff’. I’m not a
complete ..., I do get upset, I do feel things, but I’m able to let go of them so I’m not
taking stuff from place to place.”
Other people used a similar technique, though usually less consciously. For example, people
talked about having learned to “switch off” from work and thus mentally separate work from
home, and about making mental spaces in the day (or “time out”) to re-charge their batteries inbetween situations. Another head teacher also talked about pausing in the context of decisionmaking during a fraught situation:
“I suppose I was taught along the way that there a very few decisions that you make
that have to be done there and then. And so you can find stalling techniques to just
think into it a little bit, because emotion kicks in so quickly that you’ve gotta step back
and think, ‘is that the right thing I’m doing?’”
4.6.3
Concluding summary
• Some of the interviewees suggested that the heavy demands of the job had detrimental
consequences, including negative health behaviours (e.g. breaking the diet) and
outbursts of anger in the home. In some cases, the spill over of anger into the home
seemed to arise from a sort of chronic, suppressed anger, originating at work.
• Two interviewees used the ‘roller coaster’ metaphor to describe how they felt at that
time. They implied that their feelings of stress and possibly burnout incorporated
emotional ‘highs’ and ‘lows’ and also a sense of being slightly out of control. One said
she felt as if she was about to “go over the edge”.
• More optimistically, the interviews revealed some factors which seemed to buffer the
negative effects of high emotional demands on the individual. These included social
support and the tendency for negative events in the day to be offset by positive ones,
like enjoyable client contact. This latter point was particularly relevant in jobs which
centred on a caseload of clients and where the working day often consisted of several
planned interactions with clients.
• The interviewees also talked about coping. For example, in emotionally charged
interactions (e.g. situations of verbal abuse) they sometimes used ‘self-talk’ to enhance
their surface acting. Interviewees had also learnt to tactically pause within and between
88
events, and to separate things in their mind’s eye, in order to minimise emotional carryover across interactions and situations.
4.7
SUMMARY OF THE QUALITATIVE RESULTS
• More positive or rewarding events were reported in the diaries than negative, or
potentially harmful, events. In particular, the majority of diarists wrote about multiple
occasions when they had enjoyed spending time with their clients.
• Another frequently mentioned type of event was a meeting or interaction with a client
that went well and had a successful outcome. These types of event sometimes involved
dealing with a client’s negative emotions and, because the diarist put her skills to good
use to turn the situation around, they were rewarding. Such events also held a sense of
meaning and importance for these workers.
• The diaries, supplemented by the interviews, revealed a number of types of event,
which engendered negative emotions. Such types of event included: Interactions with
difficult clients; dealing with clients who were going through a negative experience; and
being subject to verbal abuse from clients. For other (non-client related) types of event,
the key themes were: Paperwork; problems with the working environment; a colleague
or subordinate’s poor handling of a situation with a client; and difficult interactions with
colleagues or third parties.
• All of these different types of event could be seen as potential ‘causes of harm’ in
human service work. However, a careful analysis of these situations also revealed that
certain over-arching themes were common to different types of event. These themes
revealed important reasons why particular situations were experienced as difficult or
emotionally demanding. They included: Not feeling listened to in your professional
role; having your actions questioned in your professional role; feeling unable to turn a
situation around in order to bring about a positive outcome; and concerns about failing a
client.
• From these data, a distinction might be drawn between events which might be harmful
simply because they generated negative emotions for the worker, and events which
might be harmful because they were emotionally demanding. Paperwork and problems
with technology, for example, were not emotionally demanding. However, paperwork
was seen as stressful, partly because it lacked the meaning and importance of other
challenges of the job. Difficult interactions with clients and co-workers were
emotionally demanding, by contrast, and were experienced as difficult for the types of
reason listed above (e.g. not being listened to in your professional role).
• Furthermore, it is also important to note that emotionally demanding situations were not
synonymous with stress and harm. Emotion work was frequently seen as both
emotionally challenging and rewarding, particularly when it resulted in desirable
outcomes. It seemed to be adaptive and functional for workers to view this aspect of
the job positively.
• That said, certain types of emotion work were seen as unequivocally difficult and
damaging. Examples of this included interactions involving the suppression of anger or
other strong negative emotions e.g. dislike and distaste. Situations in which the worker
was stonewalled and/or was unable to voice his feelings and opinions also involved
89
effortful emotion work. In all of these situations there was a marked discrepancy
between how the diarist felt and how she tried to act (emotional dissonance).
• There was also evidence, from both the diaries and interviews, that workers had an
implicit tolerance level for emotion work. When they felt tired or otherwise vulnerable,
their tolerance dipped, and emotion work became more difficult and personally taxing.
• Amongst the distressed group of diarists, certain factors seemed to be associated with
their feelings of stress and burnout. These were (i) the propensity to take things
personally and (ii) relationships with other parties (not clients) at work. Additionally,
non-work factors seemed to play an important role: Some of the seemingly most
stressed (and even burned out) of the diarists were in the grip of difficult, emotionallychallenging situations at home. From this group, it was also noticed that emotion work
seemed to become more effortful when people felt stressed and burned out. Because of
this, some features of emotion work (like surface acting) sometimes seemed to be a
consequence of burnout, rather than a cause.
• Two of the interviewees in the distressed group used the ‘roller coaster’ metaphor to
describe how they felt at that time. They implied that their feelings of stress and
possibly burnout incorporated emotional ‘highs’ and ‘lows’ and also a sense of being
slightly out of control.
• More optimistically, the interviews revealed some factors which seemed to buffer the
negative effects of high emotional demands on the individual. These included social
support and the tendency for negative events in the day to be offset by positive ones,
like enjoyable client contact. The interviewees also talked about coping. For example,
in emotionally charged interactions (e.g. situations of verbal abuse) they sometimes
used ‘self-talk’ to enhance their surface acting. Interviewees had also learnt to tactically
pause within and between events, and to separate things in their mind’s eye, in order to
minimise emotional carry-over across interactions and situations.
90
5 CONCLUSIONS AND RECOMMENDATIONS
5.1
LITERATURE REVIEW
The literature review aimed to investigate the following six areas:
1. The main correlates and causes of burnout in human service work
2. Traditional work stressors that have been found to relate to burnout (and related
outcomes)
3. The pathways through which these stressors lead to these end points
4. The extent to which and how the emotional demands of human service occupations are
different from emotional demands of other related and un-related occupations
5. The relationship between emotional demands and burnout (and related outcomes) in
customer service work and human service work
6. Interventions which may ameliorate the impacts of such emotional demands on
outcomes
A key issue which recurred throughout the literature review was the relatively small amount of
evidence in any of the six areas which allowed for causal inferences to be made. In other words,
the vast majority of the available evidence does not help us identify causal links between, say
personality and burnout or how emotional demands may relate to burnout. While this is a
common feature of most research on stress and well-being at work it is particularly problematic
here as burnout by definition is assumed to be a long-term process that develops over time.
Unfortunately, and probably for the usual reasons of cost and time, very few pieces of research
have been designed that even begin to help identify the causal processes underlying burnout in
human and customer service work. We will now go on from this general point to draw some
conclusions about our findings in each of the above areas.
It is interesting to note that there appear to be relatively few consistent relationships between
burnout and some of the most commonly studied individual level causes and correlates such as
gender, age, tenure, and occupation. This inconsistency raises important questions about what
burnout might actually be and how it works. For example, is burnout something that is
experienced later on in employees’ careers as the demands of the job take their toll or is it
something that happens relatively quickly in early career perhaps leading those people who are
particularly susceptible out of that career or occupation? The available evidence suggests both
and indeed other processes may be possible suggesting that we have some way to go to
understand the burnout process. Also, the lack of a clear cut relationship between occupation
and burnout also suggests that burnout is not a phenomenon restricted to jobs such as nursing or
social work which have traditionally been seen involving the types of work most likely to lead
to burnout. While there are some consistent correlates of burnout, such as negative affectivity,
the lack of longitudinal data means it is difficult to understand the nature of such relationships.
While emotional work demands are often studied as possible causes of burnout it is clear that
more traditional job stressors are also related to burnout and, in particular, role conflict, role
ambiguity and workload. For studies which are seeking to understand how emotional demands
may affect burnout it would therefore seem to be important, as was done in the questionnaire
survey within this project, to control for these traditional stressors. The fact that these
traditional stressors are related to burnout again draws attention to the question of how burnout
develops and what sorts of job characteristics are involved. Just how important are emotional
demands as such?
91
The third literature review area considered what is known about the pathways through which
burnout works. In particular there has been much speculation about the order in which the three
dimensions of burnout (emotional exhaustion, depersonalization, and reduced personal
accomplishment) appear as burnout develops. Both the empirical evidence and the underlying
theorizing suggest it is not possible to have much confidence in the notion that burnout as a
process is one that develops in this manner – with different dimensions appearing at different
stages or points. Other, perhaps more promising approaches to understanding burnout pathways
include those which focus on intra-personal processes of emotion regulation as a possible
mechanism underlying the development of burnout. Yet again, our knowledge of how burnout
develops is very restricted by the lack of longitudinal studies.
Two important and connected issues are the exact nature of emotional labour and whether the
nature of this emotional labour is different in human service workers and customer service
workers. Emotional labour can be viewed in different ways such as a general requirement to
regulate emotional displays, as more specific emotional display dimensions (e.g. frequency,
variety, intensity), as job characteristics which impose different ways of doing emotional labour,
or the emotional dissonance that may occur during emotional labour. Developing an
understanding of how emotional labour may affect well-being, including burnout, requires
further elaboration of the nature of the demands and tasks involved in emotional labour. In
other words, how might, in principle, doing emotional labour affect well-being?
While doing emotional labour is a feature of both HSW-Other and CSW it is thought that the
motivations behind the emotional labour in each case is somewhat different and may therefore
have different effects. HSW-Other are typically in vocational and caring professions in which
the display of emotions are presumed to be more strongly tied to identity and are more genuine
so perhaps relatively automatic. In this way HSW-Others may experience fewer of the possible
harmful effects of emotional labour compared to CSW whose displays of emotion in a customer
service setting may not be experienced by the employee as genuine and be more effortful. On
the other hand, the emotional labour involved in HSW-Other may take its toll if HSW-Other
over-identify with clients or patients.
The fifth literature review area considered the relationship between emotional demands
involved in emotional labour and burnout in HSW-Other and CSW. While HSW-Other have
more emotionally demanding jobs their levels of emotional dissonance are often lower than
those in CSW. There are surprisingly few studies of the relationship between emotional labour
and burnout in HSW-Other which may be because the emotional labour literature first
developed in the context of CSW rather than HSW-Other. For CSW there is little evidence that
the quantity of customer service is associated with burnout and other aspects of well-being.
There is some evidence of a link between emotional dissonance or surface acting and burnout.
Again, it is important to note that few of these studies allow for causal inferences to be made.
These findings again underscore the different ways in which emotional labour and emotional
demands may play a role in burnout depending on whether that emotional labour is taking place
in the context of HSW-Other or CSW.
The sixth and final literature review area explored burnout interventions. There are very few
intervention studies but those few that have been conducted suggest that they may influence
levels of burnout and, in particular, emotional exhaustion. However, the lack of evidence about
the causes of burnout or the processes through which burnout develops suggests that it is not yet
possible to develop specific and well-targeted interventions.
