The relationships betweennetwork commitment, itsantecedents and network

performanceNicholas Clarke

School of Management, University of Southampton, Southampton, UK

Abstract

Purpose – The purpose of this paper is to put forward a theoretical model grounded in the literaturethat identifies a number of antecedent conditions associated with network commitment.

Design/methodology/approach – Based upon empirical data gathered from formal healthcarenetworks in the UK, findings are presented that suggest different forms of commitment withinnetworks may have opposite effects on network performance.

Findings – This paper argues that commitment may play a significant role specifically associatedwith determining performance outcomes in networks. In so doing a theoretical model is put forwarddrawn from the literature to suggest possible antecedent conditions associated with commitment at theinter-organisational level.

Originality/value – To date few studies have been undertaken to examine the antecedents ofcommitment within networks. More importantly, the recognition of commitment as amulti-dimensional construct within networks has yet to have informed understanding of the rolecommitment may play at the inter-organisational level.

Keywords Medical care, Networking, United Kingdom

Paper type Research paper

IntroductionOrganisations in both the public and private sectors are increasingly entering intonetwork relationships with the recognition that to operate effectively in today’s businessclimate often requires the knowledge and expertise held by others (Hiscock and Pearson,1999; Mankin and Cohen, 2003; Nohria and Eccles, 1998). Within healthcare too,participation in networks is seen as a solution to meeting the multi-faceted health andsocial care needs of communities that are often too complex for any one organisation tosolve alone (Brown et al., 1992; Loxley, 1997; Ovretveit, 1993). However achievingeffective performance within network structures may often pose significant challengesfor participants, sometimes resulting in either poor performance outcomes or even abreakdown in the collaborative relationship (Himmelman, 1992; Medcof, 1997; Webb,1991). Specifically within the healthcare field, a major concern has been how to improvecollaboration between organisations that are part of both formal and informal healthcarenetworks in order to achieve specific pre-determined policy outcomes (Scott andThurston, 2004). A search for appropriate managerial responses to improvecollaboration in networks has therefore emerged as a chief concern both within socialpolicy (Lasker et al., 2001; Sullivan et al., 2002), as well as more widely within the privatesector (Anderson, 1995; Dyer and Singh, 1998). This has led many writers to suggest that

The current issue and full text archive of this journal is available at

www.emeraldinsight.com/0025-1747.htm

Networkcommitment

1183

Received March 2006Revised July 2006

Accepted July 2006

Management DecisionVol. 44 No. 9, 2006

pp. 1183-1205q Emerald Group Publishing Limited

0025-1747DOI 10.1108/00251740610707677

we need to know far more about the range of potential outcomes associated withcollaboration, as well as the associated antecedents and dynamics (Boddy et al., 2000;Hardy et al., 2003). Importantly in this respect, commitment, has been proposed as a chiefmeans through which effective performance in networks can be maximised (Clarke,2005; Gray, 1999; Huxham, 1996). It therefore follows that if we can gain a greaterunderstanding of how commitment within networks might be fostered, we may be in abetter position for guiding management decisions and specific change interventions inthis area. This necessitates a focussed effort in theorising relationships between keyvariables involved in the collaborative process. This paper aims to contribute to effortstowards theory building in this area through positing that one of the chief mechanismsthrough which effective performance outcomes are achieved in networks is throughbringing about a range of antecedent conditions which impact on network performance.These antecedent conditions are suggested to be mediated through the concept ofnetwork commitment. Network commitment is suggested here as a co-ordinatingmechanism to compensate for the lack of behavioural control processes often foundwithin such organising structures. The article begins by providing a theoretical rationalto support this key proposition and then details the findings from a study that sought totest and evaluate a proposed theoretical model.

