Are high performance work practices really necessary in family SMEs? : An analysis of the impact on employee retention

http://www.diva-portal.org

This is the published version of a paper published in The Journal of Family Business Strategy.

Citation for the original published paper (version of record):

Pittino, D., Visintin, F., Lenger, T., Sternad, D. (2016)

Are high performance work practices really necessary in family SMEs?: An analysis of the

impact on employee retention.

The Journal of Family Business Strategy, (2): 75-89

https://doi.org/10.1016/j.jfbs.2016.04.002

Access to the published version may require subscription.

N.B. When citing this work, cite the original published paper.

Open Access

Permanent link to this version:

Are

high

performance

work

practices

really

necessary

in

family

SMEs?

An

analysis

of

the

impact

on

employee

retention

Daniel

Pittino

a

,

Francesca

Visintin

b,

*

,

Tamara

Lenger

c

,

Dietmar

Sternad

c

a

JönköpingInternationalBusinessSchool,Gjuterigatan5,S-55111Jönköping,Sweden

b

UniversityofUdine,DepartmentofEconomicsandStatistics,ViaTomadini30a,I-33100Udine,Italy

c

CarinthiaUniversityofAppliedSciences,SchoolofManagement,Europastrasse4,A-9524Villach,Austria

ARTICLE INFO

Articlehistory: Received9July2015

Receivedinrevisedform24March2016 Accepted11April2016

Availableonline17May2016 Keywords:

Highperformanceworkpractices Familyfirms

Employeeretention Organizationalsize

ABSTRACT

Thepresentstudyexaminestheadoptionofhigh-performanceworkpractices(HPWPs)infamilyversus non-familyfirms,andevaluatestheeffectofsuchpracticesontheretentionofvaluableemployeesin lightofdifferentfamilycontingencies.Theanalysisiscarriedoutonasampleof232companieswith morethan10andlessthan250employees,operatinginAustriaandHungary.Theresultsindicatethat therelationalmechanismsoriginatingfromthefamilysocialcapitalmayactassubstitutesofformal practicesaimedatfosteringemployeeinvolvementandcommitmentandthatthosemechanismsdepend ontheextentfamilyinvolvementinthegovernanceofthebusiness.

ã2016TheAuthors.PublishedbyElsevierLtd.ThisisanopenaccessarticleundertheCCBY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

1.Introduction

Acquiring, developing,but particularly alsokeeping talented individuals have become someof themost important goals of humanresource management (HRM) practices in thepast two decades(Govaerts, Kyndt,Dochy, & Baert,2011; Hiltrop,1999). Organizationsthatdonotsucceedinretainingthehighperforming workforce lose their ability to remain competitive (Rappaport, Bancroft,&Okum,2003)asvoluntaryturnovercanaffectvarious dimensionsof performance, suchasproductivity andcorporate financial performance (Guthrie, 2001; Huselid, 1995; Shaw, Dineen,Fang,&Vellella,2009;Sun,Aryee,&Law,2007).

Therecentliteratureonemployeeretentionhasunambiguously shownthatoneparticularsetofHRMpractices,knownashigh performance work practices (HPWPs), plays a significant role towardstheintermediateorganizationalgoalofcurbingvoluntary turnover rates. Indeed HPWPs have been shown to positively influenceemployeeretentioninavarietyofsectors (production and service), settings(unionised and non-unionised,small and large companies, etc.) and countries (e.g. Combs, Liu, Hall, & Ketchen,2006;Karatepe,2013;Sarikwal&Gupta,2013).

Alsoinfamilyfirms,attractingandretainingqualified employ-eesandfosteringvalue-creatingbehaviourscanbecrucialfactors

for performance and long-term competitive advantage (e.g.

Chrisman,Chua,&Litz,2003;Mitchell,Morse,&Sharma,2003;

Sieger,Bernhard,&Frey,2011).Theissueinthefamilybusiness contextmightbeevenmorecriticalbecausethemajorityofthe employees usually face a particularly complex and ambiguous situation,sincetheyarepartofthebusiness,butnotofthefamily system(Barnett&Kellermanns,2006).

Despitetheevidentconceptualandpracticalrelevanceofthe topic, the theoretical and empirical contributions on HRM practices in family firms are, however, rather scarce and fragmented (Botero&Litchfield, 2013; Cruz, Firfiray,& Gomez-Mejia,2011).Further,theexistingstudieshavemainlyexamined HRMchoicesinvolvingfamilymembers,whilehaving substantial-lyoverlooked,oraddressedonlyinanindirectway,theroleof non-family personnel (e.g. Astrachan & Kolenko, 1994; Burkart, Panunzi,&Shleifer, 2003; Cruz, Gomez-Mejia,&Becerra, 2010;

Dawson,2012;Eddleston&Kellermanns,2007;Khanin,Turel,& Mahto, 2012; Lansberg, 1983; Lubatkin, Schulze, Ling, & Dino, 2005;Sharma&Irving,2005).

Theexistingevidenceshowsthattheextentof useofformal HPWPsisgenerallylowerinfamilyfirmscomparedtonon-family firms,alsowhensizedifferencesaretakenintoaccount(e.g.DeKok, Uhlaner&Thurik,2006;Pittino&Visintin,2013).Tothecontrary, familyfirmsappeartopreferinformal,oftendiscretionarymeasures, often leading (a) to the adoption of unfair compensation and performanceappraisalsystems(Barnet&Kellermanns,2006;Chua, Chrisman,&Bergiel,2009;Fiegener,Brown,Prince,&File, 1994),(b) tocasesofmanagerialentrenchment(Gomez-Mejia,Nuñez-Nickel, * Correspondingauthor.

E-mailaddresses:daniel.pittino@ju.se(D.Pittino),visintin1@uniud.it

(F.Visintin),tamara.lenger@aon.at(T.Lenger),d.sternad@cuas.at(D.Sternad).

http://dx.doi.org/10.1016/j.jfbs.2016.04.002

1877-8585/ã2016TheAuthors.PublishedbyElsevierLtd.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).

ContentslistsavailableatScienceDirect

Journal

of

Family

Business

Strategy

&Gutierrez, 2001),(c) tonepotismin personnel recruitingandcareer decisions(Spranger,Colarelli,Dimotakis,Jacob,&Arvey,2012),(d)to scapegoating of non-family executives and employees ( Gomez-Mejia,Larraza-Kintana,&Makri,2003),and,generally,(e)tolower actionsfortheengagementandworkerdevelopmentofnon-family employees (Cruz, Larraza Kintana, Garcés Galdeano, & Berrone, 2014). For the above-mentioned reasons, a growing number of studiesarguefortheneedtointroducemoreformalandequitable practices. Others argue quite the opposite, and suggest that, comparedto non-familyfirms,despite theloweruse of formal HPWPs,familycompaniesareinclinedtotreattheiremployeeswith higherconsideration.Family_firmshavebeenshowntoinvestin building cohesive internal communities (Chrisman, Chua, & Kellermanns, 2009; Miller, Le Breton-Miller, & Scholnick, 2008;

Miller,Lee,Chang,&LeBreton-Miller,2009),tobemoresensitiveto thequalityoflifeandwelfareoftheirworkers(Cennamo,Berrone, Cruz,&Gomez-Mejia,2012)andtopromoteemployees’ involve-mentandlong-termworkrelations(Kang&Kim,2014).

Thepresentworkattemptstoreconciletheseopposingviews, byexaminingaparticularoutcomeofnon-familyhumanresource management,namelytheretentionofvaluableemployees,andby assessingtowhat extent HPWPs contributetothis outcome at variouslevelsoffamilyinfluence.Indoingthis,wedrawfroma socialexchangetheoreticalperspective(Blau,1964;Cropanzano& Mitchell,2005).Asocial exchangeviewpresumesthat employ-ment practices would motivate employees to reciprocate with performance,commitmentandintentiontostay(Jia,Shaw,Tsui,& Park,2014).Ourkeyargumentisthateveniffamilyfirmsrelyless onformalHPWPs,theyachievehighretentionresultsbecausethe socio-emotionalcomponentofthefamilialmanagement philoso-phiestriggersindividualattitudes,and feelings ofreciprocation (Cennamoetal.,2012;Cropanzano&Mitchell,2005)thattranslate inlowerlevelsof turnover.Weexpectsuchpracticestobeless pronounced in non-family firms and in companies where the familyinfluenceislower.

Intheempiricalanalysis,carriedoutonanoriginalsampleof 232 small to medium-sized companies (121 family firms and 111non-familyfirms)locatedinAustriaandHungary,wecompare primarilyfamilyandnon-familyfirms.Thecomparisonisrelevant, asfamilycontrolandinfluencehavebeenshowntosignificantly affectthe adoption of informal HR practices at theexpense of formalones,eveninsmallerfirms,wherethelevelofformalization isgenerallylowercompared tolargecompanies(e.g.Aldrich& Langton,1997;DeKoketal.,2006;Kotey&Folker,2007;Pittino& Visintin,2013).Thus,weexaminewhetherthefamilyfirmstatus, asadistinctivefeatureaffectingHRMpractices,hasanimpacton employeeretention.Thisfocusgeneratesafurthermotivationfor distinguishingbetweenfamilyandnon-familyfirms,becausein family firms the majority of the employees face a particularly complex and ambiguous situation, since they are part of the business,but notof thefamily system (Barnett &Kellermanns, 2006).Ontheotherhand,innon-familyfirmswedonotobserve this distinction between social groups based on family and businesssystems_’overlap.Also,werecognizetheneedtoconsider theheterogeneityamongfamilyfirms(Chua,Chrisman,Steier,& Rau,2012),andcomparedifferentsituationsoffamilyinfluence, expressedbyfamilyversusnon-familyCEOleadership andfirst generationversussecondandlatergenerationfamilycontrol.

