the effects of ltf

Upload: tmmt

Post on 04-Jun-2018

226 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/13/2019 The Effects of LTf

    1/22

  • 8/13/2019 The Effects of LTf

    2/22

    Journalof OrganizationalBehaviorJ. Organiz.Behav.22, 849-869 (2001)DOI: 10.1002/job.116

    Explaining the effects of transformationalleadership: an investigation of the effectsof higher-order motives in multilevelmarketing organizationsJOHN R. SPARKS* AND JOSEPH A. SCHENKUniversityof Dayton,Dayton, U.S.A.

    Summary Multilevelmarketing rganizationsMLMs)area rapidlygrowingorganizationalypeenlistingnearly10millionmembersandproducing ver 20 billion dollars n salesannually.Despitetheirremarkableecentgrowth, ew studieshave examined heseunusualorganizations, ndnone ofthese have addressed ssues of transformationaleadership. nMLMs,the key leadership ela-tionshipsare those between individualmember distributors nd the memberswho recruitedthem into the organization i.e., their 'sponsors').Although sponsorsareexpectedto provideleadership o themembers heyrecruit, heypossess nodirectsupervisory esulting--authorityin an uncertain quasi-leadership'ole.Using a sampleof 736 femaleMLMmembers, he pre-sentstudyempirically estsanimportant xplanatory omponentof transformationaleadershiptheory: hatbelief in the higherpurposeof one's workis a mechanism hroughwhich transfor-mational eadershipachieves ts positiveoutcomeson cohesion,satisfaction, ffort,andperfor-mance. The results offer support to the notion that transformationaleadership indeed'transforms'ollowersby encouraginghemto see thehigherpurposes n theirwork. Addition-ally, the resultsshow positiverelationshipsbetweenbelief in a higher purposeof one's workandjob satisfaction,unitcohesion,andeffort.Copyright? 2001 JohnWiley & Sons, Ltd.

    IntroductionMulti-level marketing organizations (MLMs) have long been an active part of the American businesslandscape. These networks of member-distributors, who earn income both from selling retail productsand recruiting new members, are currently enjoying remarkable growth. According to the Direct Sel-ling Association (1999), from 1991 to 1998, MLM participation grew from 5.1 million to 9.7 millionmembers (75 per cent of whom are women) while annual sales grew from 13 billion to nearly 23 billiondollars. Traditionally limited to household cleaners (e.g., Amway) and cosmetics (e.g., Mary Kay),MLMs today sell such diverse products as jewellery, nutritional supplements, children's toys, lifeinsurance, and long distance telephone services (King, 1996; Salter, 1997). Moreover, because MLMs*Correspondenceo:JohnR.Sparks, niversityfDayton,Management/Marketingepartment,00CollegePark,Dayton,OH45469-2235,U.S.A.E-mail: [email protected]

    Received 10 June 2000Revised 1 September2000Accepted20 June2001Copyright? 2001 JohnWiley & Sons, Ltd. Publishedonline 25 October2001

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    3/22

    850 J. R. SPARKSAND J. A. SCHENKcan rapidlyand inexpensively create large distributionnetworks,these trendsseem likely to con-tinue-strongly suggestingthatan examinationof this organizationalphenomenon s long overdue.As we discuss in the following section, MLMscombine attributesof traditional ndustrialsellingwith small retailentreprenuershipndoperateusing fluidorganizational tructureshatpose complexleadership problems.Much of this complexity exists in the relationshipsbetween individualMLMmembers and the members who recruited hem into theirrespectiveorganizations i.e., their 'spon-sors'). Althoughmembers requentlycite a productivesponsor-recruit elationshipas pivotalto oper-ating a successful MLM distributorship,he nature of a sponsor'scontribution o that success isunclear.On the one hand,sponsorsreceive significant ncentivesto serve as leadersto theirrecruits,guiding and supportingrecruits'efforts in ways similarto a sales manager.On the otherhand,oncerecruited,new members ndependentlyown theirMLMdistributorships;herefore,sponsorslack theformalsupervisoryauthority hattypically accompaniesleadershippositions in traditionalorganiza-tions. Collectively,these circumstances eave the characteristicsof effective leadership n MLMs anopen question.Therefore,we begin ourstudyof MLMsby examiningthe natureand effects of trans-formational eadership n this largelyunexploredorganizational ontext.As the nameimplies, transformationaleadersostensibly 'transform' heirfollowersthroughspeci-fic behaviorsthatinspireandmotivatefollowers to 'transcend heirown self-interestsfor the good ofthe group,organization,or society' (Bass, 1990, p. 53), resulting n followers doing 'morethantheyintendedandoften morethantheyeven thoughtpossible' (Bass 1998,p. 4; see also Bass, 1985, 1997;Bass et al., 1987;Burns,1978;Podsakoffetal., 1996;Yukl,1998).Althoughtheliterature onsistentlysupportstransformationaleadership'spositive impact on attitudeand performance, ess is knownabouthow this leadershipstyle actuallyachievesthe transformation f followers.Leadership cholarspropose that followers' transformationsoccur when they begin associating work with 'higherpurposes' that extend beyond simply earning money (Bass, 1985; Burns, 1978; Podsakoffet al.,1996; Shamiret al., 1993; Yukl, 1998).Because mostempiricalresearchhasfocused on whetherand underwhat circumstances ransforma-tional leadershipproducesmore satisfied,more committed,and higherperforming ndividuals,fewstudies in any contexthave investigatedthe mechanismsby which these outcomesoccur.Moreover,we are awareof no studies investigatingthese issues in MLMs. Consequently,the purposeof thisresearch s to examine:(1) whether ransformationaleadershipbehaviorsby anMLMsponsor nvokestheirrecruits'higher-ordermotives,and(2) the degreeto which those motivesserve as a mechanismthroughwhich transformationaleadershipbringsaboutpositive effects on the recruits'effort,satis-faction,andperformance.Specifically,we developandtest a model in whichonegeneralhigher-ordermotive, 'belief in a higherworkpurpose,'mediatesthe causalsequenceleadingfrom transformationalleadership o satisfactionandperformance.Because mostresearchersikely havelittleexperiencewithMLMs, we firstbrieflydescribetheir structureandoperation,andour reasonsfor selecting one as aresearchsetting.

    The Nature and Structure of MLMSSimply put, MLMs consist of corporate-levelproducersthat market brandedgoods or services(e.g., Mary Kay Cosmetics,Amway Products) hroughnetworksof independent,member-owneddis-tributorships,whichareusually operated rom members'homes.StartinganMLMdistributorshipim-ply requires purchasing trainingand promotionalmaterials from the corporate-levelproducerandagreeingto purchasea certainamountof inventoryduringa specifiedperiod.Members henreselltheirinventoriesprimarily o retailcustomers,often beginningwith family andfriends.Copyright 2001 JohnWiley& Sons,Ltd. J. Organiz. ehav. 2, 849-869(2001)

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    4/22

    MULTILEVELMARKETINGORGANIZATIONS 851MLMdistributorshipsrovidememberswith two sourcesof income. The firstcomes fromproductsales. Relying on the same basic skills as outside industrialsalespeople, successful MLM membersidentify and qualify sales prospects,contactthem, arrange o meet them andpresentthe benefitsofthegood or service,thenclose the sale. To growtheirdistributionnetworks,corporate-levelproducersoffermembers a second sourceof income: commissionson sales by new membersthat currentmem-bers recruit.To enhance this recruiting ncentive,most MLMspay commissionson multiplelevels ofrecruits-hence, the termmulti-levelmarketing.Thatis, membersreceive commissionsnot only onsales generatedby newmembers heyrecruit,butalso on salesby members heirrecruitsbring ntotheorganization,all recruits hose membersbringin and so on. Thus,MLMmemberscould conceivablyreceive commissions from sales by hundredsof other members.AlthoughMLM members own independenthome-basedbusinesses, they do not operatethem inisolation fromone another. nfact,MLMs establishelaboratesystemsof support hroughwhichmem-bersreceiveguidanceand eadership romeachother.Twocomponentsof theseorganizational upportsystems play crucialroles in the working relationshipsamong members.First and foremost are therelationshipsbetween 'sponsors' and 'recruits.'A sponsoris an existing memberwho persuadesanew member,or recruit,to starta distributorship.Technically,all MLM membersarerecruits n thesense that some currentmemberpersuadesevery new memberto join. Thus, any member who has

