Age-related differences in the relations betweenindividualised HRM and organisationalperformance: a large-scale employer surveyP.MatthijsBal, School of Management, University of BathLucDorenbosch, TNO|ResearchInstituteforWorkandEmploymentsHumanResourceManagementJournal, Vol25, no1, 2015, pages4161The current studyaimedtoinvestigate the relationshipbetweenindividualisedHRMpractices andseveral measures of organisational performance, including the moderating role of employee age in theserelationships. A large-scale representative study among 4,591 organisations in the Netherlands showedsupport for the relationships betweenindividualisedHRpractices withorganisational performance.Employee age moderatedthe relationships betweenthe use of individualisedpractices andsicknessabsenceandturnover, suchthatorganisationswithahighpercentageof olderworkersbenetedfromworkschedule practices, andorganisations withhighpercentage of younger workers benetedfromdevelopmentpractices.Contact: P. Matthijs Bal, School of Management, University of Bath, Bath BA2 7AY, UK. Email:p.m.bal@bath.ac.ukKeywords: individualisedHRMpractices; organisational performance; employee turnover;sicknessabsence; olderworkersAs organisations increasingly stress the importance among employees to be responsiblefortheirowncareers(Greenhauset al., 2010), agrowingnumberof employeeshavebeguntonegotiate individual workarrangements withtheir employers (Rousseau,2005). Moreover, this trend of individualisation has occurred along with a decrease in collectiveagreements for employees (De Leede et al., 2004; Glassner andKeune, 2012). Accordingly,academicinteresthasbeguntofocusonhowemployeesproactivelyshapetheircareersandnegotiate individual agreements (Rousseau, 2005; Grant andParker, 2009; Bal et al., 2012).Studies on idiosyncratic deals, or i-deals, have shown that individuals who proactivelynegotiateindividual agreementsbecomemorehighlymotivated, committedandperforming(e.g. Hornunget al., 2008; Anandet al., 2010; Rosenet al., 2013). However, researchfromanorganisational perspectiveontheincreasingindividualisationofworkislacking(TaskinandDevos, 2005). This is surprising, given the strong increase in interest on outcomes of individualemployee negotiation. It is currently unknown whether this individualisation of HRM actuallyimprovesorganisationalperformance.Studies on the effects of individualisation on the employee level have shown thatrelationships with outcomes are inconsistent and differ greatly among studies (Hornung et al.,2008, 2010; Bal et al., 2012). We argue that individualisation is primarily benecial when it is inline with other aspects in the organisation (Delery and Doty, 1996; De Leedeet al., 2007), andin particular the number of older workers in an organisation is crucial in determining the effectsofindividualisationofHRM(Bal et al., 2012, 2013; Kooij et al., 2013). Becauseworkforcesareaging throughout the world (Wang and Shultz, 2010), the need for retention of older workersin organisations has become a prominent area of research (Wang and Shultz, 2010). At the sametime, it has become more difficult toretainolder workers (as well as their expertise andbs_bs_bannerdoi: 10.1111/1748-8583.12058HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 41 2014 John Wiley & Sons Ltd.Please cite this article in press as: Bal, P.M. and Dorenbosch, L. (2015) Age-related differences in the relations between individualised HRM andorganisational performance: a large-scale employer survey. Human Resource Management Journal 25: 1, 4161.knowledge)becausemanyolderworkersleavetheworkforceearly(WangandShultz, 2010).Becauseoftheincreasingagediversityintheworkplace(Schlicket al., 2013), itisimperativethat organisations implement practices that allow older workers to maintain their productivity,as performance may be decreasing after the age of 4045 (Ng and Feldman, 2008).IndividualisedHRMmayfacilitateyounger andolder workers tonegotiateindividualisedagreements to increase or maintain their performance, and hence contribute to overallperformanceoftheorganisation.The objective of this study is to test the effectiveness of individualisation on organisationalperformance, specicallywhether theeffectivenessof individualisationdependsontheagecomposition of the organisation. First, we investigate whether the availability and actual use ofindividualisedHRMcontributes toorganisational performance. Second, the studyaims todetermine the conditions under which