Thestudyhasproducedthreesetsofresults:firstitshowsthat thefamilyfirmstatusandtheconditionsstrengtheningthefamily influenceonthefirmhaveapositiveimpactonretention;secondit showsthatinthepresenceofHPWPs,the“familyeffect”combines with the formal practices to produce a positive influence on retention;finally,italsosuggeststhatsomeoftheoutcomesofthe familyinfluenceandoftheadoptionofformalpracticesmaybeso similarthattheyproducesubstitutioneffectsonretention.

Ourresultscontributetotheliteratureonbothfamilybusiness andhumanresourcemanagement.First,weaddtothedebateon the management of non-family employees in family firms by focusingnotonlyontheadoptionofhumanresourcepractices,like mostofthestudiessofar(e.g.Cruzetal.,2014;DeKoketal.,2006; Reid&Adams,2001), butalsoonasignificantoutcomeofsuch practices,namelytheretentionoftalentednon-familyemployees. Tothebestofourknowledge,thisstudyisoneofthefirstattempts toconceptualizeandempiricallyexplorethecausallinkbetween HRM practices and behaviours of non-family employees. The findingshelpustoreconcilethepreviousconflictingarguments abouttheattitude towards non-familyemployees inthefamily businessliterature(e.g.Cennamoetal.,2012;Cruzetal.,2014).

Moreover, we extend the existing work that applies social exchangetheoryin family firms,which sofarhasbeenmainly theoretical(e.g.Barnett,Long,&Marler,2012;Long&Mathews, 2011; Pearson & Marler, 2010), byempiricallytesting a typical response to the social exchange process, namely employees’ intentiontostay(Cropanzano&Mitchell,2005),anddrawingfrom theresultsfurtherconceptualimplicationsonthedimensionsof socialexchangeinthefamilybusinesssetting.

WealsocontributetotheHRMliterature,andinparticularto thestudyofHPWPsfromasocialexchangeperspective( Cropan-zano&Mitchell,2005;Shawetal.,2009)andtotheresearchonfit among HRMpractices (Delery,1998; Guest, Paauwe, &Wright, 2012),showinghowtheownershipandleadershipcontextsmay providesocialresourcesthatcombinewiththeformalpractices through additive and substitution effects to elicit positive responsesfromtheemployees.

Thearticleis structuredasfollows.Wefirstreviewthemain theoriesand empirical findingsonthe adoptionof HPWPsand their effects on organizational outcomes. Based on a social exchange viewpoint, we then develop our hypotheses on the relationship between HPWPs and the retention of valuable employees, firstin family versusnon-familyfirms, and thenin differenttypesoffamilyfirms.Thehypothesessectionisfollowed byourmethodsandfindings.Thenweconcludewithadiscussion ofimplicationsandlimitationsofthestudy.

2.EmployeeretentionandHPWPs

Asmentionedabove,activitiesaimedatacquiring,developing and keeping talented individuals have become central in the management of human resources over the last twenty years (Govaerts et al., 2011; Hiltrop,1999). HPWPs, a set of specific humanresourcespracticesdeveloped duringthe1990s (Combs et al., 2006) and including techniques ranging from selective staffingtoemployeetraining,teamworkingandother arrange-mentsillustratedbelow,haveshowntohaveapositiveimpacton employeeretentioninavarietyofsettings,includingsteel mini-mills (Arthur, 1994), automotive assembly plants (MacDuf_fie, 1995),truckingorganizations(Shaw,Gupta,&Delery,2005),call centers(Batt,2002),smallfirmsacrossavarietyofindustriesin boththemanufacturingandservicesectors(Kerr,Way,&Thacker, 2007;Patel&Conklin,2012),andpubliclyheldfirmsinallmajor USindustries(Huselid,1995),aswellasacrossdifferentnationsin NorthAmerica(e.g.Batt,2002;Kerretal.,2007),Europe(Patel& Conklin, 2012; Razouk, 2011), Asia (Su &Wright, 2012; Wang, Bruning,&Peng,2007),andOceania(Guthrie,2001).

TheimpactofHPWPsonemployeeretentionhasbeenshownto leverageonsystemsof coherentpracticesthat affectindividual level attitudes and behaviours (Jiang, Lepak,Hu, &Baer, 2012;

Sparham&Sung,2008;Su&Wright,2012).Inparticular,empirical evidence indicates that the higher the number of HPWPs employed,themorepositivetheemployeeresponses(e.g.Delery &Shaw,2001;Guestetal.,2004;Macky&Boxall,2007).

Followingarecentclassification(Lepak,Liao,Chung,&Harden, 2006),HPWPspracticescanbegroupedintothreesetsaccording totheprimarygoaltheypursue.Theempiricalevidenceshowsthat allthreesetsofpracticesdirectlyorindirectlyinfluencealsothe extentofvoluntaryturnover.Humancapital,motivationandsocial exchangetheorieshelpexplainingthephenomenon.

Skill-enhancing practices, such as selectivestaffing, intensive training,andcareerdevelopment,aimatdevelopingtheabilities andself-efficacyoftheworkforce.Suchpracticeshavebeenshown tonegativelyinfluencevoluntaryturnover(seeforexampleArthur, 1994;Guthrie,2001).Humancapitaltheorysuggeststhat employ-eeswithhighquali_ficationstendtolearnmoreeasilyatworkand asaconsequencedevelopafirm-specifichumancapitalthatwould notbevaluableinotheroccupations(Lepak&Snell,1999).Indeed, themoretalentedandvaluableemployees,eventhoughhavemore opportunities in the labour market, appear to increase their positiveemotionalresponsestotheorganizationandtobeless attractedbyalternativejobs(Jiangetal.,2012;Maertz&Griffeth, 2004;Shawetal.,2009).

Further,selectivestaffing(e.g.throughselectioncriteriasuchas person-jobfit,opennesstolearn,ortheabilitytoworkinteams) (MacDuffie,1995),canalsoenhancethecommitmentofnewbut also of existing employees, since a rigorous selection process signalsthatonlythebestarechosen,thattheorganizationoffers goodjobopportunities,thathighperformanceisrequired,andthat peoplematter(Pfeffer,1994).Likewise,extensivetraining possi-bilitiesbytheirverynatureincreasetheskillsandabilities,butcan alsohaveapositiveeffectonretentionastheyhelpemployeesto reachtheircareergoals(Batt,2002;Huselid,1995;Stovel&Bontis, 2002).Traininganddevelopmentactivitiescanalsobepartofmore comprehensivecareerdevelopmentinitiatives.Transparentcareer pathsand opportunitiescanhaveapositiveeffectonemployee intentiontostay(Jiangetal.,2012),astheyenableemployeesto extend their abilities through mastering new challenges and performing a broader set of tasks that may not be offered in differentcompanies.

Alsomotivation-enhancingpractices, which aim at increasing thewillingnessofemployees toapplytheirskills andusetheir knowledgeintheirdailyworkinglives,andtoalignindividualand organizationalinterestsand goals, haveshowntoinfluencethe extentof job satisfaction and organizational commitment, and therefore the intention to stay. Among these, the provision of generous pay and benefits are extensively used to increase employees’motivation(Middlebrook,1999).Althoughthe moti-vationaleffectofcompensationwasalsoquestioned,asextrinsic rewardscouldpotentiallyundermineintrinsic motivation(Deci, 1971;Oh&Lewis,2009),manyresearchersincludesomeformof variable compensation, particularly based on individual or company performance, aspartof their proposed setof HPWPs (Delaney&Huselid,1996; Patel&Conklin,2012;Patel, Messer-smith,&Lepak,2013;Selsetal.,2006;Wright,Gardner,Moynihan, &Allen,2005).PerformanceappraisalsareanotherHRMtoolthat canbeusedtotriggerpositiveemployeeattitudesandbehaviour, throughprovidingfeedback,clarifyingexpectations,and discus-singcareer developmentopportunities (Razouk, 2011). Huselid (1995)arguedthatappraisalsareparticularlyeffectiveinretaining individualswhentheyaretightlylinkedtoincentivecompensation systems.

Finally,opportunity-enhancingHRMpractices,aimedat increas-ing the empowerment of the employees and at creating opportunities for them to participate to the decision making process, have shown to increase the extent of employees’ organizational commitment(Lines, 2004)and intentiontostay (Yücel,2012).Moregenerally,havingavoiceindecision-making processescanbeseenasanindicatorofrespectforaperson,and canleadtoanincreaseintheidentificationwiththecompany(De

Cremer, 2002; Lind & Tyler, 1988). Instruments that offer structured opportunities to participate—for example through decentralized decision making, autonomy in organizing one’s ownwork,consultationand regularinformationexchange, self-managedteams,orformalgrievanceprocedures—arealsoincluded inthisgroupofHPWPs(e.g.Patel&Conklin,2012;Pateletal.,2013;

Prince,Katz,&Kabst,2011;Selsetal.,2006).

ThethreesetsofHRMpracticeshavebeenshowntoinfluence theextentofretentionthroughanotherimportantmechanismthat is well explained bysocial exchange theory(Blau,1964; Luna-Arocas,&Camps,2007;Patel&Conklin,2012).

Accordingtothisapproach,theemployee_–employer relation-ships are based both on economic and social exchange.Social exchangerelationshipsrelyonmutualobligations,whichinvolve resourcesand behaviours thatare oftendifficulttospecify and measure(Blau, 1964).Fromthepointofviewoftheemployer,these behaviours mayinclude, for example, theinterest in employee well-being, stability,and careeradvancement, as well as more symbolicdimensions,suchasthecommitmenttopromises and thecapacitytoensurefairnessandproceduraljustice(Cropanzano &Mitchell,2005;Guest,1999;Macky&Boxall,2007;Tsui,Pearce, Porter,&Tripoli,1997).On theotherhand, theemployeesmay reciprocatewithlong-termandopen-endedcontributions,suchas willingnesstolearnfirm-specificskills,availabilityfortasksthat may falloutside of theformal job description,a highsense of responsibility, and also a strong intention to stay with the organization(Shaw,Delery,Jenkins,&Gupta,1998).Theadoption ofHPWPsbytheorganizationhasbeenshowntostrengthenthe social dimensionoftheexchange,therebystimulatingthe long-term commitment of the employees (Macky & Boxall, 2007; Maertz&Campion,2004).