    recruitedat least one other member s at once a sponsorand a recruit.Ouruse of the terms 'sponsor'and 'recruit,'however,shouldbe thoughtof as dyadic.That s, we refer o therelationshipbetweentwomembers-one who recruitedanother.Based on extensiveinterviewswithMLMmembers,we regardsponsor-recruitdyadsto be the key leader-followerrelationshipsn MLMs.Indeed, he commissionssponsorsreceive on theirrecruits'sales serve as a powerfulincentiveto become activesourcesof lea-dershipto theirrecruits.Giventhe centralrole sponsor-recruit elationshipsplay in the operationsofMLMs,they arethe focal point for our studyof transformationaleadership.A second factor influencingthe natureof members'relationshipsare informalworkinggroupsofMLMmemberscreatedby sponsors or theirrecruits-a kindof sales teamreferred o hereas a 'unit.'Throughunits, members meet regularly o exchangeinformationand ideas (e.g., selling tips), createsharedopportunities e.g., sales workshops),and providevery practicalforms of assistanceto eachother(e.g., loaninginventoryandmaterials).Althoughmembersoperate ndependently,unitsprovidea tangibleorganization hroughwhich memberscreate andmaintaina sense of belonging,develop afeeling of espritde corps,andacquireresourceshelpfulto their ndividualdistributorships.Moreover,units serveas a vehicle for socializingmembers nto thenorms andvaluesof the MLM anda conduitthrough which members can mutually reinforce the effects of sponsors' leadership behaviors(Waldmanand Yammarino,1999; YammarinoandDubinsky,1992).Studyingtransformationaleadership n this organizational ontextis importantor severalreasons.First,transformationaleadershipcan be examined in relativeisolationfrom otherleadershipstylescommonly employed in traditionalorganizations, argely because sponsorsmay have few effectivealternatives o transformationaleadership.Members'independentownershipof theirdistributorshipsleaves MLMs without the hierarchical leader-subordinaterelationships found in most sellingorganizations.The resulting leader-follower dyads provide sponsors strong incentives to cultivatehigh-performingmembers,but few extrinsic tools to deepen the satisfactionand loyalty of theirrecruits(see Dansereauet al., 1975; Graenand Cashman,1975). In other words, sponsorsare notmanagersor supervisors,and although corporate-levelproducersprovide incentives encouragingsponsors o serve as leaders to theirrecruits,sponsorsholdnoneof the formal everageordinarilyheldby sales managers.Sponsorscannotpromote,discipline,or dismissrecruits; heycannot alterrecruits'compensation;heycannotdirectrecruits'dailyworkactivities.Theabsenceof thenormalsupervisorytools that typically accompanyleadershippositions thereforelimits sponsors'choice of leadershipstyles.Copyright 2001 JohnWiley&Sons,Ltd. J. Organiz. ehav. 2, 849-869(2001)

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    5/22

    852 J. R. SPARKSAND J. A. SCHENKSecond, althoughsome members oin MLMsprimarily or additional ncome, manyconnecttheirMLMparticipationo a varietyof 'higher-order alues' thatextendbeyond profit-making r financialsuccess.Membersexpressed hesesentimentsabout heirworknotonly duringour nterviews,butalsoin articles from the popularbusiness press (Bartlett, 1995; Dunkel, 1995) and academic research(Wotrubaand Tyagi, 1992). For example, some members identify so stronglywith the benefits ofproducts heyrepresent hatselling themassumesa specialimportance: .g., educational oys advancethe cause of children;weight managementprogrammesenhance self-confidence,and so on. Othermembers view MLMsas a way to strengthen amily ties by workingfrom home. Still others believethatMLMs are a meansfor womento control their own professionaldestiniesthroughentrepreneur-ship,opportunitiesmanyview as less available hrough raditional orporateavenues.As notedearlier,inasmuchas transformationaleadershipbehaviors 'transform' ollowersby appealing o these higherordervalues,MLMs offer a richandfavorableenvironment or exploringwhethersuch appealsarisefrom or areenhancedby a sponsor'stransformationaleadership.FinallyMLMsrepresenta growingyet underresearchedrganizational ontextin whichtransforma-tionalleadershipmay play a particularly mportant ole in explainingandpredictingperformance.Toinvestigatethese issues, this studyis guided by the following questions:(1) Arememberswho expressstrongerbeliefs in the higherpurposeof theirMLMparticipationmore likely to have sponsorswho

    exhibita transformationaleadership tyle? (2) Are these membersmoresatisfiedwith theirwork,anddo they work from withinmorecohesive and supportiveunits?(3) Do these memberssubsequentlyexertgreatereffort? And (4) do they in turnachieve superior evels of performance? n the followingsections, we examine these questionsby developing and empiricallytesting a model thatdescribestransformationaleadership'seffects onjob satisfactionandperformancehrougha series of mediatingrelationships,and introduces a key mediatingvariable-belief in a higherwork purpose-throughwhich transformationaleadershipachieves its positiveoutcomes.

    Model DevelopmentRationalefor modelFigure 1 shows the hypothesizedmodel, which posits that transformationaleadership'seffects onsatisfactionand performance s mediatedby a series of interveningvariablesincluding belief in ahigher work purpose. Two reasons promptour focus on mediationeffects. First, examinationofvariablesthat may mediate leadership-outcomerelationshipsrespondsto a need for such researchidentifiedin the literature see Conger, 1999; Shamiret al., 1993; Yukl, 1999). Indeed,Yukl (1999,p. 287) refers specifically to ambiguityabout the underlyingprocesses throughwhich transforma-tional leadershipworks andproposesmore researchon 'mediatingvariablesrelevantto taskperfor-mance' including, among other things, 'arousalof motives.' In their importanttheoreticalwork,Shamiret al. (1993) likewiseproposethat transformationaleaderships'effects occurthrougha seriesof mediating'motivationalmechanisms'thatrelateprimarily o follower self-concept.Althoughthefocus on mediation adds complexity to the model, these calls for examination of mediatingrelationshipssupportthe notion that transformationaleadership'seffects may in reality be quitecomplex.Second, our model permitstesting a hypothesiscentralto transformationaleadershiptheory.Asnoted earlier,by definition,transformationaleadershipbehaviorsarebelieved to motivatefollowersto forgotheir own needs for the sake of theorganization.Leadership cholarshavevariouslydescribedtheprocessthroughwhichtransformationaleadershipworksas 'raising'followers 'tohigher evels ofCopyright? 2001 JohnWiley& Sons,Ltd. J. Organiz. ehav. 2, 849-869(2001)

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    6/22

    MULTILEVELARKETINGRGANIZATIONS53

    A V Ir t i c u l a t i n g J S 2 J S 2

    V i s i o n J S IA V 3

    J o b S a t

    s -AM1 .113AM Appropriate 22~ M ~ \ M o d e l 92~.92s AM Pfl I8T 7 T r a n s f l i g h e r P f l o m a c

    L d e E f f o r t P e r f o r m a n c e

    H E I h PI 712paE x p e c t a t i o n s H P 2 , E f 2

    H E T18 . 0 2U n i t

    o h e s i o n

    I S 1I n d i v i d u a l

    I S 2 u p p o r t U U U C 3~ I S 3 T199UC F, F,

    Figure1. Hypothesized odelandstructural odelparameterstimates*'t*n= 736,X2(239)= 672.82(p< 0.001),GFI= 0.93,AGFI 0.91.Additionalitstatistics iven n Table3.tAllpaths ignificantop < 0.01 except 32,4 n.s. Standardizationathcoefficientshown.IFormeasurementodelcoefficients,ee Table1.