92
5.2
STUDY 1: QUESTIONNAIRE SURVEY
The questionnaire study addressed the following research questions:
Research Question 1: What is the level of burnout and well-being in key human service
occupations in the UK?
• RQ1a: How do these levels of burnout and well-being compare to other samples?
• RQ1b: How do the levels of burnout and well-being compare across different human
service occupations?
Research Question 2: Which key variables are associated with burnout and well-being?
• RQ2a: Which key variables are associated with burnout and well-being?
• RQ2b Are different key variables associated with burnout and well-being in different
human service occupations?
It should be noted that this study was cross-sectional and therefore it is not possible to make any
causal inferences about any relationships found. We have discussed the limitations of such
designs at numerous points throughout this report. These limitations are not relevant to the first
research question as this question is largely descriptive. While the second question is most
appropriately addressed in a longitudinal study, practical constraints meant that this was not
possible. In addition, the second research question has not previously been asked in this manner
and is therefore to some extent exploratory. In general, the limitations of cross-sectional design
are often seen as more acceptable in such a context but clearly future research which builds on
these findings should adopt longitudinal designs.
Before discussing the results that relate specifically to the first research question, it is worth
again noting that there were significant differences across the three groups HSW-Other, CSW
and Teachers with respect to perceived emotional demands with the exception of deep acting.
The differences found were, to some extent, similar to those found previously. In particular the
finding that HSW-Other report lower levels of some aspects of emotional labour (surface acting
or dissonance) and higher levels of other aspects (requirement to display positive emotions and
hide negative emotions). This emphasizes the point, which we will return to later, that the
nature or quality of emotional labour seems to vary across different types of job.
The results of RQ1a again showed different patterns of results across the three groups.
Teachers reported the highest levels of emotional exhaustion, while CSW reported the highest
levels of cynicism and the lowest levels of professional efficacy along with the highest levels of
intention to quit and the lowest levels of job satisfaction compared to HSW-Other and Teachers.
CSW in general therefore have the lowest levels of well-being and most negative job attitudes.
For RQ1b the results generally suggested no obvious or significant differences between the
average levels of burnout and well-being (GHQ12) across the sample as a whole and average
levels found in other similar samples. While in some cases some dimensions of burnout and
well-being scores were higher or lower than some other groups scores for this sample were
broadly comparable with the norms found elsewhere.
RQ2a explored which emotional demands were associated with the three dimensions of burnout
and well-being across the whole sample after controlling for demographic factors, personality
and traditional job stressors. Although GHQ12 scores were not related to emotional demands
each of the dimensions of burnout was. This finding suggests that different aspects of employee
well-being in general may be more or less sensitive to emotional demands. In all cases,
personality, demographics, and traditional job stressors were very strongly associated with
burnout and, though statistically significant, emotional demands were only quite weakly related
93
to burnout after controlling for these other variables. This reflects some previous research
which has suggested that while burnout is sometimes thought of as a specific consequence of
emotional labour, traditional job characteristics may actually largely account for differences
seen.
RQ2b examined differences in patterns of associations between emotional demands, burnout
and well-being across the three groups. The results showed quite different patterns across the
three groups again suggesting that any effects of emotional labour on outcomes such as burnout
may vary considerably across different occupational groups. Therefore it is not possible to
generalize about how the emotional demands associated with emotional labour impact on
employee well-being as this seems to depend crucially on the occupational context in which that
emotional labour is taking place.
5.3
STUDY 2: DAILY DIARIES AND INTERVIEWS
The research questions addressed in the daily diary and interview study were:
1. What kinds of events and situations at work are found to be particularly emotionally
demanding at work and in what ways?
2. What are the pathways and mechanisms through which specific emotional demands
lead to particular kinds of psychological health outcomes?
While the notion of emotional labour originated from qualitative research in more recent times
and particularly in organizational psychology quantitative approaches have dominated. Hence,
this qualitative study is relatively unusual and is particularly unusual in that it uses diaries as
one of the main data collection methods.
The study provided a wealth of information about the emotional experiences of teachers, social
workers, care workers, housing workers and others. One striking feature of the accounts
provided was the double-edged nature of interactions with clients which were often both highly
challenging and highly rewarding. This finding perhaps helps to explain why HSW report
higher levels of many emotional demands but not necessarily higher levels of burnout. Difficult
interactions may come to be construed very positively when HSW are able to use their skills to
manage and resolve difficult emotional situations. In addition, in such situations HSW are
enacting important and meaningful aspects of their role. Contrast this with what may be the
case in some CSW where there is usually no on-going relationship with the customer, where the
employee does not necessarily have the resources to resolve the problem, and it seems less
likely that the employee strongly identifies with the role.
While more positive than negative events were reported in the diaries these negative events
provided some insights into how specific situations and interactions may lead to harm. In
particular, emotional dissonance, in which employees display emotions they do not feel,
appeared to be particularly difficult for participants. The consistent but somewhat unexplained
finding from the literature that verbal abuse and aggression is linked to negative outcomes for
HSW was also found here. The results of this study suggest that it was often not the abuse in
itself that was difficult to take, but rather it was other aspects of the situation which were
upsetting such as the participant feeling that they were not being listened to, that they were not
able to resolve the situation, and that they were very concerned about the welfare of the client.
The results of this study suggest, as do the results of the quantitative study, that the effects of
emotional labour may be strongly shaped by context. In this case the sorts of context include
dealing with very vulnerable groups or interacting with clients who are seen as unpleasant
94
people, or engaging in emotional labour relatively effortlessly where the participant very
strongly identifies with the role. While emotion work in the study is clearly seen as an accepted
and sometimes rewarding part of the job it also reveals other contexts in which it may be
potentially harmful such as when people are already tired, unhappy or do not have extra
emotional ‘reserves’ required to engage in emotional labour.
Turning now to the processes of burnout, there was some evidence that on-going relationships
with clients, in which the client was able to express his or her feelings where the participant was
not, were found, over time, to be particularly difficult. This relates to some more recent
theorizing in the literature about the importance of intra-personal emotion management
processes. For individuals who reported strong burnout-type feelings it also seemed to be the
case that emotional demands both at work and at home were at a high level suggesting that the
lack of opportunity to recover from emotional work demands may be an important way in which
burnout develops.
5.4
RECOMMENDATIONS
Here, we will briefly summarise just some of the recommendations for both research and
practice. While the study of emotional labour and burnout is not new it is also not welldeveloped and many of the recommendations made here are inevitably similar to those which
would be made in the context of any developing field. However, wherever possible the specific
significance of these recommendations for this area will be emphasized.
5.4.1
For research
One of the most pressing research issues is the need for longitudinal designs which can start, at
least, to unpack the causal relationships between emotional and other work demands and
outcomes such as burnout. Of course this is a highly significant weakness of nearly all research
on psychosocial job conditions and well-being. However, it seems particularly significant here
as burnout appears, by definition, to be a process that unfolds over time.
On the one hand, burnout appears to be a useful idea in that it promises to shed light on a more
specific reaction to particular kinds of job demands. On the other hand, from both the literature
and the results found here, it seems that burnout does vary across occupations in the ways one
would expect and nor is it strongly related by emotional demands after controlling for
personality and traditional job stressors. Taken together, this suggests that burnout, or at least
burnout as assessed by the most widely used measures, is not necessarily a useful indicator of
more specific kinds of harm in particular occupations or as a consequence of more specific
emotional work demands. The construct validity of burnout should be examined more closely
and, at the same time, other ways of thinking about and assessing the specific forms of harm
that may result from emotional labour should be considered. In any case, it would seem
sensible when exploring how emotional demands affect outcomes such as burnout, to control
for other sorts of work demands which may also be affecting the outcomes in order to identify
the unique contribution of emotional demands over and above those that are already known to
relate to outcomes such as burnout.
From both the literature review and the research conducted here it seems reasonable to conclude
that further work is required to unpack what emotional labour is and how it works. Is it best
considered as a set of job demands or job requirements? If so, which dimensions best capture
the phenomenon? Perhaps it is not emotional job requirements as such but reactions to those
requirements, such as emotional dissonance, which are at the core of emotional labour? The
qualitative research conducted here suggests some further possibilities. Exploring emotional
95
labour as a series of events or encounters with clients that are usually embedded in episodes
seems like a potentially helpful way of thinking through what emotional labour might entail and
therefore how it may operate as a dynamic process. This perspective is also one that is
attracting interest in the wider literature on emotional and motivational processes in the
workplace (e.g., Beal, Weiss, Barros & MacDermidm 2005; Seo, Feldman Barrett & Bartunek,
2004).
What is clear from the literature review and both studies is the importance of context, in many
senses, on how emotional labour is experienced and how it may affect burnout and other aspects
of well-being. While the nature and effect of any sort of work demand is likely to be very
sensitive to context this appears to be particularly the case for emotional demands. Rather than
try to look for consistencies and commonalties across jobs and occupations in emotional labour
research it may be more fruitful to explore why and how emotional demands play out in
different and diverse ways across situations and contexts.
5.4.2
For practice
Given the quite limited level of knowledge currently available from the literature it is unwise
and simply not possible to go beyond these findings to offer practical suggestions for how
emotional demands may be better managed. In addition, there are so few studies of
interventions it is also not possible on the basis of this evidence to make recommendations for
specific techniques. However, the findings from both the literature review and two studies do
give some suggestions for ways of thinking about practice in this field.
First, picking up on the final point made in the research implications section above, the nature
of emotional labour and its effects on well-being seem to be quite context-specific. Hence, any
practical intervention needs to be based on a detailed assessment of the particular context. It
would also therefore be inappropriate to make general recommendations, which were applied
across all jobs, for practice around emotional labour.
Second, the suggestion that emotional labour can have very positive and very negative features
means that it should not be treated as a traditional ‘stressor’ or as something to be simply
controlled or reduced. Rather, it may be a case of enhancing the positive features of emotional
labour work while at the same time, if possible, finding ways of better managing or reducing the
negative features.
Third, both the quantitative and qualitative results suggest that the emotional demands of work
should be considered alongside rather than independently of other more traditional job
characteristics. It may be that traditional job demands rather than emotional demands alone
may be exerting a large influence on well-being and that the impact of emotional demands on
well-being is in various ways affected by other characteristics of the job.
Last, while emotional demands seem to be a feature of many jobs, this aspect of work does not
appear to preoccupy practitioners who tend to focus on other sorts of job features. If numerous
commentators are correct in asserting that the level of emotional demands experienced in the
workplace are likely to continue to increase then this may now be the right time for practitioners
to start thinking how their work in areas such as job design and well-being interventions can
start to incorporate this important feature of many contemporary jobs both within and beyond
human and customer service work.
96
6 APPENDICES
6.1
APPENDIX 1: MEASURES USED IN THE QUESTIONNAIRE SURVEY
Section 1
This section contains questions about basic demographic characteristics of respondents
including gender (one item, dummy coded), age (one item), marital status (one item, dummy
coded) and whether the individual has children living at home (one item, dummy coded).
Basic background job information is also collected in this section. Respondents are asked to
select the job title that most closely describes their job, from a list which is tailored for each
organisation. They are also asked to record, to the nearest year, how long they have worked in
their current job (one item) and in their current occupation (one item).
Section 2
The second section of the survey contains items measuring traditional job stressors, including
role overload/job demands, control/autonomy, role conflict and ambiguity, and social support.
