Sinking, swimming and sailing: experiences of job satisfaction and emotional exhaustion in child welfare employees

Sinking, swimming and sailing: experiences of jobsatisfaction and emotional exhaustion in childwelfare employeescfs_857 383..393

Deena Mandell*, Carol Stalker†, Margriet de Zeeuw Wright‡, Karen Frensch§ and Cheryl Harvey¶*Associate Professor, †Professor, ‡Recent Graduate, Lyle S. Hallman Faculty of Social Work, §Research Project

Manager, Partnerships for Children and Families, Lyle S. Hallman Faculty of Social Work, and ¶Associate Professor

(retired), School of Business and Economics, Wilfrid Laurier University, Kitchener, Ontario, Canada

ABSTRACT

The authors conducted a mixed-method study after a previous studyof child welfare employees revealed a subgroup exhibiting surpris-ingly high levels of both emotional exhaustion (EE) and job satisfac-tion (JS). This subgroup included direct service workers, supervisorsand managers. As these findings appeared to conflict with previousstudies, we re-reviewed the literature and undertook the currentstudy to account for the co-existence of EE and JS. We explored andcompared this subgroup with two others: workers who found theirwork satisfying without experiencing high levels of EE and thosewhose high levels of EE were associated with low JS. Using a surveythat included several standardized measures with 226 employees andsemi-structured interviews with a criteria-based subsample of 25, weexplored the role that personality, career expectations, coping styles,stage of life, education, gender and social networks play in outcomesfor individual employees. Analyses of quantitative and qualitative datayielded a profile for each subgroup, offering insights into the subjec-tive experiences of workers within individual, social and organiza-tional contexts. These findings have implications for recruitment,training and support of child welfare workers.

Correspondence:Deena Mandell,Lyle S. Hallman Facultyof Social Work,Wilfrid Laurier University,120 Duke St. W.,Kitchener, Ontario,Canada N2H 3W8E-mail: [email protected]

Keywords: child protection,child welfare,education/training/supervision,empirical research, gender,management

Accepted for publication: January2012

INTRODUCTION

In this paper, we report findings from a study designedto answer questions evoked by a previous study ofemployees’ work experience at four Ontario childwelfare agencies during a period of dramatic changes(Mandell et al. forthcoming).We conducted the initialsurvey to learn more about the complex set ofindividual and organizational characteristics that con-tribute to staff turnover, to other indicators of organi-zational health and to the overall quality of theworking environment for service providers in the dif-ficult climate then prevalent. Our aim was to offeragencies guidance in improving employee retention

and service to clients by ameliorating conditionsleading to burnout and turnover.

The initial study revealed disturbing levels of emo-tional exhaustion (EE) among many agency employ-ees; however, a subgroup consisting of direct serviceworkers (DSWs), supervisors and managers exhibitedsurprisingly high levels of both EE and job satisfaction(JS). As these findings appeared to conflict with thoseof previous studies, we re-reviewed the literature anddesigned the current study to discover how EE and JScan co-exist. We also wondered how this subgroupresembled and differed from workers who found theirwork satisfying without experiencing high levels of EEand those whose high levels of EE were associated

bs_bs_banner

doi:10.1111/j.1365-2206.2012.00857.x

383 Child and Family Social Work 2013, 18, pp 383–393 © 2012 John Wiley & Sons Ltd

with low JS. We hoped the comparison would help usunderstand whether the workers of primary interest(high EE/high JS) are ‘born’ or ‘made’ in order toidentify implications for recruitment, training andongoing support.

LITERATURE REVIEW

Relationship between EE and JS in childwelfare workers

We had previously examined studies assessing bothEE and JS in samples of child welfare workers (Stalkeret al. 2007). Of six such studies, only one (Silver et al.1997) reported findings similar to ours: namely, thatalthough the mean EE score for the sample of childwelfare supervisors was in the high range, 47% scoredin the high range on a measure of JS. A subsequentsearch for studies published since 2006 yielded twonewer studies (Mena & Bailey 2007; Cahalane & Sites2008) in which scores on JS and EE were negativelycorrelated. Thus, of eight studies of child welfareemployees measuring both EE and JS, only Silveret al. (1997) reported a substantial number scoring inthe high range on both EE and JS (see Note 1).

Other studies of child welfare workers have assessedconstructs conceptually close to JS or EE, revealingfindings that suggest that the relationship between EEand JS may be more complex than is often assumed.Anderson (2000) reported that 62% of Child Protec-tive Services workers scoring in the high range on EEscored simultaneously in the moderate range on senseof personal accomplishment. A sample surveyed byVinokur-Kaplan (1991) during ‘a time of considerableturbulence’ (p. 90) in the early 1980s reported ‘arelatively high level of overall job satisfaction’ (p. 90).Several other studies offer plausible insights into suchapparently paradoxical findings. Yoo (2002) foundthat case managers’ commitment to the families theywere serving and to the principles of family preserva-tion lessened the likelihood that low satisfaction withpay/ promotion, harsh working conditions and lack oforganizational leadership would affect job perfor-mance and client outcomes. Landsman’s (2001) studysuggested that child welfare workers are more likely toderive high satisfaction from their work when theybelieve that social work makes a difference in people’slives. Similarly, a study of compassion fatigue, burnoutand compassion satisfaction (Conrad & Kellar-Guenther 2006) found that 50% of Colorado childprotection workers scored high or extremely high oncompassion fatigue, but only 7.7% were high or

extremely high on burnout. These results supportedthe belief that compassion satisfaction may play animportant role in offsetting compassion fatigue and indecreasing burnout.

