How can managers reduce employee intention to quit

TLDR: In this article, the authors investigated the variables that may be predictive of intention to leave a job, and tested a model that includes mediating variables such as emotional support from supervisors and self-esteem.

Abstract: This paper reports on an investigation of the variables that may be predictive of intentions to leave a job, and tests a model that includes mediating variables. A total of 173 retail salespeople completed questionnaires measuring commitment to the organization for which they worked, job satisfaction, stress, supervisor support, locus of control, self‐esteem, the perceived stressors in the job and their intention to quit. Path analysis was used to test the relationships hypothesized in the model. The majority of hypotheses were supported, with the variables included accounting for 52 per cent of the variance in intention to quit. Emotional support from supervisors and self‐esteem mediated the impact of stressors on stress reactions, job satisfaction, commitment to the organization and intention to quit. It is suggested that to ameliorate intention to quit and in turn reduce turnover, managers need to actively monitor workloads, and the relationships between supervisors and subordinates in order to reduce and manage stress. Managers also need to monitor both the extrinsic and intrinsic sources of job satisfaction available to employees. These activities could assist in maintaining and increasing job satisfaction and commitment to the organization.

Full text

Deakin Research Online
This is the authors’ final peer reviewed (post print) version of the
item published as:
Firth, Lucy, Mellor, David, Moore, Kathleen A and Loquet, Claude 2004,
How can managers reduce employee intention to quit?, Journal of
managerial psychology, vol. 19, no. 2, pp. 170-187.
Available from Deakin Research Online:
http://hdl.handle.net/10536/DRO/DU:30008720
Reproduced with the kind permission of the copyright owner.
Copyright : 2004, Emerald Group Publishing

Howcanmanagersreduceemployeeintentiontoquit?
LucyFirth,DepartmentofInformationSystems,UniversityofMelbourne,Melbourne,
Australia
DavidJ.Mellor,SchoolofPsychology,DeakinUniversity,Melbourne,Australia
KathleenA.Moore,SchoolofPsychology,DeakinUniversity,Melbourne,Australia
ClaudeLoquet,SchoolofPsychology,DeakinUniversity,Melbourne,Australia
Abstract
Thispaperreportsonaninvestigationofthevariablesthatmaybepredictiveofintentions
toleaveajob,andtestsamodelthatincludesmediatingvariables.Atotalof173retail
salespeoplecompletedquestionnairesmeasuringcommitmenttotheorganizationfor
whichtheyworked,jobsatisfaction,stress,supervisorsupport,locusofcontrol,self‐esteem,
theperceivedstressorsinthejobandtheirintentiontoquit.Pathanalysiswasusedtotest
therelationshipshypothesizedinthemodel.Themajorityofhypothesesweresupported,
withthevariablesincludedaccountingfor52percentofthevarianceinintentiontoquit.
Emotionalsupportfromsupervisorsandself‐esteemmediatedtheimpactofstressorson
stressreactions,jobsatisfaction,commitmenttotheorganizationandintentiontoquit.Itis
suggestedthattoameliorateintentiontoquitandinturnreduceturnover,managersneed
toactivelymonitorworkloads,andtherelationshipsbetweensupervisorsandsubordinates
inordertoreduceandmanagestress.Managersalsoneedtomonitorboththeextrinsicand
intrinsicsourcesofjobsatisfactionavailabletoemployees.Theseactivitiescouldassistin
maintainingandincreasingjobsatisfactionandcommitmenttotheorganization.
“Whatdeterminesemployeeturnover?”Theanswertothisquestionhasgreatrelevanceto
theindividualwhomaybethinkingaboutquittingajob,andforthemanagerwhoisfaced
withlackofemployeecontinuity,thehighcostsinvolvedintheinductionandtrainingof
newstaff,and,notleast,issuesoforganizationalproductivity.Whileactualquitting
behaviouristheprimaryfocusofinteresttoemployersandresearchers,intentiontoquitis
arguedtobeastrongsurrogateindicatorforsuchbehaviour.Jobstressorsandlackofjob
satisfactionareamongthefactorsthatcontributetopeople'sintentiontoquittheirjobs
(Moore,2002);however,itisimportantbothfromthemanager'sandtheindividual's
perspectivetounderstandthefactorsthatmediatetherelationshipbetweenjobstressand
intentiontoquit.Thisstudyhadthreeaims:
1. Weaimedtoinvestigatetheeffectofjobstressorsonjobengagement(job
satisfaction,jobcommitmentandfeelingsofjobstress),andtorelateallofthese
variablestointentiontoquit.
2. Weaimedtoinvestigatewhetherpeople'sdispositionalfactors(locusofcontrol,
self‐esteemandperceptionsofsocialsupport)weremediatorsofintentiontoquit.
3. Weaimedtoprovideacomprehensivemodelofintentionstoquitthatcouldbe
extrapolatedtoinformquittingbehaviour.
Intentionsare,accordingtoresearcherssuchasAjzenandFishbein(1980)andIgbariaand
Greenhaus(1992),themostimmediatedeterminantsofactualbehaviour.Theyarealsoof