In selecting the measures for this section, we looked for scales that were as short as possible
whilst still clearly measuring the variables in question (validity) and demonstrating good
reliabilities. This was in order to keep the overall length of the survey down to a reasonable
level and to allow space to focus on the core variables of interest, i.e. emotional demands and
psychological strain.
Details of individual measures are as follows:
Role overload/job demands (3 items)
This measure is taken from the Michigan Organizational Assessment Questionnaire (Cammann,
Fichman, Jenkins & Klesh, 1979). This measure was originally used with a 7-point response
scale, but a 5-point scale (1 = strongly disagree, 5 = strongly agree) was used here to fit with the
other measures in this section. An example item is, “I have too much work to do to do
everything well”.
Control/autonomy (4 items) This measure is from Troyer et al (2000), based on earlier
scales used by Price and Mueller (1986). Responses are on a 5-point scale, as above. Troyer et
al report the alpha reliability of the scale as .85. An example item is, “I have adequate freedom
as to how to do my job”.
Role conflict (3 items) and role clarity/ambiguity (4 items)
These measures are also from Troyer et al (2000), based on earlier scales used by Price and
Mueller (1986). Responses are on a 5-point scale, as above. Troyer et al report the alpha
reliability of the role conflict scale as .79 and the role clarity/ambiguity scale as .76. An
example item from the role conflict scale is, “Sometimes I feel I have to fight the organization
to get my job done”, and from the role clarity scale is, “I have been clearly told what is
important about my job”.
97
Social support (4 items)
This measure is from the Whitehall II studies (Stansfeld, Head & Marmot, 2000): two items
measure social support from colleagues and two items measure support from supervisors. Two
items from the original measure concerning ‘information support’ were excluded as not central
to the concept of social support. All questions are answered on a four point scale assessing how
often each type of support is available to the employee (1 = often, 4 = never/almost never). An
example item is, “How often do you get help and support from colleagues?”
Section 3
The third section of the survey addresses the emotional demands of the job, specifically
measuring how often respondents engage in surface acting and deep acting, to what extent they
are required to display positive emotions or suppress displays of negative emotions as part of
their job, and how often they are exposed to aggression from the public.
We did not include a specific emotional dissonance measure, as there is a good deal of concept
overlap between emotional dissonance and surface acting, to the extent that some of the items
used by different scales to measure each of these concepts are very similar. The measures of
emotional demands used here (surface and deep acting, requirement to display positive and hide
negative emotions) are the same as those used by Brotheridge and Grandey (2002).
Details of individual measures are as follows:
Surface acting (3 items) and deep acting (3 items)
These measures are taken from the Emotional Labour Scale (ELS: Brotheridge & Lee, 1998;
2003). Surface acting items refer to modifying and faking expressions of emotions (e.g.
“Pretend to have emotions that I don’t really have”) and deep acting items tap the extent to
which the employee modifies their feelings as part of the work role (e.g. “Really try to feel the
emotions I have to show as part of my job”). Respondents are asked to rate how often each of
the statements applies to them on a five point scale (1 = never, 5 = always).
Brotheridge and Grandey (2002) report the alpha reliability of the surface acting items as .74
and the alpha reliability of the deep acting measure as .83. Brotheridge and Lee (2003) report
the alpha reliability of the surface acting items as .79 and the alpha reliability of the deep acting
measure as .83. Analyses show that the items form two independent but related factors
(Brotheridge & Lee, 1998). In terms of validity, Brotheridge and Lee (2003) found that surface
acting was significantly correlated with both emotional exhaustion and depersonalization
dimensions of burnout and significantly negatively correlated with personal accomplishment,
whereas deep acting was correlated only with personal accomplishment, with which it was
positively associated.
Requirement to display positive emotions (4 items) and requirement to hide
negative emotions (3 items)
These measures are taken from the Emotion Work Requirement Scale (EWRS: Best, Downey &
Jones, 1997), as used in Brotheridge and Grandey (2002). Respondents are asked to rate how
often their job requires them to display positive emotions (e.g. “Reassuring people who are
distressed or upset”) or hide negative emotions (e.g. “Hiding your disgust over something
someone has said or done”) on a five point response scale (1 = not at all, 5 = always).
The wording was changed slightly on some items, either to make them clearer or to make them
more widely applicable to the diverse employee sample being surveyed. For example,
considering items measuring the requirement to hide negative emotions, one item was changed
98
from, “Remaining calm even when you are astonished” to “Remaining calm even when you are
not”, and another was changed from “Hiding your fear of someone who appears threatening” to
“Hiding your fear of someone who appears or sounds threatening” (because part of the survey
sample work in call centres and interact with customers only over the telephone).
Brotheridge and Grandey (2002) report the alpha reliability of the requirement to display
positive emotions measure as .78 and the alpha reliability of the requirement to hide negative
emotions scale as .77.
Brotheridge and Lee (2003) found modest correlations between the EWRS and the ELS
subscales, indicating some degree of overlap between the measures, but argued that these
correlations were sufficiently small that divergent validity was retained. Specifically, the
requirement to hide ones emotions was more strongly associated with surface acting than deep
acting, and the requirement to show emotional support was more strongly associated with deep
acting than surface acting.
Exposure to aggression from the public (1 item)
Only a few papers have attempted to measure exposure to aggression from patients or
customers, and there are no established measures of this. The measure used here was a single
item adapted from Winstanley and Whittington (2002), who looked at the effects of exposure to
aggression from patients and their families on burnout in hospital staff. Winstanley and
Whittington looked at physical as well verbal aggression, but only the question relating to
verbal aggression was used here, due to the nature of the employee sample (i.e. including call
centre employees).
Physical aggression was also relatively rare in Winstanley and
Whittington’s study.
The wording of the original question was adapted as necessary for different employee groups,
i.e. the question referred to aggression ‘from patients or their relatives and friends’ for
healthcare workers, but ‘from the public’ for customer service workers, etc.
The response options were also changed. The original paper had only two options ‘frequent
(monthly or more frequent)’ and ‘infrequent (less than monthly or never)’. This was amended
to offer a more fine-grained analysis, with a 5-point response scale (1 = never, 5 = most days).
The high frequency options were included as studies of customer service (e.g. Grandey, Dickter
& Sin, 2004) have indicated that customer service workers in call centres are often exposed to
verbal aggression on a daily basis.
Section 4
The fourth section of the survey measures respondent burnout, job satisfaction and intention to
quit. These are described collectively to respondents as questions concerning how they feel
about their jobs. Although burnout might theoretically sit more naturally with other measures of
psychological strain (e.g. the GHQ, see next section), the Maslach Burnout Inventory manual
(Maslach et al , 1996) advises presenting the measure to respondents as a survey of job-related
attitudes, to minimise the effects of people’s beliefs and assumptions about burnout.
Details of individual measures in this section are as follows:
Burnout: Emotional exhaustion (5 items), cynicism (5 items), professional efficacy (6
items)
Burnout was assessed using the Maslach Burnout Inventory (MBI), which is the most widely
used measure of burnout. The original MBI was designed specifically for use with human
99
services professionals, and a second version was adapted for use with educators. The third and
most recent adaptation of the MBI is the General Survey, designed for use with workers in a
variety of occupations (MBI-GS: Schaufeli, Leiter, Maslach & Jackson, 1996), and this version
was used in the current study due to the diversity of the occupational groups to be sampled. The
MBI-GS measures burnout on three dimensions of emotional exhaustion, cynicism (referred to
as depersonalization in the literature review) and professional efficacy (referred to as
accomplishment in the literature review), and the three factor structure has been confirmed
across occupational samples and different nationalities. High scores on emotional exhaustion
and cynicism and low scores on professional efficacy are indicative of burnout. All items are
scored on a 7-point frequency rating scale (0 = never, 6 = every day).
Maslach et al (1996) report the alpha reliability of the emotional exhaustion subscale of the
MBI-GS as .87 to .89, the alpha reliability of the cynicism subscale as .73 to .84, and the alpha
reliability of the professional efficacy subscale as .76 to .84.
Job satisfaction (3 items)
Job satisfaction was assessed using a measure from Warr, Butcher, Robertson and Callinan
(2004). This measure of job satisfaction was chosen largely because it attempts to tap into jobrelated affect, rather than being a primarily cognitive evaluative judgement about ones job, as
are many job satisfaction measures. Items are scored on a 5-point response scale. Warr et al
(2004) reported the alpha coefficient of internal reliability for this measure was .94.
Intention to quit (1 item)
Intention to quit was measured with a single item, taken from Spector and Jex (1991). Although
this is a single-item measure, Spector and Jex demonstrated its concurrent and predictive
validity. Participants are asked to respond on a 5-point scale.
Qualitative questions about stressful and satisfying aspects of work (2 items)
Respondents were also asked two open-ended questions in this section, requiring them to
describe in their own words the most stressful aspect of their job and the most satisfying aspect
of their job. These items are very similar to those used by Lewig and Dollard (2003), who
found these open-ended questions were a useful supplement to their quantitative analysis.
Section 5
The fifth section of the survey measures psychological strain using the GHQ12.
General psychological well-being (12 items)
This section utilises the General Health Questionnaire (GHQ12: Banks, Clegg, Jackson, Kemp,
Stafford and Wall, 1980). The GHQ was developed as a self-administered screening test for
detecting minor psychiatric disorder (depression/anxiety) in community settings. The original
measure comprised 60 items (GHQ60: Goldberg, 1972), and versions have since been
developed with 30, 20 and 12 items. Banks et al (1980) state that the GHQ12 is the most
relevant version in occupational studies where mental health is only one of several variables
being measured, which was the case in the current study. Items are rated on four point response
scales, with the wordings as specified in the original 60-item measure.
Two different scoring methods have been used with the GHQ: the ‘GHQ-method’ and the
‘Likert-method’. The GHQ-method scores each answer as either 0, if one of the first two
response options is selected or 1, if either of the latter two response options is selected; this is
primarily a binary measure, identifying individuals as ‘cases’ or ‘normals’. The Likert-method
scores each item 0, 1, 2 or 3 and produces a wider and less skewed distribution of scores more
100
appropriate for most parametric statistics, including regression analyses (Banks et al, 1980).
The Likert-method was, therefore, used in this study. Banks et al (1980) report alphacoefficients in the range .82 – .90.
Section 6
The sixth and final section of the survey measures affective personality.
Trait positive affectivity (10 items) and trait negative affectivity (10 items)
Both positive and negative affectivity were measured using the Positive and Negative Affect
Schedule (PANAS: Watson, Clark and Tellegen, 1988). This is the most frequently used
measure of trait affectivity in the occupational psychology literature and has been well
validated.
The PANAS measure can be used as a measure of state or trait affect with various wordings of
the instructions, reflecting different time periods. Here people were asked to rate how they feel
generally, which is the best measure of trait affect.
Watson, Clark and Tellegen (1988) report alpha reliabilities of .88 and .87 for positive and
negative affectivity respectively, using this version of the measure. Test-retest reliability is
reported as .68 (positive affectivity) and .71 (negative affectivity). Positive and negative
affectivity show relatively low inter-correlation (-.17), indicating that the two variables are
reasonably independent of one another.
101
6.2
APPENDIX 2 : DAILY DIARY
102
Health and Safety Executive/
University of London
WORK AND WELL-BEING DIARY
This diary is about your work and your well-being. We are asking you to complete it in order to
help us to understand how events at work affect your overall well-being.