individualisation has the greatest effect onorganisational performance by investigating the moderating role of employee age in theorganisation. Basedonthenotionthatolderworkershavedifferentwork-relatedneedsfromyounger workers, we expected differences in the relationships of various types ofindividualisedHRMwithorganisationalperformance(Balet al., 2012; Kooij et al., 2013).This study contributes to research on individualisation of work arrangements by being therst to investigate the effects of individualisation on organisational-level rather than onindividual-level outcomes (Hornung et al., 2008). Demonstrating that individualisationcontributes to the bottom line not only furthers our understanding of individualised HRM, butit also investigates the effects of individualisation in a society where collective agreements areslowlydisappearing(DeLeedeet al., 2004). Moreover, thestudycontributesbyinvestigatingtheconditionsunderwhichindividualisationismosteffective.Welookattheroleoftheagecomposition within the organisation, and through this we open up new pathways for researchon HRM. Finally, this study contributes to previous research on individualisation through theinvestigation of a large-scale employer sample (including numerous organisations in differentsectors) andtherebyobtaininga comprehensive perspective onhowindividualisedHRMinuencesorganisationaloutcomes.THEORY AND HYPOTHESESThe individualisation of HRM has become increasingly common in organisations as a result ofglobalisation, the information economy and the democratisation of the workplace (Taskin andDevos, 2005). Employees are becoming more proactive in looking for opportunities to negotiateindividual agreementswiththeiremployers(Rousseauet al., 2006; Grant andParker, 2009),while at the same time organisations are also expecting employees to become more proactive.Consequently, organisations increasingly provide employees with the individual opportunity tonegotiate agreements about work arrangements. This differs fromthe traditional HRMapproach, whichis fundamentallybasedonequal treatment of all employees (Boxall andMacky, 2009).ResearchoneffectivenessofHRMhasfocusedprimarilyontheuniversalisticoutcomesofhigh-performance HRM(Delery and Doty, 1996; Boxall and Macky, 2009), based on theassumption that HR practices have a universal effect on motivation and performance among allemployees. This high-performance approach to HRM (e.g. Kehoe and Wright, 2013) postulatesthat the more HRM is available in organisations, the higher rm performance will be. However,asKaufmanandMiller(2011) argued, thisstatement maybeoversimplied. Acontingencyapproach, which stresses the idea that HRM should be in line with the goals and the contextIndividualised HRM and organisational performanceHUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 42 2014 John Wiley & Sons Ltd.of the organisation (Delery and Doty, 1996), may be necessary to understand the consequencesofHRM.IndividualisedHRMgoesbeyondthecontingencyapproachbyintroducingaperspectivebased on the individual employee rather than the goals of the organisation. While there is someresearchthatshowsthatHRMshouldbealignedwiththeneedsofemployees(Kinnieet al.,2005; Balet al., 2013), thereisnoresearchthatspecicallyfocusesontheindividualemployeeas the basis of HRM. The trend of individualisation in organisations is a reection of a broadersocietal trendof individualism(Oysermanet al., 2002), indicatedbythedeclineof collectivesocial structures, and a stronger focus on the individuals responsibility for their own welfareand well-being. This has also affected the traditional system of collective bargaining (De Leedeet al., 2004), withcollective agreements as the basis for HRMpractices slowlydecreasing(Glassner andKeune, 2012). Inresponse tothe decrease inemployee protective collectiveagreementsandadecentralisationofbargainingtotheindividual employeelevel (DeLeedeet al., 2004), individualisationhasgrownsignicantlyasthebasisfororganisational strategicHRM, and thus individualised HRM has become more common in organisations. Consequently,organisations increasingly allowtheir managers to make individual agreements with theemployees.We dene individualised HRM as an HR system where managers have the opportunity andactuallyusetheopportunitytoindividuallynegotiateagreementsabout workarrangementswithindividual employees. Inthecurrent study, weapproachindividualisedHRMas HRprogrammes that areimplementedas HRpractices