Commitment within networksEarly work on the notion of commitment within the field of organisational behaviourinitially viewed organisational commitment as a uni-dimensional construct thatrepresented the strength of identification or involvement an individual has with theorganisation, reflecting an underlying bond or attachment (Mowday et al., 1982). Itsimportance has been recognised with research implicating organisational commitmentas being related to turnover and absenteeism (Mobley, 1982; Mowday et al., 1982) andlatterly examining its links with performance (Mathieu and Zajac, 1990; Meyer andAllen, 1997). Considerable progress over the past 25 years however has led to advancesin both our understanding and conceptualisation of the commitment construct. In recentyears in particular, one of the more critical advances has been the recognition of multiplecommitments possessed by individuals within organisations. As early as 1985, Reichers(1985) suggested that individuals are likely to form a range of multiple commitments todifferent entities. Within the context of healthcare for example, Baruch andWinkelmann-Gleed (2002) demonstrated that managers and other staff within aBritish healthcare organisation possessed a variety of commitments of varying foci,each with differing antecedents and relationships including organisational, career,workgroup and occupational. Furthermore, McElroy et al. (2001) have also proposed theidea of external organisational commitment, which describes the commitment that anemployee of one organisation has for another organisation. In so doing they suggesttaking:

The idea of commitment toward a specific constituent and extending it to the forms ofcommitment beyond the boundary of a particular organisation . . . commitment to multipleorganisational targets is commonplace in many business settings, especially whereorganisational representatives serve as boundary spanners with other client organisations(p. 238).

Although initially finding particular application with the emergence of relationshipmarketing (Gundlach et al., 1995; Morgan and Hunt, 1994), the concept of external

MD44,9

1184

forms of commitment is now increasingly being recognised as having an importantrole to play at the inter-organisational level. Specifically within the context of differingforms of inter-organisational networks (Dyer and Singh, 1998; Kim and Frazier, 1997).In the field of R&D collaboration for example, commitment has been identified as a keysuccess factor upon which collaboration is dependent (Hagedorn et al., 2000; Nummela,2003; Parker, 2000). Findings from a number of studies are therefore increasinglysuggesting a prominent role for commitment as a primary force influencingcollaboration (Morgan and Hunt, 1994; Sarkar et al., 1998). To date though, much of theresearch examining commitment in networks remains embryonic, and for the mostpart has not been taken up to any significant extent within the organisational sciencefield itself, instead being confined mainly to the business or strategy arenas. A majorlimitation of much work to date is that we still know very little regarding the particularantecedent conditions that give rise to commitment in networks, nor whether these arecommon across different network contingencies.

A further limitation of previous research examining the role of commitment at theinter-organisational level is that it has to date, with few exceptions (i.e. Cullen et al.,1995; Halinen, 1997) defined it rather narrowly and tended to treat it as aunidimensional construct (Sarkar et al., 1998). Yet within the commitment literaturemore widely, commitment is now increasingly accepted as a multi-dimensionalconstruct reflecting three independent and distinguishable mind-sets or psychologicalstates (Meyer and Herscovitch, 2001). In relation to organisational commitment, thesehave been labelled as affective commitment (indicating the degree of emotionalattachment an individual has to the organisation), continuance commitment (reflectingthe strength of intention of the individual to remain with the organisation due to eithercosts of leaving or lack of alternatives on offer) and normative commitment (associatedwith a felt obligation to remain with the organisation). Commitment is therefore seen asa measure of varying degrees of all three mind-sets (Allen and Meyer, 1990). PreviouslyMeyer et al. (1993) have found evidence for the generalisability of the three-componentmodel of commitment beyond just organisational commitment. More latterly, McElroyet al. (2001) have also suggested that in relation to forms of external organisationalcommitment, these same sub-dimensions may well also apply. Given thesedevelopments, network commitment is perhaps better viewed as a psychologicalstate or force that directs individuals to adopt behaviours consistent with attainingcollaborative outcomes, comprising three independent mind-sets of affective,continuance and normative commitment. Each of these three different forms ofcommitment are characterised very differently. Affective network commitment isbased upon an emotional attachment to the network. Continuance commitment isbased upon a mind-set of being “locked-in” to the network, due to a lack of alternativesor perceived costs of leaving. While normative commitment is presumed to be basedupon a felt obligation to participate in the network.

The relationship between network commitment and performance

P1. Network commitment will be positively related to network performance.