    morality' (Bums, 1978), 'activating heirhigher-orderneeds' (Yukl,1998;also Bass 1985; Podsakoffet al., 1996), and alteringtheir 'needs, values, preferences,and aspirations'by appealingto 'ideolo-gical values' (Shamiret al., 1993). In effect, these conceptualizationsof transformationaleadershipdescribeits effects as being mediatedby the activation of these higherorderneeds. The model pro-posed hereaddresses his mostbasic issue using, as discussedearlier,a samplewe believe to be parti-cularly well-suited to the task.Figure 1 also points to another mportant eatureof the model, which is transformationaleader-ship's operationalization s a second-orderactor.This treatment xplicitlyrecognizesthattransforma-tionalleadershipencompassesseveral distinct sets of behaviors,while acknowledging hat the totalityof behaviorsassociatedwith transformationaleadershipreflect some largerlatent 'leadershipcon-struct.' The high correlationsbetween transformationaleadership'svariousdimensionsreported nnumerousstudies(see AtwaterandYammarino,1993;Bycio et al., 1995) indirectlysupport his view,while direct supportcomes from Carless(1998) who tested threeconceptualizationsof transforma-tional leadershipand found that a second-order actor model fit the data betterthan either a singlefactor model or a first-ordermulti-factormodel.ResearchhypothesesEffects of transformational leadershipSubordinateob satisfaction ranksamongthe most commonly studiedoutcomes of transformationalleadership,with significantrelationshipsbeing found among office employees (Barlinget al., 1996;Copyright 2001 JohnWiley& Sons,Ltd. J. Organiz. ehav. 2, 849-869(2001)

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    7/22

    854 J. R.SPARKS NDJ.A. SCHENKNiehoff et al., 1990), school teachers(Kohet al., 1995), hospitalnurses(Bycio et al., 1995;MedleyandLarochelle,1995),militarypersonnel Yammarino ndBass, 1990;Yukl andVanFleet, 1982),andsalespeople(Dubinskyet al., 1995).Given the strengthof thisevidence,we hypothesizea similarrela-tionshipwill hold in MLMs:

    Hypothesis 1: Sponsors' transformationaleadershipwill be positively related to recruits'jobsatisfaction.Accordingto theory,transformationaleadership ncreasesfollowers' willingnessto set aside indi-vidual needs in favor of groupor organizationalneeds (Bass, 1985, 1997; Bass et al., 1987; Burns,1978; Yukl, 1998). Shamiret al. (1993) explain this effect in terms of social identification; hat is,transformationaleadershipworks in largemeasurethrougha collective processby which followers'self-concepts become stronglyintertwinedwith their membership n a particulargroup. One likelyresult of this processis thatmemberswithina given workgroupwill experiencegreatercohesiveness(Matey,1991;WaldmanandYammarino,1999). In MLMs,the analogousworkgroupis the informalunit describedearlier.Hypothesis2: Sponsors'transformationaleadershipwill be positivelyrelatedto recruits'percep-tions of theirunit'scohesion.A centralfeatureto our model, the mediatingvariablebelief in the higherpurposeof one's work,remainslargely unexplored n the research literaturealthoughsimilar concepts have recently beenaddressedby writers n thepopularbusinesspress (e.g., Dumaine,1994).As notedearlier,MLMspro-vide numerousavenuesthroughwhichmembersmay findhigherpurposesbeyond earningmoney.Forexample,strengtheningamily ties by workingathomeor sellingaproductwhose benefitsholdspecialsocial importancemay be specific higher-orderpurposesthatmotivateMLM members.However,atthis point, our goal is not to single out any one higherorderpurposemotivatingmembers'actions.Insteadwe address the more basic question of whetherhigher orderpurposes in general motivatemembersand whethertransformationaleadershipper se encouragesMLM membersto findhigherpurposes n theirwork. To the extent that transformationaleadership s effective in 'giving meaning-

    fulness to workandinfusingworkandorganizationswith moralpurpose'(Shamiret al., 1993,p. 578),MLM recruitsshould findhighermoralpurpose n theirworkwhen theirsponsorsexhibitbehaviorsassociatedwith a transformationaltyle.Hypothesis3: Sponsors'transformationaleaderships positivelyassociatedwith recruits'belief inthe higherpurposeof theirwork.

    Effects of belief in a higher purposeWe also expect recruits'beliefs in the higherpurposeof theirwork to positively influencetheirjobsatisfactionandtheirperceptionsof unit cohesiveness.Withrespecttojob satisfaction, t seems intui-tive thatstrongaffirmations f theintrinsicvalue of one's work,alonganyof the dimensionsdiscussedearlier,shouldproduce remendouspositiveaffect whenthe tasksassociatedwith thatworkareperfor-med.Likewise,whenmembersof a workgroupsharesimilarlystrong dentificationwitha higherworkpurpose,cooperationand cohesiveness among groupmemberscontribute o the realizationof thosepurposes.Thus,we expect thatMLMmembersrecruitedby transformationalponsorswouldbelievemore stronglyin the higherpurposeof theirwork,andto subsequentlydisplayhigherlevels of iden-tification andcooperationwith othermembersof their units.

    Hypothesis4: Recruits'belief in thehigherpurposeof theirwork is positivelyassociatedwiththeirjob satisfaction.Copyright 2001 JohnWiley&Sons,Ltd. J. Organiz. ehav. 2, 849-869(2001)

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    8/22

    MULTILEVELMARKETINGORGANIZATIONS 855Hypothesis5: Recruits'belief in thehigherpurposeof theirwork is positivelyassociatedwith theirperceptionsof their units' cohesiveness.The modelpostulatesboth indirectand directeffects betweenhigherpurposeand individualeffort.Theoretically,we could not reasonably expect that group cohesion would completely explain therelationshipbetweenhigherpurposeand effort. Given that ikelihood,if higherpurposedoes in reality

    exertdirecteffects on effort,or if other variablesmediate thatrelationship,orboth,then we would stillexpect a directsignificantandpositiverelationship o exist betweenhigherpurposeand effort.Thus,we hypothesize:Hypothesis6: Recruits'belief in thehigherpurposeof theirwork is positivelyassociatedwith theirworkeffort.Withrespectto higherpurpose's ndirecteffects on effort,one routemay be throughunitcohesion.That is, as higherpurposeincreases unit cohesion, the unit's greatercohesiveness produceshigherlevels of individualeffort on thepartof its members.In supportof thisview,WaldmanandYammarino(1999) proposethatgreaterworkgroupcohesiveness resultsin greater ndividualandcollective effort(see also Lowe et al., 1996). Consequently,we makethe following prediction:Hypothesis7: Recruits'perceptionof the cohesion of theirunit is positivelyassociatedwith theirwork effort.

    Effects of effortDespite being focal variables n a large body of empiricalresearch,the natureof job satisfaction'srelationshipwith individualeffort remainsunsettled.Therefore,we acknowledgethat,shoulda statis-tical relationshipbetween effort and satisfaction exist in our data, the causal directionmay be thereverse of thatwhich we hypothesizeor the relationshipmay reflecta simple correlation.Thatsaid,we drawupon theoretical work on the intrinsic motivationof workersby Deci (1975) and Deci andRyan (1985) and empiricalwork on the effort-satisfaction relationshipfrom Brown and Peterson(1994) to posit thatgreatereffort leads to enhanced ob satisfaction.Finally,we predictthatgreatereffort will also producesuperiorperformance.

    Hypothesis8: Recruits'work effort is positively associatedwith theirjob satisfaction.Hypothesis9: Recruits'work effort is positively associatedwith theirjob performance.

    Contextual Sidebar

    OrganizationalContextMultilevel marketing nationallyMultilevelmarketingorganizations MLMs)offer an inexpensiveand attractivemeans for produ-cers or wholesalers to quicklycultivateretail channels of distributionor theirproducts,a fact thatexplains the remarkablegrowthin the numberof MLMs operatingworldwide.Similarly,MLMsoffer membersan attractiveand inexpensivemeans to starttheirown businesses. Membersearnincome by selling the producers'productsandby recruitingnew membersto the MLM.