FormalHPWPs,however,arenottheonlymechanismthatleads to the urge of reciprocating by the employees. As it will be explainedinthenextsection,alsoinformalmechanismsstemming frompeculiar organizationalenvironmentsor individual behav-ioursbyentrepreneursandtopmanagersmaytriggerthiskindof phenomenon.Inparticular,wesuggestthatsomeofthefamilial informalpracticesoperateattheindividuallevelbyreducingthe intentiontoleaveandincreasingtheretentionrate(Combsetal., 2006).

3.HPWPsandretentionofvaluableemployees:familyversus non-familyfirms

AlargepartofthescantempiricalevidenceonHRMinfamily businessshowsthat therateofadoptionofHPWPsisgenerally lowerinfamilyfirmscomparedtotheirnon-familycounterparts (e.g.Carney,2005;Cruzetal.,2014;DeKoketal.,2006;Pittino& Visintin,2013;Stewart&Hitt,2012).Familyfirmsappeartoprefer informal,oftendiscretionary,mechanismsand“familiallogics”in the management of human resources rather than formalised systems, particularly atsmall and medium organizational sizes (Pittino&Visintin,2013).

Smalltomediumsizedfirmsaregenerallylessinclinedtoadopt formalized managerial practices (including HRM techniques), regardlessoftheirfamilybusinessstatus.However,both theoreti-calargumentsandempiricalevidencesupporttheideathatfamily ownershipandmanagementareassociatedwithanevenstronger propensitytoprivilegeinformalHRMsystems(e.g.DeKoketal., 2006;Kotey&Folker,2007;Pittino&Visintin,2013;Reid&Adams, 2001).Forexample,DeKoketal.(2006),examiningtheadoptionof professionalHRMpracticesamongsmallandmediumcompanies, findthefamilybusinessstatustobesignificantlyrelatedtoalower degreeofadoptionofformalizedHRpractices,aftercontrollingfor firmsizeandvariousindexesoforganizationalcomplexity.Similar

resultsarefoundinPittinoandVisintin(2013)andReidandAdams (2001).

One of the main reasons behind this phenomenon is the decisiveinfluenceofthefamilysocialsysteminshaping,through mechanismsofisomorphism,thecharacteristicsofthe organiza-tionalsystem.The emphasisoninformality,trust,interpersonal ties, tacit and shared norms and values that characterizesthe familyrelationshipsislikelytobeimprintedintheorganizational processes and routines, influencing also the HR practices (e.g.

Arregle,Hitt,Sirmon,&Very,2007;Brannon,Wiklund,&Haynie, 2013;Marquis&Tilcsik,2013)

TheprevalenceofinformalHRsystemshasledseveralauthors toargueforaninadequacyoffamilyfirmsinattracting,managing andretainingtheemployees,particularlythetalentedones,based ontheargumentthat thelatter mayprefertoworkin a more transparent environment, where opportunities to grow and participatearewell-definedandclear-cut(Barnet&Kellermanns, 2006;Chuaetal.,2009;Fiegeneretal.,1994).

So far, studies are missing that specifically analyse the organizational outcomesof such informal practices. Outof the four mechanisms mentioned above, we argue that familial informalHRpracticesgeneratepeculiarformsofsocialexchanges, which positively influence retention, even of those valuable employees who are typically the targets of human resource managementpractices,astestifiedbytheevergrowingliterature ontalentmanagement(forareviewseeCollings&Mellahi,2009). Inthepastfewyears,anumberofstudieshaveemployedsocial exchangetheorytoexplain what mechanismsshape individual family business members’ attitudes and behaviours towards pursuing transgenerational, collective, and not only economic goals (Barnett et al., 2012; De Massis, 2012a, 2012b; Long & Mathews,2011).Inparticular,LongandMathews(2011)arguethat familyfirmsaretypicallycharacterisedbyasocalledGeneralised ExchangeSystem(GES)whichisbasedonnormsofunilateraland indirectreciprocity, whereby there are no expectations among familymembersofanimmediateanddirectreturnforanaction. Thistypeofexchangeisbasedonfriendship,kinshipandaffection andleadstobehavioursthatarecooperative,homogeneousand cohesive.Non-familyfirms,onthecontrary,tendtobeclosertoa restrictivetypeofexchange,wherebyindividualstendtopursue self-interestinformsof exchangethataretypicallymutual and direct,basedonaquidproquomentality(Uehara,1990asquoted by Long & Mathews, 2011; p. 290). Such firms most closely resembleJensenandMeckling’s(1976)nexusofcontractsinwhich relationsareofthemarketvariety(e.g.economicagency)withan emphasis ondirect/mutual reciprocity, formal/prescribed inter-actions,and hierarchicalarrangements of thelegalistic(Weber, 1978)and universalistic (Carney, 2005)kind. In thesesettings, primaryexchangesaresignificantlyinstrumental,impersonal,and time-limited,asaretheunderlyingobligationsandexpectationsof the contracted participants_’ (p. 292). On the other hand, the dominant coalition runninga family business (even though in differentdegreesaccordingtothepeculiarcharacteristicsofthe coalitionitself),shapesindividualbehaviourswithinalong-term, trust-based,non-economicframework.

InlinewithBarnettetal.(2012),andPearsonandMarler(2010), we argue that the typical GES that characterises individual attitudes and behaviours of family members extends so as to shapealsothebehavioursofnon-familyemployees.

It is wellestablished inthe literature thatthose managerial philosophies that are the basis of the familial GES aim at maximisingautilityfunctionthatincludesmanygoalsotherthan profit, typically connected with the long-term survival of the company and the overall economic, social and emotional well beingofthefamilymembersandfuturedescendants(e.g.Berrone, Cruz,&Gomez-Mejia,2012;Chrisman,Chua,Pearson,&Barnett,

2012;Gomez-Mejia,Haynes,Núñez-Nickel,Jacobson,& Moyano-Fuentes,2007;Zellweger,Nason,Nordqvist,&Brush,2013).Those same philosophies typically leverage on strategies that follow predictable trajectories of change based on cumulative, tacit knowledge developed and shared with employees, customers, suppliersandothersignificant stakeholders(LeBretonMiller & Miller,2006)andinspiredbymutualtrust(ratherthanformal non-disclosureagreements),personalrespectandintimateknowledge. Besidesthe goalof buildingandmaintaining thecompany’s strategicknowledgecapital,thedevelopment ofa cohesiveand long-lastingcommunityofemployeesalsoservestheperpetuation ofapositivefamilyimageandreputationamongthestakeholders (Dyer&Whetten,2006;Zellwegeretal.,2013)andthe enhance-ment of the family business’ social capital endowment and networkofbindingsocialties(Miller&LeBreton-Miller,2005).

Toachievethesegoals,familyfirmstendtoemphasizeloyalty andcaringforworkers(LeBreton-Miller&Miller,2006),attention for thequalityof lifeand welfare of theemployees (Stavrou & Swiercz,1998), more stableemployment (Block,2010; Stavrou, Kassinis, & Filotheou, 2007), and the implementation of more “protective”contractualfeaturesfornon-familypersonnel( Cen-namoetal.,2012).Undersuchconditionsnon-familyemployees experienceafeelingofproximitytothefamilyownersandperceive thattheyarebeinglookedafterandtakencareofwithalong-term perspective(Block,2010;LeBreton-Miller&Miller,2006),evenin periodsofcrisis(seealsoBammens,VanGils,&Voordeckers,2010;

Bernhard & O’Driscoll, 2011; Miller et al., 2008, 2009). These feelings ofloyaltyandgratitude,combinedwiththeperception thattheorganizationinvestsinemploymentrelationswitha long-timehorizon(Berroneetal.,2012;Cennamoetal.,2012;Cruzetal., 2011)arelikelytoreducetheintentionstoleavealsoamongthose valuableemployeeswhohaveotheroptionsinthelabourmarket (Shawetal.,2009).

Besides social exchange mechanisms, also processes of organizational identification (Carmon et al., 2010; Vallejo & Langa,2010)andculturalfitmay(Cruzetal.,2011)influencethe extent of retention. Family image, history and values might promoteempathywiththefamilybythenon-familyemployees (Miller&Le Breton-Miller,2005; Zellweger etal., 2013),often beinganexampletofollowandagoaltoaimfor.Theprocessof social identification favours the emergence of feelings of psychologicalownership,affectivecommitmentandjob satisfac-tion (e.g. Avey, Luthans, & Jensen, 2009; Sieger, Zellweger, & Aquino,2013;VanDyne&Pierce,2004),thusreducing employ-ees’intentiontoquit.

Non-family firms,ontheother hand,do not experiencethe translationintothebusinessofthefamilialtraitsofreciprocity, trust,sharednormsandvalues.Long-termorientationmaybealso weakerasbusinessownersandleadersmightbemorefocusedon short-termhorizons.Itis,forexample,thecaseofoutsideinvestors orserialentrepreneurs.Thus,non-family_firmsmightbelessable toleverageoncertaininformaldimensionsofthesocialexchange andneedhigherlevelsofformalpracticestoincreasethedegreeof retention(LeBretonMiller&Miller,2006).

In light of thesefeatures, we couldargue that family firms, comparedtotheirnon-familycounterparts,possessanadvantage intermsofsymbolicresourcesandbehavioursthattriggerfeelings of reciprocation by the employees that non-family firms have difficultiesinachieving.

Asmentionedbefore,theempiricalevidenceappearstoshow an additive relationship among skill enhancing, motivation enhancingandengagementenhancingHPWPspractices(Combs et al., 2006). We propose that informal familial practices complementHPWPsinretainingtheemployees,particularlyby strengthening the feelings of obligation and reciprocation describedabove.

Therefore,weexpectthat,giventhesameextentofadoptionof formalHPWPs,familyfirmswillshowahigherlevelofretentionof employees thanksto theuseof theabove-mentioned informal practices,or,inequivalentterms,thatfamilyfirmsmayneedlower levelsofHPWPscomparedtonon-familyfirmstoachievethesame levelofretentionofemployees.Theaboveargumentis summa-rizedinthefollowing:

Hypothesis1.Familyfirmswillperformbetterthannon-family firmsin the retentionof valuable employees at any level of adoptionofHPWPs.