Although to date research on the commitment-performance relationship at theorganisational level remains equivocal (Mowday, 1999; Putterhill and Rohrer, 1995), itis also postulated here that there are behavioural/performance implications arising

Networkcommitment

1185

from network commitment. Specifically, that network commitment will be related tonetwork performance (i.e. positive collaborative outcomes associated with a particularcollaborative relationship). The rationale for this is fairly straightforward. Wherenetwork members develop commitment to the network they are more likely to exertgreater efforts to work towards the goals of the network and exert greater energies tosolve particular relationship problems where they arise. Commitment should thereforesupport increased co-operative behaviours among network members leading to agreater likelihood of improved performance outcomes. From an economics perspective,the time and costs associated with recurrent disputes, posturing and renegotiationsthat take time and energy away from the real business of the network are likely to besignificantly reduced when commitment within the network is high (Milgrom andRoberts, 1992). Supporting this relationship has been a number of theoretical andempirical studies that suggest that commitment may well be a key factor in generatingfar better outcomes at the inter-organisational level (Badaracco, 1991; Ohmae, 1989;Sarkar et al., 1998). It should therefore be expected that a positive relationship shouldbe found between network commitment and network performance.

Theorising the antecedents of network commitment

P2. A belief that (1) network members are dependent upon each other in order toachieve their own goals (mutual interdependence) and (2) more can be accruedfor individual network members by participating in the network rather thanacting alone (mutual gain) will be positive associated with networkcommitment.

McElroy et al. (2001) have previously suggested that Emerson’s (1962) notion ofpower-dependence offers a basis for understanding the development of different formsof external commitment, where dependence is affected by the extent to which arelationship is characterised by factors such as uncertainty, substitutability andcentrality. Here characteristics associated with the commitment foci that increase anindividual’s dependence on it are likely to enhance the level of commitment felttowards that entity. This is achieved through their capacity to diminish the likelihoodof substituability and increase centrality. Put more simply, commitment is likely to beincreased when there are factors that make replacing the relationship with anotherpartnership a less desirable prospect or where the attraction between partners can beincreased. This being the case, mutual interdependencies between network membersand a belief that benefits will accrue to all parties should therefore serve as keyantecedents of commitment within networks, and there is some support for this fromprevious research (Dwyer et al., 1987; Holm et al., 1999; Kumar et al., 1995; Sarkar et al.,1998). Although sharing some similarity, these conditions are distinguished in thatmutual interdependence refers to the perceived need for the resources that a partnermay bring to the relationship whereas mutual gain refers to the perceived businessoutcomes as a result of the relationship.

P3. Both (3) shared values and (4) goal congruence within the network will bepositively associated with network commitment.

Emerson’s (1962) theory of power-dependence would suggest then that the morefavourably the entity (or partnership) is perceived by the individual, the stronger that

MD44,9

1186

commitment is likely to be as a result of increasing the centrality of that dependence tothe network. It would seem reasonable to suppose that factors likely to strengthen thisfavourable perception should also be associated with the development of networkcommitment (Lawler and Yoon, 1996). Elsewhere these features have also been referredto as increasing the compatibility between network members that results in enhancedcommitment (Madhok, 1995; Sarkar et al., 1998). When network members share similarvalues for example (Dwyer et al., 1987; Heide and John, 1998; Ring and Van de Ven,1992) and there is congruence between the network member’s goals and those of thenetwork (Frazier et al., 1988; Lewis, 1990; Root, 1988), the level of uncertainty betweennetwork members is likely to be diminished. According to Pennings (1970), workrelated values can be defined as constellations of attitudes which are used to evaluatethe work environment whereas a goal is what is the object or aim of an action (Lockeand Latham, 1990). In this sense shared values can be thought of as distinct fromshared goals or goal congruence.

P4. Confidence in network organisations (5) will be positively associated withnetwork commitment.

A number of authors have also stressed the importance of confidence in thecompetence of network members as a key factor in facilitating collaboration. Das andTeng (1998) for example, have argued that low levels of confidence amongst networkmembers promotes a lack of co-operation and encourages partners to behaveopportunistically or act out of self-interest. In developing this concept, they havesuggested that both trust and control are inter-related constructs underpinningconfidence in partner co-operation, and that without such confidence networks sufferdeleterious effects. The concept of possessing confidence in partners within a networkwould therefore seem to be associated with increasing the strength of attachment to thenetwork, and specifically commitment (Nummela, 2003).

P5. Involvement in decision-making (6), effective conflict resolution mechanisms(7) within the network and role clarity (8) will be positively associated withnetwork commitment.

