FILIP LIEVENS AND WILFRIED DE CORTEflievens/outsourcing.pdf · HR outsourcing takes place when a...
Transcript of FILIP LIEVENS AND WILFRIED DE CORTEflievens/outsourcing.pdf · HR outsourcing takes place when a...
Human resources outsourcing is thefastest-growing outsourcing seg-ment (Dell, 2004). HR outsourcingtakes place when a company con-tracts with an HR vendor to perform
an HR activity previously performed by thecompany (Greer, Youngblood, & Gray,1999). Forecasts predict that HR outsourcingwill continue to grow by more than 30 per-cent so that the global expenditure on HRoutsourcing in the United States will reach$14 billion by 2009 (Yankee Group, 2005).
Today, HR outsourcing involves more thanoutsourcing HR services such as payroll ad-ministration. Instead, HR outsourcing playsa strategic role as organizations are out-sourcing HR activities such as training andselection and, in some cases, even the totalHR function (Adler, 2004; M. F. Cook, 1999;Greer et al., 1999; Lepak & Snell, 1998).
Most prior studies have used transaction-cost economics theory (Williamson, 1996)for investigating specific factors that influ-ence the initial decision to outsource HR ac-
DEVELOPMENT AND TEST OF A
MODEL OF EXTERNAL
ORGANIZATIONAL COMMITMENT
IN HUMAN RESOURCES
OUTSOURCING
F I L I P L I E V E N S A N D W I L F R I E D D E C O R T E
Most prior outsourcing studies in the human resources domain have fo-cused on the initial decision for outsourcing HR activities. Hence, little isknown about HR managers’ commitment to continue an already existingoutsourcing relationship. This study constitutes a first step to increase ourunderstanding of the factors related to the continuity of HR outsourcing re-lationships. We developed and tested a model of HR managers’ (N = 186)commitment in outsourcing relationships. Affective commitment or the de-sire for the outsourcing relationship to continue was related to the depth andfrequency of HR outsourcing. Conversely, continuance commitment, whichrefers to the constraints that keep the outsourcing relationship intact, wasnot related to the continuity of HR outsourcing. Finally, affective commit-ment was fostered by HR managers’ perception that HR vendors shared thesame values for managing people. © 2008 Wiley Periodicals, Inc.
Keywords: Lievens: human resources outsourcing, external organizationalcommitment, shared values, trust, switching costs.
Correspondence to: Filip Lievens, Department of Personnel Management and Work and Organizational Psychol-ogy, Ghent University, Henri Dunantlaan 2, 9000, Ghent, Belgium, Phone: +32 9 264 64 53, Fax: +32 9 264 64 94,E-mail: [email protected].
Human Resource Management, Fall 2008, Vol. 47, No. 3, Pp. 559–579
© 2008 Wiley Periodicals, Inc.
Published online in Wiley InterScience (www.interscience.wiley.com).
DOI: 10.1002/hrm.20232
09HRM47_3lievens 8/11/08 11:07 AM Page 559
560 HUMAN RESOURCE MANAGEMENT, Fall 2008
tivities (e.g., Klaas, McClendon, & Gainey,1999, 2001; Lepak, Bartol, & Gardner, 2004).According to transaction-cost economicstheory, organizations choose governancestructures (i.e., market-based governance inthe form of HR outsourcing versus organiza-tional-based governance in the form of in-house HR) that maximize efficiency andminimize the transaction costs of perform-ing an activity. Along these lines, Klaas et al.(1999) showed that there is no simple answerto the “make or buy” decision, as organiza-
tional characteristics moderatedthe relationship between the de-gree of outsourcing and the bene-fits of outsourcing. Smaller firmsfacing more uncertainty and costpressures perceived more benefitsfrom HR outsourcing, whereasfirms with idiosyncratic HR prac-tices perceived fewer benefits.
Although prior studies haveprovided valuable insights intothe factors underlying the out-sourcing decision, there also aresome important gaps. First, priorresearch focuses on the initialoutsourcing decision, whichsheds light only on the start ofthe outsourcing relationship.However, an inherent characteris-tic of any interorganizational re-lationship is its continuity (Fich-man & Levinthal, 1991; Ring &Van De Ven, 1994). For example,HR best practices prescribe that“rather than being a Band-Aid so-lution, outsourcing needs to beviewed as a long-term strategy”(Maurer & Mobley, 1998, p. 10).So far, little is known about HR
managers’ commitment to continue the HRoutsourcing relationship. Conceptually, it isdoubtful that the same factors leading to theinitial outsourcing decision also will be re-lated to its continuance, because such a viewdisregards the duration of the relationship.In fact, research on interorganizational rela-tionships has shown that “the conditionsand processes associated with the formationof exchange relationships are not the same
as those associated with their persistence orseverance” (Seabright, Levinthal, & Fich-man, 1992, p. 123).
Second, an economic theory such astransaction-costs economics theory typi-cally ignores that outsourcing decisions aretaken by individuals (HR managers) whohave individual desires and agendas whencreating, developing, and maintaining HRoutsourcing relationships. As outlined by so-cial exchange theory (Blau, 1964; Emerson,1962; Gainey & Klaas, 2003), the continuityof HR outsourcing relationships might alsobe related to the social and affective ties thatthe partners develop over time. In researchon interorganizational relationships, it isimportant not to attribute individual moti-vations to organizations (Klein, Palmer, &Conn, 2000; Seabright et al., 1992; Zaheer,McEvily, & Perrone, 1998). In the end, indi-viduals (not organizations) are committedto staying or quit working with a partnerfirm and its people.
To start filling these gaps in the HR out-sourcing literature, we believe a psychologi-cal, in addition to a purely economicalmodel must be used to understand the HRoutsourcing relationship. Along these lines,this study builds on the psychological litera-ture on external organizational commit-ment. Specifically, we conceptualize bothparties’ commitment to each other as a bind-ing mechanism based on a mind-set of either“perceived cost” (continuance commitment)or “desire” (affective commitment). We ex-amine how these binding mechanisms arerelated to a quantitatively and qualitativelystrengthened HR outsourcing relationship.So, this study constitutes a first step to in-crease our conceptual understanding of thefactors related to the continuity of HR out-sourcing relationships.
In this study, we focus on one type of HRoutsourcing—namely, outsourcing recruit-ment and selection activities. Thus, we fo-cused on category 18 (“Recruiting, staffing,and search”) of Adler’s (2004) classificationof HR outsourcing. These services were de-fined as a wide variety of activities includingthe design (e.g., development of recruitmentcampaigns, development of selection and as-
HR best practices
prescribe that
“rather than being a
Band-Aid solution,
outsourcing needs
to be viewed as a
long-term strategy”
(Maurer & Mobley,
1998, p. 10). So far,
little is known about
HR managers’
commitment to
continue the HR
outsourcing
relationship.
Human Resource Management DOI: 10.1002/hrm
09HRM47_3lievens 8/11/08 11:07 AM Page 560
A Model of External Organizational Commitment in Human Resources Outsourcing 561
sessment tools), delivery (e.g., applicantsourcing; résumé processing; administering,scoring, and profiling tests; assessment of in-dividual capabilities; candidate report writ-ing; provision of feedback), and evaluation(e.g., validation and transportability studies,adverse impact analyses) of recruitment andselection-related services.
We focus on outsourcing recruitmentand selection activities for two reasons. First,no prior research has examined this type ofoutsourcing, even though recruitment andselection are key strategic HR functions. Sec-ond, these HR outsourcing activities arecharacterized as the outsourcing of discrete,repetitive, and high-volume HR services(Adler, 2004). Contrary to other types of HRoutsourcing (e.g., outsourcing the total HRfunction), HR managers might switch fre-quently among HR vendors when outsourc-ing these activities. Accordingly, the out-sourcing of recruitment and selectionactivities is ideally suited for examining thecontinuity of the outsourcing relationship.
Model Development andHypotheses
Relationship of Different Forms ofCommitment to Key OutsourcingOutcomes
HR managers’ commitment to continue theHR outsourcing relationship can be framed inthe context of external organizational com-mitment. McElroy, Morrow, and Laczniak(2001) defined external organizational com-mitment as the psychological bond of an in-dividual boundary spanner of a particular or-ganization toward an external organization.Similar to organizational commitment, weposit that external organizational commit-ment is a multidimensional construct con-sisting of two distinguishable forms of com-mitment—namely, affective and continuancecommitment (Meyer & Allen, 1984; Meyer,Allen, & Gellatly, 1990; Meyer, Paunonen,Gellatly, Goffin, & Jackson, 1989).1
According to Meyer and Herscovitch(2001), continuance commitment refers tocommitment to a course of action on the
basis of cost avoidance and of the threat oflosing investments. This form of commitmentmaps well into the aforementioned notions oftransaction-cost economics theory (Geyskens,Steenkamp, & Kumar, 2006; Williamson,1996). According to this theory, organizationswill choose market-based governance (HRoutsourcing) over organizational-based governance (in-house HR)when the former maximizes effi-ciency and minimizes the transac-tion costs of performing the activ-ity. So, transaction-cost economicsconceptualizes continuance interms of the costs and benefits ofstaying in the relationship versusleaving it.