inanorganisation(Arthur andBoyles,2007). Hence, insteadof astandardisedapproachbasedonequal treatment of employees,individualisedHRMrefers totheextent that managers andemployees areempoweredtonegotiatearrangementsthattthespecicneedsandpreferencesoftheindividualemployee(De Leede et al., 2004). Individualised HRM includes customised work arrangements or i-dealsonissuessuchasworkinghours, rewards, trainingandcareerdevelopment(Rousseau, 2005;Hornunget al., 2009).IndividualisedHRMis not a newphenomenoninorganisations (Rousseau, 2005), butindividual dealshavetraditionallybeennegotiatedundertheradarandthusoutsideoftheorganisations control and agreement. This may have contributed to perceptions of unfairnessandcronyism(Bricket al., 2006). IndividualisedHRM, however, concernstheorganisationsexplicit approval of individual negotiations of employees with managers as a strategic meansof achievingthegoalsof theorganisation(DeLeedeet al., 2004; Rousseau, 2005). Moreover,individualised HR should be in line with existing law and collective agreements, which everyorganisationhastoadhereto.Individualised HRM is similar, yet different, from i-deals, or idiosyncratic deals employeesnegotiate with their employers (Rousseau, 2005; Rousseau et al., 2006). I-deals arenegotiationsofindividual employeeswiththeirorganisations, andprimarilyinitiatedbytheemployee, whileindividualisedHRMreferstoaformalisedapproachbytheorganisationtocustomise workarrangements. IndividualisedHRMthus makes individualisationof workarrangements available to all employees, contributing to higher fairness (Greenberg et al.2004).We make a further distinction between the availability and actual use of individualised HRM.This is in line with the strategic HR literature, which distinguishes between HR practices thatareavailabletolinemanagersfromtheactualuseofthesepractices(e.g. ArthurandBoyles,2007). We applythe same logic toindividualisedHRM: inthe former case, organisationsprovide leewayfor line managers tonegotiate if andwhenemployees askfor individualagreements(i.e. availability), whereasinthelatter, linemanagersactuallyusethisleewaytoP. Matthijs Bal and Luc DorenboschHUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 43 2014 John Wiley & Sons Ltd.negotiate agreements with employees. Both availability and actual use may inuence positiveoutcomes, althoughviadistincttheoreticalprocesses.First, the effects of availability of individualised HRM on organisational performance can beexplainedwithsignallingtheory. Signallingtheory(Casper andHarris, 2008) proposesthatbecauseemployeeshaveincompleteinformationabouttheorganisationsintentions,theyusesignals fromthe organisationtodrawconclusions about anorganisations intentions andactions. As such, the availability of individualised HRMfunctions as signals of theorganisations benevolent intentions towards employees (e.g. Takeuchi et al., 2009). Socialexchange theory(Blau, 1964) proposes inturnthat employees will reciprocate these goodintentions through increased commitment to the organisation, and consequently higherperformanceandretention(Balet al., 2013).Second, the effects of the use of individualised HRM on organisational performance can beexplainedusing the normof reciprocity, whichis also relatedto social exchange theory(Gouldner,1960;Blau,1964). Accordingtosocialexchangetheory,whenanemployeeandanemployer commit to each other in an exchange relationship, reciprocal obligations between thetwo parties drive their behaviour. Individualised HRM serves as a basis for reciprocity betweenthe employee and the organisation because the mutual obligations that have been agreed uponstrengthen the employment relationship. More specically, the organisation negotiatesindividual deals withemployees, andinreturn, employees become more attachedtotheorganisation(Hornunget al., 2008; NgandFeldman, 2010)andcontributetoahigherdegree(Hornung et al., 2008). Hence, the use of individualised HRM in organisations is expected to bepositively related to organisational performance. In this study, we adopt a broadconceptualisation of organisational performance that includes three distinct performanceindicators: operational performance growth, sickness absence and voluntary employee turnover(PeretzandFried, 2012).Individualised HRMpractices may entail various types of agreements, but previousresearch has shown that the most common agreements are aimed at development (i.e.training and career development), exibility in