    Copyright 2001 JohnWiley&Sons,Ltd. J. Organiz. ehav. 2, 849-869(2001)

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    9/22

    856 J. R. SPARKSAND J. A. SCHENK

    Despitetheirrapidgrowth,MLMsoftenattract ontroversy n partbecausetheyareassociatedwithchain letters,pyramidschemes, and other fraudulentbusiness practices.WhatdistinguishestrueMLMsfromclassic pyramidschemes is whethermembers'earningscome primarily romproductsales to ultimateconsumersinstead of from recruitingnew members.Importantly,n legitimateMLMs,earningsgeneratedby recruitingmust be limitedto commissions from recruits' sales andnot fromrecruiting tself.MLMs featureanunusualandfluidorganizational tructurewithlittleformalhierarchydue to mem-bers' independentownershipof theirMLM distributorships.Oldermemberscan advise andassistthe new membersthey recruit,but cannot directand supervisethe new members'workactivities.Organizational factorsThe firm n thisstudy s relativelynew-less than25 yearsold,and s experiencingveryrapidgrowthin both sales and membership.The firmserves membersin all 50 states and is developingdistri-butorshipsnternationally.Currently,heMLM faces no directcompetitive hreat romotherMLMs,making his firm herecognizedmarket eader or its product ine and distribution hannel.However,the firm s facing new competitionfromtraditional etailerswho appear o have a priceadvantage.Because members independentlyown their distributorships,he firm lacks a single identifiable'managementclimate.' However, member socialization focuses on developing and maintainingenthusiasm orsellingtheproduct ine throughorganizationalolkloreaboutthefoundingmembersand otherparticularly uccessful members. Stories of the company'sfounding and early growthprovidemembersa sense of sharedhistoryas well as an emotionaltool forbuildingmember oyaltyand for recruitingnew members.Worker and job factorsBecause each member s essentiallyan independentcontractor,heirjob responsibilities endto bebroad,thoughthey controlthe amountof time they devote to theirdistributorships.Membersholdresponsibility or all aspectsof productsales fromlocatingprospects o closing eachsale. Addition-ally, membersreceive powerful financial incentives to recruitnew membersand help these newmembers develop successful distributorships f their own. Thus, in additionto managingtheirown distributorships,membersassume the tasks of socializing, training,and motivatingthe newmembersthey recruit.

    MethodMemberinterviewsMLMs' unusualorganizational haracteristicsed us to begin with a qualitativeexplorationof issuesrelatedto leadershipandperformanceby conductinga series of four groupinterviewsat a medium-sized (in terms of totalmembers)MLM's nationalconvention.The MLMcorporateproducer electedfocus groupparticipants rom its list of conventionregistrantswith the goal of avoiding excessivehomogeneitywithrespectto demographicanddistributorshipharacteristics.Selected memberswerecontactedby letterpriorto the convention;most of those contactedagreedto participate.A total ofthirty-onemembers-all women with distributorshipshatvariedin lengthof operationandaveragemonthlysales-attended each one- to two-hour nterview.Copyright? 2001 JohnWiley & Sons, Ltd. J. Organiz.Behav.22, 849-869 (2001)

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    10/22

    MULTILEVELMARKETINGORGANIZATIONS 857Theseinterviews nformedandguidedthisresearch n severalrespects.First,memberson bothsidesof the sponsor-recruitdyadrepeatedlystressedthe centralrole these relationshipsplay in developingsuccessful distributorships.Membersdiscussed at length importantdimensionsof theirsponsorrela-tionships, including mutualexpectationsof sponsorsand recruits,the natureof typical interactionbetween sponsorsandrecruits,specificways sponsorsexhibitleadership,andthe similaritiesanddif-ferences between these relationshipsand traditionalmanager-employeerelationships.Second, weasked members to commentgenerallyon their motives for joining the MLM and, when necessary,promptedmembers o commentspecificallyaboutthedegreeto which factorsother thanmoneyinflu-enced their decisions. Finally,we used these discussions as an opportunity o familiarizeourselveswith typicalMLM terminology,which facilitated the wordingof questionnairetems.

    SurveysampleWe subsequentlyenlistedthe cooperationof the corporate-levelproducer o administera briefsurveyof its nationalmembership.The corporate-levelproducerrandomlyselected a sampleof respondentsfrom a computerizeddatabaseof currentmembers withoutregardto attendanceat the nationalcon-ference.However,we imposedtwo limitingcriteria orinclusionin the sample.First,membersshouldreside in thevicinityof theirsponsor-a distinctionmade onthebasis of zip code. Occasionally,mem-bers move greatdistancesfrom theirsponsors,which generallyaltersthe dynamicsof the sponsor-recruitrelationship.These 'remote' membersoftenrelyless ontheirsponsors(whocontinueto receivecommissions fromthe members'sales) because face-to-face interactionwith them is no longerpossi-ble. Of the sponsorswe interviewedwho hadremoterecruits,most indicatedthat,although hey con-tinueto communicatewith them,theirrelationshipssignificantlychangeafterthe recruitsmove. Thecorporate-levelproducer ould notprovideestimateson the proportionof remotemembersrelativetototalmembers,however,focus groupparticipantsndicatedthatin theirexperiences,remotemembersareby far the exception.Thus,the numberof potential responsesaffectedby this criterion s likely tobe quite small.Second, because ourprimary ocus is on the effects of transformationaleadership,we wished toreceiveresponsesfrommemberswho are morefrequently argetsof leadershipbehaviors hansourcesof them.Leadershipresearchersypicallymake suchdistinctionseasily on the basisof organizationalrankor title. As notedearlier,however,MLMsdo not consist of traditionalhierarchical upervisor-subordinate elationships.As a practicalmatter,memberswith substantialnumbersof recruits ikelyturn ess frequently o theirown sponsors orleadership;nstead,theirprimary ole becomesprovidingleadershipto theirrecruits.Therefore,we limited our sample to memberswho had fewer than fiverecruitsand whose sponsorshad more than five recruits.Five was selected as the cut-off because,according o the focus groupparticipants,hetransitionrom 'primarilyollower'to 'primarilyeader'generallybecomes mostpronouncedas a sponsorapproaches iverecruits. ndeed,this MLM'scorpo-rateproducerrecognizesmembers with five recruitsby awarding hemwith the title, 'unitmanager.'Afterimposingthese criteriaandaccounting ornon-deliverables,hefinalsampleconsistedof 1883members,all of whom arewomen. (Over99 percent of all membersof this MLM arewomen.)Eachreceiveda questionnaire,a cover letterexplainingthe purposeof the research,anda letterof endorse-ment from the corporate-levelproducer.A totalof 886 respondents eturned heirquestionnaires;150were excluded fornotreportingperformancedata.Theremaining736 usableresponses yielded a finalresponserateof 39 per cent.To check for non-responsebias,we compared he demographic ompositionof oursampleto a cen-sus of themembership akentwo yearspriorby thecorporateproducer.No statisticallysignificantdif-ferenceswere foundin memberage, education,maritalstatus,or outsideemploymentstatus.BecauseCopyright 2001JohnWiley& Sons,Ltd. J. Organiz. ehav. 2, 849-869(2001)

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    11/22

    858 J. R. SPARKSAND J. A. SCHENK

    impositionof theselectioncriteriaproduceda stratified ample, ts comparison o a two-yearold mem-bershipcensusdoes notconstitutea direct test of non-responsebias.However,the fact thatno signifi-cant differencesemergedamongthe tested variablesprovidessome evidence thatnon-responsebiaswas notproblematic.Mostrespondents tarted heirdistributorships etweenone and threeyearsago,and operatetheir distributorships art-time, workingbetween 10 and 30 hoursper week. Notably,almost half do not have otheremployment;only 22 per cent have full-timejobs apart romtheirdis-tributorships.Most members arebetween 25 and44 yearsold; 94 percent aremarried;over half arecollege graduates.