Based on the work of Lawler and Yoon (1996) and their theory of relational cohesion inexchange relationships, equal power is likely to increase the propensity of mutualexchange in a network which in turn should lead to positive emotions that result inenhancing commitment. Elsewhere both the literature on the antecedents oforganisational commitment (Mathieu and Zajac, 1990; Meyer and Allen, 1988) aswell as that on inter-organisational collaboration provides support for the importanceof shared power and involvement in decision-making as key determinants ofcommitment (Bucklin and Sengupta, 1993; Huxham and Vangen, 2000a). At theinter-organisational level, reciprocal commitment between network members has alsobeen suggested as arising from mitigating conflict (Yan and Gray, 1994), and again thishas its basis in the literature relating to the balance of power and its effect on conflictwithin exchange-based relationships (Emerson, 1962). It would therefore seemreasonable to propose that effective conflict resolution is a necessary activity thatfacilitates the emergence of network commitment (Sayeed, 2001). Roles clearly definedbetween members participating in a network has also been suggested as contributingto commitment through diminishing potential conflict and reducing uncertainty and

Networkcommitment

1187

ambiguity, as well as clarifying mutual expectations (Sarkar et al., 1998). In so doingrole clarity should therefore diminish uncertainty and increase centrality resulting ingreater commitment.

P6. Effective performance feedback (9) within the network will be positivelyassociated with network commitment.

Considerable effort is required to sustain a network over time, which is at risk ofsimply grinding down without a capacity for ongoing monitoring and review to assistregeneration (Sink, 1991; Ring and Van de Ven, 1994). Sink (1991) for example,suggests that in order to avoid atrophy or death, networks must establish “evaluativecapacity” whereby evaluative processes underpin remedial strategies for the ongoingsustenance of inter-organisational relationships. Effective performance feedbackmechanisms should therefore generate processual and outcome data, which can then befeedback to stakeholders regarding task achievement and the extent to whichpartnership objectives have been achieved. This gives rise to the final antecedentcondition that is posited here to affect network commitment:

MethodologySampling and respondentsIn order to test the above ten propositions, data was collected from all 61 statutorydrug misuse networks in England, responsible for formulating and implementingregional drug misuse prevention and treatment plans across the country. Each of thesenetworks consists of senior representatives (usually the director, CEO or equivalent)from key community organisations involved in the delivery of drug misuse orprevention services (such as the health service, social services, police, probation, and arange of community drug misuse services). One of these representatives (normallyfrom the statutory sector) also takes on the role of chairperson of the network. Inaddition, each network possesses a co-ordinator who works closely with thechairperson and other network members to facilitate joint working betweenorganisations in the network in order to meet set performance targets. Aquestionnaire to collect data on commitment, antecedent conditions, andcollaborative outcomes (network performance) was sent by post to 122 members(chairs and co-ordinators from each network) of the 61 healthcare networks and a totalof 99 questionnaires were completed and returned, a response rate of 81 percent. A totalof 57 (56 per cent) of respondents were chairpersons of these networks, while 42 (44 percent) were co-ordinators. Females comprised 54 per cent of respondents, and theaverage age was 39. Participation in the study was voluntary, and an accompanyingletter explained the aims of the study, contact details for further information andguaranteed respondent anonymity. A pilot study of the questionnaire amongst a groupof 22 drug service professionals not taking part in the research did not indicate anyproblems with either the content, length or instructions for completion.

In previous studies examining measures of inter-organisational commitment, singlemeasures of commitment often have been obtained from an individual memberrepresenting a particular network or dyad relationship (Daniel et al., 2002; Sarkar et al.,1998; Walter and Ritter, 2003). Sarkar et al. (1998) for example, utilisedself-administered questionnaires to collect data from 129 executives representingconstruction projects that involved collaborative ventures with foreign partners. More

MD44,9

1188

recently, Daniel et al. (2002) also adopted a similar approach in collecting data fromrepresentatives of 58 research centres who had responsibility for maintainingcollaborative networks. Typically then in previous studies, one senior networkrepresentative has often only provided a measure of network commitment. In thisstudy however measures of network commitment were obtained from bothchairpersons and co-ordinators from each of the networks. The chairpersons andco-ordinators from each network were considered to possess a more in-depthunderstanding of the relationships amongst network members as well as antecedentconditions since their role required them to liase on a more regular basis with allorganisations in the network. Both these key individuals have responsibility for tryingto bring about the type of conditions conducive to collaboration and are believed topossess a more strategic view concerning network functioning and “health” (DrugStrategy Directorate, 2004). Importantly however a number of writers have suggestedthat it is the commitment of leadership roles such as these that are of pivotalsignificance in determining collaborative outcomes (Huxham and Vangen, 2000b;Mulroy and Shay, 1998; Purdue and Razzique, 1999; Scott and Thurston, 2004).