work schedules (i.e. working hours) andnancialagreements(i.e. salary; DeLeedeet al., 2004, 2007; Rosenet al., 2013). Hence, inthisstudy,wedifferentiateamongdevelopment,workschedulesandpayarrangementspractices,and we expect that these three practices will be differentially related to types oforganisational performance. Development practices motivateandrewardhighperformance(Hornung et al., 2008, 2011). Through development, including training and specialopportunitiesforskill development, employeesmayenhancetheirownperformance. InlinewiththeAMO-model (Appelbaumet al., 2000), development provides employees withtheabilities andmotivationtoperform. Hence, development practices motivate employees toperformbutalsotostaywithintheorganisation. Inlinewiththenormofreciprocity, whenemployeesreceivedevelopment,theybecomemorecommittedtotheorganisationandhencemorelikelytostay. Thus, weexpectthatdevelopmentpracticesarerelatedtoorganisationalperformanceas well as turnover.Individualisedworkschedulepracticeswillenhanceemployeemotivationinlinewiththework adjustment model; through negotiation of personalised work schedules, greatercorrespondenceisachievedbetweentheemployees abilitiesandtherequirementsofthejob(Balteset al., 1999). Whenjobrequirementsareadaptedtoindividualabilities, employeesarebetter able to full their job role; for example, exible work schedule practices allow employeesto arrange their work hours to better align with their personal situation. Consequently,employeesarebetterabletodotheirjobwithout droppingout (e.g. throughburnout), andhence work schedule practices are expected to contribute to lower employee sickness absence.Individualised HRM and organisational performanceHUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 44 2014 John Wiley & Sons Ltd.Previous research has shown that exible work arrangements indeed tend to reduceabsenteeisminorganisations(DeMenezesandKelliher, 2011).Finally, individualisedpaypracticesreect theeconomicconditionsof ajob(Rosenet al.,2013)andsignaltoemployeesthatthecurrentorganisationvaluesandwantstoretainthem.Paypractices alsoincreasecontract unreplicability(Rosenet al., 2013) becausethenancialbenetsof ajobcanbeeasilycomparedwithotherjobsandhencedecreasethelikelihoodthat employees will turnover. Thus, nancial practices areexpectedtoberelatedtoloweremployeeturnover becauseemployersarelikelytooffer special compensationpackagesorincentivestotheirvaluedemployeesinordertoretainthem. Insum, bothavailabilityanduseof individualisedHRMinorganisationareexpectedtobepositivelyrelatedtotypesoforganisational performance. Tosummarisethearguments above, our rst threehypothesesare:Hypothesis1: Availabilityanduseof individualiseddevelopment practicesarepositivelyrelatedto(a)performancegrowthandnegativelyrelatedto(b)employeeturnover.Hypothesis 2: Availability and use of individualised work schedule practices are negativelyrelatedtosicknessabsence.Hypothesis 3: Availability and use of individualised pay practices are negatively related toemployeeturnover.EMPLOYEE AGE AND EFFECTIVENESS OF INDIVIDUALISED HRMWe argue that the effectiveness of individualised HRM is dependent upon the context and inparticular the composition of the employee population (Bal et al., 2013; Kooij et al., 2013). Socialexchange theory predicts that the strengths of the effects of social exchanges between employeeandorganizationaredeterminedbytheutilityemployeesattachtoresources. Accordingly,previous studies have shown that individualisedagreements are more likely to producepositiveoutcomeswhentheyareinlinewithemployeeneeds(Anandet al., 2010; Bal et al.,2012). We argue that the relationship between the use of individualised HRMandorganisational performancewill dependupontheextent towhichit tstheneedsof olderworkers.WeexpectthistobethecaseforuseofindividualisedHRMratherthanavailabilitybeccause the utilityof individualisedHRMis manifestedparticularlywhenworkers haveactuallynegotiatedagreementswiththeirorganisation(Rousseau, 2005).Lifespan psychology has shown that aging is associated with changes in needs andpreferences (Baltes and Baltes, 1990; Kooij et al., 2011). Socioemotional selectivity theory(CarstensenandMikels, 2005; Carstensen, 2006)hasbeenusedextensivelyinunderstandinghowolder people differ fromyounger people inmotivationandbehaviour, as well as inexplaining the impact of age on work behaviours (Kooij et al., 2011). Socioemotionalselectivitytheorystatesthatinyoungadulthood,timeisperceivedasexpansive(Carstensen,2006). Young