At some point during the evening please think back over the things that you did and the things
that happened over the previous 24 hours and answer the following questions. Filling in this
diary should probably take about 15 minutes - but please feel free to spend more or less time on
it as you wish.
Please note that some questions within the diary use the term ‘client’. By ‘client’ we mean a
student, resident, child, patient or any other member of the public with whom you interact as
part of your job. In this context, ‘client’ would also include a friend or relative of a member of
these groups.
It is not necessary to answer all of the questions each day: If you cannot think of anything to say
in response to a particular question, just leave it blank.
The completed diary will be considered confidential.
Research participant’s code:
Today’s date and time (i.e. at what time is this diary being completed?):
Is today a work day, weekend or day of annual leave? (Please click)
Work day
Weekend
Day of annual leave
Other:
SECTION ONE
1. Firstly, please note down briefly the main things that happened to you or that you did over the
last 24 hours. Please include both work and non-work events and activities.
a. Yesterday evening/last night
(Please start typing here)
b. This morning
(Please start typing here)
c. This afternoon
103
(Please start typing here)
d. So far this evening
(Please start typing here)
104
SECTION TWO – Your emotions today
2. Secondly, how have you felt today? Listed below are a number of emotions, some of them
positive (e.g. happy, confident) and some negative (e.g. angry, disappointed). If you can
remember feeling any of the following emotions and feelings (positive, negative or both) please
click the box:
happy
angry
confident
empathetic
cheerful
sad
upset
burned out
discouraged
disinterested
stressed
excited
guilty
worried
ashamed
scared
nervous
disappointed
proud
calm
frustrated
emotionally
optimistic
drained
irritated
hassled
cynical
helpless
pessimistic
hurt
SECTION THREE – Today’s work events
In this part of the diary we would like you to think about events, either at work today or to do
with work, that made you feel either positive or negative emotions. We are interested in both
events that involved interaction with a client and events that did not involve client interaction.
For each of the boxes below, please consider whether you experienced such an event today. If
there was such an event in the day, please tell us about it, giving as much detail as possible. In
particular, we would like to know who was involved (e.g. a client, a colleague, no-one else),
what happened (including where and when it happened, if relevant), how you felt (e.g. what
emotions did you feel and how strongly did you feel them) and why you think you felt this way.
If you cannot think of any relevant event for a particular box, please leave it blank. You can also
tell us about more than one event, if you wish.
3a. An interaction with a client that made you feel positively
(please start typing here)
105
3b. An interaction with a client that made you feel negatively
(please start typing here)
3c. A work-related event that made you feel positively and that did not involve a client
interaction
(please start typing here)
3d. A work-related event that made you feel negatively and that did not involve a client
interaction
(please start typing here)
106
SECTION FOUR – At home
In this part of the diary we would like you to think about events that have happened at home or
to do with your personal life that have made you feel either better or worse overall.
Please consider whether you experienced any such events today. If so, please tell us about them,
giving as much detail as possible. In particular, we would like to know who was involved (e.g.
your partner, your children, no-one else), what happened (including where and when it
happened, if relevant), how you felt (e.g. what emotions did you feel and how strongly did you
feel them) and why you think you felt this way.
If you cannot think of any relevant events, please leave this section blank. You can also tell us
about more than one event, if you wish.
4. A home-related event that made you feel either better or worse, overall
(please start typing here)
Thank you for filling out this diary today.
If you have any queries or comments, please contact any of the researchers: Dr Sarah
Poppleton (tel: XXXXX) or email: XXXXX; Dr Tina Kiefer (email: XXXXX) or Professor Rob
Briner (tel: XXXXX or email: XXXXX).
107
6.3
APPENDIX 3: INTERVIEW SCHEDULE TOPIC GUIDE
108
TOPIC GUIDE FOR HSE INTERVIEWS ABOUT WORK EVENTS (DIARIST 11)
Introduction
Study is about how daily events at work affect our well-being, with a particular emphasis on our
emotional responses to events. The study will be written up into a report for the HSE.
Purposes of interview: To clarify and perhaps enlarge on some of the diary entries.
Confidential.
Anonymity will be preserved in the report (and any other written or presentation materials
arising from the project). Long quote – ask permission.
Tape recorded for analysis purposes.
Background
• How long as a head teacher?
• At this school? At other schools?
• Other career information (e.g. have you always been in teaching?)
• Can you tell me a bit about the type of school this is? (compare with previous schools?)
• 1b and 3b. Liam. Speaking to teacher ‘trying to get her on board to empathise with
need’. Teacher as focus?
• The SEF? Paperwork ‘sword of damacles’
DAY 2
• 1b.TA – confrontational – need to talk about manner
• 3b and 3d. Yr 3 child returned from exclusion. Frustration with SS ‘the play is so
cautious’?
DAY 3
•
•
•
•
Just section one. Very busy.
Caretaker’s future.
Critical Child Protection issue.
Removing child with behavioural problems.
DAY 4
• Sect 1 only: Philosophy group. Not to attached to role, separate one task…  needed?
Why?
• Another really challenging day. CP referral. Time with children in difficulties…(more?
E.g. Having to deal with the children’s emotions…)
109
• Calm and cheerful?
DAY 5
• 1b: Illegal substance.
• 3a. Modelling a lesson. Also positive one to one (parent and child learned). Lots of
positive interaction with children – immediacy, rewarding.
• 3b. Teacher with toothache (poor peforming  procedure) ‘children are not getting a
good deal’ – but uncomfortable although ‘keeping needs of children in mind makes it
tolerable’
• 3c. Consultant leadership programme – again modelling.
• 3d. Paperwork. And the contrast with the interactions with the children (3a – ‘a lot of
the role…’)
DAY 8
• 1b. Difficult day. ‘Just kept the place going…’
• More negative emotions than previous diaries.
• 3a. Meeting with teachers. Changed their feelings – managing their emotions
(annoyance and rejection)
• 3b. Red cards for children. Some problems with teachers handling of situations.
• Emotionally drained? (volume of problems etc)
• Reference to not talking about work at night. Reasoning? (compartmentalisation? Not
internalising?) Link to Day 13 – member of staff rang…didn’t return …importance of
weekends)
DAY 9
•
•
•
•
1b. Very difficult morning with y5 class, distressed teacher sent home.
1c. Row between parents.
Again, negative emotions.
3b. More analysis of the above. Didn’t really achieve what Liz wanted to achieve with
the mother (accepting son’s responsibility). Row ‘depressing examples’.
• Paperwork.
• 1d. ‘Don’t feel well disposed…need to find a quiet space (mental? In time? Place?)
(Probe: feeling that emotions might hinder doing a good job or affect meeting?) E.
Labour.
DAY 10
• 1b and c.
• 3b. Speaking to the two parents (same as day 9) awful having to speak to them…in
what way? Felt awful? Shouldn’t happen?
DAYS 13 AND 14
• Phone call and paperwork.
Themes
• Empathetic – often with children. Teaching others to be empathetic?
• Modelling behaviour (onerous - feel inside?)
110
•
•
•
Importance of saying and doing the right thing for the children.
Compartmentalization. Space. Meditation ie. coping methods.
Few negative emotions in spite of circumstances.
111
6.4
APPENDIX 4: THE DIARY CODING TEMPLATE
112
CODING TEMPLATE FOR HSE DIARIES – VERSION TWO
Codes: D = Descriptive Code i.e. theme of event (1 = positive client interaction, 2 = other
positive work event, 3 = negative client interaction, 4 = other negative work event, 5 =
positive non-work, 6 = negative non-work).
M = Meaning Code i.e. meaning attributed, including reasons for seeing it as
positive/negative and reasons for emotions.
3A. POSITIVE CLIENT INTERACTION
Code
Descriptive Code
Meaning Code
D1.1
Enjoyable client contact
•
•
•
•
•
D1.2
Client contact
feedback
with
positive
•
•
•
•
•
D1.3
Positive meeting/Interaction with
client
•
Success/Going well – self as central
Most important part of the job/Meaning of
the job
Helping client
Seeing through client’s viewpoint/Empathy
with their feelings
Good relationship with client (self) or
amongst clients
Client expressing gratitude or appreciation
Client’s praise/+ve feedback re. self
Turning a situation around
Offset against the demands/Difficulties of
the job
Show of affection from a client
•
•
•
Effective management of a
problem/situation by (a) self and/or (b)
organization
Engendering (more) positive feelings in a
client as a result of actions or words (+ve
client feelings)
Better than expected
Optimism about a good outcome
Using specialist helping skills
•
D1.4
Getting positive action (from a
third party) to help resolve a client’s
situation/Problem
•
Helping and effective management above
D1.5
Employee/subordinate’s successful
performance
•
•
Having a positive impact on client’s
development /Progress/Learning
Giving positive feedback
D1.6
Client making progress towards a
goal
•
•
Empathy with client’s positive feelings
+ve client feelings
D1.7
Absence of ‘hassle’ from clients or
•
A good working atmosphere/Positive mood
113
absence of problems with clients
D1.8
amongst clients and/or colleagues
•
Irate client
+ve client feelings and turning a situation
around
3B. OTHER POSITIVE WORK EVENT
Code
Descriptive Code
Meaning Code
D2.1
Recruiting someone in difficult
circumstances
•
Valuable person
D2.2
Positive feedback (about self) from
(a) a colleague or (b) a superior
•
•
Successful leadership/Management
Giving support
D2.3
Positive interaction with a
representative(s) from a third
party/outside organization
•
Working towards making a positive change
in the organization
Feelings associated with status and
privilege
Feelings associated with community and cooperation
Social support (outside)
Also success, positive feedback and
leadership
•
•
•
•
D2.4
Making good progress with, or
completing (a) admin task/list of
admin tasks (b) another work
task/tasks
•
•
•
Success with the bureaucratic process
Offset against sense of this ‘hanging over’
you
Tasks which enable others
•
Better
D2.5
Sorting out problems
D2.6
Client’s or
Achievement
D2.7
Attending a course or conference
•
Excitement over future opportunities
D2.8
Positive feelings about working
with a colleague/colleagues on a
project (working w. colleague +ve)
•
Pride
D2.9
Enjoyable
and/or
productive
interaction with colleague(s), either
(a) work-related (e.g. meetings) or
(b) social (e.g. lunch) (+ve
colleague interaction)
•
•
•
Success
Leadership
Morale (+ve)
D2.10
Success with IT
•
Feelings of competence
clients’
success/
114
•
D2.11
Receiving support from (a) a
supervisor or (b) someone else in
the organization (social support)
D2.12
Being physically removed from
work ( absence of hassle from
clients)
D2.13
Good news about a client
D2.14
Personal,
achievement
Success
work-related
3C. NEGATIVE CLIENT INTERACTION
Code
Descriptive Code
Meaning Code
D3.1
Client who has been or is going
through a negative experience
•
•
•
•
Compassion/Sympathy for client’s
predicament
Empathy with client
Concern over a client’s welfare
Concern about ‘failing’ the client
•
Disagreement with client
D3.3
Client’s choice of action as wrong
or inappropriate
Problem with communication
•
D3.4
Difficult client(s)
•
•
Discomfort with the process of
communication
Number of or volume of
Not being listened to or having your actions
questioned in your professional role
Making demands on valued time
Questioning the necessity of your
involvement (cf. someone further down the
hierarchy)
Handling
Client welfare
Staff welfare
Fear re. future events (e.g. escalation of
problem(s)
Also complaint higher
D3.2
•
•
•
•
•
•
•
D3.5
D3.6
Verbal abuse from a client(s),
either (a) face to face or (b) over
the phone
Client’s complaint
official process
about
an
•
•
Threat of client complaining to someone
above you in the organizational hierarchy
(complaint higher)
Indirect (aimed at staff/colleague)
•
Client’s lack of understanding
115
D3.7
Teaching/modelling behaviour to
a subordinate, to benefit a client(s)
•
•
D3.8
Client who cannot cope in a
situation
•
•
+ve (e.g. modelling specific skills)
-ve (e.g. feeling you have to model empathy
when it is lacking)
Frustration over another agency’s poor
handling of and/or lack of progress with a
client
Concern over lack of resources or capacity
to help (organization or other agency)
D3.9
Client complaining about your
handling of a situation (client
complaint – self)
•
•
Complaint higher and social support.