Two early qualitative studies examining workers’perseverance despite stressful working conditions con-cluded that child welfare workers who believe they canmake a difference in people’s lives and are committedto the goals of child welfare were more likely to stay(Reagh 1994; Rycraft 1994). More recent qualitativestudies have supported these conclusions. Ellett et al.(2007) found that personal and professional commit-ment to clients and the desire to ‘make a difference’were among the personal factors identified by childwelfare professionals as contributors to employeeretention.

A national American study examining characteris-tics of child welfare workers in relation to JS reportedthat those with a Bachelor of Arts were less satisfiedthan those with a Bachelor of Social Work (BSW;Barth et al. 2008). The researchers speculated thatthose with the commitment and preparation involvedin completing a BSW ‘might be more willing toendure a mediocre sense of satisfaction’ (p. 207).Including personal characteristics in her examinationof predictors of child welfare employee retention,Ellett (2009) reported that ‘human caring’ was themost heavily weighted variable differentiating workerswho intended to leave child welfare from those whointended to stay. Similarly, a study of Master of SocialWork graduates of the US Title IV-E Programs(O’Donnell & Kirkner 2009) reported that the ‘mostconsistent pre-employment predictor of [employeecontinuation beyond] their stipend commitmentperiod was commitment to the agency and . . . tochild welfare in general’ (p. 254).

These studies suggest that, even when EE is high,certain individual characteristics may allow childwelfare workers to remain satisfied with their jobs.These characteristics include personal commitment tothe mission of child welfare, a belief that social workerscan make a positive difference in people’s lives, com-mitment to social work values and a personality that issupportive and sensitive to the needs of others.

EE and JS in social work settings outsideof child welfare

Many studies of social workers in non-child welfaresettings have also reported the expected negative cor-relation between EE and JS; only a few reported highlevels of JS despite high levels of EE (e.g. Jayaratne &

Sinking, swimming and sailing D Mandell et al.

384 Child and Family Social Work 2013, 18, pp 383–393 © 2012 John Wiley & Sons Ltd

Chess 1983; Evans et al. 2006). Several studies havenoted that the relationship between EE and JS is mod-erated or mediated by variables such as social supportand organizational support (Koeske & Kelly 1995;Martin & Schinke 1998; Acker 1999; Takeda et al.2005), supervisor support (Um & Harrison 1998),coping strategy (Koeske et al. 1993), internal locus ofcontrol (Koeske & Kirk 1995) and job autonomy(Kim & Stoner 2008). The unexpected relationshipfound in our study might be explained by interactionswith these mediating or moderating variables or withother individual variables such as causal attribution(Moore 2000), goal orientation (Van Yperen &Janssen 2002) or gender.

Because women predominate in most samples ofchild welfare workers, we were particularly interestedin whether gender might influence the relationshipbetween EE and JS. Most studies have not foundsignificant differences in burnout levels between maleand female social workers (Hansung & Stoner 2008;Boston 2009). Himle et al. (1987) reported that formen, increased contact with clients was the mostimportant predictor of EE, whereas the relevant pre-dictor for women was dissatisfaction with supervisorsupport, co-worker support and working conditions.To our knowledge, these findings have not beenreplicated.

In summary, workers who simultaneously score highon EE and high on JS may not be as unusual as we

initially believed. Diverse variables have been found tomoderate or mediate the relationship between EE andJS. No previous studies have captured the subjectiveexperience of child welfare workers known to havescored high on both EE and JS.

METHOD

In this study, we pursued two questions: (i) How canone be simultaneously highly exhausted and highlysatisfied with the job? (ii) How do these workers differfrom those low on EE and high on JS, and those highon EE and low on JS? We sought to understand therole that personality, career expectations, copingstyles, stage of life, education, gender and social net-works play in outcomes for individual employees. Itseemed most fruitful to study two other groups aswell: those finding satisfaction in their work withoutexperiencing high levels of EE and those for whomhigh levels of EE were accompanied by low JS scores.We therefore focused on three categories of interest,which we named Sinking: high EE, low JS; Swimming(upstream): high EE, high JS; and Sailing: low EE, highJS, as illustrated in Fig. 1.

Phase One: survey

We administered a shortened version of our originalWorkplace Study survey (Mandell et al. forthcom-ing) to 226 employees in four Ontario child welfare

Figure 1 Categories of interest.

Sinking, swimming and sailing D Mandell et al.

385 Child and Family Social Work 2013, 18, pp 383–393 © 2012 John Wiley & Sons Ltd

organizations (three of the original four and oneadditional). Measures included the Maslach BurnoutInventory (MBI), Overall JS scale (Quinn & Shepard1974; Quinn & Staines 1979) and other measures ofworkplace experience. Participants were asked if theyconsented to be contacted for an interview and tohave their individual scores from the first surveylinked to their scores from the second survey forcomparison.