    MeasurementAlthoughmanyof theconceptsusedin thisstudyoriginate n well-establishedresearch, hescale itemsdevelopedto measure hemtypicallyreferto managerbehaviorsorwork situations hatsimplydo notapplyin MLMs.Therefore,using publishedmeasuresas a guide, we adaptedor createditems to suitthe presentresearchcontext. Unless otherwisenoted,each item was measuredon a seven-pointscalerangingfrom 'stronglydisagree' (1) to 'stronglyagree' (7).Transformational leadership behaviorsBased on the member nterviews,we focus on four transformationaleadershipbehaviors(Podsakoffet al., 1990):articulatinga vision, providingan appropriatemodel,providing ndividualizedsupport,and settinghigh performance xpectations.Afterreadinganddiscussingitems fromthe Transforma-tionalLeadership nventory TLI;Podsakoffet al., 1990), focus groupparticipantsndicatedthatfos-teringthe acceptanceof group goals andintellectualstimulationseemedlimitedin theirapplicabilityto the particularsof sponsor-recruit elationships.Regarding osteringthe acceptanceof groupgoals, focus groupparticipants xpressedsome confu-sion over the groupto which the goals would apply.Although sponsorsfrequentlyassist members nsettinggoals for theirindividualdistributorships ndencouragecooperationamongmembers n theirunits, they rarely, f ever,establishunit-widegoals or promoteallegianceto the unit. Withrespecttointellectualstimulation, ocus groupparticipants eported hatonly one of the threeitems in the TLI('has stimulatedme to thinkabout old problemsin new ways') seemed to fit the normsof sponsor-memberrelationships.As a rule,membersseek problem-solvingassistancefromsponsorsaboutverypracticalmatters that fit more closely with whatLord(1977) refersto as 'functional'dimensionsofleadership.In light of these commentsand the corporateproducer'srequestthatthe questionnairebebrief, we omitted items from these two dimensionsof transformationaleadership.Belief in a higher work purposeOursearch of the literaturedid not uncoverany previousempiricalresearchon this construct.Thus,based on our memberinterviews,we developedthreeoriginalitems to measurethe degreeto whichMLM membersbelieve theirworkis partof a 'cause' thathas purposes'moreimportant' hansimplymaking money. Consistentwith our broadconceptualizationof this construct,we deliberatelywrotethe items to reference no specific higher purposes, leaving that to the interpretation f individualrespondents.Job satisfactionJobsatisfaction s an emotional reaction that stems from the belief that one's job fulfils certainvaluesandneeds (Locke, 1976).Althoughindividualsmay be satisfiedor dissatisfiedwith an almost infiniteCopyright 2001 JohnWiley& Sons,Ltd. J. Organiz. ehav. 2, 849-869(2001)

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    12/22

    MULTILEVELMARKETINGORGANIZATIONS 859

    varietyof individual acets to aparticularob (pay,coworkers,etc.), theyalsopossess a globalaffectiveresponseto their ob (c.f., Lucaset al., 1990;McFarlinandRice, 1992; Sujan,1986).To measure hisoverall response,we adapted our items from the general job satisfactionsubscaleof HackmanandOldham's(1975) JobDiagnosticSurvey.Unit cohesionTo measure he degreeto whichMLMmembersbelieve theirworkgroupsactedas cohesive units,weadapted tems fromPodsakoffandMacKenzie(1994). The four items used in thisstudyoperationalizecohesiveness in terms of trust,dependability, ooperation,andrespect.Effort and performanceFrom ourdiscussions with MLMmembers,we concluded thatmembereffortis best reflectedby twobasicmeasures.First s thenumberof hoursperweek membersdevoteto theirdistributorships,whichvaryfromonly a few hoursto morethan40. We measured his usinga fivecategoryscale (10 hoursorless, 11 to 20 hours,21 to 30 hours,31 to 40 hours,more than40 hours).Secondis thenumberof sales'workshops'held with prospectiveand currentcustomers each month.This was measuredby askingmembers 'On average,how manyworkshopsdo you hold each month?'Weused bothas indicantsofeffort.Members ypicallyview performancebothintermsof product ales andnewmembersrecruited.Forsales, we askedrespondents or the 'dollarvalue of last month'ssales;' forrecruiting,we askedhowmanynew membersthey hadrecruited n total. Because the numberof new membersone recruits scumulative,members'totalnumberof recruitswas dividedby the numberof yearsthey hadoperatedtheirdistributorshipso producethe averagenumberof new membersrecruitedper year.Both salesandrecruitsper year were used to measureperformance.

    Assessment of construct validityTo assess the performanceof individualscale items and evaluatethe degree to which the datafit thehypothesizedfactorstructure,we estimateda measurementmodel using the 26 items as indicantsofnine totalconstructs.Following recommendationsby Breckler(1990), the data were split into deriva-tion andcross-validation amples.The derivation ampleis usedto fita favorablemeasurementmodelto the data,which is thenvalidatedusing the cross-validation ample.Using the derivation ampletoestimatea model containingall 26 items, we found two items-one each from the higherworkpur-pose and unitcohesivenessscales-did not load as anticipated.Afterdeletingthese items, the modelwas re-estimatedusing the cross-validation ample,which yielded X2(216)= 340.80 (p < 0.001). ThesignificantX2statistic ndicatesthatthe outputcorrelationsproducedby the modeldo not fit the inputsample correlations within sampling error.However, as recommendedby Andersonand Gerbing(1988) and Bagozzi and Yi (1988), we examined alternative it indices, which suggest satisfactorymodel fit:GFI= 0.93, AGFI= 0.90, NFI = 0.95, RMSEA= 0.04 (p= 0.98).Finally,usingthe entiresample,model estimationyieldedX2(216)= 463.53 (p

  • 8/13/2019 The Effects of LTf

    13/22

    860 J.R.SPARKS NDJ.A. SCHENKTable 1. Assessment of measurement

    StandardizedConstructand items* M SD LISRELestimatet P, Pc(7)111. Articulatinga vision 0.82 0.85 0.66AV1:My sponsoris able to get me committedto 4.71 1.61 0.81

    a positive vision of the futurewith this MLMAV2:My sponsorpaintsa positive pictureof 4.75 1.67 0.80my possible futurewith this MLMAV3:My sponsorhelps me see a positive future 4.98 1.65 0.83for myself as a member of this MLM2. Providingan appropriatemodel 0.90 0.91 0.78AMI: My sponsorleads 'doing' rather han 'telling' 4.49 2.43 0.77AM2: My sponsorleads by example 4.80 1.67 0.95AM3: I look to my sponsoras an exampleof how 4.65 1.77 0.91to be an effective MLMmember3. Settinghigh performanceexpectations 0.74 0.78 0.65HEI: My sponsoractively encouragesall members 4.39 1.75 0.93in herunit not to settle for less thanoutbest performanceHE2:My sponsorshows the members n her 3.46 1.62 0.65unit thatshe expects a lot from us as membersof this MLM4. Providing ndividualizedsupport 0.96 0.96 0.90ISI: My sponsorshows respectfor my 5.41 1.60 0.95personalfeelingsIS2: My sponsorbehavesin a manner hat is 5.29 1.68 0.97thoughtful o my personalneedsIS3: My sponsorcares aboutmy feelings 5.32 1.62 0.935. Belief in higherpurposeof one's work 0.80 0.81 0.68HPI: In my workwith this MLM,I feel I'm part 5.35 1.23 0.73of a 'cause' that'sabout more thanearning moneyHP2:While earning money is nice, there 5.45 1.24 0.91are othermore important easonsfor my workwith this MLM5. Job satisfaction 0.78 0.74 0.42JS1: In general,I get a greatdeal of personal 5.29 1.27 0.79satisfactionfrombeing a memberof this MLMJS2:My workas a memberof this MLM is 5.62 1.65 0.59personally very fulfillingJS3:I considermy work as a MLM member 4.97 1.22 0.62to be very rewardingJS4:I am generallysatisfied with the work I 5.11 1.36 0.57do as a memberof this MLM6. Unit cohesion 0.92 0.92 0.78UCI: There is a greatdeal of trustamong 4.80 1.63 0.87fellow members n my unitUC2: The membersof my unitarevery 5.14 1.39 0.90cooperativewith each otherUC3: The members of my unitbelieve we can 5.09 1.48 0.89dependon each other for advice and supportconcerningourdistributorships7. Effort 0.68 0.69 0.50Efl: Hoursper week devoted to 1.26 0.78 0.67distributorshipworkOEf2: Number of sales workshopsheld per month 2.36 1.73 0.75

    (Continues)Copyright( 2001 JohnWiley & Sons, Ltd. J. Organiz.Behav.22, 849-869 (2001)

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    14/22

    MULTILEVELMARKETINGORGANIZATIONS 861Table 1. (Continued)

    StandardizedConstructand items* M SD LISRELestimatet Pc 7 Pvc(7)i8. Performance 0.77 0.71 0.56Pfl: Productsales $668 576 0.85

    Pf2: Numberof recruitsper 0.51 0.75 0.63year of membership*Wordingof some items are alteredslightly from the questionnaire o conceal the MLM's identity.Except for Effort andPerformance, esponsesare measuredon 7-point scales rangingfrom 'stronglydisagree' (1) to 'stronglyagree' (7).tAll paths significant at p

  • 8/13/2019 The Effects of LTf

    15/22

    o0?)Ct0

    CD20

    t-

    CO

    00O-

    t.)