A total of 42 dyads consisting of chairpersons and co-ordinators from the samenetwork were obtained. In order to determine the extent of agreement between scoreson commitment and antecedents correlation scores were undertaken for all antecedentvariables, measures of antecedent and continuance commitment and networkperformance scores. The range of r values obtained for each dyad is shown for eachvariable in Table I. These show a high level of correlation on all antecedent conditionsapart from confidence in network organisations (r ranges from 0.58 to 0.68 for alldyads) and Involvement in decision making (r ranges from 0.57 to 0.64 for all dyads)which suggests that items on these particular variables are somewhat less reliablemeasures of these variables compared to measures of the other variables obtained.However there was a good level of agreement on scores obtained on all other variablessuggesting some consistency in beliefs regarding antecedent conditions and

Variable Pearson’s r range

AntecedentsMutual interdependence 0.69-0.80Mutual gain 0.67-0.82Shared values 0.60-0.73Goal congruence 0.72-0.76Confidence in network organisations 0.58-0.62Conflict resolution 0.72-0.88Role clarity 0.74-0.78Involvement in decision making 0.57-0.64Effective performance feedback 0.68-0.76

Network commitmentAffective commitment 0.69-0.80Continuance commitment 0.68-0.78

Network performance 0.76-0.92

Note: n ¼ 42 dyads

Table I.Range (min, max) ofcorrelations between

chairperson andcoordinator scores from

same network onantecedents, commitment

and performancemeasures

Networkcommitment

1189

commitment in the network between these two individuals. On this basis one mightreasonably accept these measures as significant indicators of network commitment,albeit partial measures (Nummela, 2003).

Development of measuresAntecedent conditions. Nine scales containing three items each, posited as theantecedent conditions of network commitment were generated based on previousmeasures used in the literature, and in two instances devised by the author (appendix).Items were amended in a number of instances to reflect the network as the focushowever without altering the underlying meaning. Items were randomly distributed inthe questionnaire and respondents asked to indicate on a seven-point likert scale(1 ¼ strongly disagree, 7 ¼ strongly agree) the extent to which the posited antecedentconditions were present in their network. Items for the mutual interdependence (alpha0.65), and goal congruence (0.64) scales were adapted from those used by Sarkar et al.(1998). The scale for shared values was based on that previously used by Caldwell et al.(1990) (alpha 0.68). Items for the role clarity scale were based on those previously usedby Rizzo et al. (1970) (alpha 0.68) and those for measuring Involvement indecision-making were adapted from the scale used by Morris and Koch (1979) (alpha0.82); items for the performance feedback scale were derived from those developed byHimmelman (1992) (alpha 0.80). The effective conflict resolution scale was derived fromitems from Rahim’s (1983) measure of organisational conflict. The two remainingscales, mutual gain (alpha 0.79) and confidence in network members (alpha 0.78), weredeveloped by the author. Alpha values of 0.70 are considered acceptable with 0.60acceptable for new scales (Churchill, 1979). Reliability scores for both mutualinterdependence and goal congruence were considered satisfactory if somewhatmodest compared to the other scales and this may have been due to changes in thewording of the scales in adapting to the study.

Network commitmentScale development and construction. Measures of affective, continuance and normativecommitment were initially included in the study. Items for inclusion in the threecommitment scales were based on items generated from a focus group held with 22healthcare professionals who had experience of working in these networks (althoughdid not take any further part in the study), and items initially derived from a previousmeasure of organisational commitment (Allen and Meyer, 1990). This resulted in a poolof 17 items. The order of the items on the questionnaire was random and four questions(2, 3, 11 and 15) were reverse scored (Table II). All 17-items comprising these threescales were subjected to a factor analysis (principal factor analysis), extracted androtated to a varimax criterion. Six factors were subsequently extracted witheigen-values greater than 1, although an examination of the scree point suggested thatfive factors were the most significant, accounting for 17.5, 13.5, 11.9, 8.3 and 7.3 percent of the total variance, respectively (Table II). This was compared with the results ofa forced factor analysis where the items in the three scales were forced to rotate to theeefactors corresponding with the three scales. This yielded three factors witheigen-values of 2.97, 2.30 and 2.02 accounting for 17.5, 13.5 and 11.9 per cent of thevariance, respectively. A comparison was made between these two results to informjudgements concerning which factors to retain as scale items. Specifically, items for