Participants’ scores on the EE scale of the MBI andthe JS scale were calculated and those whose scoreswere classified within the three categories of interestwere identified. The ranges of EE scores were: low:�16; med: 17–27; high: �28. JS scores were: low:�1.73; med: 1.74–2.47; high: �2.48. Participantswere compared by category on variables of gender,age, level of professional education, position andtenure in the organization. This yielded similaritiesand differences across the three categories that weexpected might influence employees’ experience oftheir work.

Phase Two: selection of participants for interview

We conducted qualitative interviews with a purposivesubsample of workers whose survey scores fell withinone of our three categories of interest. Because wewere interested in workers reliably representative ofthe categories (see Note 2), the original criteria forselection of interviewees were (i) completion of thequestionnaire at time 1 (original study) and time 2and (ii) scores placing the participant in one of thecategories of interest at both times. We had hoped tointerview 30 participants, 10 from each category butmodified the criteria when we discovered only 87 ofthe participants completing the questionnaire at time2 had also completed it at time 1. Of these 87 par-ticipants, only 69 consented to be contacted for aninterview and only 16 of the 69 completed bothquestionnaires and were in a category of interest onboth. Another nine respondents had completed bothsurveys and while they did not meet the criteria fora category of interest at time1, they did at time 2and so were included in the interview sample. Weinterviewed only five workers in the Sinking cate-gory because, of the 13 participants in the time 2survey who were in that category, only 11 consentedto be contacted and only 5 of the 11 agreed to beinterviewed.

Semi-structured interviews of approximately 1.5hours took place at a location chosen by partici-pants. Interviews were audio-taped and transcribed.

Participants were told which of the three categoriestheir survey scores placed them in, and then wereasked to comment on how this fit with their employ-ment realities at the time of survey completion.Participants elaborated on what contributed to a senseof gratification in their work and what prevented it,sources and manifestations of stress, and whether andwhere they sought support during stressful times.Finally, participants were asked to reflect on whetherthey thought men and women might approach orexperience child welfare work differently.

Analysis

The following questions guided the analysis of theinterviews: What are the elements employees identifyas being associated with EE and with JS?What are thepersonal qualities, workplace/organizational contextand coping skills they identify as being involved intheir experiences of EE and JS? How do employeesunderstand the relationship between EE and JS?Whatare the relationships among personal qualities andexperiences, workplace context and coping strategiesthat help to understand their differing experiences?How are various theories about EE reflected in theways employees talk about their work? These ques-tions were informed by the literature and reflect ourresearch sub-questions.

Initial coding identified specific elements of theworkers’ narratives. Second-level coding organized theinitial codes into 10 themes discussed below. Withineach of the three employee categories, the codes werecompiled across interviews to yield an initial compos-ite profile for each category. We then looked for simi-larities and differences across the categories. Thecoding at every stage was shared among the authorsand reviewed to ensure reliability.

FINDINGS

As expected, a substantial proportion of the childwelfare employees surveyed (75 of the 211 respon-dents who completed the EE scale or 35.5%)reported being highly emotionally exhausted; 69(32.7%) scored in the moderate range; and 67(31.7%) in the low range. On overall JS (n = 226),103 (46.3%) scored in the high range, 84 (37.8%) inthe moderate range and 15 (6.7%) in the low range.Twenty-four (11.4%, n = 211) respondents weresimultaneously highly emotionally exhausted andhighly satisfied with their jobs.

Sinking, swimming and sailing D Mandell et al.

386 Child and Family Social Work 2013, 18, pp 383–393 © 2012 John Wiley & Sons Ltd

Participants meeting criteria for an interview

Of the 226 employees who returned the Phase Onesurvey, 87 (38%) met the final criteria for possibleinterview because their scores at time 2 permittedassignment to one of the three categories of interest.Seventy-two (83%) were female, four (5%) were maleand the remainder did not indicate their gender. Forty(46%) were 39 years or less and 47 (52%) werebetween 40 and 59 years; 36 (41%) had a Master’sdegree or higher and 50 (57%) had a high school orcollege diploma or Bachelor’s degree; 57 (66%) wereDSWs and the remainder were managers or supervi-sors.The mean year of tenure in the organization was6.3, the mean tenure in their current position was 3.3years and the mean tenure in the profession was 11.4years. Fifty-three (61%) met the criteria for the Sailingcategory, 13 (15%) met the criteria for the Sinkingcategory and 21 (24%) met the criteria for the Swim-ming category.

Interview participant characteristics

Twenty (84%) of the 25 interviewees were female and13 (52%) were 39 years or younger. Nine (37%) hada Master’s degree, and 15 (63%) had a Bachelor’sdegree or college diploma. Sixteen (64%) were DSWsand the remaining nine (36%) were managers orsupervisors. The mean year of tenure in the organiza-tion was 6.2, the mean tenure in their current positionwas 2.9 years and the mean year in the profession was12.96.The interview sample was representative of thelarger sample on most variables.

Comparison of categories by age, position and tenure

Table 1 illustrates that those in the ‘Sailing’ categorytended to be older than employees in the other twocategories.