HR practices could also produce substitute effects (the organizationalbenefitsarisingfromtheadoptionoftwopractices isrespectivelythesameasthebenefitsthatwouldarisefromthe adoption of a single practice) (Delery, 1998). We propose that informal familial practices maycombine with HPWPs through partialsubstituteeffectsinretainingtheemployees,asfarasthe social exchange determinants of retention is concerned. In particular,thefamilyfirms’attentiontowardstrans-generational sustainability (e.g. Zellweger et al., 2013), the willingness to establishandpreserveacohesivecommunityofemployees(e.g.

Cruzetal., 2011;Miller &LeBreton-Miller, 2005),and alsothe tendency to promote employees’ spontaneous involvement in informal innovation activities(e.g. Bammens, Notelaers, & Van Gils,2015)canbeseenassubstitutesofthoseHPWPsaimedat increasingtheperceptionoforganizationallong-terminvestments inemployees’skills (e.g.Huselid,1995; Pfeffer,1994)andthose aimed at strengthening employees’ participation and sense of belonging(e.g.Patel&Conklin,2012).Thus,infamily firmsthe informalHRpracticesgroundedontheemotionalbondwiththe familymaymake redundant sometheformalHPWPsaimedat affectingtheperceptionofsocialexchangebytheemployees.In otherwords,anincreaseinHPWPsadoptionmaytranslateinlower positiveeffectsthaninnon-familyfirms.

Hypothesis2.Infamilyfirmsthemarginaleffectofanincreased adoption of HPWPs may be lower compared to non-family firms.

4.Theeffectsindifferenttypesoffamilyfirms

Familyfirmsshouldnotbetreatedasahomogeneous group (Chrisman,Chua,&Sharma,2005;Zahra,Hayton,&Salvato,2004), especiallyregardingtheiremphasisonsocioemotionaldimensions (e.g.,Berroneetal.,2012).Inlightofourtheoreticalframeworkitis thereforeappropriatetoexplorehowexistingdifferenceswithin familyfirmsinfluencetherelationshipsbetweenHPWPsandthe retention of valuable employees. We focus on two sources of heterogeneitythathavebeenindicatedasbeingstronglyrelatedto theimportanceofsocioemotional goalsin familyfirms,namely family CEO versus non-family CEO leadership and the family generationinvolvedinthebusiness(Berroneetal.,2012; Gomez-Mejiaetal.,2007).

4.1.Familyvs.non-familyCEOfamilyfirms

ThepresenceofafamilyCEOisusuallyassociatedwithahigher emphasisontheprotectionandenhancementofthefamilyfirm’s socioemotionalwealth,incomparisontosituationswherefamily ownersdelegate thecompany leadershiptoaprofessional non-familyexecutive (e.g. Berrone et al., 2012;DeTienne &Chirico, 2013).

Itiscertainlypossible(andevencommon)thatalsonon-family CEOsexhibithighlevelsofculturalcongruencewiththefamilyand areloyaltoitsvaluesandgoals(e.g.Blumentritt,Keyt,&Astrachan, 2007;Hall&Nordqvist,2008).However,afamilyCEO,byvirtueof

thekinshipties,thesharedfamilyname,andthereferencetothe common family history is more likely to promote a common identity and a stronger “familyvision” withinthe organization (Arregleetal.,2007;Barnettetal.,2012).Ithasbeensuggestedthat astrongfamilyvision,definedasthesharedintentionwithinthe dominantcoalitionforthetransgenerationalsustainabilityofthe firm(Chrismanetal.,2012),isassociatedtotheestablishmentof normsofreciprocitythatbroadenthescopeofthesocialexchange. Inparticular,asthefamilyvisionbecomesstronger,theexchange systemoftheorganizationmovesfromarestrictedtype,whichis characterizedbyatransactionalandbilateralviewofreciprocity,to ageneralizedtype,whichinvolvesa_“serial”viewofreciprocityand a focus onthe collectiveinterest (Long&Mathews,2011).The exchange system affectsboth familyand non-familyemployees (Pearson & Marler, 2010). Thus, if non-family employees are involved in a generalized exchange system that leverages on group-based and long-term obligations among organizational members,theyaremorelikelytoexperiencefeelingsof commit-mentandlowerintentionstoleave(Barnettetal.,2012).

Moreover, as observed by Gomez-Mejia et al. (2007), the strongertheroleofthefamilyintheownershipandmanagement, the more the company will adopt strategies to satisfy its socioemotional priorities, including the preservation of long-lasting relationships with the employees (Berrone, Cruz, & Gomez-Mejia,2014).Thedirectinvolvementofafamilymember inthemostimportanttopexecutivepositionalsoreinforcesthe interactionbetweentheowningfamilyandthecompany stake-holders (e.g. Block & Wagner, 2010). Moreover, it has been suggestedthatstronglevelsoffamilycommitmentandidenti fi-cationwiththefirm,expressedbythepresenceofafamilyCEO, cancreatesimilarfeelingsamongemployees.Zahraetal.(2008), forexample,reportempiricalevidenceofa“socialcontagion”of commitment attitudes: when one member of an organization behavesinacommittedfashion,itincreasestheprobabilitythat other members will also experience higher identification and commitment totheorganization(Barsade, 2002).Accordingto

Rhoadesand Eisenberger(2002),theorganization’sleadersare the main influencers of employee commitment. Whereas the socialcontagioneffectcanoccuralsoinpresenceofanon-family CEO,whocanbeagoodrolemodelforthenon-familymembers, weassumethatafamilyCEOusuallyexhibitsthestrongestlevels ofcommitmenttothefamilyvisionofthebusiness,andthisisthe main trigger of imitative attitudes by the employees. Hence, familyCEOleadershipis likelytoproduce an overall improve-mentofnon-familyemployees’feelingsofpsychological owner-shipandaffectivecommitment,therebyreducingtheintentions toleave.

Consistentwiththeargumentsalreadydevelopedinsupportto

Hypothesis1andHypothesis2,wethereforesuggestthatatany givenlevelofadoptionofHPWPs,familyfirmsledbyafamilyCEO have an advantage in the retention of valuable employees compared to family firms led by a non-familyCEO, thanks to strongerexpectedfeelingsofreciprocationandhigherextent of commitmentandidenti_fication.

Hypothesis3.ForagivenlevelofadoptionofHPWPs,family firmswithafamilyCEO achievehigherretention ofvaluable employeescomparedtofamilyfirmswithanon-familyCEO.

However, in the situations of family CEO leadership, the immaterialandsymbolicresourcesexchangedwiththeemployees maypresentahigheroverlapwiththesocial-exchangeeffectsof someHPWPs.Thus,asa consequenceofa possiblesubstitution effect,themarginaleffectofanincreaseduseofHPWPsonthe retentionofvaluableemployeesmaybelesspronouncedinfamily firmsledbyafamilyCEO incomparisontothoseledbya non-familyCEO.Thereforeitfollowsthat:

Hypothesis4.ThemarginaleffectofanincreaseinHPWPson theretention of valuableemployees maybe lowerin family firmswithafamilyCEOincomparisontofamilyfirmswitha non-familyCEO.

4.2.Firstgenerationvs.latergenerationfamilyfirms

Thereiswidespreadagreementinthefamilybusinessliterature thatfamily identification, family influenceand personal invest-mentsinthecompanyareattheirhighestlevelsinthefounding, first generation stage, and decrease as the firm moves into subsequent stages of family ownership and leadership (e.g.

Gersick,Lansberg,Desjardins,&Dunn,1999;Gomez-Mejiaetal., 2007;Le-BretonMiller&Miller,2013;Ward&Aronoff,1994).In first generation family firms, the social systems of family and organizationarelargelyisomorphic(Arregleetal.,2007;Sirmon& Hitt,2003), sincetheactorsinvolved areusuallymembersof a nuclearfamily and there is a predominantrole of the founder (Schein, 1983). The founding families have a deep emotional attachmenttothecompaniestheyhavebuiltandstrongaffective tieswiththenon-familyemployees,whichhavehelpedthefirmto surviveand grow throughout theinitialdifficulties (Le Breton-Miller&Miller,2013).In firstgenerationfamilycompanies,the mainsocioemotionalgoalisthetransmissionofthebusinessto later generations, and this results in an enhanced long-term perspectiveofbusinessstrategiesandstakeholders’relationships (Berrone et al., 2014; Zellweger et al., 2013). Although in first generationcompanies theinfluenceof thefamily is limited in comparison tothe paramount role of thefounder (e.g.Cruz & Nordqvist,2012),the“dynasticpriorities”ofthefamilybusiness areattheirhighestlevels(e.g.Arregleetal.,2007).Foundersfeel the responsibility to provide long-term benefits for family members, mainly by securing the family control of the firm beyondthefoundinggeneration.Asfounderstendtoviewtheir businessesasextensionsofthemselves(Schein,1983),theystrive fortheperpetuationoftheirentrepreneurialeffortbytransferring abusinesslegacytofuturegenerations,(Dyer&Whetten,2006).By virtue of theirroles as heads of families and business leaders, foundersaim at protecting and promoting the family business reputation, and the continuation of their business lineage (

Athanassiou,Crittenden,Kelly,&Marquez,2002;Lim,Lubatkin,& Wiseman,2013).Thesefactorsareconsistentwithastrongfamily visionwithinthebusinessandthusreinforcethebroadscopeand thecollective dimensionofsocial exchange,involvingalso non-familyemployees(Barnettetal.,2012)

As the family firm proceeds tosecond and later generation stages,theoverlapbetweenfamilyandorganizationdiminishes, the family members’ emotional attachment and identification becomeprogressively weaker, the socioemotional prioritiesare generallylesspronounced,morefragmentedandmorefocusedon short-termneedsofindividualfamilymembers(e.g.Gomez-Mejia etal.,2007;Miller,Minichilli,&Corbetta,2013).