Central concepts in transac-tion-cost economics theory aredependence and switching costs(the costs of switching from oneHR service provider to anotherone). According to transaction-cost economics theory, partnerswill prolong their relationshipwhen switching costs are high(e.g., investments in specializedtechnology and expertise). Asnoted above, Klaas et al. (1999)successfully applied transaction-cost economics theory to explainunder which conditions out-sourcing might be beneficial fororganizations.
Compared to continuancecommitment, affective commit-ment is a more positive form ofcommitment. Affective commitment repre-sents the “desire” for a relationship to con-tinue and reflects a feeling of emotional at-tachment to an organization. This form ofcommitment parallels assumptions underly-ing social exchange theory (Blau, 1964;Emerson, 1962). Social exchange theorycomplements transaction-cost economicstheory by emphasizing factors that developon the basis of continuing relations betweenpartners over time and that turn a purelyeconomic exchange into a social exchange(Gainey & Klaas, 2003). These factors refer topersonal relationships, affective ties, high-
These two
perspectives (i.e.,
transaction-cost
economics theory
and social
exchange theory)
are complementary
instead of
contradictory. That
is, what begins as
an arm’s-length
exchange might
gradually evolve
into a stronger
affective bond.
Human Resource Management DOI: 10.1002/hrm
09HRM47_3lievens 8/11/08 11:07 AM Page 561
562 HUMAN RESOURCE MANAGEMENT, Fall 2008
quality communication patterns, and trustamong the social exchange partners (in thiscase, the HR manager and the outsource di-rector). In a seminal paper, Cook and Emer-son (1978) argued that such affective com-mitments reduce the calculus of resourcepower and dependency.
These two perspectives (i.e., transaction-cost economics theory and social exchangetheory) are complementary instead of con-
tradictory. That is, what begins asan arm’s-length exchange mightgradually evolve into a strongeraffective bond. Hence, we expectthat in external organizationalcommitment, both continuanceand affective commitment will berelated to the likelihood of con-tinuing the HR outsourcing rela-tionship. In other words, bothcontinuance and affective com-mitment will be related to thelength of the HR outsourcing re-lationship. The length of the HRoutsourcing relationship refers tothe number of years the partnershave been working together. Out-
sourcing length is not a qualitative feature ofthe continuity of an outsourcing relation-ship as it does not denote the level of invest-ment in the relationship.
We do expect divergent relationships ofthe types of commitment with respect toqualitatively strengthening the outsourcingrelationship. An HR outsourcing relationshipmight become qualitatively stronger whenan organization increases the depth of out-sourcing. This refers to the extent to whichan organization outsources a higher portionof that activity on average (Gilley & Rasheed,2000). Similarly, an organization might in-crease the frequency of outsourcing (i.e., thenumber of times a firm relies on the servicesof an HR vendor). These two outcomes aredifferent from outsourcing longevity, as afirm has the freedom to gradually outsourcemore activities within an existing HR outsourc-ing relationship. Similarly, a firm mightchoose to gradually rely more frequently onthe services of a specific HR vendor within anexisting relationship. So, increased outsourc-
ing frequency and outsourcing depth arequalitative features of the continuity of anoutsourcing relationship that might resultfrom the external commitment relationshipevolving into more than a purely economicexchange. Therefore, we expect that only thestronger affective form of commitment willbe related to these two outcome variablesthat reflect a qualitatively strengthened out-sourcing relationship (see Figure 1).
Conceptually, these divergent relation-ships of the types of commitment with out-sourcing frequency and depth also are consis-tent with the organizational commitmentliterature. There is ample evidence that affec-tive commitment constitutes a stronger bondthan continuance commitment. A meta-analysis of Meyer et al. (2002) showed that af-fective commitment correlated significantlywith a wider range of outcomes and corre-lated more strongly with any specific out-come. Although these relationships typicallyare examined at the individual level in the or-ganizational commitment literature, we positthat these relationships also will hold at thelevel of external organizational commitment.
In sum, we hypothesize the following:
H1a: Continuance commitment will be signifi-cantly and positively related to outsourcinglength.
H1b: Affective commitment will be significantlyand positively related to outsourcing length,depth, and frequency.
Bases of Continuance and AffectiveCommitment
Although the relationship between the differ-ent forms of commitment and outsourcingoutcomes constitutes the central part of ourmodel, we also formulate hypotheses aboutpossible driving forces of each of the forms ofcommitment because they might provideclues for fostering commitment in HR out-sourcing relationships. To develop these hy-potheses, we used two sources of informa-tion. First, we drew on the organizationalcommitment (Meyer & Allen, 1997; Meyer &Herscovitch, 2001) and external organiza-tional commitment literatures (McElroy et
Continuance
commitment
develops when
someone has made
investments that
could be lost when
the activity is
discontinued.
Human Resource Management DOI: 10.1002/hrm
09HRM47_3lievens 8/11/08 11:07 AM Page 562
A Model of External Organizational Commitment in Human Resources Outsourcing 563
al., 2001; Siders, George, & Dharwadkar,2001). Second, we conducted semistructuredinterviews with ten experienced HR man-agers (six men, four women, aged between 32and 38 years) and senior consultants of HRservice firms (six women and four men, agedbetween 30 and 39 years). Apart from servingas a basis to determine possible variables ofinterest, these interviews also provided inspi-ration to construct survey questions.
Continuance commitment developswhen someone has made investments thatcould be lost when the activity is discontin-ued (Cohen & Lowenberg, 1990; Meyer &Herscovitch, 2001). Our interviews con-firmed that a bond between a company andan HR vendor on the basis of perceived costwas relevant in the context of outsourcingHR activities. It was mentioned that out-sourced recruitment and selection servicescould be placed on a continuum in terms oftheir degree of customization. On one end ofthe continuum, one finds, for example, stan-
dardized (“prepackaged”) stand-alone testingprograms that are applicable across variouscompanies and that do not require substan-tial investments on the part of the HR ven-dor. Conversely, other services need to be tai-lored to a company’s specific needs such asthe development of company-specific com-petency models or recruitment campaigns. Itis difficult or impossible to redeploy suchidiosyncratic or asset-specific investments toanother relationship (Gainey & Klaas, 2003;Klaas et al., 1999). The greater these idiosyn-cratic and highly specialized investmentsare, the more one party might perceive thatit would be costly to stop working with theother party, leading to dependence and vul-nerability to opportunistic behavior (e.g.,disproportionate price increases, chargingfor modifications during the contract period,or reducing quality in areas where outcomesare not specified; Klaas et al., 1999).
In the organizational commitment litera-ture, perceived lack of alternatives consti-
Human Resource Management DOI: 10.1002/hrm
FIGURE 1. Conceptual Model of External Organizational Commitment in HR Outsourcing Relationships
09HRM47_3lievens 8/11/08 11:07 AM Page 563
564 HUMAN RESOURCE MANAGEMENT, Fall 2008
tutes another base for continuance commit-ment (Meyer & Herscovitch, 2001). The ra-tionale is that people commit to a course ofaction on the basis of perceived costs whenfew alternative actions are available (McGee& Ford, 1987). Our interviews confirmedthat the number of alternatives available isindeed relevant in an HR service context.Specifically, HR managers reported that theyfelt more dependent on a specific HR vendorwhen they were relying on only one HR ser-vice firm. Conversely, HR managers arguedthat reliance on many HR vendors made
them less dependent on any oneHR vendor, as they could “playthe game harder” (if necessary).
In sum, this discussion leadsto the following hypotheses:
H2a: Specific investments will besignificantly and positively related tocontinuance commitment.H2b: The number of comparable al-ternative partners will be signifi-cantly and negatively related to con-tinuance commitment.
The development of affectivecommitment is based on valuecongruence, emotional involve-ment, and identification (Meyer& Herscovitch, 2001). Our inter-
views confirmed that perceived value con-gruence (i.e., the perception of sharing thesame values, norms, and goals about manag-ing people) as a basis of affective commit-ment was relevant in an HR service context.Some interviewees referred to this perceivedvalue congruence as a good “cultural fit” or“chemistry.” These notions parallel the largebody of research on person-organization fit(see Kristof, 1996; Schneider, 1987). This areaof research has consistently found evidencethat people are more likely to stay in organi-zations that are perceived to be similar totheir own values and personality. To a cer-tain extent, this tenet might also hold for ex-ternal organizations with which these indi-viduals work. That is, boundary spannersmight decide to stay working with peoplefrom an external partner organization be-
cause they have developed a sense of unitybased on perceived shared values.
Besides perceived value congruence, trustalso has been linked to the development ofaffective commitment in the organizationalcommitment literature (Z. X. Chen, Aryee, &Lee, 2005; Watson & Papamarcos, 2002).Similarly, research in the marketing domainhas related trust to affective commitment(Bansal, Irving, & Taylor, 2004; Morgan &Hunt, 1994; Yilmaz, Sezen, & Ozdemir,2005). Trust is crucial to developing a senseof unity with the other party, because it re-flects a willingness to make oneself vulnera-ble to another party with the expectationthat the other party will not act opportunis-tically (Gainey & Klaas, 2003; Mayer, Davis,& Schoorman, 1995; McAllister, 1995). In-deed, a recurring theme throughout the in-terviews was that trustworthiness played aprevalent role in the development of affec-tive and personal ties.