    4Ioo

    too

    Table 2. Interconstruct orrelations*'t1 2 3 4 5 6

    1. Performance 1.002. Effort 0.79 1.000.043. Job satisfaction 0.36 0.43 1.000.03 0.034. Unit cohesion 0.00 0.05 0.42 1.000.03 0.03 0.035. Higher purpose 0.11 0.20 0.67 0.34 1.000.03 0.02 0.04 0.036. Articulatinga vision 0.11 0.10 0.45 0.62 0.33 1.000.03 0.03 0.03 0.04 0.037. Providing appropriatemodel 0.03 0.01 0.35 0.57 0.28 0.920.03 0.02 0.03 0.03 0.03 0.048. Setting high expectations 0.03 0.03 0.38 0.45 0.25 0.860.02 0.03 0.03 0.04 0.03 0.049. Providing ndividualsupport 0.01 0.00 0.30 0.68 0.26 0.810.03 0.03 0.03 0.04 0.03 0.04*Standardrrors ivendirectly elowcorrelations.tCorrelationshownareinterconstructorrelationsstimated romconfirmatoryactoranalysis.n= 736. X2(216)=463.53p < 0.001),RMSEA 0.039(p= 1.0),absoluteGFI= 0.85(seeMcDonaldndMarsh, 990).

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    16/22

    MULTILEVELMARKETINGORGANIZATIONS 863commonmethodfactor added to the items of the multi-itemlatent constructs(job satisfaction,unitcohesion, belief in a higherwork purpose,and the transformationaleadershipbehaviors).Becausemost constructsare measuredusing two- or three-itemscales, the additionof a methodfactorto all24 itemsproducedanunderidentifiedmodel. To overcomethisproblem,we excludedthe measures oreffort and performance rom the method factor on the groundsthat they measurequantifiableandobjective concepts, and are thereforeless susceptibleto method variance.No paths were renderednon-significantby the inclusion of the method factor.Based on this evidence, we concluded thatthe datalikely do not sufferfromcommon method variance.

    ResultsInthis section,we describe theresultsof fittingthe datato the modelshownin Figure1. In addition oreporting heresults of individualhypothesestests,we also compare hehypothesizedmodelto severalalternativemodels in order o provideevidencethat no reasonablerivalmodelsprovidesuperior it tothe data.Estimationof the hypothesizedmodel produceda significantX2 statistic:X2(239)= 672.82(p

  • 8/13/2019 The Effects of LTf

    17/22

    864 J. R. SPARKSAND J. A. SCHENK

    leadershipbehaviorsrelatepositivelyto members' ob satisfactionandunitcohesion.Inhypothesis3,we predicta positiverelationshipbetweentransformationaleadershipandbelief in a higherworkpur-pose. Supporting hisprediction hepathfrom transformationaleadership o higherpurpose s signifi-cant (Ys5,= 0.32, p < 0.01).The next threehypothesesmakepredictionsabout the effects of higher purpose;positiverelation-shipsareexpectedbetweenhigherpurposeandjob satisfaction hypothesis4), unit cohesion (hypoth-esis 5), and effort (hypothesis 6). Results supportedall three hypotheses:03,5= 0.55, /4,5 = 0.16,/2,5 = 0.18 (all p < 0.01). Hypothesis7 predictsa positive relationshipbetween unit cohesion andeffort; however, the results did not supportthis hypothesis (02,4=- 0.02). Finally, hypotheses 8and 9 respectivelypredictpositiverelationshipsbetweeneffortand ob satisfaction,and effortandper-formance.Both hypotheseswere supported 03,2 = 0.33, 01,2 = 0.80, bothp < 0.01).

    Testof alternativemodelsThe fact that data fit one model, of course, does not precludethe possibility that the data might fitalternativemodelsas well. Topermitdirectcomparisonof ourhypothesizedmodelto alternativemod-els, we estimateda series of less restrictivenested modelsby freeingone theoreticallyplausiblepathata time. We chose this particularapproach n orderto assess the extent of mediationrelativeto ourhypothesesandto betterframethe results.Does transformational leadership directly relate to effort and performance?Given the centralrole thatmediationby higherpurpose plays in the model's underlyingthesis, twocriticalquestionsthatshouldbe addressedarewhether ransformationaleadershipexhibitsdirectrela-tionshipswith effortandperformance.Withrespectto effort,afterfreeingthe associatedpath,modelestimationproduceda non-significantdifference n X2(X2diff= 0.27); transformationaleadershipdidnot directlyassociate with effort.Similarly,freeingthe pathfromtransformationaleadership o per-formanceresulted in an almostequally small change in X2(X2diff= 0.30). Collectively,these resultsprovideadditionalsupport or the hypothesized mediatingrole of belief in a higherworkpurpose.Do higher purpose, job satisfaction, and unit cohesion directly relate to performance?The model also hypothesizesthat effort mediatesthe effects on performanceby higherpurposeandunit cohesion. To test these mediatingeffects, we estimated alternativemodels that included therespectiverelationships.First,we allowed the directpathfromhigherpurpose o performanceo esti-mate.Relaxingthisrestrictiondid notyield a significantchangeinX2 (X2diff 1.48).Likewise,freeingthe pathfrom unit cohesion to performanceproduced ittle change in X2(X2diff=1.58). Finally,wefreed thepathfrom ob satisfaction o performance; stimationresulted n a nearnegligibledifferencein X (X2diff= 0.02). These results furthersupport he mediatingeffects of efforthypothesizedin themodel.

    DiscussionThis study presents an initial examination of leadership in multilevel marketing organizations(MLMs), a unique, growing,and understudiedorganizationalcontext. The results of this study,webelieve, make two distinctbutrelatedcontributionso the organizationaliterature.First,theyprovideCopyright 2001 JohnWiley& Sons,Ltd. J. Organiz. ehav. 2, 849-869(2001)

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    18/22

    MULTILEVELMARKETINGORGANIZATIONS 865tentative insights into the nature of MLMs, and in particular, he importanceof sponsor-recruitrelationships o the success of MLMdistributorships. econd, they supportan importantcomponentof transformationaleadership heory,belief in a higherworkpurpose,as a mechanism hroughwhichtransformationaleadershipmay achieve satisfactionandperformance.With respect to our understandingof MLMs, one noteworthy nsight is the apparent mportanceof sponsor-recruit relationships to the performance of recruits' distributorshipsand recruits'satisfaction with their work.The positive direct and indirectrelationshipsbetween transformationalleadershipand effort, job satisfaction, and performancesuggest that sponsors may exert positiveeffects on theirrecruitsand theirrecruits'distributorships espite the fact thatsponsorscannotexertdirect supervisorycontrol over their recruits' work activities. Beyond the individualrelationshipsbetween sponsorsandrecruits,sponsors'transformationaleadershipbehaviorsmay produceeffectsat the grouplevel as well. In supportof this inference, the positive directand indirectassociationsbetween leadershipbehaviors and unit cohesion suggest that members' transformationsmay leadto cooperative and cohesive behaviors among members within the sponsor-formedunits. Thesubsequentresults of this cohesiveness, however, remain unclear in light of the non-significantrelationshipbetween unit cohesion and effort. To the extent that the true relationshipis indeednon-significant,it could be due to mitigatingfactors such as social loafing (Karauand Williams,1993). The independentnatureof MLM distributorshipsmakes this possibility plausible and per-haps worthy of further nvestigation.An additional nsightinto the characterof MLMs comes frommembers'ascriptionof higherpur-poses to theirwork,whichappears o be associatedwith theirperformance.This resultmay hold par-ticular interest to MLM corporate producers, who frequently refer to higher purposes incommunicationswith theirmembership King, 1996). To the extentthatmembersdo achievegreaterperformanceandsatisfactionthrough heirbeliefs in the higher good of theirwork,corporateprodu-cers wouldbe well servedto learnmoreaboutwhich higher purposesbest motivateandsatisfy theirmembers.