Wemightthereforeexpectthatfirstgenerationfamilyfirms, comparedtopost-founderandlatergenerationones,haveamore intense exchange of immaterial resources with non-family employeesintheformofmoreintenseandmeaningfulinformal practices.Thisislikelytocreateanadvantageinthecapacityto retainvaluablenon-familypersonnel.Fromthisitfollows:

Hypothesis5.Atany givenlevelof adoptionofHPWPs,first generationfamilyfirmsachieveahigherretentionofvaluable employees compared to second and later generation family firms.

Also,asaconsequenceofapossiblesubstitutioneffectbetween theuseofHPWPsandtheinformalrelationalsystemandsymbolic

involvementexperienced bythenon-familyemployees in first-generationcompanieswemayexpectthefollowing:

Hypothesis 6. The effect of an increase in HPWPs on the retention of valuable employees may be lower in first generation family _firms in comparison to second and later generationfamilyfirms.

5.Sampleandmethod 5.1.Sample

To test our hypotheses, we conducted a quantitative study basedona sampleof companiesfromtwoEuropeancountries, AustriaandHungary.Thesamplewastakenfromthesametwo industries,“manufactureof foodproducts”and“manufactureof computer,electronicandopticalproducts/manufactureof electri-calequipment”inbothcountries.TheCMDcomplete(Austria)and OPTEN(Hungary)databaseswereusedtoidentifyexecutiveboard members and managing directors of companies within these industries.Asmicroenterpriseswereconsideredasbeingunlikely tohavesystematicHPWPsinplace,businesseswithlessthanten employeeswereexcludedfromthelist.

Wesentpersonalizede-mailswithalinktotheonlinesurvey instrument to1649 managers(917in Austria,732in Hungary), guaranteedanonymityfollowingPodsakoff,MacKenzie,Lee,and Podsakoff’s (2003) advice,and useda multiplecontact method (withtworemindersafterthefirste-mail)assuggestedbyDillman, Smyth, and Christian (2009), yielding a gross response rate of 20.3%. Of the 335 respondents, those who were not in a top managementpositionorinsufficientlyfilledoutthequestionnaire wereexcluded.Eventually,232questionnaireswereusedinthe analysis,representinganetresponserateof14.1%overall(15.5%in Austriaand12.3%inHungary),whichissomewhatlowerthanin prior empirical studies related to HPWPs while comparing favourablytothetypical10–12%responserateofsurveysamong top managers (Geletkanycz, 1997). 97 companies in our final samplehavebetween10and50empoyees,135havemorethan 50employees.

5.2.Instrumentdesign

Thesurveyinstrumentwasdevelopedfollowingathree-step process.First,theliteraturewasreviewedforexistingandtested scalesformeasuringthemainconstructs.Thedraftquestionnaire waspeer-reviewedbyresearcherswithexperienceinthefieldof HRM.InlinewiththesuggestionofDillmanetal.(2009),apre-test was then conductedamong approximately tenper cent of the targetgroup(17companies)followingasimilarprocessasthelater main study. The purpose of the pre-test was to test the questionnaire,bothregardingrespondents_’answeringbehaviour (e.g.break-uppoints)aswellasregardingtheinternalconsistency oftheconstructs.Followingthepre-testandapreliminaryfactor analysis,thequestionnairewasslightlyadapted(someitemsthat did not contribute tothe factor loading were removed,others reformulated). In order to explore whether the items used appropriatelycontributetotheconstructstheyoughttomeasure, rotated principal component analysis (PCA) was conducted for eachconstruct.Asageneralrule,itemswithfactorloadingsabove 0.5werekept.FollowingBackhaus,Erichson,Plinke,andWeiber’s (2003) recommendation, we also tested whetherthe datawas actually suitable for PCA. The results of Kaiser-Meyer-Olkin measureof samplingadequacy, Bartlett’s testof sphericity,and the anti-image correlation matrix indicated that PCA was an adequate tool. Based on the findings from the test of the

instrument,inparticularregardingtheinternalconsistencyofthe scales,thequestionnairewasadaptedwithminorchanges.

TheinstrumentwasoriginallydevelopedinEnglish.Thefinal questionnairewas thentranslated intoGermanand Hungarian. Standardback-translationprocedureswereusedfollowedbysmall adaptationstoensurethatthequestionshadthesamemeaningin alllanguages.

5.3.Variables

Employeeretentionwasmeasuredbytherespondents’ subjec-tiveassessmentoftheir_firm’srelativeperformancein retaining whatthetopmanagementconsidersasvaluableemployeesover thelasttwoyears,comparedtofirmsofsimilarsalesvolumein theirindustry on a five-pointLikert-type scale ranging from1 (significantly below industry average) to 5 (significantly above industryaverage). Althoughan objective measurefor retention rateswouldbepreferable,topmanagersareoftenunawareofthe exactnumbersatthetimewhentheyfilloutthesurvey.Retention research commonly uses employee turnover as a measure for studying retention (Guthrie, 2001; Patel & Conklin, 2012). However,since there areseveral ways of calculating employee turnover,comparabilitycouldnotbeensured.Therearereports from prior research about significant correlations between objective and subjective measures of performance (Dess & Robinson, 1984). We therefore followed the example of other authors(Delaney&Huselid, 1996;Hau-siuChow,2004;Shawetal., 2009)inusingasubjectiveassessmentofthefirm’sperformancein retainingvaluableemployeesoverthelasttwoyearscomparedto similarfirmsintheindustry.To establishtherobustness ofour results,however,weperformedadditionaltestsofourhypotheses employingtheoverallturnoverrateasdependentvariable(inthe questionnaire,wealsoincludedaquestionaboutthepercentageof employeeswholeftthecompanyduringthepreviousyear).

TheadoptionofHighPerformanceWorkPracticeswasassessed calculating the average values of the scores for six HPWP categories, obtainedas averagesof the equallyweighted items thattogetherconstitutetheindividualcategories.Theselectionof high-performance work practices items was based on the classification in: selective staffing, intensive training, career development,extensivecompensationandbenefits,performance appraisals,andemployeeparticipation.Theindividualitemswere adaptedfrompriorHPWPstudiesofvariousauthors(e.g.Patel& Conklin,2012;Selsetal.,2006)(seetheAppendixAforadetailed overview of the scales and their

a

-levels that range between

a

=0.698and

a

=0.836)andweremeasuredonafive-point Likert-type scale (from 1=very little or no importance to 5=very important).

GiventhenatureofourHPWPsand retentionmeasures,we assessed the potential of common method bias in the data

performing Harman’s one-factor test (Harman,1967; Podsakoff etal.,2003),whichisregularlyusedindealingwiththisissuealso infamilybusinessstudies(e.g.Siegeretal.,2013).Weenteredall theitemsofourstudyintoafactoranalysis,extractingan8-factor solution,accountingfor65.48%oftotalvariance.Thefirstfactor explained 14.85%of variance, suggesting that commonmethod biasisnotamajorproblembecausenosinglefactoraccountsfor themajorityofvariance.

Familyfirmstatusisadichotomousvariableassumingvalue1if afamilyisreportedasthecontrollingownerofthecompanywith the absolute majority of ownership stakes, and 0 otherwise (Andres, 2008). In the case of family _firms, we also surveyed whethertheCEOisafamilymember(dichotomousvariableFamily CEOassumingvalue1ifafamilymemberholdstheCEOposition and0otherwise)andwhetherthefirmisinfirstgenerationorlater generation family ownership (First Generation, dichotomous variable,assumingvalue1ifthefamilyisinthefirstgeneration ofinvolvementinthecompany,and0otherwise).

FollowingpreviousstudiesofHPWPsandemployeeretention, weincludedascontrolvariables:(1)abinarymeasureaccounting for companies belonging to high technology industries, (2) a measure of financial performance over the past three years (in terms of ROA), (3) company size, operationalized through a dichotomousvariableassumingvalue0ifthefirmhasanumber ofemployeesbetween10and49,andvalue1ifithasmorethan 50 employees),(4)companynationality, assumingvalue1ifthe companybelongstothesamplefromAustriaand0ifthecompany belongstothesamplefromHungary.

Descriptivestatisticsandcorrelationforallthestudyvariables arereportedinTable1.

5.4.Estimationtechnique

Weusedhierarchicalordinaryleastsquaresregressionmodels totestthehypotheses.Instep1ofeachhierarchicalmodel,we enteredthecontrolvariables;instep2weaddedtheHPWPsindex. Instep3,weaddedthefamilyfirmstatus(inthemodeltesting

Hypothesis1),presenceofafamilyCEO(Hypothesis3),andfirst generation familyfirm (Hypothesis 5). In step 4we added the interaction terms between HPWPs and family status (model testingHypothesis2),HPWPsandfamilyCEO(Hypothesis4)and HPWPs and firstgeneration stage(Hypothesis6).We usedthe significance of the family firm status, family CEOs and first generationcoefficientsplustheexplainedvariancechangesfrom step2tostep3todeterminesupportforHypothesis1,Hypothesis 3,Hypothesis5,respectively.Theexplainedvariancechange,the significanceoftheinteractiontermandtheplottedformofthe interaction effect (if significant) were used to determine the supportforHypothesis2,Hypothesis4,Hypothesis6.

Table1

Means,standarddeviationsandcorrelations.