Table 2 illustrates that while DSWs dominated theSwimming category, half of those in the Sailing cat-egory were managers or supervisors. Table 3 showsthat, on average, Sailors had been with the organiza-tion and in their current position longer than those inthe Swimming or Sinking categories. Sailors were alsothe longest-tenured in the profession, followed bythose in the Sinking category.

Findings from analysis of interview data

Ten predominant themes emerged from analysis ofthe data: the nature of child welfare, relationships withclients, what contributes to high stress levels, impactof EE on the worker, gender, degree of fit betweenthe individual and the job, personal characteristics,balance, boundaries and making a difference. By com-paring the specific content of these themes across thethree categories of interest, we developed a compositeprofile of ‘what it takes’ to manage the work success-fully, cope with the multiple stresses and feel goodabout the job. The qualitative data confirmed thatmany, including some who were experiencing highlevels of EE, found the work gratifying and believed inits importance; they had apparently found a fit thateluded others.The findings shed light on how high EEand high JS can co-exist for some workers. We alsoexamined how the themes differed across categories,

Table 1 Categories of interest byage (interviewed sample) Category 25–30 years 31–39 years 40–49 years 50–59 years

Swimming (n = 10) 3 (30%) 3 (30%) 4 (40%) 0Sinking (n = 5) 1 (20%) 3 (60%) 1 (20%) 0Sailing (n = 10) 0 3 (30%) 4 (40%) 3 (30%)Total 4 (16%) 9 (36%) 9 (36%) 3 (12%)

Table 2 Categories of interest byposition (interviewed sample) Category Direct service workers Managers/supervisors

Swimming (n = 10) 7 (70%) 3 (30%)High EE & high JS

Sinking (n = 5) 4 (80%) 1 (20%)High EE & low JS

Sailing (n = 10) 5 (50%) 5 (50%)Low EE & high JS

Total 16 (64%) 9 (36%)

EE, emotional exhaustion; JS, job satisfaction.

Sinking, swimming and sailing D Mandell et al.

387 Child and Family Social Work 2013, 18, pp 383–393 © 2012 John Wiley & Sons Ltd

noting what might account for the three differentcombinations of EE and JS scores.

Ontario’s child welfare mandate promotes a rela-tively high degree of consistency in terms of agencystructure, policies and practices; therefore, despitecoming from four different agencies, the participantsaddressed similar issues and themes within the pro-vincial context, suggesting a relatively constant orga-nizational context.

Across categories, the interviewees generally agreedon what constitutes a successful child welfare worker.The composite profile indicates that workers per-ceived success as the ability to derive satisfaction fromtheir work, despite the challenges. They claimedworkers can derive this satisfaction to the extent thatthey are able to maintain firm but flexible boundaries;periodically achieve balance between energy devotedto work and attention paid to personal restoration;believe they are contributing in valuable ways to indi-viduals and/or society; resist feeling personally vulner-able; and draw on personal and/or professionalsupport.

We focus here on DSWs to ensure the relevance ofour interpretations for the subsample that appeared tobe most at risk for EE. At the broadest level, therewere some differences in the profiles of DSWs in thethree categories. In general, the Sailors (low EE, highJS) did not appear to be struggling. They exhibited astable comfort level in terms of perceiving challenge,ability to derive satisfaction from their work and todisengage constructively in the service of self-care.Those in the Sinking category (high EE, low JS) con-veyed a sense of defeat. Workers in the Swimmingcategory (high EE, high JS), however, actively soughtways to cope with the challenges – which they openlyacknowledged – and to find positive meaning in thework. They recognized contradictions in the job,remained hopeful and were conscious of what wasrequired to stay afloat. While workers in all three cat-egories had a concept of what constitutes the ‘right fit’

between an individual and child welfare work, those inthe Sinking category saw themselves as lacking what ittakes to achieve this fit.

A central difference between workers in the Swim-ming and Sinking categories, compared with theSailors, was that they were much more vocal abouthow hard it was to believe one actually can make adifference in the lives of client families.This may havebeen related, in part, to the lofty goals that Sinkingand Swimming workers expressed. They wanted tomake a difference in the child welfare system as awhole and help people to change their lives.They saidlittle about deriving satisfaction from small gainsmade in the lives of client families.

Sinking: high EE, low JS

The workers in this category described themselves asoverwhelmed, defeated and out of steam.The percep-tion that the problems were relentless was palpable.

There was a sense of ‘I cannot keep doing this. I cannot meet

the response times that’s necessary.’ . . . [I]t’s the constant –

it’s like a wheel. And it just goes and it goes and it goes, and

there’s no stopping it.

Those in the Sinking category saw themselves as nothaving the capacity to survive these perceived condi-tions. Lack of time for their personal lives was a majorconcern. They experienced multiple physical effects,including sleeplessness and deterioration in diet andphysical fitness, and relatively fewer of the satisfac-tions that Sailors and Swimmers described. Like theSwimmers, they mentioned barriers to their desireto make a difference in the organization, indicatingawareness of difficulties beyond themselves in the joband in the system. They acknowledged not knowinghow to cope with these challenges.