Our study focuses on knowledge-basedtrust instead of identification-based trust be-cause identification-based trust is analogousto perceived shared values (Lewicki, McAllis-ter, & Bies, 1998; McAllister, 1995; Sheppard& Tuchinsky, 1996). Knowledge-based trustrefers to the expectancy that a partner’s wordmay be relied upon and the belief that thevendor is able to perform the job effectivelyand reliably (Moorman, Zaltman, & Desh-pande, 1992; Zaheer et al., 1998).
In the interviews, three other factors(communication, reputation, and approach-ability) with respect to affective commit-ment were mentioned. We believe that thesefactors are precursors of people’s trustworthi-ness rather than antecedents of affectivecommitment because they reflect patterns ofbehaviors signaling that people from a firmare trustworthy. First, studies in various do-mains have documented that effective com-munication, defined as the exchange of for-mal as well as informal (tacit) information,serves as a key vehicle to build trust amongparties. For example, Gainey and Klaas(2003) showed that open and full communi-cation was a significant determinant of trustin outsourcing training activities. In an orga-nizational context, McAllister (1995) found
In the interviews,
three other factors
(communication,
reputation, and
approachability)
with respect to
affective
commitment were
mentioned.
Human Resource Management DOI: 10.1002/hrm
09HRM47_3lievens 8/11/08 11:07 AM Page 564
A Model of External Organizational Commitment in Human Resources Outsourcing 565
that interaction frequency was a significantpredictor of trust. Finally, in the marketingdomain, effective communication betweentwo parties was found to increase confidencein the continuity of the relationship (Kim &Frazier, 1997) and to dampen conflict (E. An-derson & Weitz, 1989; J. C. Anderson &Narus, 1990; Morgan & Hunt, 1994).
A second factor consists of the approach-ability of the people from the respectivefirms. Approachability can be defined as theextent to which people of the respectivefirms are easily reachable and whether theirresponses are timely. Although we did not re-trieve this factor in models of trust or com-mitment, it emerged repeatedly throughoutour interviews. We posit that approachabilityis an antecedent of trust because being ap-proachable and reachable is a concrete ex-ample of the reliability and predictabilitycomponents that are essential to knowledge-based trust.
Third, HR managers do not rely solely onfirsthand experience about the vendor’strustworthiness (signaled by the approacha-bility and effective communication of peopleat the HR vendor). They also tend to use lesscostly secondhand information (Barthélemy,2003). In a business context, a primarysource of secondhand information abouttrustworthiness is a firm’s reputation forbeing fair (cf. Fombrun, 1996). We posit thatHR managers will use others’ assessment ofthe fairness of an HR service firm as an im-portant informational element in assessingthe trustworthiness of people from that firm.
In sum, the following hypotheses arestated:
H3a: Perceived shared values will be signifi-cantly and positively related to affectivecommitment.
H3b: Knowledge-based trust will be significantlyand positively related to affective commit-ment.
H4a: Communication will be significantly andpositively related to knowledge-based trust.
H4b: Approachability will be significantly andpositively related to knowledge-based trust.
H4c: Company reputation will be significantly andpositively related to knowledge-based trust.
Method
Sample and Procedure
The sample consisted of HR professionalsworking in large private firms in the Flemishpart of Belgium. On the basis of data pro-vided by the National Institute of Statistics,there were 1,249 large private firms (withmore than 200 employees) in the Flemishpart of Belgium. A stratified (in terms of re-gion and industry) random sample of 478firms was drawn from this list. Research as-sistants made telephone calls tothe HR department of these com-panies and asked whether theyoutsourced some of their recruit-ment and selection-related activi-ties to an HR vendor.
Next, they asked the HR pro-fessional who was responsible foroutsourcing (HR manager ormanager recruitment and selec-tion) to complete the survey. Atotal of 348 respondents agreed toparticipate in the study. Apartfrom time constraints, the pri-mary reason for not participatingin the study was that the com-pany did not outsource recruitment and se-lection–related activities. Participating HRmanagers received the hyperlink to the sur-vey, wherein they were told to complete thesurvey with one specific HR vendor in mind.To increase the variation in the sample of HRvendors and avoid positive response bias, HRmanagers were prompted to think of either ashort-term or a long-term partnership. Wedid not study commitment to a group of HRvendors or to the “average” HR vendor be-cause commitment is a specific construct. Asmentioned by McElroy et al. (2001), individ-uals who deal with multiple client organiza-tions show some level of commitment to-ward each of them but not toward clientorganizations in general.
We received completed and usable re-sponses of 195 HR professionals (57% fe-males, 43% males), yielding a response rateof 56% (as compared to the sample of HRprofessionals who agreed to participate). The
The sample
consisted of HR
professionals
working in large
private firms in the
Flemish part of
Belgium.
Human Resource Management DOI: 10.1002/hrm
09HRM47_3lievens 8/11/08 11:07 AM Page 565
566 HUMAN RESOURCE MANAGEMENT, Fall 2008
distribution of the companies across indus-tries was as follows: manufacturing (22%),other services (13%), food (12%), chemicals(11%), textile (10%), distribution (9%), con-struction (7%), transportation (5%), profes-sional services (5%), information systems(4%), pharmaceuticals (1%), and energy(1%). As suggested by Rogelberg and Stanton(2007), we compared this distribution to theoriginal distribution of firms across industryto assess potential nonresponse bias. Thisdistribution significantly differed from theoriginal distribution, χ2 (11) = 24.72, p < .05.This change was due to the overrepresenta-
tion of textile companies in oursample (10% versus 3% in thepopulation distribution).
In our sample, HR profession-als consisted of HR managers(70%) and manager recruitmentand selection (30%). Their meanage was 37 (SD = 9) and meanwork experience in the HR do-main was 10 years (SD = 7 years).On average, HR professionalswere working for 6.5 years to-gether with the HR vendor theyrated. Length of the outsourcingrelationship varied across thesample (minimum length: 0.5years, maximum length: 30
years). Besides the HR vendor they rated, HRprofessionals worked with a median of threeother HR vendors.
Check of Respondents’ Involvementin HR Outsourcing
To check whether respondents could becalled key informants (M. J. Chen, Farh, &MacMillan, 1993; Kumar, Stern, & Anderson,1993), we measured their involvement in HRoutsourcing. Respondents indicated whetherthey could influence the continuity of theHR outsourcing relationship on a six-pointscale, ranging from 1 (no influence at all) to6 (a lot of influence). The mean rating onthis scale was high: 5.31 (SD = .71). However,nine HR representatives indicated they hadrelatively little voice (i.e., rating < 4). In linewith key informant methodology (Kumar et
al., 1993), these respondents were removed,resulting in a final sample of 186 HR profes-sionals.
Measures
The survey consisted of three parts. The firstpart contained the items and rating scales.To increase the response rate, we used abbre-viated versions of existing scales. Unless oth-erwise stated, respondents rated all items ofthese scales on a five-point rating scale, rang-ing from 1 (strongly disagree) to 5 (stronglyagree). The Appendix presents the itemsused.
Perceived Shared Values
We measured whether HR managers per-ceived that HR vendors shared the same val-ues, goals, and norms about managing peo-ple on the basis of two items. Similar to Klaaset al. (2001), only HR managers’ perceptionsof shared values were measured, as the actualperceptions of HR vendors were not assessed.These items also came from Klaas et al.(2001). The internal consistency of thisscale’s ratings was .76.
Communication
HR managers’ perceived level of informaland formal information sharing in the out-sourcing relationship was measured withthree items. These items were taken fromYoung-Ybarra and Wiersema (1999). Theinternal consistency of this scale’s ratingswas .74.
Approachability
HR managers’ reports of the degree to whichpeople were easily approachable and reach-able were measured with three items adaptedfrom LaBahn and Kohli (1997). The internalconsistency of this scale’s ratings was .74.
Reputation for Fairness
We measured how HR managers perceivedthe reputation of each other with two items
We measured
whether HR
managers perceived
that HR vendors
shared the same
values, goals, and
norms about
managing people…
Human Resource Management DOI: 10.1002/hrm
09HRM47_3lievens 8/11/08 11:07 AM Page 566
A Model of External Organizational Commitment in Human Resources Outsourcing 567
that were adapted from Ganesan (1994). Theinternal consistency of this scale’s ratingswas .82.
Knowledge-Based Trust
As argued above, knowledge-based trust wasmeasured in this study. HR managers’ per-ceptions of the degree to which HR vendorswere fair and reliable in their business deal-ings were measured on the basis of threeitems taken from scales developed by Gane-san (1994) and Siguaw, Simpson, and Baker(1998). The internal consistency of thisscale’s ratings was .76.