Mean SD 1 2 3 4 5 6 7 8 9 1Retention 4.12 0.80 1.00 2HPWPs 3.64 0.62 0.33 1.00 3Pastperformance 3.43 1.05 0.35 0.28 1.00 4Sizeb 0.58 – 0.18 0.46 0.22 1.00 5Hightechnologyb 0.38 – 0.07 0.15 0.09 0.13 1.00 6Companynationalityb 0.62 –
0.08
0.01
0.15
0.10 0.06 1.00 7Familyfirmb 0.52 – 0.02
0.16 0.15
0.12
0.08
0.12 1.00 8FamilyCEOb _{0.39 –
0.01
0.16 0.08
0.11
0.13
0.17 0.77 1.00}

9FirstGenerationa,b _{0.48 – 0.02 0.01
0.06
0.07 0.35 0.30 – 0.06 1.00}

N=232;Correlationshigherthan0.15andlowerthan
0.14aresignificantat0.05level.Themeanincaseofbinaryvariablesindicatestheratioofobservationswherethe variableassumesvalue1.

a

N=121.

b

6.Results

6.1.Descriptiveanalysis

After splitting our sample and breaking down the overall indicatorofHPWPsinitssinglecomponents(Table2),weobserve that:(1)familyfirmspresentalowerdegreeofadoptionofHPWPs comparedtotheirnon-familycounterparts;(2)thelowerintensity of adoption is homogeneous across the different subsets of practices;(3)theindexofvaluableemployees’retentionishigher infamilyfirmscomparedtonon-familyones; (4)withinfamily firms,companiesledbyanon-familyCEOandcompanieswhere secondor laterfamilygenerationsareinvolvedexhibitahigher intensityofHPWPsuse.

6.2.Regressionanalyses

6.2.1.Hypothesis1andHypothesis2

It was hypothesized that family businesses achieve better performancesintheretentionofvaluableemployeescomparedto non-familybusinesses,holdingfixedtherateofHPWPsadoption.

Table3presentstheresultsofthehierarchicalregressionmodel with the retention of valuable employees as the dependent variableandthestepwiseinclusionoftheexplanatoryterms.

Our analysis shows that HPWPs are strongly and positively related with the retention of valuable employees (

b

=0.21, p<0.001). Whenwe add the family firm status in step 3, we observe that the direct average effect of HPWPs remains substantially unchanged and the family business status has a positiveandsignificantimpactonthedependentvariable(

b

=0.12, p<0.05). This provides support to our Hypothesis 1: after controllingfortheintensityofHPWPsadoption,familyfirmson averageexhibit ahigher capacity toretainvaluable employees. Hypothesis 2 stipulates that family firm status and HPWPs adoptioninteract in sucha waythat in family firmsthere is a lowersensitivityofvaluableemployees’retentiontotheincrease ofHPWPs.Step4inthehierarchicalregression(Table3)includes theinteractiontermbetweenfamilyfirmstatusandHPWPs.The significant improvement of the R-squared indicates that the inclusionoftheinteractiontermenhancestheexplanatorypower ofourmodel.Theinteractiontermisalsosignificantandnegative (

b

=
0.19, p<0.10), indicating that the effect of HPWPs is weakenedbythefamilybusinessstatus.

The relationships proposed in Hypothesis 1 and 2 can be appreciated in the plot of Fig. 1 depicting the significant interaction.Itcanbeobservedthat:(1)onaveragefamilyfirms haveahighercapacitytoretainvaluableemployeesgivenacertain level of HPWPs, and this indicatesthat there is some kind of additiveeffectbetweenformalpracticesandinformaldimensions oftheexchange, but (2)theeffectof anincrease inHPWPs on retentionislowerinfamily_firmscomparedtonon-family_firms,

andthissuggeststhatthereisalsoapartialsubstitutioneffectthat makesthecombinationbetween HPWPsandfamily firmstatus “sub-additive.”

Theseresultsarerobustacrossdifferentspecificationsofthe model.Inparticular,theeffectsremainunchangedifweusethe percentageofoverallemployeeturnoverasourdependentvariable (Table5).TheuseofHPWPsandthefamilybusinessstatushavea negativeimpactontheoverallturnoverrate,andtheeffectonthe reductionof turnoverrate ofanincrease inHPWPsadoptionis lowerinfamilyfirms.Furtherresultsinsupportofourhypotheses are found in additional analyses that assess the effect of each subset of practices in the family and non-family business subsample. In the family business subsample, the HPWPs especially aimed at increasing employee participation and involvementdonothaveanimpactontheretentionofvaluable employees(Table6).

6.2.2.Hypothesis3andHypothesis4

Hypothesis3suggeststhat,withinfamilyfirms,companiesled byafamilyCEOachieveabetterperformanceinvaluableemployee retentioncomparedtothosewhereanon-familyCEOisincharge (givenacertainlevel ofHPWPs).Regressionresultsreportedin

Table4 do notprovidesupporttothehypothesis.Theeffectof HPWPsonretentionisconfirmed butthereis noevidenceofa “familyCEO”effect(

b

=
0.01,notsignificant).Howeverthereis someevidenceinsupportofasubstitutioneffect(Hypothesis4). Theinteractiontermenteredinstep4aisnegativeandmarginally significant(

b

=
0.13,p<0.10),andthissuggeststhatHPWPshave astrongerincrementaleffectwhenanon-familyCEOisincharge, although,onaveragethere arenodifferencesin thedegreesof retentionbetweenfamilyandnon-familyCEOfamilyfirms. 6.2.3.Hypothesis5andHypothesis6

We hypothesized (Hypothesis 5) that, within family firms, companieswherethefoundinggenerationisinvolvedachievea higherretentionofvaluableemployeescomparedtosecondand later generation family firms. We find some support to this hypothesis:thecoefficientoftheFirstGenerationtermenteredat step3bissignificantataborderlinelevel(

b

=0.15,p<0.10).Onthe other hand, the substitution argument of Hypothesis 6 is not supported:thecoefficientoftheinteractiontermenteredinthe step4bisoftheexpectedsign,butnotsignificant(

b

=
0.13,not significant).

7.Discussion

The literature onhuman resource management practices in family firmsdisagrees abouttheattitude of the owningfamily towardsnon-familyemployees:itisunclearwhetherfamilyfirms “caremore”or“careless”abouttheirpersonnelcomparedto non-family_firms(e.g.Cennamoetal.,2012;Cruzetal.,2014).

Table2

HPWPsandretention:descriptiveanalyses.

Familyfirms Non-familyfirms FamilyCEO Non-familyCEO Firstgeneration Secondorlatergeneration Mean S.D. Mean S.D. Mean S.D. Mean S.D. Mean S.D. Mean S.D. Selectivestaffing 3.46 0.91 3.86 0.91 3.44 0.57 3.52 0.63 3.55 0.61 3.53 0.56 Intensivetraining 3.58 0.76 3.76 0.85 3.52 0.92 3.74 0.87 3.46 0.92 3.46 0.90 Careerdevelopment 3.65 0.73 3.84 0.80 3.62 0.77 3.73 0.71 3.63 0.76 3.53 0.75 Extensivecompensation 3.23 0.82 3.43 0.89 3.23 0.72 3.21 0.76 3.28 0.78 3.66 0.67 Performanceappraisals 3.65 0.76 3.83 0.84 3.59 0.79 3.79 0.91 3.68 0.76 3.16 0.85 Employeeparticipation 3.70 0.68 3.79 0.69 3.69 0.77 3.74 0.68 3.61 0.72 3.60 0.75 TotalHPWPs 3.54 0.58 3.75 0.65 3.54 0.76 3.62 0.74 3.61 0.67 3.79 0.62 Retention 4.19 0.81 3.88 0.85 4.17 0.83 4.20 0.76 4.23 0.78 4.11 0.78 N 121 111 30 91 59 62

Ourresultscontributeatreconcilingtheseopposingviews,by observingtherelationshipbetweentheadoptionofHPWPsand theretentionof valuableemployees, througha social exchange framework. This perspective allows us to consider the actual response of non-family employees to the activation of certain humanresourcepractices,andreveals thatemployeesinfamily firmsmightbenefitfromthehighqualityofrelationshipsenabled bythefamily influence(Barnettet al.,2012; Pearson &Marler,

2010), that combines with the formal HPWPs in stimulating retention.Therefore,thelowerdegreeofadoptionofHPWPsdoes notmeanthatfamilyfirmscarelessabouttheirworkforce.

Inparticular,insupportofhypothesis1,weshowthatateven very low levelsof HPWPs adoption, thedegree of retention of valuableemployees byfamilyfirmsishigherthanthat by non-family firms. This means that, notwithstanding the potential negativeimpactsofanumberoffamilialdiscretionarypractices

Table3

OLSregressionanalysiswithdependentvariableRetentionofvaluableemployees:comparisonbetweenfamilyandnonfamilyfirms. (1)

ControlVariables (2)

HPWPandControlVariables (3)

DirectEffectFamilyvs.Non-familyFirms (4) Two-wayinteraction PreviousPerformance 0.21*** (0.05) 0.21*** (0.05) 0.20*** (0.05) 0.19*** (0.05) Size 0.00 (0.05) 0.00 (0.05) 0.00 (0.05) 0.00 (0.04) Hightechnologyindustry 0.02

(0.10) 0.02 (0.10) 0.03 (0.10) 0.03 (0.10) CompanyNationality
0.06 (0.10)
0.06 (0.10)
0.06 (0.10)
0.06 (0.10) HPWPs 0.21*** (0.09) 0.20*** (0.12) 0.32*** (0.08) Familyfirm 0.12* (0.07) 0.21y (0.18) HPWPFamilyfirm
0.19** (0.12) Constant 3.99*** 3.95*** 3.96*** 3.96*** (0.22) (0.24) (0.24) (0.24) N 232 232 232 232 R-sq 0.14 0.24 0.29 0.34 AdjustedR-sq 0.10 0.20 0.26 0.31 Fstatistic 7.38*** 8.38*** 7.93*** 7.19*** R-sqchange 0.10*** 0.15*** 0.20*** FforR-sqchange 8.33*** 4.21*** 6.63***

Note:Standarderrorsinparentheses;***p<0.001,**p<0.01,*p<0.05,yp<0.1.