Internalisation and/or self-doubt
D3.10
Concern that client won’t get the
help he/she needs (concern re.
help)
•
•
•
•
Because of the system
Because of client’s behaviour
Also self doubt, unable to turn situation
round
Helpless
•
•
•
Having to reprimand
Having to exclude
Concern over staff welfare
D3.11
Dealing
with
client’s
behaviour (child)
poor
D3.12
Questioning own action in a
situation
D3.13
Client contact lacking
feedback (but not negative)
D3.14
Client’s over-dependence
D3.15
Feeling
unsupported
colleagues/Superior
D3.16
Missing client contact
+ve
by
3D. OTHER NEGATIVE EVENT
Code
Descriptive Code
Meaning Code
D4.1
Difficulties with staffing
•
Feeling unable to turn a situation round
D4.2
Concern
over
employee/Subordinate’s general
attitude or level of knowledge
D4.3
Concern
over
employee/subordinate’s
performance on task or handling
•
•
Affecting client
Affecting organization/other staff
(including self)
116
D4.4
of a situation
•
Also self doubt
Admin/Paperwork
•
•
Too much
Unnecessary/generated by
bureaucracy/waste of anyone’s time
Waste of your time (e.g. caused by staff
ineptitude or lack of anyone else to do it)
Relief from more important/consequential
activities
Hindered by interruptions
Lack of progress/target or deadline unmet
Tiring
Having to take it home
Extra to usual
•
•
•
•
•
•
•
D4.5
Problem
with
working
environment e.g. premises, IT,
organizational property
•
•
•
•
•
Sense of responsibility for client welfare
Concern for health, safety or welfare of staff
(including self)
Unnecessary/Bureaucracy
Someone’s fault
Lack of confidence
D4.6
Being away from the organization
> unable to deal with the problems
or dealing with them at arm’s
length
D4.7
Dwelling on an issue
•
•
•
•
Self doubt
Client related
Career related
Other work related
D4.8
Colleague’s problem(s)/needs
•
•
Difficulty listening and/or helping
Morale
D4.9
Training which isn’t very useful
•
•
Anticipating
Experiencing
D4.10
Time-wasting meeting
•
•
Unproductive
Unnecessary
D4.11
Difficult
interaction
with
colleague(s)
(-ve
colleague
interaction) or difficult aspect of
colleague interaction
Tension over roles and responsibilities
Negative reflection on self (e.g. --> feeling
undermined, under suspicion, not accepted,
unimportant)
D4.12
Negative interaction with third Unnecessary and lack confidence
party
Handling
Being let down
Having to fit too much into
working day
D4.13
117
D4.14
Busy and/or hard work time
Anticipating
D4.15
Out of hours work
D4.16
Disagreement with organizational
policy
D4.17
Boredom at work
4. NON-WORK EVENT: BETTER
Code
Descriptive Code
Meaning Code
D5.1
General, positive feelings about
home/being with family
•
Safety, security
D5.2
Completion of domestic project
D5.3
Own child’s success
D5.4
Holiday/Travel plans
•
Lack of progress (negative), progress
(positive)
D5.5
Show of affection from family
member or friend
Good time with friends
•
•
Changing mood/re-energising
Safety/security
•
•
Emotional
Practical
D5.6
D5.7
Support from a family member
(practical and/or emotional)
D5.8
Participating in a leisure activity or •
outing
•
Changing mood
Success/self
D5.9
D5.10
Catching up on sleep
Quality non-work time
purposefully not working
Interruption
D5.11
+ve health behaviours
D5.12
Lessening of worry about family
member
Work blocking out health-related
anxieties
Good news re. family member
D5.13
D5.14
–
•
4. NON-WORK EVENT: WORSE
Code
Descriptive Code
D6.1
A worry about your child
Meaning Code
118
D6.2
Unfinished domestic task(s)
•
Interruptions
D6.3
Health problem, either (a) self or
(b) friend or family member
Forgetting/Failing
to
do
a
personal
task/attend
personal
matters in the day because of work
•
Effortful emotion work
•
Personal life suffers because of work/ work
negatively affecting family or personal life (ve effect personal life)
D6.5
Tension/Argument with a family
member
•
•
Negative personal life
Not work related
D6.6
Needing to rush to fit in domestic
matters
Missing out on contact with
family members or friends/
missing family
Sad family occasion
•
Negative personal life
•
•
Negative personal life
Not work related
D6.4
D6.7
D6.8
D6.9
D6.10
Concern
over
mood/
psychological state of family
member/friend
Negative health behaviours
D6.11
Disappointing social event
D6.12
General negative feelings re.
home e.g. bored, tired, demotivated
Other domestic worry
D6.13
119
6.5
APPENDIX 5: ADDITIONAL INTERVIEW CODES
120
ADDITIONAL CODES FOR THE INTERVIEWS
CLIENT RELATED
Emotionally demanding situations
•
•
•
•
•
•
•
•
•
•
•
Client’s pressure
Boundaries/Clarity over job and responsibility
Tug of war
Detachment – not taking on people’s anxieties
Role of ‘expert’ and ‘helper’
Depletion
Responsibility
Supporting staff
Urgency
Low expectations
Success
Emotion work interactions
•
•
•
•
•
•
•
•
•
•
•
Acting
Emotional dissonance
Role model, self awareness and reflexivity
Anger
Guilt
Personality
Effects
Emotional roller coaster – events --> large ups and downs, taking you with them.
Cynicism
Success
Managing the emotions of staff (protective?)
Fear of failing client
•
•
•
•
Concern about conveying the clients needs effectively
Concern about doing the right thing for the client
Tension between client’s wants versus needs
Tension between family’s demands versus client’s feelings
Concern that client won’t get the help he needs
•
•
•
Lack of skill and other demands on teacher
Lack of experience
Lack of understanding own emotions
Questioning own actions in a situation
•
•
Right thing
Reflecting on own/organization’s actions
121
Middle class parents
•
•
Actions questioned in professional role
Litigation
Difficulty withdrawing
•
Handling being a carer in other situations (e.g. family, colleagues)
Coping
•
•
•
•
•
•
•
•
•
•
Boundaries
Not over-promising ( personal responsibility)
Training
Experience
Self talk
Talk/Counselling
Exercise
Flexibility/Flexible hours
Speaking out about feelings
Pause/Separation
NOT CLIENT RELATED
Admin/Paperwork
•
•
•
•
Confidence
Unrewarding, thankless
Range of responsibilities
After the children have gone
Busy, hard work, time
•
•
Meaning of burnout in diary
Emergencies/Urgency
Health problem – Self or family member
•
Affecting how you feel about your clients – reduced sympathy
Effect of work on home life
•
•
•
Avoiding confrontation, disliking conflict
Combination of work and non-work --> Burnout? Depression?
Parallels in work and non-work life
Taking things personally
•
Professional detachment
Difficult interaction with colleague/staff
•
Discomfort/Putting off
122
•
•
Lack of support
Racism
123
7 REFERENCES
Abraham, R. (1998). Emotional dissonance in organizations: Antecedents, consequences and
moderators. Genetic, Social and General Psychology Monographs, 124, 229-246.
Aiken, L.H. & Sloane, D.M. (1997). Effects of organizational innovations in AIDS care on
burnout among urban hospital nurses. Work and Occupations, 24(4), 453-477.
Alexander, D.A. & Klein, S. (2001). Ambulance personnel and critical incidents – Impact of
accident and emergency work on mental health and emotional well-being. British Journal of
Psychiatry, 178, 76-81.
Ashforth, B.E. & Humphrey, R.H. (1993). Emotional labor in service roles: The influence of
identity. Academy of Management Review, 18, 88-115.
Bakker, A.B., Killmer, C.H., Siegrist, J. & Schaufeli, W.B. (2000a). Effort-reward imbalance
and burnout among nurses. Journal of Advanced Nursing, 31(4), 884-891.
Bakker, A.B., Schaufeli, W.B., Sixma, H.J., Bosveld, W. & Van Dierendonck, D. (2000b).
Patient demands, lack of reciprocity, and burnout: A five-year longitudinal study among general
practitioners. Journal of Organizational Behavior, 21(4), 425-441.
Banks, M.H., Clegg, C.W., Jackson, P.R., Kemp, N.J., Stafford, E.M. & Wall, T.D. (1980). The
use of the General Health Questionnaire as an indicator of mental health in occupational studies.
Journal of Occupational Psychology, 53, 187-194.
Barber, C. E. & Iwai, M. (1996). Role conflict and role ambiguity as predictors of burnout
among staff caring for elderly dementia patients. Journal of Gerontological Social Work, 26(12), 101-116.
Beal, D.J., Weiss, H.M., Barros, E. & MacDermid, S.M. (2005). An Episodic Process Model of
Affective Influences on Performance. Journal of Applied Psychology, 90, 1054-1068.
Beehr, T. A., Jex, S. M., Stacy, B. A. & Murray, M. A. (2000). Work stressors and coworker
support as predictors of individual strain and job performance. Journal of Organizational
Behavior, 21(4), 391-405.
Ben-Zur, H. & Yagil, D. (2005). The relationship between empowerment, aggressive
behaviours of customers, coping and burnout. European Journal of Work and Organizational
Psychology, 14(1), 81-100.
Best, R.G., Downey, R.G. & Jones, R.G. (1997, April). Incumbent perceptions of emotional
work requirements. Paper presented at the 12th annual conference of the Society for Industrial
and Organizational Psychology, St. Louis, MO.
Bolton, S.C. (2000). Who cares? Offering emotion work as a ‘gift’ in the nursing labour
process. Journal of Advanced Nursing, 32(3), 580-586.
Bolton, S.C. & Boyd, C. (2003). Trolley dolly or skilled emotion manager? Moving on from
Hochschild’s managed heart. Work, Employment and Society, 17(2), 289-308.
124
Bond, F. W. & Bunce, D. (2003). The role of acceptance and job control in mental health, job
satisfaction, and work performance. Journal of Applied Psychology, 88(6), 1057-1067.
Boyd, C. (2002). Customer violence and employee health and safety. Work, Employment and
Society, 16(1), 151-169.
Briner, R. B. (1999). The neglect and importance of emotion at work. European Journal of
Work and Organizational Psychology, 8(3), 323-326.
Briner, R. B. & Reynolds, S. (1999). The costs, benefits, and limitations of organizational level
stress interventions. Journal of Organizational Behavior, 20(5), 647-664.
Briner, R. B. & Totterdell, P. (2002). The experience, expression and management of emotion at
work. In Warr, P. (ed) Psychology at Work (5th Edition). New York: Penguin Press.