Specific Investments
This three-item measure was adapted fromGanesan (1994). The internal consistency ofthis scale’s ratings was .77.
Perceived Number of Available Alternatives
As this was a factual antecedent of continu-ance commitment (besides the perception ofswitching costs; see Figure 1), we used a sin-gle item for measuring this antecedent. Thisitem asked HR managers about other possi-ble HR vendor choices.
Continuance Commitment
Three items were adapted from Meyer andAllen (1997). These items tapped both thehigh-sacrifice and low perceived alternativesaspects of continuance commitment (McGee& Ford, 1987). The internal consistency ofthis scale’s ratings was .76.
Affective Commitment
Three items were adapted from Meyer andAllen (1997). The internal consistency of thisscale’s ratings was .89.
The second part of the survey asked forfactual information about the quantitativeand qualitative nature of the outsourcingrelationship. Three dependent variableswere measured. First, respondents indicatedthe length of the outsourcing relationship. To
this end, they indicated in years how longthey had outsourced (or had been outsourc-ing) recruitment/selection activities to theHR service provider that they chose to rate.Second, they provided information aboutthe depth of outsourcing within this specificHR outsourcing relationship (i.e., the per-centage of recruitment/selection activitiescurrently outsourced to this specific HRvendor). Third, the frequency of outsourcingwithin this specific HR outsourcing rela-tionship (i.e., the number of times they hadoutsourced recruitment/selection activitiesto this HR vendor since the out-sourcing relationship started)was measured. As these three de-pendent variables asked aboutfactual information, each ofthem was measured with a sin-gle-item measure. Although weacknowledge the provisional na-ture of the items used, similaritems have been used in prior re-search in other domains (market-ing) for measuring factual infor-mation about outsourcing (seeGanesan, 1994; Jap & Ganesan,2000; Kim & Frazier, 1997). TheAppendix provides the exactwording of the dependent meas-ures used.
The third part of the survey asked aboutcompany information (size, type of industry,region, and distance between company andHR vendor). We also included questionsabout demographic information (gender,age, managerial level, and experience in HRdomain).
Analyses and Results
Descriptive Statistics
Descriptive statistics of the study variablesare presented in Table I. Higher scores on theindependent variables indicate that respon-dents perceived, for instance, their level ofcommunication as satisfactory. Table I fur-ther shows that the highest ratings were forthe approachability of HR vendors. Con-versely, perceived specific investments were
The second part of
the survey asked for
factual information
about the
quantitative and
qualitative nature of
the outsourcing
relationship.
Human Resource Management DOI: 10.1002/hrm
09HRM47_3lievens 8/11/08 11:07 AM Page 567
568 HUMAN RESOURCE MANAGEMENT, Fall 2008
Human Resource Management DOI: 10.1002/hrm
MS
D1
23
45
67
89
10
11
12
1.
Nu
mb
er o
f o
ther
HR
ven
do
rs3.
021.
79--
2.
Sp
ecif
ic in
vest
men
ts2.
58.7
8–.
02.7
7
3.
Sh
ared
val
ues
3.48
.64
–.06
.27
.76
4.
Co
mm
un
icat
ion
3.76
.64
.10
.23
.29
.74
5.
Ap
pro
ach
abili
ty4.
09.5
4.0
5.0
8.1
8.2
2.7
4
6.
Rep
uta
tio
n3.
79.5
3.0
0.1
3.4
3.4
0.1
4.8
0
7.
Tru
st3.
93.5
0–.
16.0
9.4
2.2
9.4
6.2
7.8
1
8.
Co
nti
nu
ance
co
mm
itm
ent
2.07
.67
.02
.26
.01
.04
–.02
.09
–.07
.76
9.
Aff
ecti
ve c
om
mit
men
t2.
62.8
3.0
1.3
7.3
4.2
4.2
3.2
7.3
5.3
3.8
9
10. D
epth
of
ou
tso
urc
ing
(p
erce
nta
ge
ou
tso
urc
ed)
48.9
232
.21
–.10
.13
.06
.14
.12
.06
.16
.18
.27
--
11.
Len
gth
of
ou
tso
urc
ing
re
lati
on
ship
a3.
561.
22.0
8.1
2.0
3.0
7.0
6.0
7.0
5.0
6.1
4.1
9--
12. F
req
uen
cy o
f o
uts
ou
rcin
gb
2.61
1.29
.14
.26
.10
.28
.11
.11
.09
.03
.29
.33
.46
--
Not
e. In
tern
al c
on
sist
enci
es a
re b
old
ed o
n t
he
dia
go
nal
. Co
rrel
atio
ns
≥.1
5 ar
e si
gn
ific
ant
at p
< .0
5, c
orr
elat
ion
s ≥
.19
are
sig
nif
ican
t at
p <
.01.
Var
iab
les
2 to
9 w
ere
mea
sure
d w
ith
a f
ive-
po
int
rati
ng
scal
e, r
ang
ing
fro
m 1
= s
tro
ng
ly d
isag
ree
to 5
= s
tro
ng
ly a
gre
e.
aT
his
var
iab
le w
as m
easu
red
on
a f
ive-
po
int
scal
e: 1
= le
ss t
han
1 y
ear;
2 =
eq
ual
or
mo
re t
han
1 y
ear–
less
th
an 2
yea
rs; 3
= e
qu
al o
r m
ore
th
an 2
yea
rs–l
ess
than
4 y
ears
; 4 =
eq
ual
or
mo
re t
han
4ye
ars–
less
th
an 7
yea
rs; 5
= e
qu
al o
r m
ore
th
an 7
yea
rs.
bT
his
var
iab
le w
as m
easu
red
on
a f
ou
r-p
oin
t sc
ale:
1 =
bet
wee
n 1
an
d 1
0 ti
mes
; 2 =
bet
wee
n 1
1 an
d 2
0 ti
mes
; 3 =
bet
wee
n 2
1 an
d 5
0 ti
mes
; an
d 4
= m
ore
th
an 5
0 ti
mes
.
TA
BL
E
ID
escr
iptiv
e St
atis
tics
of S
tudy
Var
iabl
es (N
= 1
86)
09HRM47_3lievens 8/11/08 11:07 AM Page 568
A Model of External Organizational Commitment in Human Resources Outsourcing 569
rated lowest. With respect to the dependentvariables, outsourcing depth correlated .19with the length of the outsourcing relation-ship and .33 with outsourcing frequency.The largest correlation (r = .46) was obtainedbetween outsourcing length and outsourcingfrequency. The finding that outsourcinglength and frequency were more highly cor-related with each other than with outsourc-ing depth confirms what one would expect,providing some evidence about the con-struct validity of these measures. In addition,these results show that the three outcomevariables had moderate correlations, indicat-ing that they were relatively distinct con-structs.
We used the statistical program EQS(Bentler, 1995) to test the fit of the measure-ment model. This is a confirmatory factoranalysis (CFA) model in which each indica-tor variable was specified to load only on thelatent variable it was purported to measure.There were eight correlated latent variablesspecified (see the Measures section).2 Nostructural relationships between these latentvariables were specified. More informationabout CFA can be found in Hoyle (1995) andSchumacker and Lomax (1996).
We used different fit indices to assesshow the CFA model represented the data.First, we used two absolute fit indices: χ2 andthe ratio of the χ2 to its degrees of freedom(χ2/df). Second, we used two relative fit in-dices—namely, the Tucker-Lewis index (TLI)and Bollen’s incremental fit index (IFI). Fi-nally, two fit indices that are based on thenoncentrality parameter were used, namelythe comparative fit index (CFI) and the rootmean square error of approximation(RMSEA).
The criteria for evaluating these indiceswere for the χ2 to be nonsignificant and theχ2/df to be small approaching unity (Bentler,1995). For the TLI, IFI, and CFI, values > .95constitute good fit and values > .90 accept-able fit (Medsker, Williams, & Holahan,1994). For the RMSEA, values < .05 consti-tute good fit, values in the .05 to .08 rangeacceptable fit, values in the .08 to .10 rangemarginal fit, and values > .10 poor fit(Browne & Cudeck, 1992).
Mardia’s normalized multivariate kurto-sis was 15.07 (p < .01). As the assumption ofmultivariate normality was violated, we usedmaximum likelihood estimation but addedthe robust option in EQS that corrects the χ2
and standard errors for non-normality(Satorra & Bentler, 1994).
The goodness-of-fit indices showed thatthe CFA model produced an acceptable fit tothe data, χ2 (181) = 253.13 (p < .00), χ2/df =1.40, TLI = .92, IFI = 94, CFI = .94, andRMSEA = .05. Inspection of the factor load-ings showed that each variable had a highlysignificant loading on the factorit was purported to measure, indi-cating satisfactory convergent va-lidity. In addition, the discrimi-nant validity of the measures wasalso good, since a one-factormodel produced a poor fit to thedata.