Table4

OLSregressionanalysiswithdependentvariableRetentionofvaluableemployees:differenttypesoffamilyfirms. (1)Control

Variables

(2)HPWPand ControlVariables

(3a)Directeffect FamilyCEO

(4a)Two-wayinteraction (3b)DirecteffectFirstGeneration (4b)Two-way interaction PreviousPerformance 0.19*** (0.06) 0.19*** (0.06) 0.18*** (0.06) 0.18*** (0.06) 0.19*** (0.06) 0.18*** (0.06) Size 0.12 (0.08) 0.12 (0.08) 0.12 (0.08) 0.11 (0.08) 0.12 (0.08) 0.13y (0.08) Hightechnologyindustry
0.04

(0.15)
0.04 (0.15)
0.04 (0.15)
0.04 (0.15)
0.09 (0.15)
0.11 (0.16) CompanyNationality
0.01 (0.10)
0.01 (0.10)
0.01 (0.10)
0.01 (0.10)
0.06 (0.15)
0.04 (0.15) HPWPs 0.20*** (0.08) 0.20*** (0.05) 0.14** (0.05) 0.20*** (0.08) 0.18** (0.09) FamilyCEO
0.01 (0.12) 0.01 (0.16) FirstGeneration 0.12y (0.10) 0.15y (0.15) HPWP*FamilyCEO
0.13y (0.10) HPWP*FirstGeneration
0.13 (0.14) Constant 3.82*** 3.78*** 3.78*** 3.70*** 3.96*** 3.96*** (0.24) (0.24) (0.24) (0.24) (0.24) (0.24) N 121 121 121 121 121 121 R-sq 0.08 0.19 0.21 0.21 0.19 0.21 AdjustedR-sq 0.06 0.15 0.16 0.16 0.14 0.16 Fstatistic 3.52*** 5.52*** 4.57*** 4.46*** 4.00*** 4.30*** R-sqchange 0.09 0.00 0.01
0.01 0.01 FforR-sqchange 6.32*** 0.92y 1.01 0.54 0.88

highlightedby theliterature, such asnepotism, familial scape-goating and so on, the positive ones connected with the GeneralisedExchangeStructurestypicaloffamilyfirmsareeven stronger(Long&Mathews,2011).Therefore,althoughfamilyfirms relylessthantheirnon-familycounterpartsonformalHPWPs(as alsoconfirmedbyourdescriptivestatistics),theirrateofretention isnotsystematicallylowerbutmaybesimilarorevenhigher.By clarifyingthecausallinkbetweenHRMpractices,familyinfluence and employee retention in family firms, this result provides indirectsupporttothe ideathat the pursuit of socioemotional prioritiesandgoalsatthefamilylevelimprovesthequalityofthe socialexchangewithnon-familyemployees,andthisisinlinewith previous conceptual arguments on the social exchange in the familybusiness context (Barnett et al.,2012; Long& Mathews, 2011;Pearson&Marler,2010).

As a further specification of this argument, the empirical evidenceinsupportofHypothesis2showsthattheimpactofthe adoption of HPWPs onretention is marginally lower in family businessesatanylevelofadoptionofformalpractices.Thisresult can be interpreted by viewing family influence as a source of informalHRMpracticesthatarecommittedinthesocialexchange

process.Infamilyfirmsthefamilialinformalpracticesmayaccount fortheretentiondifferential withnonfamilyfirmsatanygiven level of HPWPs adoption. While some informal practices may impact on retention also through skill-enhancing, motivation-enhancing, and participation-enhancing effects, the difference withnon-familyfirmsliesprimarilyinthereciprocationeffects.

Thisdependsonthestrategichorizonadoptedbyfamilyfirms as a consequenceof thegoalof maximising thefamilial socio-emotional wealthand in theassociated, close, trust-basedand long-termrelationshipsestablishedwiththeemployees.Inlightof socialexchangetheory,thesetypeofinvestmentsbythe employ-ersmayinducereciprocationbehavioursintheemployees.Indeed, theresultsshowthatastheintensityofHPWPsadoptionincreases, thedifferenceintheextentofretentionbecomeslower.Infamily firmstheeffectsarisingfromthereciprocationneedarealready satisfiedbytheinformaldimensionsofthesocialexchangeandthe employeesinfamilyfirmswillbelesssensitivetoanincreasein HPWPsintensitycomparedtonon-familyfirms’employees,whoin turnaremorelikelytovalueformalmechanismsofinvolvement, astheinformalonesareabsentorlessrelevant.Thisobservation corroboratesthetheoreticalpropositionadvancedbyPearsonand Marler(2010),accordingtowhichfamilyfirmsaremorelikelyto establishastrongerreciprocalstewardshipculturecomparedto non-familyfirms.

Apossiblecomplementaryexplanationofourfindings,inline with the theory of person-organization fit (Memili & Barnett, 2008;O’Reilly,Chatman,&Caldwell,1991),couldbethatfamily firms attract people who have a motivational profile which is consistentwiththeculture andthe valuesystemof thefamily businessorganizationalsetting.

Regardingthedifferencesamongdifferenttypesoffamilyfirms (Hypotheses3–6), ourarguments findonly partial support. No significant“familyCEO”effecthasbeenobservedinadditionwith HPWPs,and onlymarginalsupportis providedtotheidea ofa moderatingeffectoffamilyCEOstatusontheimpactofHPWPs. Thissuggeststhattheconsequencesoffamilybusinesses’informal HR practices on retention depend on family involvement, regardlessof thefamilystatusoftheorganization’sleader.This couldbeexplainedin light ofthesmall tomediumsizeof our companies. In such contexts, the proximity of the controlling

3 3.5 4 4.5 5 High HPWPs Low HPWPs

Retention of valuable employees

Non family firms Family firms

Fig.1.EffectofHPWPsontheretentionofvaluableemployeesinfamilyandnon familyfirms.

Table5

Robustnesscheck:OLSregressionanalysiswithdependentvariableOverallturnoverrate.

(1)ControlVariables (2)HPWPandControlVariables (3)DirectEffectFamilyvs.Non-familyFirms (4)Two-wayinteraction PreviousPerformance
1.19*** (0.44)
1.19*** (0.44)
1.25*** (0.44)
1.20*** (0.05) Size 1.92*** (0.51) 1.91*** (0.51) 1.94*** (0.51) 1.99*** (0.52) Hightechnologyindustry
1.75y

(0.94)
1.75y (0.94)
1.71y (0.94)
1.87* (0.94) CompanyNationality 0.27 (0.93) 0.27 (0.93) 0.32 (0.94)
0.06 (0.10) HPWPs
1.43*** (0.52)
1.39*** (0.52)
2.12*** (0.69) Familyfirm 0.51* (0.35) 0.50y (0.38) HPWPFamilyfirm 1.54* (0.81) Constant 10.01*** 10.10*** 9.93*** 10.09*** (2.22) (2.20) (2.23) (2.22) N 224 224 224 224 R-sq 0.07 0.11 0.13 0.18 AdjustedR-sq 0.05 0.09 0.12 0.16 Fstatistic 4.44*** 5.62*** 4.72*** 4.49*** R-sqchange 0.04 0.03** 0.07*** FforR-sqchange 2.89** 1.21 3.63**

family is perceived independently from the formal positions occupiedbyfamilymembers.

Theeffectofthefoundinggenerationispositiveasexpected, butitsmagnitudeisrathersmall,witha borderlinesignificance level.Thiscouldbeexplainedbythepresenceoffamilyreputation concernsbothinthefirstgenerationandinlatergenerations.In first generation the “sense of dynasty” and the willingness to establishafamilylegacymightprevail,whereasfamilyimageand prestigeofthefamilynamebecomeincreasinglyimportantinthe latergenerations(e.g.Zellwegeretal.,2013).Boththeseforcesare positivelyrelatedwiththe establishmentof cohesive employee communities.

Overallourresultsaddalsotothehumanresourcemanagement literature,andinparticulartothesocialexchangeperspectivein thestudyofHRMpractices(Cropanzano&Mitchell,2005;Jiang, etal.,2012),bysuggestingthatinformalpracticesmighthavean important role in the social exchange between employers and employees and byshowing howthe ownershipand leadership contextmaypromotefeelingsofreciprocationthatcombinewith theindividualeffectsofformalpracticesontheemployeestoelicit positiveorganizationaloutcomes.

Moreover,astheexistingliteratureonHPWPshasidentifieda prevalence of additive effects among sets of practices (Chuang etal.,2012;Delery,1998; Jiangetal.,2012)eventhoughearlier theoreticalcontributionshadadmittedthepossibilityofdifferent typesofrelationships(forexamplesynergisticorsubstitution)no studyhasconsideredthatinformalpracticesmaycombinewith theformalonestoproducepositiveorganizationaloutcomes.The analysisofthecombinationbetweenthefamilyeffectfosteringthe establishmentofinformalpracticesandHPWPsthusallowsusto extendtheknowledgeofthewaysHRMpracticesinteract.

WhereasDelery(1998)andothers(e.g.Macky&Boxall,2007) suggest that HRM practices interact either in an additive, synergisticorsubstitutiveeffect,withonealternativeexcluding theother, our resultsshow that the interactionbetween HRM practices(inourcase,formalandinformalones)couldleadtoboth additiveandsubstitutiveeffects.Thiscombinationofadditiveand

interactiveeffectsmaygeneratesuper-additiveor,asinourcase, sub-additiveeffects.

Ourstudyhasalsosignificantmanagerialimplications.Family firmsareoftencriticisedfornotinvestingenoughinHPWPs;our results, while not denying the need to establish management practicesaimedatincreasingproceduraljusticeandreducingthe perceptions ofnepotism(Barnett&Kellermanns,2006)suggest thatHPWPsinvestmentsbeyondacertainlevelwouldactuallybe superfluous,giventhattheinformalmechanismsoriginatingfrom the family influence may act as substitutes of various formal techniques.

8.Conclusions,limitationsandfuturedevelopments

Inthis studyweexaminedtowhatextentHighPerformance Work Practices(HPWPs)contributetotheretentionof valuable employees in family versus non-family firms and in different situationsoffamilyinfluence.

Overalltheresultsindicatethat:(a)thefamilyfirmstatusand theconditionsstrengtheningthefamilyinfluenceonthefirmhave apositiveimpactonretention;(b)inthepresenceofHPWPs,the “familyeffect” combineswiththeformalpracticestoproducea positiveinfluenceonretention;and(c)someoftheoutcomesof thefamilyinfluenceandoftheadoptionofformalpracticesmaybe sosimilarthattheyproducesubstitutioneffectsonretention.