Brotheridge, C.M. & Lee, R.T. (1998). On the dimensionality of emotional labour:
Development and validation of the Emotional Labour Scale. Paper presented at the First
Conference on Emotions in Organizational Life, San Diego.
Brotheridge, C.M. & Grandey, A.A. (2002). Emotional labor and burnout: Comparing two
perspectives of “people work”. Journal of Vocational Behavior, 60(1), 17-39.
Brotheridge, C.M. & Lee, R.T. (2003). Development and validation of the emotional labour
scale. Journal of Occupational and Organizational Psychology, 76, 365-379.
Büssing, A. & Glaser, J. (1999). Work stressors in nursing in the course of redesign:
Implications for burnout and interactional stress. European Journal of Work and Organizational
Psychology, 8(3), 401-427.
Büssing, A. & Glaser, J. (2000). Four-stage process model of the core factors of burnout: the
role of work stressors and work-related resources. Work and Stress, 14(4), 329-346.
Burisch, M. (2002). A longitudinal study of burnout: the relative importance of dispositions and
experiences. Work and Stress, 16(1), 1-17.
Burke, R.J. (1997). Toward an understanding of psychological burnout among police officers.
International Journal of Stress Management, 4(1), 13-27.
Burke, R.J. & Greenglass, E. (1995). A longitudinal study of psychological burnout in teachers.
Human Relations, 48(2), 187-202.
Cammann, C., Fichman, M., Jenkins, D. & Klesh, J. (1979). The Michigan Organizational
Assessment Questionnaire. University of Michigan, Ann Arbor, MI. In Cook, J.D., Hepworth,
S.J., Wall, T.D. & Warr, P.B. (1981). The Experience of Work. New York: Academic Press.
Carson, J., Maal, S., Roche, S., Fagin, L., De Villiers N., O’Malley, P., Brown, D., Leary, J. &
Holloway, F. (1999). Burnout in mental health nurses: Much ado about nothing? Stress
Medicine, 15(2), 127-134.
Coffey, M. & Coleman, M. (2001). The relationship between support and stress in forensic
community mental health nursing. Journal of Advanced Nursing, 34(3), 397-407.
125
Cordes, C.L., Dougherty, T.W. & Blum, M. (1997). Patterns of burnout among managers and
professionals: a comparison of models. Journal of Organizational Behavior, 18(6), 685-701.
Côté, S. (2005). A social interaction model of the effects of emotion regulation on work strain.
Academy of Management Review, 30(3), 509-530.
Côté, S. & Morgan, L. M. (2002). A longitudinal analysis of the association between emotion
regulation, job satisfaction, and intention to quit. Journal of Organizational Behavior, 23(8),
947-962.
Cranswick, K. (1997). Burnout: A study of levels of burnout and factors responsible for burnout
in rehabilitation workers. Journal of Rehabilitation Administration, 21(2), 119-134.
Cropley, M. & Millward Purvis, L.J. (2003). Job strain and rumination about work issues during
leisure time: A diary study. European Journal of Work and Organizational Psychology, 12(3),
195-207.
Deary, I.J., Watson, R. & Hogston, R. (2003). A longitudinal cohort study of burnout and
attrition in nursing students. Journal of Advanced Nursing, 43(1), 71-81.
Deery, S., Iverson, R. & Walsh, J. (2002) Work relationships in telephone call centres:
Understanding emotional exhaustion and employee withdrawal. Journal of Management
Studies, 39(4), 471-496.
De Jonge, J. & Schaufeli, W. B. (1998). Job characteristics and employee well-being: A test of
Warr’s vitamin model in health care workers using structural equation modelling. Journal of
Organizational Behavior, 19(4), 387-407.
De Rijk, A.E., LeBlanc, P.M., Schaufeli, W.B. & de Jonge, J. (1998). Active coping and need
for control as moderators of the job demand-control model Effects on burnout. Journal of
Occupational and Organizational Psychology, 71, 1-18.
Demerouti, E., Bakker, A.B. & Bulters, A.J. (2004). The loss spiral of work pressure, workhome interference and exhaustion: Reciprocal relations in a three-wave study. Journal of
Vocational Behavior, 64(1), 131-149.
Diefendorff, J.M. & Gosserand, R.H. (2003). Understanding the emotional labor process a
control theory perspective. Journal of Organizational Behavior, 24(8), 945-959.
Diefendorff, J.M. & Richard, E.M. (2003). Antecedents and consequences of emotional display
rule perceptions. Journal of Applied Psychology, 88(2), 284-294.
Dollard, M. F., Dormann, C., Boyd, C. M., Winefield, H. R. & Winefield, A. H. (2003). Unique
aspects of stress in human service work. Australian Psychologist, 38(2), 84-91.
Dollard, M.F., Winefield, H.R., Winefield, A.H. & de Jonge, J. (2000). Psychosocial job strain
and productivity in human service workers: A test of the demand-control-support model.
Journal Occupational and Organizational Psychology, 73(4), 501-510.
Dormann, C. & Zapf, D. (2004). Customer-related social stressors and burnout. Journal of
Occupational Health Psychology, 9(1), 61-82.
126
Duquette, A., Kerouac, S., Sandhu, B.K., & Beaudet, L. (1994). Factors related to nursing
burnout: a review of empirical knowledge. Issues in Mental Health Nursing, 15(4). 337-58.
Ekstedt, M. & Fagerberg, I. (2005). Lived experiences of the time preceding burnout. Journal
of Advanced Nursing, 49(1), 59-67.
Erickson, R.J. & Ritter, C. (2001). Emotional labor, burnout, and inauthenticity: Does gender
matter? Social Psychology Quarterly, 64(2), 146-163.
Erickson, R.J. & Wharton, A.S. (1997). Inauthenticity and depression – Assessing the
consequences of interactive service work. Work and Occupations, 24(2), 188-213.
Ewers, P., Bradshaw, T., McGovern, J. & Ewers, B. (2002). Does training in psychosocial
interventions reduce burnout rates in forensic nurses? Journal of Advanced Nursing, 37(5), 470476.
Fujiwara, K., Tsukishima, E., Tsutsumi, A., Kawakami, N. & Kishi, R. (2003). Interpersonal
conflict, social support, and burnout among home care workers in Japan. Journal of
Occupational Health, 45(5), 313-320.
Gerzina, M. A. & Drummond, P. D. (2000). A mulitmodal cognitive-behavioural approach to
anger reduction in an occupational sample. Journal of Occupational and Organizational
Psychology, 73(2), 181-194.
Giebels, E. & Janssen, O. (2005). Conflict stress and reduced well-being at work: The buffering
effect of third party help. European Journal of Work and Organizational Psychology, 14(2),
137-156.
Golembiewski, R. T., Boudreau, R. A., Munzenrider, R. F. & Luo, H.(1996). Global burnout: A
worldwide pandemic explored by the phase model. Monographs in organizational behavior and
industrial relations, Vol.21.Greenwich, CT: JAI Press.
Golembiewski, R. T., Munzenrider, R. F. & Stevenson, J. G. (1986). Phases of burnout:
Developments in concepts and applications. New York: Praeger.
Goldberg, D. (1972). The Detection of Psychiatric Illness by Questionnaire. London: Oxford
University Press.
Grandey, A.A. (2000). Emotion regulation in the workplace: a new way to conceptualize
emotional labor. Journal of Occupational Health Psychology, 5(1), 95-110.
Grandey, A. A. (2003). When “the show must go on”: Surface acting and deep acting as
determinants of emotional exhaustion and peer-rated service delivery. Academy of Management
Journal, 46(1), 86-96.
Grandey, A. A., Dickter, D. N. & Sin, H. P. (2004). The customer is not always right: Customer
aggression and emotion regulation of service employees. Journal of Organizational Behavior,
25(3), 397-418.
Greenglass, E.R., Burke, R.J. & Fiksenbaum, L. (2001). Workload and burnout in nurses.
Journal of Community Applied Sociology, 11(3), 211-215.
127
Greenglass, E.R. & Burke, R.J. (2002). Hospital restructuring and burnout. Journal of Health
and Human Services Administration, 25(1/2), 89-115.
Greenglass, E.R., Burke, R.J. & Koniarski, R. (1997). The impact of social support on the
development of burnout in teachers: examination of a model. Work and Stress, 11(3), 267-278.
Guerrier, Y. & Adib, A. (2003). Work at leisure and leisure at work: A study of the emotional
labour of tour reps. Human Relations, 56(11), 1399-1417.
Hardy, G.E., Shapiro, D.A. & Borill,C.S. (1997) Fatigue in the workforce of the national health
service trusts: levels of symptomatology and links with minor psychiatric disorder,
demographic, occupational and work role factors. Journal of Psychosomatic Research
Research, 43, 83-92.
Hart, P. M., Wearing, A. J. & Headey, B. (1995). Police stress and well-being – integrating
personality, coping and daily work experiences. Journal of Occupational and Organizational
Psychology, 68, 133-156.
Hayter, M. (1999). Burnout and AIDS care-related factors in HIV community Clinical Nurse
Specialists in the North of England. Journal of Advanced Nursing, 29(4), 984-993.
Heaney, C. A., Price, R. H. & Rafferty, J. (1995). Increasing coping resources at work – a field
experiment to increase social support, improve work team functioning and enhance employee
mental health. Journal of Organizational Behavior, 16(4), 335-352.
Heuven, E. & Bakker, A..B. (2003). Emotional dissonance and burnout among cabin attendants.
European Journal of Work and Organizational Psychology, 12(1), 81-100.
Higgins, N. C. (1986). Occupational stress and working women: The effectiveness of two stress
reduction programs. Journal of Vocational Behavior, 29, 66-78.
Himmelweit, S. (1999). Caring labor. Annals of the American Academy of Political and Social
Science, 561, 27-38.
Hochschild, A. R. (1983). The Managed Heart: Commercialization of human feeling. Berkeley:
University of California Press.
Holman, D., Chissick, C. & Totterdell, P. (2002). The effects of performance monitoring on
emotional labor and well-being in call centers. Motivation and Emotion, 26(1), 57-81.
Houkes, I., Janssen, P. P. M., de Jonge, J. & Nijhuis, F.J.N. (2001). Specific relationships
between work characteristics and intrinsic work motivation, burnout and turnover intention: A
multi-sample analysis. European Journal of Work and Organizational Psychology, 10(1), 1-24.
Houkes, I., Janssen, P.P.M., de Jonge, J. & Bakker, A.B. (2003). Specific determinants of
intrinsic work motivations, emotional exhaustion and turnover intention: A multisample
longitudinal study. Journal of Occupational and Organizational Psychology, 76, 427-450.
Huby, G., Gerry, M., McKinstry, B., Porter, M., Shaw, J. & Wrate, R. (2002). Morale among
general practitioners: Qualitative study exploring relations between partnership arrangements,
personal style, and workload. British Medical Journal, 325(7356), 140-142B.
128
Iverson, R. D., Olekalns, M. &Erwin, P. J. (1998). Affectivity, organizational stressors, and
absenteeism: A causal model of burnout and its consequences. Journal of Vocational Behavior,
52(1), 1-23.
Jamal, M. & Baba, V.V. (2001). Type A behavior, job performance and well-being in college
teachers. International Journal of Stress Management, 8 (3), 231-240.