To further investigate the dis-criminant validity of our model,we compared the proposed eight-factor model to several modelswherein we combined conceptu-ally related constructs into onefactor. Specifically, we tested (a) aseven-factor model wherein wecollapsed the two forms of commitment (af-fective and continuance commitment) intoone factor, (b) a seven-factor model whereinwe collapsed the two antecedents of affectivecommitment (perceived shared values andknowledge-based trust) into one factor, and(c) a six-factor model wherein we collapsedthe three antecedents of knowledge-basedtrust (reputation, communication, and ap-proachability) into one factor. None of thesemodels provided a good fit to the data (seeTable I). These results provide evidence thatthe eight constructs measured were relativelydistinct from each other. On average, in theeight-factor model, the eight factors corre-lated .29.
Controlling for Common MethodVariance
In this study, all variables were measured onthe basis of a single survey at the same time
These results
provide evidence
that the eight
constructs
measured were
relatively distinct
from each other.
Human Resource Management DOI: 10.1002/hrm
09HRM47_3lievens 8/11/08 11:07 AM Page 569
570 HUMAN RESOURCE MANAGEMENT, Fall 2008
using a single source. Therefore, our resultsmight be prone to common method vari-ance. To statistically control for method vari-ance, we followed a structural equation mod-eling approach (i.e., controlling for theeffects of a single unmeasured latent methodfactor) recommended by Podsakoff, MacKen-zie, Lee, and Podsakoff (2003). This meantthat we specified a model that included theconstructs of our conceptual model and oneadditional method factor. All indicator vari-ables loaded on this method factor.
The fit statistics of this new model were χ2
(159) = 211.05 (p < .00), χ2/df =1.33, TLI = .95, IFI = .97, CFI = .97,and RMSEA = .04. Thus, the addi-tion of a method factor to the con-structs-only model significantlyimproved the fit of the model overthe constructs-only model, differ-ence in χ2 (22) = 42.08, p < .01. Yet,all construct factor loadings re-mained significant in the methodand construct factors model (thus,even when the method variancewas factored out). In addition, theconstruct factors on average ac-
counted for 46.2% of the variance, whereasthe method factor accounted for only 15.4%of the variance. So, the average method vari-ance is clearly lower than the average methodvariance (26.3%) reported in the large-scalereview of Podsakoff et al. (2003, p. 880).
In short, these analyses indicate thatmethod bias has some, but no substantial in-fluence on our data. However, in all our fur-ther analyses, we specify models with themethod factor included so that we ade-quately control for method bias.
Test of Model and Hypotheses
Given that the measurement models produceda good fit to the data, we tested the structuralmodel. In this model, structural relationshipsbetween the latent variables were specified (seeFigure 1). In case of one-item measures, wefixed the factor loading to 1.0 and the meas-urement error to 0.0. Note that on the basis ofprior research (Klaas et al., 1999), we also con-trolled for organizational size. However, for
clarity reasons, this measure was not includedin Figure 1. The fit of the structural model wasacceptable, χ2 (274) = 366.75 (p < .00), χ2/df =1.34, TLI = .91, IFI = .93, CFI = .93, and RMSEA= .05. Standardized parameter estimates for thecoefficients related to our hypotheses are pre-sented in Figure 2.
Our first set of hypotheses dealt with theeffects of the two forms of commitment onoutsourcing outcomes. Figure 2 shows thataffective commitment was significantly re-lated to outsourcing depth (.19, p < .05, R2 =.05) and frequency (.29, p < .01, R2 = .09).The coefficient between affective commit-ment and length of the outsourcing relation-ship was .13 (p < .10, R2 = .02). This resultprovides partial support for Hypothesis 1b.There was no support for Hypothesis 1a be-cause continuance commitment was not re-lated to outsourcing longevity.
Our second set of hypotheses was relatedto the bases of continuance commitment. Asshown in Figure 2, we found partial supportfor this set of hypotheses. The coefficientfrom specific investments to continuancecommitment toward HR vendors was signifi-cant (.37, p < .01, R2 = .14), lending supportto Hypothesis 2a. Yet, we found no supportfor Hypothesis 2b. The coefficient fromnumber of alternatives to continuance com-mitment was not significant.
Hypotheses 3a and 3b dealt with thebases of affective commitment. We foundsupport for Hypothesis 3a, as perceivedshared values (.34, p < .01, R2 = .18) was sig-nificantly linked to affective commitment.Knowledge-based trust was not significantlyrelated to affective commitment, providingno support for Hypothesis 3b.
Finally, we found partial support for ourset of hypotheses related to knowledge-basedtrust. Approachability was significantly re-lated to knowledge-based trust (.54, p < .01,R2 = .39), providing support to Hypothesis 4b.Reputation and communication were not sig-nificantly related to knowledge-based trust.
Test of Rival Models
Our theoretical model (Figure 1) is a fullymediated model because the antecedents of
…affective
commitment was
significantly related
to outsourcing
depth and
frequency…
Human Resource Management DOI: 10.1002/hrm
09HRM47_3lievens 8/11/08 11:07 AM Page 570
A Model of External Organizational Commitment in Human Resources Outsourcing 571
the two forms of commitment are hypothe-sized to have only indirect effects (throughcontinuance and affective commitment, re-spectively) on the relevant HR outsourcingoutcomes. Apart from this model, we testedtwo rival models that proposed different the-oretical relationships between the commit-ment antecedents, the two forms of commit-ment, and the outsourcing outcomes (seeTable II). To compare the fit of the models,we used χ2 difference tests.
A first rival model hypothesized only di-rect effects of the antecedents of the twoforms of commitment on the HR outsourc-ing outcomes. Although this rival modelprovided a relatively good fit, χ2 (274) =389.31 (p < .00), χ2/df = 1.42, TLI = .88, IFI =.91, CF1 = .91, RMSEA = .05, its fit was worsethan the fit of the theoretical model (fit in-dices, see above and Table II). This was alsoshown at the parameter level because noneof the coefficients associated with the directeffects proposed by this rival model weresignificant.
A second rival model was a partiallymediated model. Essentially, this secondrival model hypothesized that the relation-ship between the antecedents of the twoforms of commitment and the relevant HRoutsourcing outcomes was partially medi-ated by the two forms of commitment. So,in this model, we supplemented the indi-rect effects that were already specified inFigure 1 with direct effects from the an-tecedents of the two forms of commitmentto the relevant HR outsourcing outcomes.As our theoretical model (Figure 1) isnested in this rival model, it is possible tocompare these models to each other. The fitof this rival model was good, χ2 (266) =363.30 (p < .00), χ2/df = 1.36, TLI = .90, IFI= .93, CF1 = .92, RMSEA = .50. However, itdid not provide a significantly better fitthan our theoretical model, difference in χ2
(8) = 3.45, p = ns. Thus, adding direct ef-fects to our theoretical model did not leadto a significant improvement of fit. Indeed,at the parameter level, none of the direct
Human Resource Management DOI: 10.1002/hrm
FIGURE 2. Model of External Organizational Commitment in HR Outsourcing Relationships With Parameter Estimates
09HRM47_3lievens 8/11/08 11:07 AM Page 571
572 HUMAN RESOURCE MANAGEMENT, Fall 2008
Human Resource Management DOI: 10.1002/hrm
χχ2d
fχχ2
/df
TLI
IFI
CFI
RM
SE
A
Measu
rem
en
t m
od
els
Eig
ht-
fact
or
mo
del
25
3.13
181
1.40
.92
.94
.94
.05
Sev
en-f
acto
r m
od
el (
affe
ctiv
e an
d c
on
tin
uan
ce c
om
mit
men
t ar
e o
ne
fact
or)
343.
4318
81.
83.8
4.8
8.8
7.0
7
Sev
en-f
acto
r m
od
el (
two
an
tece
den
ts o
f af
fect
ive
com
mit
men
t ar
e o
ne
fact
or)
326.
9418
81.
74.8
6.8
9.8
8.0
7
Six
-fac
tor
mo
del
(th
ree
ante
ced
ents
of
kno
wle
dg
e-b
ased
tru
st a
re o
ne
fact
or)
448.
1019
42.
31.7
5.8
0.8
0.0
9
Nin
e-fa
cto
r m
od
el (
eig
ht-
fact
or
mo
del
plu
s o
ne
fact
or
den
oti
ng
co
mm
on
met
ho
d v
aria
nce
)21
1.05
159
1.33
.95
.97
.97
.04
Str
uctu
ral
mo
dels
Hyp
oth
esiz
ed m
od
el (
Fully
med
iate
d m
od
el, F
igu
re 1
)36
6.75
274
1.34
.91
.93
.93
.05
Dir
ect
effe
cts
mo
del
s38
9.31
274
1.42
.88
.91
.91
.05
Part
ially
med
iate
d m
od
el36
3.30
266
1.36
.90
.93
.92
.05
Not
e:T
LI =
Tu
cker
-Lew
is in
dex
; IFI
= B
olle
n’s
incr
emen
tal f
it in
dex
; CFI
= c
om
par
ativ
e fi
t in
dex
; RM
SE
A =
ro
ot
mea
n s
qu
are
erro
r o
f ap
pro
xim
atio
n.