Sailing: low EE, high JS

Sailing workers tended to articulate realistic andattainable objectives: commitment to the protection of

Table 3 Mean tenure inorganization/position/profession(interviewed sample)Category

Mean tenure inorganization (years)

Mean tenure inposition (years)

Mean tenure inprofession (years)

Swimming (n = 10) 4.46 2.62 8.66High EE & high JS

Sinking (n = 5) 4.18 2.12 11.94High EE & low JS

Sailing (n = 10) 9.95 3.98 18.28Low EE & high JS

EE, emotional exhaustion; JS, job satisfaction.

Sinking, swimming and sailing D Mandell et al.

388 Child and Family Social Work 2013, 18, pp 383–393 © 2012 John Wiley & Sons Ltd

children, bettering children’s lives and helping parentsto be better at parenting; nevertheless, they appreci-ated small changes.They identified and valued specifickinds of changes in parents, children or families’ lives;what is more, they observed these changes happeningfrequently. They referred to a good fit between them-selves and the work, even though 8 out of 10 expressedsome criticism of their organizations.They perceived asupportive work environment and were less inclinedto identify the high volume of work as a problem.

Sailors came across as strong individuals with clearpersonal boundaries and a sense of perspective; theywere relatively easy going on several levels. Possiblyrelated to their reported ability to disengage emotion-ally was the ability to tolerate not being in control.They saw themselves as ‘people persons’ who relatedwell to clients; they did not take personally the fail-ures, difficult decisions or client anger frequentlyencountered. They did not seem to internalize thefrustrations and negative run-ins with clients and col-lateral contacts as workers in the other categories did.They enjoyed opportunities for personal learning anddevelopment that their work afforded them.

Sailors mentioned efforts to avoid taking cases tocourt; fewer court cases meant less work and lesspressure. They emphasized the importance of a non-authoritarian stance with parents, describing them-selves as taking time to listen and negotiate. Theyarticulated the value of being able to disengage emo-tionally: ‘Not feeling indispensable’ allowed them toset aside their work at day’s end and carry on withtheir own lives. One’s construction of the work andone’s role thus seemed to shape one’s ability to main-tain self-protective boundaries.

Um, being very clear about your role – being able to say ‘I

can’t do that.’You know, that’s what you have to do in order

for you to just breathe . . . to survive. . . .

‘What doesn’t get done today will get done tomor-row’ was a Sailors’ approach quite different from thatdescribed by workers in the other two categories, whospoke of working endless hours and sometimesneglecting their personal lives.What may have enabledconstructive disengagement for Sailors was their beliefthat they had done their very best for the client fami-lies and their recognition that they were not respon-sible for the choices families ultimately make.

. . . I don’t own those clients, and I don’t own those kids. And

if people make bad decisions, it’s not my problem, it’s their

problem. I’ll pick up the pieces, but it’s not my fault.

We consider it significant that the Sailing grouptended to be older than the other two categories. All of

the Sailors reported having supportive partnersoutside of work and all talked about using co-workers’support to help them cope with the job’s frustrations.Most had extensive child welfare experience.This sug-gests that maturity, life stage and personal supportwithin and outside the workplace contributed to theirability to manage the work’s challenges.They had timeto hone their skills, develop perspective on their clientsand on their own role. Over time, they had developedways of managing workload and achieving a self-protective balance between work and personal life.

. . . I’ve learned to pace myself better. Um, and I’ve learned

to stand up for myself more . . . I’ve learned the word ‘no’,

which was difficult for me at first.

The confidence they expressed in their ability tomake good decisions may have been related to theirlife and work experience. All were parents themselves,which they saw as helping them do the work well.

Swimming: high EE, high JS

DSWs with high scores on both EE and JS tended tobe younger than those in the other categories, particu-larly the Sailors. Some had young children anddescribed the effort to balance competing demands offamily and work as a major challenge. They empha-sized the parallel between themselves as parents(mostly mothers) and their role as child welfareworkers in ways that suggested they saw the latter asan extension of the former. This implied a personalinvestment in and identification with the work, remi-niscent of a kind of calling or mission, not just a job.This may have made it more difficult to disengageemotionally in order to maintain protective bound-aries and sustainable balance, to avoid taking failuresand public criticism personally, and to derive satisfac-tion from small instances of clients’ progress. It mayalso have been what enabled the Swimmers to persistdespite the challenges, whereas workers in the Sinkinggroup, who did not express this sense of mission,experienced a sense of defeat.

Swimmers saw the job as impossible to get donebecause of sheer volume and competing demands.They were the most vocal about documentationrequirements and anxiety regarding court prepara-tion. Balancing these conditions with family and socialrelations appeared to be a constant struggle for them,especially for women who were primary caregivers athome. They reported that leisure time and self-carewere often minimal as work and family responsibilitiesconsumed their lives. Several mentioned a general

Sinking, swimming and sailing D Mandell et al.

389 Child and Family Social Work 2013, 18, pp 383–393 © 2012 John Wiley & Sons Ltd

organizational culture that expected considerableovertime, taking work home and deferring holidays.

Swimmers identified several external conditionsthat ameliorated or prevented EE, none of whichemerged in the Sinking group. Although Swimmerssaw the system as inadequate, within the organizationthey referred to positive relationships with co-workers,supervisor or administrative staff as sources of JS.Satisfying aspects of the work itself related to positiveexperiences with clients and, in a couple of instances,to specific work conditions such as time off and arelatively autonomous position.