people have an open future time perspective and prepare for a long andunknown future, and therefore primarily focus on growth and knowledge-related goals.Older people, however, increasingly experience time as running out. For them, theexperience of approaching the endtheir careers andlife causes a shift towards present-related emotional goals over knowledge goals and a focus on emotional well-being(CarstensenandMikels, 2005). Because younger people have broader time horizons, theyprepare for a long and unknown future by learning and seeking growth opportunities. Olderpeople, however, increasinglyexperiencetimeasrunningout, andhenceperceivelessfutureP. Matthijs Bal and Luc DorenboschHUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 45 2014 John Wiley & Sons Ltd.intheir organisation, causingthemtoprioritise present-relatedgoals over future-orientedgoals.Moreover, the lifespanselection, optimisationandcompensation(SOC)-model of Baltes(1997; Baltes & Baltes, 1990) proposes that throughout life, people experience gains and lossesin physical and mental capabilities, and they are in general focused on maximising the benetsof thesechanges whileminimisingtheir losses (Kanfer andAckerman, 2004). Tominimiselosses in outcomes due to the losses in abilities aging people experience, they select fewer goalsand refrain from learning so that they do not have to spread their diminished resources overtoomanygoalsandcanthusremainhealthyandproductivecontributorsintheorganisation(Baltes and Baltes, 1990; Baltes, 1997). Hence, the SOC-model also predicts that while youngerworkershavehighergrowthneeds, thereisadeclineintheseneedsoverthelifespan. Olderworkers cope withage-relatedlosses, suchas declines inhealth, physical capabilities andmemory, and become more focused on maintaining what they have and minimising the effectsofthelossestheyexperience.Insum, theselifespantheories suggest that whileyounger workers aregenerallymorefocused on building their careers, learning and growth, older workers employ strategies to copewith age-related losses. Hence, the utility of different types of individualised HR practices willaccordingly vary depending on the age of the workers. Individualised pay arrangements anddevelopment will be more important for younger workers because these facilitate career growthandlearning. Individualisedwork schedules, however, will be more important for olderworkersbecausetheyfacilitateamoreexiblewayofcopingwithage-relatedlossesandthedemands at work. Thus, the possibility for older workers to negotiate an individualised workschedulewiththeir employer enables themtoremainproductiveandprevents themfromhighersicknessabsence. Weexpectthereforethatinorganisationswithmanyolderworkers,use of individualised work schedules will be more strongly related to performance and sicknessabsence.In contrast, we expect individualised development and pay practices to be more importantamong younger workers (Ebner et al., 2006). Younger workers primarily tend to seek tooptimise resources or maximise economic gains and career development, enhancing their statusand advancement within their organisation and career (Maurer et al., 2003). Recentmeta-analytic work has indeedshownthat growthandextrinsic work motives are moreimportant for younger workers than for older workers (Kooij et al., 2011). Extending this logic,we propose that the use of development and pay agreements are more important for youngerworkers, and hence in organisations with many younger workers, individualised developmentand pay practices will be more strongly related to performance growth and turnover. Based ontheabove, weproposethefollowinghypotheses:Hypothesis 4: Employee age moderates the relationship between use of individualiseddevelopment practices and (a) performance growth and (b) employee turnover, such that therelationisweakerfororganisationswithahighpercentageofolderworkers.Hypothesis 5: Employee age moderates the relationship between use of individualised workschedule practices and sickness absence, such that the relation is stronger for organisationswithahighpercentageofolderworkers.Hypothesis6: Employeeagemoderatestherelationshipbetweenuseof individualisedpaypracticesandemployeeturnover,suchthattherelationisweakerfororganisationswithahighpercentageofolderworkers.Individualised HRM and organisational performanceHUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 46 2014 John Wiley & Sons