    Apartfrom what may be learned about MLMs as an organizationalcontext, this study tests animportant xplanatorycomponentof transformationaleadership heory:thatthe ascriptionof highermoralpurpose o one's workis atleast one mechanism hroughwhichtransformationaleadershippro-duceshigherperformingandmore satisfied ollowers.Ourresultsaresuggestiveof thatbasicpremise.Significantrelationshipsbetween higherpurposeandjob satisfaction,unit cohesion, and effort andsubsequentrelationshipswith satisfactionand performancesupportthe model's basic hypothesizedmediatedsequence.Thus,one might reasonably nferthattransformationaleadershipappears o trans-formfollowers.However,muchremains o be discoveredabout hattransformation.nparticular,whatspecific higher purposes does transformationaleadership produce, and which produce the bestresponsesin effort and satisfaction?The member interviewsmay proveinstructive n this regard.A post hoc examinationof intervie-wees' commentsseem to indicate that a useful way of conceptualizinghigherpurposes s in termsofbenefitsaccrued o particularndividualsorgroups.Forexample,some membersspokeof exceptionalproductbenefits; heproducts heyrepresentedmproved helives of theircustomers n someuniqueorexceptional way. Some even extendedthis notion of individualcustomerbenefit to society at large.Othermembersnoted the enjoyment they derivedfrommeetingnew people andmakingnew friendseitherthroughcustomersorMLMmembers n theirunits.Thus,they realizesocial benefitsfromtheirMLMparticipation.Futureresearchcoulddevelopa taxonomyof higherpurposesandprovidea basisfor explicatingthis constructandfurtherexaminingits nomologicalvalidity.Our resultsmustbe considered n light of the study'slimitations,each of which suggestdirectionsfor futureresearch.Asjust discussed,one is ourbroadconceptualizationof belief in a higherpurpose,whichmay be composedof manytypes or facets. Another imitationarisesfromour focus on four ofCopyright 2001JohnWiley& Sons,Ltd. J. Organiz. ehav. 2, 849-869(2001)

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    19/22

    866 J. R. SPARKSAND J. A. SCHENKthe six transformationaleadershipbehaviors.Althoughwe believe warranted y the unique qualitiesof sponsor-recruit elationshipsn MLMs,theirexclusionfrom this studyconstrains onclusionsaboutthe universeof transformationaleadershipbehaviors.A third imitationcomes from the marginalper-formanceof thejob satisfaction tems, whichdidnot sharemore thanhalf theirtotalvariance.Fourth,to the extent thatnon-responsebias influencedourresults,the responserateof 39 percent shouldbeviewed as a studylimitation.A series of potential imitationsalso arisesfrom our use of two- andthree-itemmeasuresandself-reportsas our datasource.Althoughour use of two- andthree-itemmeasuresresulted n part romtheMLMcorporateproducer's equest hat the questionnairebe as briefas possible,the use of such mea-suresmaynonethelessgive somepause.Certainly heiruse decreasesthe likelihoodof adequately ap-turingthe theoreticaldomain of a given construct.To the extentthatsuch measuresaretheoreticallyproblematic,we acknowledgethem as a limitationof ourstudy.The use of two- andthree-itemmea-sures also occasionally presentsmethodologicalchallenges. In our case, this challenge arose in theformof modelunderidentification henusinga methodfactorto test for samesource bias.As a result,we limited the applicationof the methodfactorto the latentconstructs,andthuscouldnotconductourpreferred est for methodbias.On this point, however, we offer two comments. First, Cramptonand Wagner's(1994), meta-analysisconcludes that 'percept-percept nflationmay be more the exceptionthan the rule' (p. 72).Second, Singh (2000, p. 31) notes that to the extent that self-reportsproducebias at all, it is 'morelikely to bias the [construct]meanvalues (upward)but less likely to bias theircorrelationswith otherconstructs.'Thus,while we cannotignorethe possibilitythatself-reportsbiasedthe data usedin thisstudy,evidence suggests that this likelihood is small.Additionally,the uniquecontextand sampleused in this study,while offeringcertainadvantages,may inhibit the generalizabilityof the results,as does the use of a single MLM as a sample frame.However,we see good reason to expect thatthese resultswill generalizebeyondthe presentcontext.Totheextent thatmanyorganizations ontinuetheir trends oward latter tructures, igherrelianceonmultidisciplinaryeams,encouragementof entrepreneurial ehavior,andless clearlydefined ines ofauthority,henmanycharacteristics f MLMspertinent o leadershipmay applyin a varietyof orga-nizationalsettings. By samplingorganizations hatpermita varietyof leadershipbehaviors,futureresearchcould examine the interrelationships etween transformationaleadershipand other leader-ship styles in theirabilities to evoke followers' higherordermotives for working.Finally,we note that ourmodel, like all theoreticalmodels, does not includepossible explanatoryvariables. n terms of the futureresearch,we suggestthreebasic domainsof variables hatmay fitwellwithin the generalmediational ramework ollowed here.First,researchersmightexamine other med-iatorsof higherpurposeandeffort.Beyondunitcohesion, theorganizationaliteratures a richsourceof possibilities includingorganizational ommitment,motivationto work, or work confidence.Sec-ond, as discussedearlier,we believe thatexplicationof the higherpurposeconstructholds greatpro-mise for deepeningourunderstanding f the mechanismsthroughwhich transformationaleadershipworks. Third,beyond higherpurpose,other variablesmay serve to explain the mechanismthroughwhich transformationaleadershiptransforms.As noted earlier,Shamiret al. (1993) suggest self-esteem as just one possibility.In conclusion,this researchprovidesa glimpse into a seldom-studiedandcomplex organizationalcontext,multilevelmarketingorganizations. nparticular,he studyexaminesand findssupport ortheimportanceof a key dyadicMLMrelationship, hatwhich exists between sponsorsand recruits.Theevidence suggests that sponsors' transformationaleadershipmay be effective in bringing aboutgreatereffort, performance,and satisfactionfrom her recruits.Moreover,this researchinvestigatesa fundamentalexplanatory componentof transformationaleadershiptheory;that one mechanismthroughwhich leadersobtainsuperioroutcomesfrom theirfollowersis by appealing o theirfollowers'Copyright 2001 JohnWiley&Sons,Ltd. J. Organiz. ehav. 2, 849-869(2001)

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    20/22

    MULTILEVELMARKETINGORGANIZATIONS 867

    higherordermotives.Our resultsoffersupport or thatexplanatorycomponentandprovidea startingpoint for its understanding.

    Author biographiesJohn R. Sparks is an AssociateProfessorof Marketingat the Universityof Dayton.He receivedhisPh.D. fromTexas TechUniversityin 1994. His research nterests nclude leadership,ethics, andper-suasion. His research has appeared in the Journal of Marketing, Communication Monographs, and theJournal of the Academy of Marketing Science among others.Joseph A. Schenk is anAssociateProfessorof Managementat theUniversityof Dayton.He receivedhis D.B.A. from Kent State University in 1976. His research interests include business strategy, leader-ship, and entrepreneurship.

    ReferencesAndersonJC,GerbingDW. 1988. Structural quationmodelingin practice:a reviewand recommended wo-stepapproach.PsychologicalBulletin 103: 411-423.AtwaterLE,YammarinoFJ. 1993. Personalattributes s predictorsof superiors'andsubordinates'perceptionsofmilitary academyleadership.HumanRelations 46: 645-668.Bagozzi RP,Yi Y. 1988. On the evaluationof structural quationmodels. Journalof theAcademy of MarketingScience 16: 74-94.BarlingJ,WeberT,KellowayKE. 1996. Effectsof transformationaleadership rainingon attitudinal ndfinancialoutcomes: a field experiment.Journalof Applied Psychology 81: 827-832.Bartlett RC. 1995. MaryKay's foundation.Journalof BusinessStrategy16: 16-19.Bass BM. 1985. Leadershipand PerformanceBeyond Expectations.Free Press:New York.Bass BM. 1990. Bass andStodgill'sHandbookof Leadership:Theory,ResearchandManagerialApplications,3rd