practicalmeritfromaresearchperspective,asoncepeoplehaveactuallyimplementedthe
behaviourtoquit,thereislittlelikelihoodofgainingaccesstothemtounderstandtheir
priorsituation.Thevalidityofstudyingintentionsintheworkplacecanalsobedrawnfrom
Sager's(1991)longitudinalstudyofsalespeople,inwhichintentiontoquitwasfoundto
discriminateeffectivelybetweenleaversandstayers.However,whileitisreasonableto
arguethatintentionsareanaccurateindicatorofsubsequentbehaviour,westilldonot
knowwhatdeterminessuchintentions.
Numerousresearchers(e.g.Bluedorn,1982;KalliathandBeck,2001;Krameretal.,1995;
Petersetal.,1981;Saks,1996)haveattemptedtoanswerthequestionofwhatdetermines
people'sintentiontoquitbyinvestigatingpossibleantecedentsofemployees’intentionsto
quit.Todate,therehasbeenlittleconsistencyinfindings,whichispartlyduetothe
diversityofconstructsincludedbytheresearchersandthelackofconsistencyintheir
measurementsbutalsorelatestotheheterogeneityofpopulationssampled.Further,some
authorshavereportedvalidityco‐efficients(readfactorloadings)forthevariablesunder
investigationbut,whilestatisticallysignificant,theseareoftenoflittlepracticalutility.Such
authorshavealsofailedtoreporttheamountofvarianceinintentiontoquitexplainedby
thefactorsintheirmodels.
Despitetheselimitations,amongthevariablesconsistentlyfoundtorelatetointentionto
quitare:
 theexperienceofjob‐relatedstress(jobstress);
 therangeoffactorsthatleadtojob‐relatedstress(stressors);
 lackofcommitmenttotheorganization;and
 jobdissatisfaction(e.g.GreenhausandBeutell,1985;IgbariaandGreenhaus,1992;
Kahnetal.,1964;Leongetal.,1996;Lumetal.,1998;Petersetal.,1981;Rahimand
Psenicka,1996).
Thesevariablescanbemediatedbypersonalordispositionalfactorsandbyenvironmental
ororganizationalfactors.Amongthepersonalfactorsthatmediatebetweenstressorsand
intentiontoquitareaspectsofpersonalagency,self‐esteemandsocialsupport(e.g.Avison
andGotlib,1994;Coyneetal.,1990;CoyneandDowney,1991;TurnerandRoszell,1994).
Personalagencyreferstoconceptssuchasasenseofpowerlessness,locusofcontroland
personalcontrol.Researchfindingsstronglysuggestthatagreatersenseofpersonalagency
isassociatedwithareducedriskofnegativeoutcomesfollowingmajornegativelifeevents
androle‐relatedstress(seeTurnerandRoszell,1994).Oneofthepersonalagencyvariables,
locusofcontrol,whichreferstotheextenttowhichpeoplebelievetheyorexternalfactors
suchaschanceandpowerfulothersareincontroloftheeventsthatinfluencetheirlives
(Levenson,1974;Rotter,1966),hasbeenstudiedinrelationtotheworkplace.Rahimand
Psenicka(1996)foundthataninternallocusofcontrolmediatedtherelationshipbetween
jobstressorsandtheintentiontoleaveajob.Morespecifically,internallocusofcontrolwas
foundtobepositivelyrelatedtojobsatisfaction(Sujan,1986)andtobeingemployed
(WatersandMoore,2002),whileanexternalorientationwasfoundtobenegativelyrelated
tojobsatisfaction(BehrmanandPerreault,1984)andwashigheramongunemployed
versusemployedpeople(WatersandMoore,2001).