Ltd.METHODSSample and procedureTheNetherlandsEmployersWorkSurvey(NEWS; Oeij et al., 2011)isastudyofemploymentarrangementsinorganisationsintheNetherlandsandwascarriedout in2010. NEWSisarepresentativesurveyamongmorethan5,000for-prot as well as non-prot organisationscountingtwoormoreemployees. Becausethecurrent studywaspart of alargerstudyonemployer policies and conditions, the survey included various other questions. Therefore, thelikelihood of respondents being aware of the aims of the current study would be minimal. Thesampleselectedwasastratiedsamplebasedonsectorandorganisationsize. Organisationswere approached by mail and telephone to participate in the research at the establishment level.Thismeansthatforlargerorganisationswithmultiple(regional)establishments, respondentswereapproachedat alower hierarchical level wheretheycouldmoreaccuratelyjudgetheactual useof individualisedHRMpractices. Thefocusof thecurrent studyisthusontheestablishmentlevel.Respondents (company owners, management team members or HR managers) were able toparticipate throughllingout either a paper-and-pencil or a digital questionnaire. It wasdeemed appropriate to ask company owners or HR managers to act as organisationalrepresentatives and to ll out the survey because they would be aware of the policies of theirorganisation, aswellastheextenttowhichindividualisedHRMpracticeswouldbeactuallyused(Arthur andBoyles, 2007). Moreover, onaveragetheestablishments consistedof 161employees, becauseof whichit is likelythat HRmanagers wereabletoaccuratelyassessavailability and use of HR practices. If they were unaware of the use of individual agreementsin their organisation, it would be likely that the existence of these individual agreements wasinfact cronyismor favouritismrather thanorganisationallyapprovedindividualisedHRM(Rousseau et al., 2006). The initial response was from 5,518 establishments (37 per cent responserate). Three thousand three hundred seventeen (60 per cent) represented independentcompanies that were not part of a larger rm. One thousand four hundred seventeen responseswere based on establishments of Dutch companies or multinational companies (with a foreignowner), and784werebasedonseparateheadofficesof Dutchormultinational companies.Thereisnoindicationthat oneparticularorganisationwasoverrepresentedinthedataset.After deletingparticipants withmissingresponses, we obtaineda nal response of 4,591organisations (31 per cent response rate). Thirty-eight per cent of the respondents were directororowner,36 per centwereHRmanagers,14 per centestablishmentmanagersand12 per centhad another function in the organisation. Seventy-one per cent of the organisations werefor-prot rms, 22 per cent non-prot and 7 per cent had both for-prot and non-protactivities.MeasuresIndividualisedHRpracticesweremeasuredinlinewithpreviousresearchonHRMaswell asi-deals(e.g. CasperandHarris, 2008; Hornunget al., 2008, 2009). Availabilityof individualisedHRM was measured by asking respondents the extent to which in their organisation differentagreements couldbemadewithindividual employees. Responses couldbeprovidedonave-point scale (1 = not available at all; 5 = available to a great extent). Availability wasmeasured with one-item scales for development (development/education of employees), workschedules(workinghoursofemployees)andpayarrangements(salaryofemployees). Useofindividualised HRM was measured by asking respondents to indicate the extent to which in theirorganisation supervisors actually negotiated individualised agreements with employeesP. Matthijs Bal and Luc DorenboschHUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 47 2014 John Wiley & Sons Ltd.(1 = not at all; 5 = to a very great extent). Use of individualised HRM was measured with thesame items as availability. One-item scales were used because of restrictions on survey length.Eventhoughthereliabilityofone-itemscalescouldnotbeassessedinthisstudy, thescaleshave strong practical relevance to the participants, as individualised HRMwas widelyacknowledgedtobeanimportanttopic(Balet al., 2012).Organisational performance was conceptualised broadly as the effectiveness of the organisationto perform, ensure employee well-being and retention. It was measured using three indicators.Performance growth ( = 0.72) was measuredthroughthree items referring to performancegrowth during the last 2 years. We chose performance growth because many organisations fromvarious sectors took part in