Workers experiencing both high EE and high JSperformed ongoing conscious analyses of their situa-tion in an effort to maintain satisfaction. They indi-cated awareness of EE fluctuating according to thetime of week, time of year and time off; thus, there wasa sense of ‘it will get better’ during difficult times andsometimes they experienced some measure of relief.This appeared to reflect a level of resilience notevident among the workers in the Sinking group.

There’s times that ya get really frustrated with the job.There’s

times that you’re satisfied in the job . . . I don’t think you

would have really high job satisfaction the same time that

you’re highly emotionally exhausted. . . . if you’re- you’re

looking at balances, . . . over time . . . you still have job sat-

isfaction ‘cause it’s not always like that – always exhausted.

The struggle for balance was spoken of frequentlyamong Swimmers.

It is a roller coaster. Um, at times, you know, you wanna get off

because it’s overwhelming. It’s . . . like I have to draw on all

my coping skills . . . at times, to be able to function

because . . . the workload is just so overwhelming – along with

the emotional impact of some of the things that you see . . . .

Swimmers – more explicitly than Sailors – talkedabout the nature and necessity of boundaries toprevent or manage EE, emphasizing the difficulties inmaintaining them. This suggests awareness of anongoing struggle; Sailors apparently maintainedboundaries with less effort. While acknowledging theneed for boundaries, the Sinking workers describedthemselves as unable to maintain them.

Swimmers were the most expressive with regard torelationships with client families. They were inclinedto focus on the importance of these relationships tothemselves as individuals.This differs from the empha-sis Sailors put on these relationships, which was theirimportance to the children. This suggests that Swim-mers’ investment in these relationships may have beenassociated with their own identities as successfulhelpers; if so, the satisfaction they experienced may

rely upon perceiving clients as successful and appre-ciative of the worker’s help. It could render boundarymaintenance more difficult.

Swimmers shared with Sailors some similar viewsabout personal characteristics perceived to be salientfor child welfare work; these included the ability to beflexible and deal with the unexpected, to communi-cate clearly and directly with parents, to remain calmand to make difficult decisions despite personal sym-pathies. Overall, though, they did not express the samesense of personal confidence and clear boundariesthat the Sailing workers did. In fact, only Swimmersmade statements about putting the job first and being‘150%-ers’. Swimmers wanted to make major differ-ences in the lives of families and did not speak ofnoticing or deriving satisfaction from very modestgains made.

The question of gender

In response to the question about whether gendermight play a part in how the job is experienced, inter-view participants seemed to think it does not. Somesaid they had not thought about it much; others saidthey thought men and women handle the job in afairly similar way. This is consistent with the limitedresearch, although, in light of challenges raised regard-ing what it might mean to be a man in the predomi-nately woman-dominated field of child welfare(Gillingham 2006), this issue warrants further inves-tigation from a critical perspective.

Measuring EE

We note that our measure of EE was only one of thethree scales included in the MBI. While others haveargued that EE is the essence of burnout (e.g. Coxet al. 1993), it may actually measure a more temporarystate. Swimmers’ references to the fluctuation ofintensity in the stress they experienced, to practicalways of mitigating it and to the balance they soughtbetween stress and gratification suggest that highscores on EE do not necessarily predict a lineardecline towards defeat.

IMPLICATIONS FOR SUPERVISION,MANAGEMENT AND EDUCATION

We speculate that some Swimmers may eventuallyevolve into Sailors. Several Sailors referred to havinglearned to cope with the job better over time and tostrengths developed through experience. This raises

Sinking, swimming and sailing D Mandell et al.

390 Child and Family Social Work 2013, 18, pp 383–393 © 2012 John Wiley & Sons Ltd

several questions: Does length of time in the job forgethe Sailing worker or must the ‘right fit’ be presentinitially? Could workers in the Sinking category beindividuals who were not originally committed tocareers in child welfare? Does a Swimmer’s fatedepend on personal resilience? Will some remainSwimmers throughout long careers, withstandingongoing tension between high EE and high JS? Willothers drift into the Sinking category and somemanage the shift to a more comfortable Sailing status?These are questions that our data cannot answer, butthey warrant further study. Our findings do, nonethe-less, point to some directions for improved recruit-ment, selection, training and supervision practices.

The profiles described here could inform the devel-opment of specific workplace scenarios to be posed tojob candidates at the hiring stage. Their responses tothe scenarios might offer insights into their readinessto deal with the challenges that child welfare workpresents. Given the likelihood that some Swimmerscould ultimately grow into the Sailing profile, wesuggest that new workers need help to identify spe-cific, manageable goals and to recognize the value tochildren and families of small changes. The key toencouraging these perceptions is ongoing supportiveand reflective supervision/consultation (see Barthet al. 2008; Dill & Bogo 2009).