TA
BL
E
II
Sum
mar
y of
Fit
Stat
istic
s of
Mea
sure
men
t and
Str
uctu
ral M
odel
s Te
sted
09HRM47_3lievens 8/11/08 11:07 AM Page 572
A Model of External Organizational Commitment in Human Resources Outsourcing 573
effects specified in this rival model weresignificant.
In sum, investigation of these two rivalmodels provides further support for the rela-tionships between the commitment an-tecedents, the two forms of commitment,and the outsourcing outcomes that werespecified in our theoretical model in Figure 1.
Discussion
The rapid growth of HR outsourcing exem-plifies that for many organizations the deci-sion to either “make” or “buy” HR activitiesswings toward the latter. Hence, it is pivotalthat researchers also shift their attentionfrom identifying conditions related to theinitial outsourcing decision to psychologicalfactors related to the continuity of the HRoutsourcing relationship. This study takes afirst step in this direction by examiningwhich factors are related to HR managers’commitment to continue the HR outsourc-ing relationship.
Our conceptual model distinguished be-tween continuance and affective commit-ment. These two forms of commitment aregrounded in different theoretical frame-works—namely, transaction-cost economicstheory and social exchange theory, respec-tively. One key conclusion of our study wasthat positive, affective-based motivations forcontinuing the outsourcing relationship (i.e.,the desire for a relationship to continue) wererelated to a qualitatively strengthened out-sourcing relationship instead of negative mo-tivations (i.e., constraints that keep it intact).In fact, affective commitment based on per-ceived value congruence was related to thedepth and frequency of outsourcing. Con-versely, continuance commitment based onspecific investments was unrelated to out-sourcing outcomes. Thus, dependency andconstraints that keep the HR outsourcing rela-tionship intact do not seem to be related tothe length of this relationship. Another keyconclusion was that an interindividualphenomenon (boundary spanners’ affectivecommitment to continue HR outsourcing)was significantly linked to interorganizationaloutcomes (outsourcing depth and frequency).
On a conceptual level, our results con-firm predictions made by social exchangetheory (Blau, 1964; Emerson, 1962; Gainey& Klaas, 2003). They are not in line with as-sumptions underlying transaction-cost eco-nomics theory. According to social exchangetheory, the development of affec-tive commitment and personalrelationships provides a mecha-nism other than dependency andswitching costs to prolong the re-lationship between partners. Ourfindings are also consistent withrecent research revealing that ashared vision can help partner or-ganizations to believe that theirgoals are cooperatively related sothat they do not feel threatenedby opportunistic behavior (Wong,Tjosvold, & Yu, 2005).
In fact, perceived shared val-ues were significantly related tothe development of an affectivebond with people from the part-ner firm. Probably, HR managersattach importance to the culturalfit of the HR vendor and the com-pany because a good cultural fitmight enable an HR service firmto factor in the company’s uniqueculture when selecting new per-sonnel for that specific company(Klaas et al., 2001). Conversely,perceived value incongruence inthe context of HR outsourcingmight disrupt organizational cul-ture (Greer et al., 1999).
As only six of the eleven hy-pothesized relationships were sig-nificant, it is important to discussin more detail some of the rela-tionships that were not sup-ported. First, an explanation forthe fact that continuance com-mitment was not related to out-sourcing outcomes might be thatin selection-related services, HRmanagers typically make human capital in-vestments instead of physical capital invest-ments (cf. Seabright et al., 1992). In addition,the outsourcing of recruitment and selection
The rapid growth of
HR outsourcing
exemplifies that for
many organizations
the decision to
either “make” or
“buy” HR activities
swings toward the
latter. Hence, it is
pivotal that
researchers also
shift their attention
from identifying
conditions related to
the initial
outsourcing
decision to
psychological
factors related to
the continuity of the
HR outsourcing
relationship.
Human Resource Management DOI: 10.1002/hrm
09HRM47_3lievens 8/11/08 11:07 AM Page 573
574 HUMAN RESOURCE MANAGEMENT, Fall 2008
activities often is classified as the outsourc-ing of repetitive, high-volume, and discreteservices (Adler, 2004). In other contexts,more support might be found for transac-tion-cost economics theory when technolog-ical investments play a more important role,leading to greater dependency. Clearly, fu-
ture research is needed to test ourhypotheses in HR outsourcing do-mains other than outsourcing se-lection and recruitment activitiesand in the context of the out-sourcing of the total HR function.
We believe that the nonsignif-icant relationship betweenknowledge-based trust and affec-tive commitment might be ex-plained by the cognitive-basedand short-term nature of thisform of trust (Gainey & Klaas,2003; Lewicki et al., 1998). Alongthese lines, it is generally ac-knowledged that socially basedforms of trust (also known asidentification-based trust) mightexert more powerful effects. Asnoted above, our study attests tothis, as our measure of perceivedshared values maps well ontosuch forms of identification-based trust. In addition, it wasnoteworthy that only approacha-bility was significantly related toknowledge-based trust. Clearly,the approachability of HR ven-dors is a very direct index ofwhether HR vendors “keep theirpromise” (as compared to reputa-tion and communication), whichunderscores the short-term andcognitive-based nature of knowl-edge-based trust.
A final nonsignificant findingthat deserves attention was that
our factual measure of the number of alter-natives was not related to continuance com-mitment. Given that there has been somecontroversy as to whether continuance com-mitment is unidimensional versus multidi-mensional (high sacrifice as one dimension
and perceived alternatives as the other; seeMcGee & Ford, 1987), future research mightinclude separate measures of these dimen-sions.
These results also suggest some research-based clues for HR vendors for ensuring thecontinuity of HR outsourcing relationships.One conclusion is that HR managers do notbelieve there is a link between continuancecommitment and key HR outsourcing out-comes. Thus, locking in HR managers withspecific investments (the switching cost ar-gument) does not seem to be a useful strat-egy for persuading HR managers to build aqualitatively strengthened HR outsourcingrelationship. Conversely, this study demon-strates that in the mind-set of HR managers,affective commitment is related to perceivedshared values. Hence, HR vendors might de-vote attention to this factor to show an af-fective bond with HR professionals of organ-izations. Similarly, HR vendors might ensurethat they are reachable and approachable, asHR managers report these factors are impor-tant signals for trusting HR vendors. In turn,these efforts might be associated with bene-ficial outcomes for HR vendors, such as alarger percentage of selection activities beingoutsourced to them and a higher frequencyof being given the opportunity to provide se-lection-related services.
In terms of future research, we need toexamine the perceptions of the other partyin the HR outsourcing relationship—namely, HR vendors. Insight in the agendaof HR vendors might reveal possible percep-tual differences with one of the HR man-agers. If possible, HR manager and HR ven-dor dyads might be studied. Such a dyadicapproach would also enable a study ofwhether some form of mutual commitmentprocess operates in the HR service relation-ship. For example, there might be an inter-action between affective/continuance com-mitment on the part of HR managers andaffective/continuance commitment on thepart of HR service providers. It also would beinteresting to examine whether individualswho made the initial decision are morecommitted to the outsourcing relationship
HR managers do not
believe there is a
link between
continuance
commitment and
key HR outsourcing
outcomes. Thus,
locking in HR
managers with
specific investments
(the switching cost
argument) does not
seem to be a useful
strategy for
persuading HR
managers to build a
qualitatively
strengthened HR
outsourcing
relationship.
Human Resource Management DOI: 10.1002/hrm
09HRM47_3lievens 8/11/08 11:07 AM Page 574
A Model of External Organizational Commitment in Human Resources Outsourcing 575
and/or whether this relationship changeswhen either the HR manager or the HR ven-dor (or possibly both) changes. A final in-triguing direction for future research is tofocus on other consequences of external or-ganizational commitment in HR outsourc-ing. One possibility is to examine the linkbetween external organizational commit-ment and objective measures of firm per-formance (e.g., profitability, service quality;cf. Seabright et al., 1992).
This study is not without limitations.First, we conducted the study in Belgium,which might impact on the generalizabilityof our findings. Generally, we believe that re-cruitment and selection practices are fairlysimilar in Belgium and the United States. Forexample, a comparative study of interna-tional selection practices did not revealmarked differences between the UnitedStates and Belgium (Ryan, McFarland, Baron,& Page, 1999). In addition, many multina-tionals and HR service providers have divi-sions in both the United States and Belgium.Nonetheless, recent reports show some re-gional differences in HR outsourcing (Dell,2004). Whereas 87% of executives at U.S.companies surveyed reported that they out-sourced key HR functions, these numberswere only 71% in Canada and 57% in Eu-rope. Hence, future research is needed to testour results in other populations, settings,and countries.
Second, only one key informant com-pleted the survey. Hence, use of multiple keyinformants might have provided more reli-able data. However, an assumption underly-ing the use of multiple informants is thatthey are equally knowledgeable and (in thecase of our study) equally committed (Kumaret al., 1993). Conceptually, this is at oddswith the commitment literature, becausecommitment is an individual-level phenom-enon (i.e., an individual is committed to acourse of action and/or entity), and hencethe assumption that boundary spannersshare the same level of commitment to agiven external organization might be unten-able. Another problem with using multipleinformants from the same firms in our study
was that it required identification of the HRvendor to be rated. Prior research has shownthat few respondents are willing to partici-pate in a study where a particular HR vendoris named and evaluated (Gainey & Klaas,2003).