Self‐esteemreferstotheevaluationthatindividualsmakeandcustomarilymaintainwith
regardtothemselves.ReviewsbyKaplan(1975,1980)andsubsequentstudiesbyother
researchers(e.g.Silverstone,1991;WatersandMoore,2002)haveconsistentlyshownthat
lowself‐esteemisrelatedtopsychologicalproblems,unemploymentandmaladaptive
behaviours.Anumberofotherstudieshaveprovidedsupportforthecontentionthathigh
self‐esteemiscorrelatedwithjobsatisfaction(e.g.GreenhausandBadin,1974;Inkson,
1978;Kohli,1985;Teas,1981,1982).InastudybyMoore(2002),arelatedconstruct,self‐
efficacy(assessedspecificallyamongnursesinrelationtotheirprofessionalabilities),was
associatedwithreducedintentiontoquit(β=−0.23).
Socialsupporthasbeenshowntoplayanimportantroleinmitigatingintentiontoquit,
althoughnotallfindingshavebeeninagreement.Forexample,Moore(2002)foundthat
socialsupportfromsupervisorsreducedthelevelofnurses’burnoutandindirectly,through
reducedlevelsofburnout,reducednurses’intentiontoquit.Asimilarresultwasreported
byKalliathandBeck(2001)whentheytestedtheimpactofsocialsupportontwo
componentsofburnout,namelydepersonalizationandemotionalexhaustion,andfound
thatsupervisorysupportreducednotonlythosesymptomsofburnoutbutalsodirectlyand
indirectlynurses’intentiontoquit.Ontheotherhand,Munn etal.(1996),inastudyof
Americanchildlifespecialists,foundlackofsupervisorsupportwasthebestpredictorofjob
dissatisfactionandintentiontoleaveajob,whileHattonandEmerson(1998)foundthat
actualstaffturnoverwaspredictedinpartbylowlevelsofsupportfromsuperiors.
However,otherstudies(e.g.RahimandPsenicka,1996)havefailedtofindamoderating
effectforsocialsupportintherelationshipbetweenjobstressorsandintentiontoquit.
Otherresearchers(e.g.Coghlan,1984;KellyandCross,1985)havefoundthatratherthan
supervisors’support,itisthesupportgainedfromtalkingwithpeers,familyandfriendsthat
isfrequentlycitedasasourceofstressreduction.Consistentwiththesefindings,Freddolino
andHeaney(1992)foundthatpeersocialsupportwasassociatedwithhigherjob
satisfactionamongdirectcarestaffandhomemanagersforintellectuallydisabledclients,
whileintentiontoquitwasassociatedwiththepresenceofsocialunderminingbyco‐
workersandprovideragencies.However,thebulkofevidencesuggeststhatitissituation‐
specificsupport,thatis,work‐supervisor/home‐family(e.g.TinkerandMoore,2001),
thatis
mosteffective.Accordingly,inthisstudyweexaminedtheroleofsupervisors’supportin
employees’intentiontoquit.
Othervariablesimportantinthecontinuityofemploymentareemployees’jobsatisfaction
andtheircommitmenttotheorganization.Therelationshipsbetweenthesevariablesand
intentiontoquithavebeenfoundto
prevailacrossarangeofoccupations.Forinstance,
Wunderetal.(1982)foundthatjobstressorshadadirect,negativeeffectonjobsatisfaction
amongmanagersofalargeinternationalmanufacturer,whichledtoareducedcommitment
totheorganizationandtointentiontoquitandactualquittingbehaviour.However,several
otherstudies(e.g.Armstrong‐Stassenetal.,1994;IgbariaandGreenhaus,1992;Koeskeand
Koeske,1993;TinkerandMoore,2001)havefoundnodirecteffectofstressorsonintention
toquit,butratherindirecteffectsthroughtheexperienceofjobstress,socialsupport,job
satisfactionandlackofcommitmenttotheorganization.Forexample,Igbariaand
Greenhaus(1992)reportedthatamongmanagementinformationsystemspersonnel,the
mostimmediatedeterminantsofturnoverintentionswerelackofjobsatisfactionandlack
ofcommitmenttotheorganization.Itisimportant,therefore,toevaluateboththedirect
andindirecteffectsofjobstressorsonintentiontoquit.