Managers should model and proactively supportreasonable boundaries between work and personallife, entailing intentional modification of workplaceculture. This includes encouraging workers andsupervisors to set limits on the work and to acceptthe limits of one’s role in the lives of clients. Systemicchanges should be initiated and supported at theorganizational and community levels so that a senseof contributing to social change and social justice isrecognized as a responsibility the organization, thecommunity, and government share with the indi-vidual worker. We understand that such recommen-dations may be seen as ‘band-aid’ solutions for achild welfare system requiring radical change. Weconcur with some calls in the literature forre-thinking the child welfare paradigms in NorthAmerica (Freymond & Cameron 2006). However,such change will require time to be accepted andimplemented. In light of ample evidence of highlevels of stress and its implications for workers andclients, changing the workplace culture of childwelfare seems to be an achievable shorter termethical imperative.

Such efforts may not enhance the retention of thosewho enter child welfare without deep commitment to

the profession. Consistent with research on the appar-ent buffering effects of such commitment and wantingto ‘make a difference’ (Reagh 1994; Rycraft 1994;Ellett et al. 2007), a relative lack of commitmentclearly has implications for recruitment and initialtraining. Educators know from experience that manyyoung social work graduates enter the field of childwelfare precisely because high turnover means entrylevel positions are often available and the pay is rela-tively good where unionized environments exist. Atbest, many of these individuals see child protectionservices work as a career stepping stone and a place todevelop skills. At worst, some see it as a necessary evilwhen desired positions are unattainable. Social workeducation has a role to play here.Teaching about childwelfare work must include a strong critical analysis,for the health of the system and the benefit of itsservice users. Discussion of power and ethics,however, can be accompanied by discussion of values,goals, boundaries and self-care in order to preparefuture graduates to manage the rigours and find therewards in child welfare work. Educators can alsoencourage students to explore ‘making a difference’ ata collective level.

DIRECTIONS FOR RESEARCH

One question worthy of further study is: Whatbecomes of Swimmers over time? A longitudinal studyof child welfare employees in this category, examiningwhether they continue to work at high levels of EE andJS over long periods or eventually shift into a differentcategory – and what contributes to a shift – woulddeepen understanding of the interplay between work-place conditions, personal characteristics and copingskills in the trajectories of child welfare careers. Evalu-ation studies of organizational changes introducedregarding supervision, hiring practices and ongoingtraining would help confirm the effectiveness of thechanges our findings suggest. Comparison of maleand female child welfare workers’ motivation, values,JS and responses to stressors would also contribute tounderstanding of workers’ experiences in the childwelfare context.

REFERENCES

Acker, G.M. (1999) The impact of clients’ mental illness on

social workers’ job satisfaction and burnout. Health & Social

Work, 24, 112–119.

Anderson, D.G. (2000) Coping strategies and burnout among

veteran child protection workers. Child Abuse & Neglect, 24,839–848.

Sinking, swimming and sailing D Mandell et al.

391 Child and Family Social Work 2013, 18, pp 383–393 © 2012 John Wiley & Sons Ltd

Barth, R.P., Lloyd, E.C., Christ, S.L., Chapman, M.V. &

Dickinson, N.S. (2008) Child welfare worker characteristics

and job satisfaction: a national study. Social Work, 53 (3),199–209.

Boston, T.M. (2009) The Impact of stress, burnout and job satisfac-

tion on rural social workers. (ProQuest Information & Learn-

ing). (Electronic; Print). (AAI3352263; 2009-99190-016).

Cahalane, H. & Sites, E. (2008) The climate of child welfare

employee retention. ChildWelfare, 87 (1), 91–114.

Conrad, D. & Kellar-Guenther, Y. (2006) Compassion fatigue,

burnout and compassion satisfaction among Colorado

child protection workers. Child Abuse & Neglect, 30, 1071–

1080.

Cox, T., Kuk, G. & Leiter, M.P. (1993) Burnout, health, work

stress and organizational healthiness. In: Professional Burnout:

Recent Developments inTheory and Research (eds W.B. Schaufeli,

C. Maslach & T. Marek), pp. 177–193. Taylor & Francis,

Washington, DC.

Dill, K. & Bogo, M. (2009) Moving beyond the administrative:

supervisors’ perspectives on clinical supervision in child

welfare. Journal of Public ChildWelfare, 3, 87–105.

Ellett, A.J. (2009) Intentions to remain employed in child

welfare: the role of human caring, self-efficacy beliefs, and

professional organizational culture. Children andYouth Services

Review, 31 (1), 78–88.

Ellett, A.J., Ellis, J.I., Westbrook, T.M. & Dews, D. (2007) A

qualitative study of 369 child welfare professionals’ perspec-

tives about factors contributing to employee retention and

turnover. Children and Youth Services Review, 29 (2), 264–

281.

Evans, S., Huxley, P., Gately, C., Webber, M., Mears, A., Pajak,

S. et al. (2006) Mental health, burnout and job satisfaction

among mental health social workers in England andWales. The

British Journal of Psychiatry:The Journal of Mental Science, 188,75–80.

Freymond, N. & Cameron, G. (2006) Towards Positive Systems of

Child and Family Welfare: International Comparisons of Child

Protection, Family Service, and Community Care Models. Univer-

sity of Toronto Press, Toronto.

Gillingham, P. (2006) Male social workers in child and family

welfare: new directions for research. Social Work, 51 (1), 83–

85.

Hansung, K. & Stoner, M. (2008) Burnout and turnover inten-

tion among social workers: effects of role stress, job auto-

nomy and social support. Administration in SocialWork, 32 (3),5–25.