Third, our results are based on cross-sec-tional self-reports to a survey. Both inde-pendent and outcome variables were gath-ered with the same survey. Although westatistically controlled for common methodvariance, the cross-sectional nature of ourstudy precludes ruling out causality. Finally,the dependent measures were sin-gle-item measures. Althoughthese measures asked about fac-tual information and were usedin prior studies in the marketingdomain (e.g., Ganesan, 1994; Jap& Ganesan, 2000; Kim & Frazier,1997), we acknowledge the provi-sional nature of these measures.The measurement error presentmight have contributed to thelack of significant relationshipsfound for some dependent meas-ures (e.g., length of outsourcing).
In conclusion, this study provides a firststep in examining factors related to commit-ment to continue HR outsourcing relation-ships. Affective commitment rather thancontinuance commitment was related to keyoutsourcing outcomes. In addition, affectivecommitment was primarily linked to per-ceived shared values. Future longitudinal re-search is needed to further explore the dy-namics of HR outsourcing relationships.
Acknowledgment
We would like to acknowledge Michael Harrisand Christian Vandenberghe for their valuablesuggestions on previous versions of this article.We would also like to thank Veroniek De Neve,Jessie Morre, Sarah Laureys, Liesbeth Roose, andBram Timperman for their help in collecting thedata. Portions of this manuscript were presentedat the Annual Conference of the Society of In-dustrial and Organizational Psychology,Chicago, IL (2004).
Future longitudinal
research is needed
to further explore
the dynamics of HR
outsourcing
relationships.
Human Resource Management DOI: 10.1002/hrm
09HRM47_3lievens 8/11/08 11:07 AM Page 575
576 HUMAN RESOURCE MANAGEMENT, Fall 2008
NOTE
1. Although the literature on organizational commit-ment often discusses normative commitment (i.e.,commitment based upon perceived obligation;Meyer & Herscovitch, 2001), this form of commit-ment was not included in our conceptual modelbecause our semistructured interviews with HRmanagers and HR vendors indicated that it wasnot relevant in the context of HR outsourcing. Ourdecision to exclude normative commitment is alsoconsistent with the literature on external organiza-tional commitment (see McElroy et al., 2001) andwith the marketing literature on commitment indistribution channel relationships (Bansal et al.,2004; Kim & Frazier, 1997).
2. The factual single-item measure of perceived alter-natives was not included.
REFERENCES
Adler, S. (2004, April). HR outsourcing: Trends andpatterns. In S. Adler (Chair), HR outsourcing: Therole of I-O psychologists. Symposium presentedat the 19th Annual Conference of the Society forIndustrial and Organizational Psychology,Chicago, IL.
Anderson, E., & Weitz, B. (1989). Determinants of con-tinuity in conventional industrial channel dyads.Marketing Science, 8, 310–323.
Anderson, J. C., & Narus, J. A. (1990). A model of dis-tributor firm and manufacturer firm working part-nerships. Journal of Marketing, 54, 42–58.
Bansal, H. S., Irving, G. P., & Taylor, S. F. (2004). Athree component model of customer commitmentto service providers. Journal of the Academy ofMarketing Science, 32, 234–250.
Barthélemy, J. (2003). The seven deadly sins of out-sourcing. Academy of Management Executive, 17,87–100.
Bentler, P. M. (1995). EQS: Structural equations programmanual. Encino, CA: Multivariate Software, Inc.
Blau, P. M. (1964). Exchange and power in social life.New York: Wiley.
Browne, M. W., & Cudeck, R. (1992). Alternative waysof assessing model fit. Sociological Methods andResearch, 21, 230–258.
Chen, M. J., Farh, J. L., & MacMillan, I. (1993). An ex-ploration of the expertness of outside informants.Academy of Management Journal, 36, 1614–1632.
Chen, Z. X., Aryee, S., & Lee, C. (2005). Test of a medi-ation model of perceived organizational support.Journal of Vocational Behavior, 66, 457–470.
Cohen, A., & Lowenberg, G. (1990). A reexaminationof the side-bet theory as applied to organizationalcommitment: a meta-analysis. Human Relations,43, 1015–1050.
Cook, K. S., & Emerson, R. M. (1978). Power, equity
Human Resource Management DOI: 10.1002/hrm
FILIP LIEVENS is a professor in the Department of Personnel Management and Work andOrganizational Psychology at Ghent University, Belgium. In 1999, he earned his PhD inwork and organizational psychology from the same university. His current research in-terests focus on selection procedures (e.g., assessment centers, situational judgmenttests) and organizational attractiveness. He has published in the Journal of Applied Psy-chology, Personnel Psychology, the Journal of Organizational Behavior, the Journal ofOccupational and Organizational Psychology, Applied Psychology: An International Re-view, and the International Journal of Selection and Assessment. In 2006, he receivedthe Distinguished Early Career Award from the Society of Industrial and OrganizationalPsychology.
WILFRIED DE CORTE is a professor in the Department of Data Analysis at Ghent Univer-sity, Belgium. He earned a PhD in work and organizational psychology from the sameuniversity. His current research focuses on integrating results from order statistics andoperations research within current psychometric theory to address selection decisionproblems. He has published in the Journal of Applied Psychology, Personnel Psychology,the Journal of Occupational and Organizational Psychology, the Journal of Applied Sta-tistics, Applied Psychological Measurement, and the British Journal of Mathematical andStatistical Psychology.
09HRM47_3lievens 8/11/08 11:07 AM Page 576
A Model of External Organizational Commitment in Human Resources Outsourcing 577
and commitment in exchange networks. AmericanSociological Review, 43, 721–739.
Cook, M. F. (1999). Outsourcing human resourcesfunctions. New York: American Management Asso-ciation.
Dell, D. (2004). HR outsourcing: Benefits, challengesand trends. New York: The Conference Board.
Emerson, R. M. (1962). Power dependence relations.American Sociological Review, 27, 692–703.
Fichman, M., & Levinthal, D. A. (1991). Honeymoonsand the liability of adolescence: A new perspectiveon duration dependence in social and organiza-tional relationships. Academy of Management Re-view, 16, 442–468.
Fombrun, C. J. (1996). Reputation: Realizing valuefrom the corporate image. Cambridge, MA: Har-vard Business School Press.
Gainey, T. W., & Klaas, B. S. (2003). The outsourcing oftraining and development: Factors impacting clientsatisfaction. Journal of Management, 29, 207–229.
Ganesan, S. (1994). Determinants of long-term orien-tation in buyer-seller relationships. Journal of Mar-keting, 58, 1–19.
Geyskens, I., Steenkamp, J. E. M., & Kumar, N. (2006).Make, buy, or ally? A meta-analysis of transactioncost theory. Academy of Management Journal, 49,519–543.
Gilley, K. M., & Rasheed, A. (2000). Making more bydoing less: an analysis of outsourcing and its ef-fects on firm performance. Journal of Manage-ment, 26, 763–790.
Greer, C. R., Youngblood, S. A., & Gray, D. A. (1999).Human resource outsourcing: The make or buy deci-sion. Academy of Management Executive, 13, 85–96.
Hoyle, R. (1995). Structural equation modeling: Con-cepts, issues, and applications. Thousand Oaks,CA: Sage.
Jap, S. D., & Ganesan, S. (2000). Control mechanismsand the relationship life cycle: implications forsafeguarding specific investments and developingcommitment. Journal of Marketing Research, 37,227–245.
Kim, K., & Frazier, G. L. (1997). On distributor commit-ment in industrial channels of distribution: A multi-component approach. Psychology & Marketing, 14,847–877.
Klaas, B. S., McClendon, J. A., & Gainey, T. W. (1999).HR outsourcing and its impact: The role of transac-tion costs. Personnel Psychology, 52, 113–136.
Klaas, B. S., McClendon, J. A., & Gainey, T. W. (2001).Outsourcing HR: The impact of organizational char-
acteristics. Human Resource Management, 40,125–138.
Klein, K. J., Palmer, S., & Conn, A. B. (2000). Inter-orga-nizational relationships: A multilevel perspective. InS. W. J. Kozlowski & K. J. Klein (Eds.), Multileveltheory, research, and methods in organizations:Foundations, extensions, and new directions (pp.267–307). San Francisco, CA: Jossey-Bass.
Kristof, A. L. (1996). Person-organization fit: An integra-tive review of its conceptualizations, measurement,and implications. Personnel Psychology, 49, 1–49.
Kumar, N., Stern, L. W., & Anderson, J. C. (1993). Con-ducting interorganizational research using key in-formants. Academy of Management Journal, 36,1633–1651.
LaBahn, D. W., & Kohli, C. (1997). Maintaining clientcommitment in advertising agency-client relation-ships. Industrial Marketing Management, 26,497–508.