Thisstudyassessedtheimpactofjobstressorsonintentiontoquitusingthedispositional
factors:locusofcontrol,self‐esteem,andperceptionsofsupervisorsupport;andmediating
variablesthatarerelatedtojobengagement:commitmenttotheorganization,job
satisfaction,andfeelingsofstress.ThehypothesizedmodelispresentedinFigure1.
Method
Participants
Atotalof173salespeoplewererecruitedfromtheclothingsectionsofalargedepartment
storeinAustralia.Thesampleconsistedof164femalesandninemales,withameanageof
35.95years(SD=12.66,R=18‐61).Educationlevelsvariedfromcompletingsometimeathigh
schooltocompletingadegree.Intermsofyearsofservice,11participantshadworkedat
thestoreforlessthanoneyear,33foronetothreeyears,60forfourtosixyears,27for
seventotenyears,and42formorethantenyears.
Measures
AquestionnaireadaptedfromthecomprehensiveworkplacescaledevelopedbyTateetal.
(1997)fortheirtri‐nationstudywasusedinthisstudy(seetheAppendix).Tateand
colleaguesreportedadequateinternalreliabilityforeachfactorofthescale(α=0.61to
0.91).Thefactorsassessedwerestressors,jobstress,jobsatisfaction,commitmenttothe
organization,locusofcontrol,self‐esteem,supportofferedbysupervisorsandintentionto
quit.Eachofthefactorsisexplainedbelow.
Stressorsmeasuredfouraspectsofstress.Threeitemsmeasuredeachofthefollowing
stressors:
 roleambiguity(e.g.myjobresponsibilitiesarenotcleartome);
 roleconflict(e.g.tosatisfysomepeopleatmyjob,Ihavetoupsetothers);
 work‐overload(e.g.itseemstomethatIhavemoreworkatmyjobthanIcan
handle);and
 work‐familyconflict(e.g.myworkmakesmetootiredtoenjoyfamilylife).
Itemswereansweredonafive‐pointLikertscale,fromagreetodisagree.
Jobstresswasmeasuredwiththreeburnoutitems(e.g.Ifeelemotionally‐drainedbymy
job)andfiveitemsrelatedtoanxietyandsomatic
complaints(e.g.job‐relatedproblems
keepmeawakeatnight;Ifeeltenseatmyjob).Participantsindicatedonasix‐pointscale
thedegreetowhichtheyexperiencedeachofthesesymptoms.
Jobsatisfactionwasmeasuredusingafive‐pointLikertscaletoassessparticipants
agreementwithfourstatementsrelatingtoextrinsicfactors(e.g.jobsecurity,physical
conditions),andfourstatementsrelatingtointrinsicfactors(e.g.therecognitionreceived
forworkdone,thefreedomgiventodoone'sbestatthejob).
Commitmenttotheorganizationwasassessedviafiveitemsratedonafive‐pointscale(e.g.
Ireallycareaboutthefateofthisstore).

Frequently asked questions

Q1. What are the contributions in this paper?

This paper reports on an investigation of the variables that may be predictive of intentions to leave a job, and tests a model that includes mediating variables. This study had three aims: 1. the authors aimed to investigate the effect of job stressors on job engagement ( job satisfaction, job commitment and feelings of job stress ), and to relate all of these variables to intention to quit. 2. the authors aimed to investigate whether people 's dispositional factors ( locus of control, self‐esteem and perceptions of social support ) were mediators of intention to quit. 3. the authors aimed to provide a comprehensive model of intentions to quit that could be extrapolated to inform quitting behaviour. One of the personal agency variables, locus of control, which refers to the extent to which people believe they or external factors such as chance and powerful others are in control of the events that influence their lives ( Levenson, 1974 ; Rotter, 1966 ), has been studied in relation to the work place. In a study by Moore ( 2002 ), a related construct, self‐ efficacy ( assessed specifically among nurses in relation to their professional abilities ), was associated with reduced intention to quit ( β=−0. 23 ). On the other hand, Munn et al. ( 1996 ), in a study of American child life specialists, found lack of supervisor support was the best predictor of job dissatisfaction and intention to leave a job, while Hatton and Emerson ( 1998 ) found that actual staff turnover was predicted in part by low levels of support from superiors. Consistent with these findings, Freddolino and Heaney ( 1992 ) found that peer social support was associated with higher job satisfaction among direct care staff and home managers for intellectually disabled clients, while intention to quit was associated with the presence of social undermining by co‐ workers and provider agencies. Accordingly, in this study the authors examined the role of supervisors ’ support in employees ’ intention to quit. This study assessed the impact of job stressors on intention to quit using the dispositional factors: locus of control, self‐esteem, and perceptions of supervisor support ; and mediating variables that are related to job engagement: commitment to the organization, job satisfaction, and feelings of stress. The hypothesized model is presented in Figure 1. It is suggested that to ameliorate intention to quit and in turn reduce turnover, managers need to actively monitor workloads, and the relationships between supervisors and subordinates in order to reduce and manage stress. Further, some authors have reported validity co‐efficients ( read factor loadings ) for the variables under investigation but, while statistically significant, these are often of little practical utility. Research findings strongly suggest that a greater sense of personal agency is associated with a reduced risk of negative outcomes following major negative life events and role‐related stress ( see Turner and Roszell, 1994 ). However, the bulk of evidence suggests that it is situation‐ specific support, that is, work‐supervisor/home‐family ( e. g. Tinker and Moore, 2001 ), that is most effective.