Withthisresearchweaddtotheknowledgeonhumanresource practicesinfamilyfirms,atopicthatisstillinitsinfancyinthe family business studies(Cruz et al.,2011), and inparticular, as explainedabove,weshednewlightonthecontrastingarguments regardingHRMpractices targetedatnon-familyemployees. We alsocontributetothesocialexchangeperspectiveinthestudyof familyfirmsandtotheHRMliteratureconcerningthewaysHRM practicesinteract.Ourstudymaysuggestseveralissuesforfuture research.Firstofall,wemeasuredtheintensityofuseofHPWPs, but we make only conjectures about the use of informal and relational practices, assumingthat a lower adoption of HPWPs correspondstoanhigherrelianceonsocial mechanisms.Future research should directly identify the informal dimensions of humanresource management practicesthat leverage onfamily socialcapital.

Amajorlimitationinourdatasetisthatwedonothaveaccurate measures of different degrees of family involvement in the management (except for the CEO family status), and this is probablyoneofthemainreasonswhywedidnotfindsignificant differenceswithinthefamilyfirmssubsample.Therefore,future studiesshouldfocusonmorefine-graineddifferentiationwithin family firmsaccordingtovariousdimensions offamily involve-ment.Thiscouldstimulatefurthercontributionsalsotothegeneral HRMliterature,forexamplebyexploringhowdifferentcognitive orientations of family versus non-family business managers impactontheadoptionofHRMpracticesandonthe employee-leveloutcomes.

Although statisticaltests suggest thatcommon methodbias shouldnotbeaproblem,theuseofmoreobjectivemeasuresand/ ormultiplerespondentswouldimprovethequalityofempirical results.Inparticular,itwouldbeinterestingtomeasuredirectly the perceptions of the employees in connectionas a mediator betweenHPWPsandturnoverintentions.

Moreover,qualitativetechniquescouldbemoreappropriateto investigate in-depth thephenomena under scrutiny,since they involverelationalprocessesamongfamilymembersand employ-ees. For example, a qualitative analysis of configurations of practices could lead to the identification of trade-off effects betweensingleinformalandformalpractices,sheddingmorelight on the combinatorial dynamics of HR practices and offering

Table6

Effectsofthedifferentpracticesontheretentionofvaluableemployeesinfamily andnon-familyfirms.

(1)FamilyFirms (2)Non-familyFirms PreviousPerformance 0.17** (0.06) 0.26*** (0.08) Size 0.10 (0.08)
0.14y (0.08) Hightechnologyindustry
0.06

(0.14) 0.06 (0.16) CompanyNationality
0.02 (0.17)
0.01 (0.17) Selectivestaffing
0.07 (0.11) 0.38*** (0.13) Intensivetraining 0.04 (0.14) 0.19y (0.14) Careerdevelopment 0.29** (0.13) 0.14y (0.12) Extensivecompensation 0.07 (0.11) 0.10** (0.07) Performanceappraisals 0.11** (0.06) 0.13** (0.07) Employeeparticipation
0.02 0.13*** (0.12) (0.04) Constant 2.35*** 1.91*** (0.59) (0.59) N 121 111 R-sq 0.23 0.25 AdjustedR-sq 0.17 0.17 Fstatistic 3.41*** 3.34***

suggestions to family firms about the most effective formal practicestobeimplemented.

AppendixA.

Items for measuring high-performance work practices (HPWPs).

Constructand resulting Cronbach’salpha level

Originalitemsfromprevious literature

Adapteditems

Selectivestaffing (a=0.822)

‘Wehavegonetogreat lengthstoestablishthebest staffingprocedurespossible’ (Patel&Conklin,2012:227).

SS1:“Theselectionofnew stafffollowsastructured process.”

‘Predictorswithhigh predictivevalidityareused (worksampletests, assessmentcentre, biographicalquestionnaire)’ (Selsetal.,2006:330) ‘Applicantsundergo structuredinterviews[...] beforebeinghired.’(Wright etal.,2005:425)

SS2:“Techniques,suchas structuredinterviews,work sampletests,orassessment centres,areusedtoassistin theselectionprocess.”

‘[...]thehiringcriteria usedtoselectemployeesin threecategories[...]’ (MacDuffie,1995:207)

SS4:“Hiringcriteriaareused toselectnewemployees.”

‘Therecruitmentand selectionactivitiesofthe companyaresystematically evaluated’(Selsetal.,2006: 330)

SS5:“Recruitmentand selectionactivitiesofthe companyaresystematically evaluated.”

Intensive training (a=0.836)

‘Istrainingahighpriority?’ (Patel&Conklin,2012:227)

IT1:“Trainingisahigh priority.”

‘[...]theleveloftraining providedtonewlyhired productionworkers, supervisors,andengineersin thefirstsixmonthsof employment[...]’ (MacDuffie,1995:208)

IT2:“Trainingisprovidedto newlyhiredstaff.”

‘Aredifferentkindsof trainingavailable?’(Patel& Conklin,2012:227)

IT3:“Thecompanyoffers differentkindsofformal trainingtoexistingstaff.” ‘Thecompanyhasastrategic

trainingplan’(Selsetal., 2006:330)

IT4:“Thecompanyhasa strategictrainingplan.” ‘Extenttowhichreactions,

learning,behaviouraland performanceeffectsafter companytrainingare measured.’(Selsetal.,2006: 330)

IT5:“Reactions,learning, behaviouraland performanceeffectsafter companytrainingare measured.” Career

development (a=0.776)

‘Thecompanyoffers operationalstaffthe possibilityofattaininga higherhierarchicallevel’ (Selsetal.,2006:330)

CD1:“Thecompanyoffers non-managementstaffthe possibilityofattaininga higherhierarchicallevel.” ‘Thecompanyoffers

operationalstaffthe possibilityofbecoming activeinotherfunctional domainsatthesamelevel’ (Selsetal.,2006:330)

CD2:“Thecompanyoffers non-managementstaffthe possibilityofbecoming activeinotherfunctional domainsatthesamelevel.” ‘Appraisalsystemrelatedto

successionplanning, concernedwithwhatan individualwillbecapableof doinginthefuture’(Sels etal.,2006:330)

CD3:“Theappraisalsystem isrelatedtosuccession planningandconcerned withwhatanindividualwill becapableofdoinginthe future.”

‘Employeeshaveclearcareer pathsinthisorganization’ (Patel&Conklin,2012:227)

CD4:“Intheorganization differentcareermodelsare availabletoemployees.” Extensive

compensation

‘Thelevelofemployees’total compensation’(Kerretal., 2007:60),rangingfrom

EC1:“Employeesreceivea highleveloftotal

(Continued) Constructand resulting Cronbach’salpha level

Originalitemsfromprevious literature

Adapteditems

andbenefits (a=0.698)

below,around,toabove industrystandard

compensation(comparedto theindustryaverage).” ‘Partofthewageofblueand/

orwhite-collarworkers dependsonindividual performancesormerit’(Sels etal.,2006:330)

EC2:“Salariesandwages partlydependonindividual performanceormerit.”

‘Numberofextrabenefits whichthecompanyoffersits employees’(Selsetal.,2006: 330)

EC3:“Thecompanyoffers fundedextrabenefitstoits employees(e.g.childcare facilities,companycars).” ‘Profitsharing(orsome

othertypeofcompanybased rewardsystem)’(Patel& Conklin,2012:227)

EC4:“Theorganizationhas someformof company-basedrewardsystem(e.g. profit-sharing).” Performance

appraisals (a=0.835)

‘Whatproportionofthe workforcereceivesformal performanceappraisals?’ (Huselid,1995:646) ‘Employeesinthisjob regularly(atleastoncea year)receiveaformal evaluationoftheir performance.’(Wrightetal., 2005:425)

PA1:“Thecompanyregularly conductsformal

performanceappraisalswith itsemployees.”

‘Useofasystemwhich specifiesprocedureand criteriafortheappraisal process’(Selsetal.,2006: 330)

PA2:“Performance appraisalsfollowaspecific procedure,andcriteriafor theappraisalprocessare available.”

‘Appraisalprocedurethat relatestotheallocationan attributionofawards, rewardsandbenefits’(Sels etal.,2006:330)

PA3:“Appraisalsarerelated totheallocationofawards, rewards,andbenefits.”

‘Appraisalprocedureaimed atthedevelopmentand motivationofstaffby lookingathowwellheorshe isdoing’(Selsetal.,2006: 330)

PA3:“Appraisalsareaimed atgivingfeedbackonthe developmentofemployees.”

‘Employeeappraisals emphasizelongtermand group-basedachievement.’ (Patel&Conklin,2012:227)

PA5:“Employeeappraisals emphasiseanemployee’s pastperformanceaswellas theirfuturepotential.” Employee

participation (a=0.786)

‘[...]degreeofinfluence overtasks,tools,work methods,paceofwork, schedules,vacations,and technologydesign[...]’ (Batt,2002:591)

EP1:“Employeesarefreein organisingtheirwork(e.g. tasks,tools,speedofwork, workschedules,holiday).”

‘Employeerepresentationat board/seniormanagement meetings’(Patel&Conklin, 2012:228)

EP2:“Employeesare representedatboard/general managementmeetings.” ‘Degreetowhich

non-supervisoryemployees participatein“management” decisionsregarding investment,workflow, productdevelopment,and productivitymonitoring.’ (Arthur,1994:495)

EP3:“Employeescan influencemanagement decisionsoninvestments, workflows,product development,and productivitymonitoring.”

‘Extenttowhich,e.g.via consultation,consideration isgiventoemployees’ opinions’(Selsetal.,2006: 330)

EP4:“Consideration,e.g.via consultationorsuggestion systems,isgivento employees’opinions.” ‘Whatistheproportionof

theworkforcewhoare includedinaformal informationsharing program(e.g.,a

EP5:“Thecompanyregularly sharesinformationon strategy,performance outcomesorthe