    edn. Free Press:New York.Bass BM. 1997. Personalselling and transactional/transformationaleadership.Journalof PersonalSelling andSales Management17: 19-28.Bass BM. 1998. Transformational eadership: ndustrial,Military,and EducationalImpact.LawrenceErlbaumAssociates:Mahwah,New Jersey.Bass BM, Avolio BJ, Goodheim L. 1987. Biographyand the assessmentof transformationaleadershipat theworld-classlevel. Journalof Management13: 7-19.Breckler SJ. 1990. Applications of covariance structuralmodeling in psychology: cause for concern?Psychological Bulletin 107: 260-273.BrownSP,PetersonRA. 1994. The effect of efforton salesperformance nd ob satisfaction.Journalof Marketing58: 70-80.BurnsJM. 1978. Leadership.Harperand Row: New York.Bycio P,HackettRD, Allen JS. 1995. Furtherassessmentsof Bass's 1985 conceptualizationof transactional ndtransformationaleadership.Journalof Applied Psychology80: 468-478.Carless SA. 1998. Assessing the discriminantvalidityof transformationaleaderbehaviouras measuredby theMLQ.Journalof Occupationaland OrganizationalPsychology71: 353-358.CongerJA. 1999. Charismaticandtransformationaleadership n organizations:an insider'sperspectiveon thesedevelopingstreamsof research.LeadershipQuarterly10: 145-179.CramptonSM, WagnerJA. 1994. Percept-percept nflation n microorganizationalesearch:an investigationofprevalenceandeffect. Journalof AppliedPsychology79: 67-76.Dansereau F, Graen G, Haga WJ. 1975. A vertical dyad linkage approach to leadership within formalorganizations:a longitudinal investigationof the role making process. OrganizationalBehavior & HumanPerformance13: 46-78.

    Copyright? 2001 JohnWiley & Sons, Ltd. J. Organiz.Behav.22, 849-869 (2001)

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    21/22

    868 J. R.SPARKS NDJ. A. SCHENKDeci EL. 1975. IntrinsicMotivation. Plenum Press:New York.Deci EL, RyanRM. 1985. IntrinsicMotivationand Self Determination n HumanBehavior. Plenum Press: NewYork.DirectSelling Association.1999. 1999 DirectSelling Growthand OutlookSurvey.The DirectSelling Association:Washington,DC.Dubinsky AJ, YammarinoFJ, Jolson MA, Spangler WD. 1995. Transformational eadership: an initialinvestigation n sales management.Journalof PersonalSelling and Sales Management15: 17-31.DumaineB. 1994. Why do we work? Fortune130: 197-204.Dunkel T. 1995. The Tupperwaready is an MBA. WorkingWoman 0: 45-49.FornellC, LarckerDE 1981.Evaluating tructural quationmodelswith unobservable ariablesandmeasurementerror.Journalof MarketingResearch18: 39-50.Graen G, Cashman JF. 1975. A role-makingmodel of leadershipin formal organizations:a developmentalapproach.Organizationand AdministrativeSciences 6: 143-165.HackmanJR, OldhamGR. 1975. Developmentof thejob diagnosticsurvey.Journalof AppliedPsychology60:159-170.JdreskogKG, Sorbom D. 1993. LISREL8User's ReferenceGuide.Scientific SoftwareInternational:Chicago.KarauSJ, Williams KD. 1993. Social loafing: a meta-analyticreview and theoretical integration.Journal ofPersonality& Social Psychology65: 681-706.King CW. 1996. Services: the New Front n the MLMRevolution.Success 43: 72-75.KohWL,SteersRM,Terborg R. 1995.Theeffects of transformationaleadershiponteacherattitudesand studentperformance n Singapore.Journalof OrganizationalBehavior 16: 319-333.Locke EA. 1976. The natureand causes of job satisfaction.In Handbook of Industrial and OrganizationalPsychology,DunnetteMD (ed.). RandMcNally:Chicago; 1297-1350.Lord RG. 1977. Functional eadershipbehavior: measurementand relationshipto social power and leadershipperceptions.AdministrativeScience Quarterly22: 114-133.Lowe KB, KroeckKG, SivasubramaniamN. 1996. Effectivenesscorrelatesof transformation nd transactionalleadership:a meta-analyticreview of the MLQliterature.LeadershipQuarterly7: 385-425.LucasGH,BabakusE, IngramTN. 1990.Anempirical est of thejob satisfaction-turnoverelationship. ournaloftheAcademy of MarketingScience 18: 199-209.MacKenzie SB, Podsakoff PM, Fetter R. 1993. The impact of organizationalcitizenship behavior on theevaluationsof salespersonperformance.Journalof Marketing57: 70-80.Matey DB, 1991. Significanceof transactionaland transformationaleadershiptheoryon the hospitalmanager.Hospital & Health ServicesAdministration 6: 600-606.McDonaldRP,MarshHW.1990. Choosinga multivariatemodel:noncentrality ndgoodnessof fit.PsychologicalBulletin 107: 247-255.McFarlinDB, Rice RW.1992.Theroleof facet importanceas a moderatornjob satisfactionprocesses.JournalofOrganizationalBehavior 13: 41-54.Medley F, Larochelle DR. 1995. Transformationaleadershipand job satisfaction.Nursing Management26:64JJ-64NN.Niehoff BP,Enz CA, GroverRA. 1990.The impactof top-management ctionson employeeattitudes.Group&OrganizationManagement15: 337-352.Podsakoff PM, MacKenzie SB. 1994. An examination of the psychometric properties and nomologicalvalidity of some revised and reduced substitutesfor leadershipscales. Journal of Applied Psychology 79:702-713.PodsakoffPM, OrganDW. 1986. Self-reports n organizationalresearch:problemsand prospects.Journal ofManagement12: 531-544.PodsakoffPM,MacKenzieSB, MoormanRH,FetterR. 1990.Transformationaleaderbehaviorsand theireffectson followers' trustin leader, satisfaction,and organizationalcitizenshipbehaviors.LeadershipQuarterly1:107-142.Podsakoff PM, MacKenzie SB, Bommer WH. 1996. Transformationaleader behaviors and substitutes forleadership as determinants of employee satisfaction, commitment, trust, and organizationalcitizenshipbehaviors.Journalof Management22: 259-298.SalterJ. 1997. MultilevelMarketingGoes Mainstream.MarketingNews 31: 1-5.ShamirB, House RJ,ArthurMB. 1993. The motivationaleffects of charismatic eadership:a self-conceptbasedtheory. OrganizationScience 4: 577-594.Singh J. 2000. Performanceproductivityand qualityof frontlineemployees in service organizations.JournalofMarketing64: 15-34.

    Copyright0 2001 JohnWiley & Sons, Ltd. J. Organiz.Behav.22, 849-869 (2001)

    This content downloaded from 109.100.54.2 on Sat, 18 Jan 2014 06:54:47 AMAll use subject to JSTOR Terms and Conditions

    http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
  • 8/13/2019 The Effects of LTf

    22/22

    MULTILEVELARKETINGRGANIZATIONS 869SujanH. 1986.Smarterversusharder: nexploratoryattributionalnalysisof salespeople'smotivation.JournalofMarketingResearch 23: 41-49.WaldmanDA, YammarinoFJ. 1999. CEOcharismatic eadership: evels-of-managementand levels-of-analysiseffects. Academy of ManagementReview24: 266-285.WotrubaTR,TyagiPK. 1992.Motivation o become a directsalespersonand ts relationshipwith workoutcomes.Journalof MarketingChannels1: 41-56.YammarinoFJ,Bass BM. 1990.Long-term orecastingof transformationaleadershipand ts effects amongnavalofficers.InMeasurescaLeadership,ClarkKE,ClarkMB (eds). LeadershipLibraryof America:WestOrange,NJ; 151-170.YammarinoFJ, DubinskyAJ. 1992. Superior-subordinateelationships:a multiple levels of analysisapproach.HumanRelations 45: 575-600.YuklG. 1998. Leadership n Organizations,6th edn. Prentice-Hall:UpperSaddleRiver,NJ.YuklG. 1999. An evaluationof conceptualweaknesses in transformational ndcharismatic eadership heories.LeadershipQuarterly10: 285-305.Yukl GA, Van Fleet DD. 1982. Cross-situational,multimethod research on military leader effectiveness.OrganizationalBehaviorand HumanPerformance30: 87-108.

    Copyright 9 2001 JohnWiley & Sons, Ltd. J. Organiz.Behav.22, 849-869 (2001)