Himle, D.P., Jayaratne, S. & Chess, W.A. (1987) Gender differ-

ences in work stress among clinical social workers. Journal of

Social Service Research, 10, 41–56.

Jayaratne, S. & Chess, W.A. (1983) Job satisfaction and burnout

in social work. In: Stress and Burnout in the Human Service

Professions (ed. B. Farber), pp. 129–141. Pergamon Press,

New York.

Kim, H. & Stoner, M. (2008) Burnout and turnover intention

among social workers: effects of role stress, job autonomy and

social support. Administration in SocialWork, 32 (3), 5–25.

Koeske, G.F. & Kelly, T. (1995) The impact of overinvolvement

on burnout and job satisfaction. The American Journal of Ortho-

psychiatry, 65, 282–292.

Koeske, G.F. & Kirk, S.A. (1995) Direct and buffering effects of

internal locus of control among mental health professionals.

Journal of Social Service Research, 20, 1–28.

Koeske, G.F., Kirk, S.A. & Koeske, R.D. (1993) Coping with job

stress: which strategies work best? Journal of Occupational and

Organizational Psychology, 66, 319–335.

Landsman, M.J. (2001) Commitment in public child welfare.

The Social Service Review, 3, 386–419.

Mandell, D., Stalker, C., Harvey, C., Frensch, K. & Ringrose, P.

(forthcoming) When the going gets tough: a workplace study

of four southern Ontario Children’s Aid Societies. In: Towards

Positive Systems of Child and Family Welfare: Families’ Realities

and Service Responses (eds G. Cameron, M. Fine & N. Frey-

mond et al.), University of Toronto Press, Toronto.

Martin, U. & Schinke, S.P. (1998) Organizational and individual

factors influencing job satisfaction and burnout in mental

health workers. SocialWork in Health Care, 28, 51–62.

Mena, K.C. & Bailey, J.D. (2007) The effects of the supervisory

working alliance on worker outcomes. Journal of Social Service

Research, 34 (1), 55–65.

Moore, J.E. (2000) Why is this happening? A causal attribution

approach to work exhaustion consequences. Academy of Man-

agement Review, 25, 335–349.

O’Donnell, J. & Kirkner, S.L. (2009) A longitudinal study of

factors influencing the retention of title IV-E master’s of social

work graduates in public child welfare. Journal of Public Child

Welfare, 3 (1), 64–86.

Quinn, R.P. & Shepard, L.J. (1974) The 1972–73 Quality of

Employment Survey. Survey Research Center, Institute for

Social Research, University of Michigan, Ann Arbor, MI.

Quinn, R.P. & Staines, G.L. (1979) The 1977 Quality of Employ-

ment Survey: Descriptive Statistics, with Comparison Data from

the 1969–70 and the 1972–73 Surveys. Survey Research Center,

Institute for Social Research, University of Michigan, Ann

Arbor, MI.

Reagh, R. (1994) Public child welfare professionals – those who

stay. Journal of Sociology and SocialWelfare, 21, 69–78.

Rycraft, J.R. (1994) The party isn’t over: the agency role in the

retention of public child welfare caseworkers. SocialWork, 39,75–80.

Silver, P.T., Poulin, J.E. & Manning, R.C. (1997) Surviving the

bureaucracy: the predictors of job satisfaction for the public

agency supervisor. The Clinical Supervisor, 15, 1–20.

Stalker, C., Mandell, D., Frensch, K., Harvey, C. & Wright, M.

(2007) Child welfare workers who are exhausted yet satisfied

with their jobs: how do they do it? Child & Family SocialWork,

12 (2), 182–191.

Takeda, F., Ibaraki, N., Yokoyama, E., Miyake, T. & Ohida, T.

(2005) The relationship of job type to burnout in social

workers at social welfare offices. Journal of Occupational Health,

47, 119–125.

Um, M.-Y. & Harrison, D.F. (1998) Role stressors, burnout,

mediators, and job satisfaction: a stress-strain-outcome

Sinking, swimming and sailing D Mandell et al.

392 Child and Family Social Work 2013, 18, pp 383–393 © 2012 John Wiley & Sons Ltd

model and an empirical test. Social Work Research, 22, 100–

115.

Van Yperen, N.W. & Janssen, O. (2002) Fatigued and dissatisfied

or fatigued and satisfied? Goal orientations and responses to

high job demands. Academy of Management Review, 45, 1161–

1171.

Vinokur-Kaplan, D. (1991) Job satisfaction among social

workers in public and voluntary child welfare agencies. Child

Welfare, 70, 81–91.

Yoo, J. (2002) The relationship between organizational variables

and client outcomes: a case study in child welfare. Administra-

tion in SocialWork, 26, 39–61.

NOTES

1 The data from these studies may have revealed find-ings similar to our own had the relationship betweenJS and EE been the focus of the study; however, onlySilver et al. ‘s study commented on the unexpectedrelationship between these variables.2 Meeting criteria for the same category of interest onboth surveys would suggest the scores reflected anongoing response to the workplace and not a reactionto a recent situation.

Sinking, swimming and sailing D Mandell et al.

393 Child and Family Social Work 2013, 18, pp 383–393 © 2012 John Wiley & Sons Ltd