Janssen, O. (2004). How fairness perceptions make innovative behavior more or less stressful.
Journal of Organizational Behavior, 25(2), 201-215.
Janssen, P.P.M., de Jonge, J. & Bakker, A.B. (1999). Specific determinants of intrinsic work
motivation, burnout and turnover intentions: a study among nurses. Journal of Advanced
Nursing, 29(6), 1360-1369.
Janssen, P.P.M., Peeters, M.C.W., de Jonge, J., Houkes, I. & Tummers, G.E.R. (2004). Specific
relationships between job demands, job resources and psychological outcomes and the
mediating role of negative work-home interference. Journal of Vocational Behavior, 65(3),
411-429.
Janssen, P.P.M., Schaufeli, W.B. & Houkes, I. (1999). Work-related and individual
determinants of the three burnout dimensions. Work and Stress, 13(1), 74-86.
Jenkins, R. & Elliott, P. (2004). Stressors, burnout and social support: nurses in acute mental
health settings. Journal of Advanced Nursing, 48(6), 622-631.
Johnson, L.B. (1991). Job strain among police officers: Gender comparisons. Police Studies,
1991(1), 12-16.
Johnson, C.N.E. & Hunter, M. (1997). Vicarious traumatization in counsellors working in the
New South Wales Sexual Assault Service: an exploratory study. Work and Stress, 11(4), 319328.
Karasek, R.A. (1979). Job demands, job decision latitude and mental strain: Implications for job
redesign. Administrative Science Quarterly, 24, 285-308.
King, N. (1998). Template analysis. In Symon, G. and Cassell, C. (eds) Qualitative Methods
and Analysis in Organizational Research: A Practical Guide. London: Sage.
Klein, D. J. & Verbeke, W. (1999). Autonomic feedback in stressful environments: How do
individual differences in autonomic feedback relate to burnout, job performance, and job
attitudes in salespeople? Journal of Applied Psychology, 84(6), 911-924.
Koniarek, J. & Dudek, B. (1996) Social support as a buffer in the stress burnout relationship.
International Journal of Stress Management, 3(2), 99-106.
Kop, N., Euwema, M. & Schaufeli, W. (1999). Burnout, job stress and violent behaviour among
Dutch police officers. Work and Stress, 13(4), 326-340.
Kruml, S.M. & Geddes, D. (1998). Catching fire without burning out: Is there an Ideal way to
perform emotional labor? Paper presented at the first conference on Emotions in
Organizational Life, San Diego (August, 1998).
129
Kushnir, T., Malkinson, R. & Ribak, J. (1998). Rational thinking and stress management in
health workers: A psychoeducational program. International Journal of Stress Management,
5(3), 169-178.
Lawoko, S., Soares, J.J.F. & Nolan, P. (2004). Violence towards psychiatric staff: a comparison
of gender, job and environmental characteristics in England and Sweden. Work and Stress,
18(1), 39-55.
LeBlanc, M. M. & Kelloway, E. K. (2002). Predictors and outcomes of workplace violence and
aggression. Journal of Applied Psychology, 87(3), 444-453.
Lee, R.T. & Ashforth, B.E. (1993) A longitudinal study of burnout among supervisors and
managers: Comparisons between the Leiter and Maslach (1988) and Golembiewski et al (1986)
models. Organizational Behavior and Human Decision Processes, 54(3), 369-398.
Lee, R. T. & Ashforth, B. E. (1996). A meta-analytic examination of the correlates of the three
dimensions of job burnout. Journal of Applied Psychology, 81(2), 123-133.
Leiter, M. P. (1993). Burnout as a developmental process: Consideration of models. In
Schaufeli, W. B., Maslach, C. and Marek, T. (eds) Professional burnout: Recent developments
in theory and research. Taylor and Francis: Washington.
Leiter, M. P. & Durup, M. J. (1996). Work, home and in-between: A longitudinal study of
spillover. Journal of Applied Behavioral Science. 32(1), 29-47.
Leiter, M.P., Frizzell, C., Harvie, P. & Churchill, L. (2001). Abusive interactions and burnout:
Examining occupation gender and the mediating role of community. Psychology and Health,
16(5), 547-563.
Leiter, M. P. & Maslach, C. (1988). Impact of interpersonal environment on burnout and
organizational commitment. Journal of Organizational Behavior, 9, 229-243.
Lewig, K.A. & Dollard, M.F. (2003). Emotional dissonance, emotional exhaustion and job
satisfaction in call centre workers. European Journal of Work and Organizational Psychology,
12(4), 366-393.
Macdonald, J. & Morley, I. (2001). Shame and non-disclosure: A study of the emotional
isolation of people referred for psychotherapy. British Journal of Medical Psychology, 74, 121.
Malkinson, R., Kushnir, T. & Weisberg, E. (1997). Stress management and burnout prevention
in female blue-collar workers: Theoretical and practical implications. International Journal of
Stress Management, 4, 183-197.
Mann, S. (2004). ‘People-work’: emotion management, stress and coping. British Journal of
Guidance and Counselling, 32(2), 205-221.
Maslach, C. (1982). Burnout: The Cost of Caring. Englewood Cliffs, NJ: Prentice-Hall.
Maslach, C. & Leiter, M.P. (1997). The truth about burnout: How organizations cause personal
stress and what to do about it. Jossey-Bass.
130
Maslach, C. & Jackson, S.E. (1981). The measurement of experienced burnout. Journal of
Occupational Behaviour, 2,99-113.
Maslach, C. & Jackson, S.E. (1985). The role of sex and family variables in burnout. Sex Roles,
12, 837-851.
Maslach, C., Jackson, S.E. & Leiter, M.P. (1996). Maslach Burnout Inventory Manual (3rd
Edition). Mountain View, CA: CPP, Inc.
Maslach, C., Schaufeli, W.B. & Leiter, M.P. (2001). Job Burnout. Annual Review of
Psychology, 52, 397-422.
McQueen, A.C.H. (2004). Emotional intelligence in nursing work.
Nursing, 47(1), 101-108.
Journal of Advanced
Mikkelsen, A., Saksvik, P. O. & Landsbergis, P. (2000). The impact of a participatory
organizational intervention on job stress in community health care institutions. Work and Stress,
14(2), 156-170.
Miles, M. and Huberman, M. (1994). Qualitative Data Analysis: An Expanded Sourcebook. 2nd
Edition. California: Sage.
Miller, K.I., Stiff, J.B. & Ellis, B.H. (1988).Communication and empathy as precursors to
burnout among human service workers. Communication Monographs, 55, 250-265.
Miller, K., Birkholt, M, Scott, C. & Stage, C. (1995). Empathy and burnout in human service
work: An extension of a communication model. Communication Research, 22(2), 123-147.
Moore, J.E. (2000). Why is this happening? A causal attribution approach to work exhaustion
consequences. Academy of Management Review, 25(2), 335-349.
Morris, J. A. & Feldman, D. C. (1996). The dimensions, antecedents, and consequences of
emotional labor. Academy of Management Review, 21(4), 986-1010.
Muncer, S., Tayler, S., Green, D.W. & McManus, I.C. (2001). Nurses’ representations of the
perceived causes of work-related stress: a network drawing approach. Work and Stress, 15(1),
40-52.
Omdahl, B.L. & O’Donnell, C. (1999) Emotional contagion, empathic concern and
communicative responsiveness as variables affecting nurses’ stress and occupational
commitment. Journal of Advanced Nursing, 29(6), 1351-1359.
Payne, N. (2001) Occupational stressors and coping as determinants of burnout in female
hospice nurses. Journal of Advanced Nursing, 33(3), 396-405.
Peeters, M. C. W. & Le Blanc, P. M. (2001). Towards a match between job demands and
sources of social support: A study among oncology care providers. European Journal of Work
and Organizational Psychology, 10(1), 53-73.
Pines, A. & Aronson, E. (1988). Career Burnout: Causes and Cures. The Free Press: New York.
Pines, A.M. (2004) Adult attachment styles and their relationship to burnout: A preliminary,
cross-cultural investigation. Work and Stress, 18(1), 66-80.
131
Poulin, J.E. & Walter, C.A. (1993). Burnout in gerontological social work. Social Work, 38(3),
305-310.
Price, J.L. & Mueller, C.W. (1986). Absenteeism and turnover of hospital employees.
Greenwich, CT: JAI.
Price, L. & Spence, S.H. (1994). Burnout symptoms amongst drug and alcohol service
employees: Gender differences in the interaction between work and home stressors. Anxiety,
Stress and Coping, 7(1), 67-84.
Pugliesi, K. & Shook, S. L. (1997). Gender, jobs and emotional labor in a complex organization.
In Erickson, R. J. and Cuthbertson-Johnson, B. (eds) Social Perspectives on Emotion (Volume
4). London: Sage.
Rafaeli, A. & Sutton, R.I. (1988). Untangling the relationship between displayed emotions and
organizational sales: The case of convenience stores. Academy of Management Journal, 31(3),
461-487.
Rafaeli, A. & Sutton, R.I. (1990). Busy stores and demanding customers: How do they affect
the display of positive emotion? Academy of Management Journal, 33(3), 623-637.
Rafaeli, A. & Sutton, R.I. (1991). Emotional contrast strategies as means of social influence:
Lessons from criminal interrogators and bill collectors. Academy of Management Journal,
34(4), 749-775.
Rafferty, Y., Friend, R. & Landsbergis, P.A. (2001) The association between job skill
discretion, decision authority and burnout. Work and Stress, 15(1), 73-85.
Ramirez, A.J., Graham, J., Richards, M.A., Cull, A. & Gregory, W.M. (1996). Mental health of
hospital consultants: The effects of stress and satisfaction at work. The Lancet, 347(9003), 724728.
Reynolds, S. (1997). Psychological well-being at work: Is prevention better than cure? Journal
of Psychosomatic Research, 43,93-102.
Salmela-Aro, K., Naatanen, P. & Nurmi, J. E. (2004). The role of work-related personal
projects during two burnout interventions: A longitudinal study. Work and Stress, 18(3), 208230.
Savicki, V. & Cooley, E.J. (1994). Burnout in child protective service workers – a longitudinal
study. Journal of Organizational Behavior, 15(7), 655-666.
Schaubroeck, J., Ganster, D.C., Sime, W.E. and Ditman, D. (1993). A field experiment testing
supervisory role clarification. Personnel Psychology, 46, 1-25.
Schaubroeck, J. & Jones, J.R. (2000). Antecedents of workplace emotional labor dimensions
and moderators of their effects on physical symptoms. Journal of Organizational Behavior, 21,
163-183.
Schaufeli, W. B. & Bakker, A. B. (2004). Job demands, job resources, and their relationship
with burnout and engagement: A multi-sample study. Journal of Organizational Behavior,
25(3), 293-315.
132
Schaufeli, W. B. & Enzmann, D. (1998). The burnout companion to study and practice: A
critical analysis. London: Taylor & Francis Ltd.
Schaufeli, W.B., Leiter, M.P., Maslach, C. & Jackson, S.E. (1996). Maslach Burnout Inventory
– General Survey (MBI-GS). In C. Maslach, S.E. Jackson and M.P. Leiter, MBI Manual (3rd
ed.). Mountain View, CA: CPP, Inc.
Sczesny, S. & Stahlberg, D. (2000). Sexual harassment over the telephone: occupational risk at
call centres. Work and Stress, 14(2), 121-136.