the study, and objective indicators such as sales rates, prots or ROIarenot applicabletoeveryorganisation(suchasnon-prot organisations; PeretzandFried,2012). Ratings of organisational performance growth have been estimated as valid and reliableindicatorsoforganisational performance(Gonget al., 2009; Ngoet al., 2009). Theitemswere:Over the last two years, the labour productivity in our organisation has . . ., The quality of ourproductsand/orserviceshas . . . andthesatisfactionof thecustomersof ourorganisationhas . . .. Answers could be provided on a ve-point scale ranging from 1 = strongly decreasedto 5 = strongly increased. Sickness absence was measured by asking respondents the percentageof sicknessabsenceduringthepreviousyear (2009), excludingpregnancyleave. Themeanpercentage (M) was 3.51 per cent, [standard deviation (SD) = 3.67]. Employee turnover(M = 5.06 per cent, SD = 13.35) was measuredbyaskingthenumber of contracts that werevoluntarily ended by employees themselves during the last year. This number was divided bythetotal numberof employeeswithapermanent contract intheorganisationtoobtainthepercentage of employee turnover. The moderator employee age was measured by indicating thepercentage of employees older than 45 years in the organisation (M = 39.6 per cent, SD = 24.78).Forty-veyearsisgenerallyconsideredtobetheageafterwhichemployeesareregardedasolder workers and from that age experience increasing problems with their (physical) abilitiesto do their jobs (Kooij et al., 2008; Ng and Feldman, 2008). While there is no strong theoreticalcut-off point for distinguishing younger and older workers, age causes gradual changes in howpeople experience their work (Kooij et al., 2011). In the meta-analysis of Ng and Feldman (2008)on the relationship between age and job performance, it was estimated that the relationship ofagewithjobperformancewaspositiveuntil theageof 40, afterwhichit becamenegative.Moreover, in the review of Kooij et al. (2008), it was shown that the effect of chronological ageonworkmotivationchangedaftertheageof4045. Hence, thereisageneralconsensusthataftertheageof 4045, peopleexperienceage-relatedchangesandperceivechangesintheirmotivationintheirworkasaresultofaging. Moreover, researchshowsthataftertheageof4045, peoplehavehigherriskforwork-relateddiseases(Alaviniaet al., 2009).Control variablesIn the analyses, we controlled for a range of factors that could possibly inuence the outcomevariables (see also Gong et al., 2009; Ngo et al., 2009). Education was measured (using dummycoding) by the percentage of employees who had lower education (M = 30.84 per cent,SD = 31.09 per cent), vocational education(M = 40.46 per cent, SD = 28.54 per cent)andhighereducation (M = 28.73 per cent, SD = 31.42 per cent). Gender was measured as the percentage ofmale employees (M = 58.38 per cent, SD = 30.67 per cent). Moreover, we controlled for thepercentageof employeeswithatemporarycontract (M = 10.49 per cent, SD = 14.78 per cent)and the percentage of employees working part time (M = 37.89 per cent, SD = 31.74 per cent) torule out alternative explanations, such as that turnover rates are inuenced by the percentageof employees with a temporary contract. Furthermore, we controlled for sector (using dummyIndividualised HRM and organisational performanceHUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 48 2014 John Wiley & Sons Ltd.coding; industry/agricultural: 26 per cent; service: 47 per cent; government: 4 per cent;education: 9 per cent; health care: 8 per cent; other sectors: 6 per cent). Finally, we controlled fororganisation size because larger rms may have more resources and market power (Gong et al.,2009). Organisationsize(M = 162, SD = 546.98) wasmeasuredbythenumber of employeesworking for the organisation. For multinational organisations, respondents indicated thenumberofemployeeswithintheNetherlands.AnalysisBecausesomeofthevariableswerenon-normallydistributed, weappliedlogtransformationto the variables education, gender, percentage of temporary employment, percentage ofpart-time workers, organisation size, percentage of employees above 45 years, sickness absenceandemployee turnover (Finchet al., 1997). The hypotheses were testedusing moderatedhierarchical regressionanalyses. Independentvariableswerestandardisedbeforeinteractionswerecalculated(AikenandWest, 1991). Intherststep, controlvariableswereaddedtothemodel (not shown in table). For categorical variables, we createddummy