Lepak, D. P., Bartol, K. M., & Gardner, S. (2004,May/June). Understanding the strategic motiva-tions for outsourcing HR activities. IHRIM Journal,pp. 29–38.
Lepak, D. P., & Snell, S. A. (1998). Virtual HR: Strategichuman resource management in the 21st century.Human Resource Management Review, 8, 24–59.
Lewicki, R. J., McAllister, D. J., & Bies, R. J. (1998).Trust and distrust: New relationships and realities.Academy of Management Review 23, 438–459.
Maurer, R., & Mobley, N. (1998). Outsourcing: Is it theHR department of the future? HR Focus, 75(11),9–10.
Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995).An integrative model of organizational trust. Acad-emy of Management Review, 20, 709–734.
McAllister, D. J. (1995). Affect- and cognition-basedtrust as foundations for interpersonal cooperationin organizations. Academy of Management Jour-nal, 38, 24–59.
McElroy, J. C., Morrow, P. C., & Laczniak, R. N. (2001).External organizational commitment. Human Re-source Management Review, 11, 237–256.
McGee, G. W., & Ford, R. C. (1987). Two or more? Di-mensions of organizational commitment: Reexami-nation of the affective and continuance commit-ment scales. Journal of Applied Psychology. 72,638–641.
Medsker, G. J., Williams, L. J., & Holahan, P. J. (1994).A review of current practices for evaluating causalmodels in organizational behavior and human re-sources management research. Journal of Man-agement, 20, 439–464.
Human Resource Management DOI: 10.1002/hrm
09HRM47_3lievens 8/11/08 11:07 AM Page 577
578 HUMAN RESOURCE MANAGEMENT, Fall 2008
Meyer, J. P., & Allen, N. J. (1984). Testing the “side-bettheory” of organizational commitment: Somemethodological consideration. Journal of AppliedPsychology, 69, 372–378.
Meyer, J. P., & Allen, N. J. (1997). Commitment in theworkplace. Thousand Oaks, CA: Sage.
Meyer, J. P., Allen, N. J., & Gellatly, I. R. (1990). Affec-tive and continuance commitment to organiza-tions: Evaluation of measures and analysis of con-current and time-lagged relations. Journal ofApplied Psychology, 75, 710–720.
Meyer, J. P., & Herscovitch, L. (2001). Commitment inthe workplace: Toward a general model. HumanResource Management Review, 11, 299–326.
Meyer, J. P., Paunonen, S. V., Gellatly, I. R., Goffin, R.D., & Jackson, D. N. (1989). Organizational commit-ment and job performance: It’s the nature of thecommitment that counts. Journal of Applied Psy-chology, 74, 152–156.
Meyer, J.P., Stanley, D. J., Herscovitch, L., & Topolnyt-sky, L. (2002). Affective, continuance, and norma-tive commitment to the organization: a meta-analy-sis of antecedents, correlates, and consequences.Journal of Vocational Behavior, 61, 20–52.
Moorman, C., Zaltman, G., & Deshpande, R. (1992).Relationships between providers and users of mar-ket research: The dynamics of trust within and be-tween organizations. Journal of Marketing Re-search, 39, 314–328.
Morgan, R. M., & Hunt, S. D. (1994). The commitment-trust theory of relationship marketing. Journal ofMarketing, 58, 20–38.
Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Pod-sakoff, N. P. (2003). Common method biases in be-havioral research: A critical review of the literatureand recommended remedies. Journal of AppliedPsychology, 88, 879–903.
Ring, P. S., & Van De Ven, A. H. (1994). Developmentalprocesses of cooperative interorganizational rela-tionships. Academy of Management Review, 19,90–118.
Rogelberg, S. G., & Stanton, J. M. (2007). Under-standing and dealing with organizational surveynonresponse. Organizational Research Methods,10, 195–209.
Ryan, A. M., McFarland, L., Baron, H., & Page, R.(1999). An international look at selection practices:Nation and culture as explanations for variabilityin practice. Personnel Psychology, 52, 359–391.
Satorra, A., & Bentler, P. M. (1994). Corrections to teststatistics and standard errors in covariance structureanalysis. In A. Von Eye & C. C. Clogg (Eds.), Latent
variables analysis: Applications for developmentalresearch (pp. 399–419). Thousand Oaks, CA: Sage.
Schneider, B. (1987). The people make the place. Per-sonnel Psychology, 40, 437–453.
Schumacker, R. E., & Lomax, R. G. (1996). A begin-ner’s guide to structural equation modeling. Mah-wah, NJ: Erlbaum.
Seabright, M. A., Levinthal, D. A., & Fichman, M.(1992). Role of individual attachments in the disso-lution of interorganizational relationships. Acad-emy of Management Journal, 35, 122–160.
Sheppard, B. H., & Tuchinsky, M. (1996). Interfirm rela-tionships: A grammar of pairs. Research in Organi-zational Behavior, 18, 331–373.
Siders, M. A., George, G., & Dharwadkar, R. (2001).The relationship of internal and external commit-ment foci to objective measures of job perform-ance. Academy of Management Journal, 44,570–579.
Siguaw, J. A., Simpson, P. M., & Baker, T. L. (1998). Ef-fects of supplier market orientation on distributormarket orientation and the channel relationship:The distributor perspective. Journal of Marketing,62, 99–111.
Watson, G., & Papamarcos, S. (2002). Social capitaland organizational commitment. Journal of Busi-ness and Psychology, 16, 537–552.
Williamson, O. E. (1996). The mechanisms of gover-nance. New York: Oxford University Press.
Wong, A., Tjosvold, D., & Yu, Z. (2005). Organizationalpartnerships in China: Self-interest, goal interde-pendence, and opportunism. Journal of AppliedPsychology, 90, 782–791.
Yankee Group. (2005). New Yankee Group research re-veals enterprise adoption of HR BPO acceleratingat rampant pace. Retrieved September 16, 2005,from http://www.yankeegroup.com/public/news_re-leases/news_release_detail.jsp?ID=PressReleases/news_04042005_HRBPO.htm
Yilmaz, C., Sezen, B., & Ozdemir, O. (2005). Joint andinteractive effects of trust and (inter) dependenceon relational behaviors in long-term channel dyads.Industrial Marketing Management, 34, 235–248.
Young-Ybarra, C., & Wiersema, M. (1999). Strategicflexibility in information technology alliances: Theinfluence of transaction cost economics and socialexchange theory. Organization Science, 10,439–459.
Zaheer, A., McEvily, B., & Perrone, V. (1998). Doestrust matter? Exploring the role of interorganiza-tional and interpersonal trust on performance. Or-ganization Science, 9, 141–159.
Human Resource Management DOI: 10.1002/hrm
09HRM47_3lievens 8/11/08 11:07 AM Page 578
A Model of External Organizational Commitment in Human Resources Outsourcing 579
Human Resource Management DOI: 10.1002/hrm
Perceived Shared Values
• We agree with this HR service provider’s philosophy about managing people.• People from this HR service provider share our values about managing people.
Communication
• We always keep each other informed about events or changes that may affect the other party.• It is expected that any information that might help the other party will be provided to them.• It is expected that proprietary information will be shared if it can help the other party.
Approachability
• The key people from this HR service provider are always very easy to reach.• The key people from this HR service provider are often too busy to respond to our requests (Rev).• The people from this HR service provider are always ready for answering our questions.
Reputation for Fairness
• Most people think that this HR service provider has a reputation of being fair.• This HR service provider is held in high regard.
Knowledge-Based Trust
• Promises made by people from this HR service provider are reliable.• People from this HR service provider do not make false claims. • If problems arise, people from this HR service are honest about the problems.
Specific Investments
• We have made significant investments (e.g., in selection procedures, software installed) dedicatedto our relationship with this HR service provider.
• If we switched to a competing HR service provider, we would lose a lot of the specific investments(e.g., in selection procedures, software installed) made.
• We have invested substantially in personnel dedicated to this HR service provider.
Perceived Number of Available Alternatives
• How many other HR service providers are you considering to work with?
Continuance Commitment
• I feel that I have too few options to consider leaving this HR service provider.• One of the few serious consequences associated with quitting working with this HR service
provider would be the scarcity of available alternatives.• Our HR activities would be disrupted if we decided to quit working with this HR service provider.
Affective Commitment
• I do not feel “emotionally attached” to this HR service provider (Rev).• This HR service provider has a great deal of personal meaning for me. • I do not feel like “part of the family” of this HR service provider (Rev).
Length of the Outsourcing Relationship
• How long have you outsourced (or have you been outsourcing) recruitment/selection activities tothis HR service provider?
Depth of Outsourcing Within This Specific HR Outsourcing Relationship
• What percentage of your recruitment/selection activities is currently outsourced to this HR serviceprovider?
Frequency of Outsourcing Within This Specific HR Outsourcing Relationship
• How many times have you outsourced recruitment/selection activities to this HR service provider(since you set up your outsourcing relationship with this HR service provider)?
Overview of Survey ItemsA P P E N D I X
09HRM47_3lievens 8/11/08 11:07 AM Page 579