Seo, M-G., Feldman Barrett, L., & Bartunek, J.M. (2004). The role of affective experience in
work motivation. Academy of Management Review, 29, 423-439.
Shimizu, T., Mizoue, T., Kubota, S., Mishima, N. & Nagata, S. (2003). Relationships between
burnout and communication skill training among Japanese hospital nurses: A pilot study.
Journal of Occupational Health, 45(3), 185-190.
Siegrist, J. (1996). Adverse health effects of high effort-low reward conditions. Journal of
Occupational Health Psychology, 1, 27-41.
Söderfeldt, M., Söderfeldt, B., Ohlsson, C. G., Theorell, T. and Jones, I. (2000). The impact of
sense of coherence and high-demand/low control job environment on self-reported health,
burnout and psychological stress indicators. Work and Stress, 14, 1-15.
Spector, P. E., & Jex, S. M. (1991). Relations of job characteristics from multiple data sources
with employee affect, absence, turnover intentions, and health. Journal of Applied Psychology,
76, 46– 53.
Staden, H. (1998). Alertness to the needs of others: a study of the emotional labour of caring.
Journal of Advanced Nursing, 27(1), 147-156.
Stansfeld, S., Head, J. & Marmot, M. (2000) Work related factors and ill health: The Whitehall
II Study. HSE Contract Research Report 266/2000.
Steinberg, R.J. & Figart, D.M. (1999). Emotional demands at work: A job content analysis.
Annals of the American Academy of Political and Social Science, 561, 177-191.
Stephens, C. & Long, N. (2000). Communication with police supervisors and peers as a buffer
of work-related traumatic stress. Journal of Organizational Behavior, 21(4), 407-424.
Steptoe, A., Cropley, M. & Joekes, K. (1999). Job strain, blood pressure and response to
uncontrollable stress. Journal of Hypertension, 17(2), 193-200.
Stordeur, S., D’hoore, W. & Vandenberghe, C. (2001). Leadership, organizational stress, and
emotional exhaustion among hospital nursing staff. Journal of Advanced Nursing, 35(4), 533542.
Sutton, R.I. (1991). Maintaining norms about expressed emotions: The case of bill collectors.
Administrative Science Quarterly, 36, 245-268.
133
Sutton, R.I. & Rafaeli, A. (1988). Untangling the relationship between displayed emotions and
organizational sales: The case of convenience stores. Academy of Management Journal, 31(3),
461-487.
Symon, G. (1998). Qualitative research diaries. In Symon, G. and Cassell, C. (eds) Qualitative
Methods and Analysis in Organizational Research. London: Sage.
Tang, C. S. K., Au, W. T., Schwarzer, R. & Schmitz, G. (2001). Mental health outcomes of job
stress among Chinese teachers: Role of stress resource factors and burnout. Journal of
Organizational Behavior, 22(8), 887-901.
Taylor, S. & Tyler, M. (2000). Emotional labour and sexual difference in the airline industry.
Work, Employment and Society, 14(1), 77-95.
Thompson, B.M., Kirk, A. & Brown, D.F. (2005). Work based support, emotional exhaustion,
and spillover of work stress to the family environment: A study of policewomen. Stress and
Health, 21(3), 199-207.
Tolich, M. B. (1993). Alienating and liberating emotions at work. Journal of Contemporary
Ethnography, 22(3), 361-381.
Toppinen-Tanner, S. Kalimo, R. & Mutanen, P. (2002). The process of burnout in white-collar
and blue-collar jobs: Eight-year prospective study of exhaustion. Journal of Organizational
Behavior, 23(5), 555-570.
Totterdell, P. & Holman, D. (2003). Emotion regulation in customer service roles: Testing a
model of emotional labor. Journal of Occupational Health Psychology, 8(1), 55-73.
Troyer, L., Mueller, C.W. & Osinsky, P.I. (2000). Who’s the boss? A role-theoretic analysis of
customer work. Work and Occupations, 27(3), 406-427.
Tummers, G.R., Landeweerd, J.A. & van Merode, G.G. (2002). Work organization, work
characteristics, and their psychological effects on nurses in the Netherlands. International
Journal of Stress Management, 9(3), 183-206.
Van Dierendonck, D. & Mevissen, N. (2002) Aggressive behaviour of passengers, conflict
management behaviour, and burnout among trolley car drivers. International Journal of Stress
Management, 9(4), 345-355.
Van Dierendonck, D., Schaufeli, W. B. & Buunk, B. P. (1998). The evaluation of an individual
burnout intervention program: The role of inequity and social support. Journal of Applied
Psychology, 83(3), 392-407.
Van Dierendonck, D., Schaufeli, W.B. & Buunk, B.P. (2001a). Burnout and inequity among
human service professionals: A longitudinal study.
Journal of Occupational Health
Psychology, 6(1), 43-52.
Van Dierendonck, D., Schaufeli, W.B. & Buunk, B.P. (2001b). Toward a process model of
burnout: Results from a secondary analysis. European Journal of Work and Organizational
Psychology, 10(1), 41-52.
Van Gorp, K. & Schaufeli, W. B. (1996). Een gezonde geest in een gezond lichaam (A healthy
mind in a healthy organization). The Hague: VUGA.
134
Van Horn, J.E., Schaufeli, W.B. & Taris, T.W. (2001) Lack of reciprocity among Dutch
teachers’ Validation of reciprocity indices and their relation to stress and well-being. Work and
Stress, 15(3), 191-213.
Van Maanen, J. V. & Kunda, G. (1989). Real feelings: Emotional expression and organizational
culture. In Cummings, L. L. and Staw, B. M. (eds) Research in Organizational Behaviour, Vol
11, 43-104. Greenwich, CT: JAI Press.
Van Vegchel, N., de Jonge, J., Meijer, T. & Hamers, J.P.H. (2001). Different effort constructs
and effort-reward imbalance: effects on employee well-being in ancillary health care workers.
Journal of Advanced Nursing, 34(10), 128-136.
Van Vegchel, N., de Jonge, J., Bakker, A. B. & Schaufeli, W. B. (2002). Testing global and
specific indicators of rewards in the effort-reward imbalance model: Does it make any
difference? European Journal of Work and Organizational Psychology, 11(4), 403-421.
Van Yperen, N. W. (1996). Communal orientation and the burnout syndrome among nurses: A
replication and extension. Journal of Applied Social Psychology, 26(4), 338-354.
Wall, T. D., Bolden, I. R., Borrill, C. S., Carter, A. J., Golya, D. A., Hardy, G. E., et al (1997).
Minor psychiatric disorder in NHS trust staff: occupational and gender differences. British
Journal of Psychiatry, 171, 519 – 523.
Wade, D. C., Cooley, E. & Savicki, V. (1986). A longitudinal study of burnout. Children and
Youth Services Review, 8(2), 161-173.
Walsh, B.R. & Clarke, E. (2003). Post-trauma symptoms in health workers following physical
and verbal aggression.
Warr, P., Butcher, V., Robertson, I. & Callinan, M. (2004). Older people's well-being as a
function of employment, retirement, environmental characteristics and role preference. British
Journal of Psychology, 95(3), 297-324.
Watson, D., Clark, L.A. & Tellegen, A. (1988). Development and validation of brief measures
of positive and negative affect: The PANAS scales. Journal of Personality and Social
Psychology, 54(6), 1063-1070.
Weiss, D.S. & Marmar, C.R. (1997). The Impact of Events Scale – Revised. In Wilson, J.P. and
Keane, T.M. (eds) Assessing Psychological trauma and PTSD. New York: Guildford Press.
Westman, M., Etzion, D. & Danon, E. (2001). Job insecurity and crossover of burnout in
married couples. Journal of Organizational Behavior, 22(5), 467-481.
Wharton, A. S. (1993). The affective consequences of service work Managing emotions on the
job. Work and Occupations, 20(2), 205-232.
Wharton, A. S. (1999). The psychological consequences of emotional labor. Annals of the
American Academy of Political and Social Science, 561, 158-176.
Wharton, A. S. & Erickson, R. J. (1995). The consequences of caring: Exploring the links
between women’s job and family emotion work. Sociological Quarterly, 36, 273-296.
135
Williams, C. (2003) Sky service: The demands of emotional labour in the airline industry.
Gender, Work and Organization, 10(5), 513-550.
Winstanley, S. & Whittington, R. (2002). Anxiety, burnout and coping styles in general hospital
staff exposed to workplace aggression: a cyclical model of burnout and vulnerability to
aggression. Work and Stress, 16(4), 302-315.
Workman, M. & Bommer, W. (2004). Redesigning computer call center work: A longitudinal
field experiment. Journal of Organizational Behavior, 25(3), 317-337.
Wouters, C. (1989). The sociology of emotions and flight attendants: Hochschild’s Managed
Heart. Theory, Culture & Society, 6,95-123.
Wright, T. A. & Cropanzano, R. (1998). Emotional exhaustion as a predictor of job
performance and voluntary turnover. Journal of Applied Psychology, 83(3), 486-493.
Zapf, D. (2002). Emotion work and psychological strain: A review of the literature and some
conceptual considerations. Human Resource Management Review, 12, 237-268.
Zapf, D., Isic, A., Bechtoldt, M. & Blau, P. (2003). What is typical for call centre jobs? Job
characteristics, and service interactions in different call centres. European Journal of Work and
Organizational Psychology, 12(4), 311-340.
Zapf, D., Seifert, C., Schmutte, B., Mertini, H. & Holz, M. (2001). Emotion work and job
stressors and their effects on burnout. Psychology and Health, 16(5), 527-545.
Zapf, D., Vogt, C., Seifert, C., Mertini, H. & Isic, A. (1999). Emotion work as a source of
stress: The concept and development of an instrument. European Journal of Work and
Organizational Psychology, 8, 371-400.
Zellars, K.L. & Perrewé, P.L. (2001). Affective personality and the content of emotional social
support: Coping in organizations. Journal of Applied Psychology, 86(3), 459-467.
Zohar, D. (1997). Predicting burnout with a hassle-based measure of role demands. Journal of
Organizational Behavior, 18(2), 101-115.
Published by the Health and Safety Executive
02/08
Health and Safety
Executive
The nature, causes and consequences
of harm in emotionally-demanding
occupations
Traditional approaches to understanding psychosocial
job characteristics and well-being have been quite
general in that they explore links between general job
characteristics such as workload and control on workers
in many different sorts of occupations. One example
of a more specific approach can be found in research
into emotional labour - the requirement to regulate both
feelings and the expression of feelings for organizational
goals. Early research into emotional labour focused on
customer service workers (CSW) but has more recently
also considered human service workers (HSW) such as
nurses and social workers. A more specific approach to
thinking about the outcomes of demanding psychosocial
job characteristics can be found in research on burnout
which is thought to have three elements: emotional
exhaustion, depersonalization (also labelled cynicism), and
(low) accomplishment (also called professional efficacy).
Much recent research has started to explore the links
between emotional demands and burnout. The main aim
of this project is therefore to explore the nature of such
links through undertaking three distinct tasks. The first is a
literature review of evidence and theory while the second
two tasks comprise two empirical studies examining
several key issues in burnout research.
This report and the work it describes were funded by
the Health and Safety Executive (HSE). Its contents,
including any opinions and/or conclusions expressed, are
those of the authors alone and do not necessarily reflect
HSE policy.
RR610
www.hse.gov.uk
Fly UP