variables andincludedtheseintheanalyses.Foreducation,percentageofemployeeswithlowereducationwas thereferencegroup, andfor sector weusedindustry/agricultural as referencegroup.Subsequently, main effects were added in the second step and in the nal step the interactions.Weincludednon-hypothesisedmaineffects(e.g. ofworkscheduleandnancial practicesonperformance growth) as well as non-hypothesised interactions to rule out alternativeexplanations. Signicant interactions were plotted with slopes for one SD below and above themeanof themoderator (AikenandWest, 1991). Table 1shows thecorrelations amongthevariables.RESULTSHypothesis 1 predicted that availability and use of individualised development practices wouldbe positivelyrelatedtoperformance growthandnegativelyrelatedtoemployee turnover.Table 2showstheresultsof thehierarchical regressionanalyses. Bothavailability( = 0.079,p < 0.001) and use ( = 0.058, p < 0.01) of individualised development practices were positivelyrelatedtoperformancegrowth. Hence, Hypothesis1awasfullysupported. Bothavailabilityand use of individualised development practices for employees are related to strongerperformance growth of the organisation. However, availability of individualised developmentpractices was not related to employee turnover ( = 0.003, ns), and use of development practiceswasalsounrelatedtoemployeeturnover( = 0.031,ns).Hence,Hypothesis1bwasrejected.Hypothesis2predictedthatavailabilityanduseofindividualisedworkschedulepracticeswould be negatively related to sickness-related absence. Availability ( = 0.066, p < 0.001) butnot use( = 0.018, ns) of individualisedworkschedulepracticeswasnegativelyrelatedtosickness absence, indicating lower sickness absence in organisations where individualised workschedules are available. Hence, Hypothesis 2 was partially supported. We also found anon-hypothesised relationship between use of work schedule practices with performancegrowth ( = 0.067, p < 0.01), indicating that organisations with more employees usingindividualisedworkschedulesobtainedhigherperformancegrowth.Hypothesis 3 predicted that availability and use of individualised nancial practices wouldbe negatively related to employee turnover. This hypothesis was partially supported;availability( = 0.045, p < 0.05)butnotuse( = 0.041, ns)wasrelatedtoemployeeturnover.Hypothesis 4 predicted that employee age would moderate the relationship between use ofindividualised development practices and performance growth and employee turnover. Table 2P. Matthijs Bal and Luc DorenboschHUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 49 2014 John Wiley & Sons Ltd.TABLE1Correlationsbetweenvariablesinthestudy(N=4591)VariablesMeanSD123456789101112131415161718192021Education1%vocationaleducation40.4628.542%highereducation28.7331.420.45**3Gender(%men)58.3830.670.06**0.19**4%temporaryemployment10.4914.780.020.04**0.12**5%part-timeworkers37.8931.740.05**0.12**0.70**0.20**Sector6Service0.460.09**0.04**0.07**0.16**0.07**7Government0.040.010.05**0.020.07**0.03*0.18**8Education0.090.17**0.37**0.20**0.020.19**0.28**0.06**9Healthcare0.080.09**0.06**0.42**0.000.38**0.30**0.06**0.09**10Othersectors0.060.020.020.15**0.04**0.15**0.23**0.05**0.07**0.08**11Organisationsize(numberofemployees)161.63546.980.020.05**0.06**0.04*0.06**0.06**0.08**0.03*0.14**0.02AvailabilityofindividualisedHRM12Development3.610.800.020.14**.0230.020.030.020.020.05**0.13**0.05**0.09**13Workschedule3.231.020.010.13**0.10**0.06**0.15**0.08**0.06**0.08**0.03*0.06**0.000.28**14Payarrangements2.831.090.05**0.010.18**0.000.19**0.17**0.12**0.18**0.13**0.05**0.09**0.26**0.38**UseofindividualisedHRM15Development3.260.880.010.17**0.03*0.000.010.020.04*0.07**0.030.020.03*0.61**0.18**0.14**16Workschedule2.971.020.010.07**0.09**0.10**0.13**0.09**0.020.08**0.06**0.06**0.020.22**0.65**0.26**0.35**17Payarrangements2.601.070.04*0.010.17**0.020.19**0.16**0.09**0.17**0.12**0.04**0.04**0.22**0.31**0.72**0.29**0.40**18%ofemployees>45years(age)39.6424.780.07**0.05**0.000.21**0.04**0.23**0.11**0.16**0.04**0.03*0.11**0.05**0.04**0.15**0.03*0.05**0.14**19Performancegrowth3.460.520.010.11**0.10**0.09**0.06**0.04**0.010.020.06**0.010.020.15**0.13**0.06**0.16**0.15**0.08**0.09**20Sicknessabsence3.513.670.08**0.05**0.04*0.020.04*0.17**0.07**0.05**0.11**0.010.14**0.010.11**0.16**0.010.09**0.12**0.14**0.0221Employeeturnover5.0613.350.04**0.03*0.07**0.25**0.09**0.12**0.04**0.03*0.010.010.04**0.04**0.010.010.020.05**0.020.17**0.010.02*p