MD44,9

1190

each factor were retained (1) if loadings were greater than 0.40, (2) did not load ontomore than one factor, and (3) if items were found to be present in the same factor inboth factor analyses. This resulted in five items being retained for factor 1 (1.Collaborating with other organisations in this network has a great deal of meaning tome. 2. I enjoy discussing this network with people outside it. 3. Not participating in thisnetwork would make a difference to the success of the network. 4. Participating in thisnetwork makes a difference to the local community. 5. I believe that collaborationbetween organisations in the network will improve outcomes for service users)

Factor 1 Factor 2 Factor 3 Factor 4 Factor 5

Affective commitment scale1. Collaborating with other organisations in this

network has a great deal of meaning to mea 69 11 205 38 132. I do not feel emotionally attached to this

network 01 201 27 79 083. I do not feel a sense of belonging to this network 38 226 249 23 2144. I feel as if this network’s problems are my own 31 28 252 13 305. I enjoy discussing this network with people

outside ita 50 13 37 01 2556. Not participating in this network would make a

difference to the success of the networka 55 229 17 216 347. I believe that collaborating in this network is

important if services are to improve 61 233 226 238 2058. Participating in this network makes a difference

to the local communitya 62 205 20 14 2289. I believe that collaboration between

organisations in the network will improveoutcomes for service usersa 68 220 29 211 05

Continuance commitment scale10. It would be very difficult to leave this network

even if I wanted toa 31 47 218 207 0011. It wouldn’t be too costly for me to leave this

network nowa 14 67 222 205 23912. Right now participating in this network is a

matter of necessity as much as desire 25 04 70 07 20313. I feel I have too few options to consider leaving

this networka 13 67 18 220 0514. One of the few serious consequences of leaving

this network would be scarcity of availablealternativesa 06 66 00 225 201

Normative commitment scale15. I do not think that collaborating in this network

is necessarily in itself a good thing 42 66 200 225 20116. I feel a sense of moral obligation to participate

in this network 31 47 207 20 5317. The only reason for participating is due to

policy requirements 01 02 59 235 38

Note: aItems retained in each of the scales

Table II.Varimax rotated factor

matrix based oncorrelations among the

items of the affective,continuance and

normative commitmentscales

Networkcommitment

1191

corresponding to the affective commitment scale. Four items were retained for factor 2,corresponding to the scale for continuance commitment:

(1) It would be very difficult to leave this network even if I wanted to.

(2) It would not be too costly for me to leave this network now.

(3) I feel I have too few options to consider leaving this network.

(4) One of the few serious consequences of leaving this network would be scarcityof available alternatives.

Items pertaining to the normative commitment scale did not meet the above criteriaand as a result the scale was dropped from any further analyses in the study. The tworemaining scales had reliability scores (cronbach’s alpha) of 0.72 (affectivecommitment) and 0.71 (continuance commitment).

Network performanceA scale containing three items relating to specific Health Department performanceindicators for these networks were used to measure network performance (appendix).From 2003, these healthcare networks are required to audit and submit data to theDepartment of Health on the extent to which they achieve government set performancetargets. Annual monitoring data collected by the network requires that the networkindicate the extent to which they have achieved these performance targets.Respondents were therefore asked to indicate on a nine-point likert scale the extentto which each of the three outcomes had been achieved ranging from 1 ¼ not at all to9 ¼ comprehensively, based on their annual monitoring data (which was not publiclyavailable). Although performance outcomes are therefore ostensibly self-reportmeasures, they are informed by the objective performance data previously collected bythe network. The particularly high levels of agreement obtained on networkperformance scores between dyads in Table I would appear to lend some support to theextent to which these scores were based upon objective annual performance data thatwas available to these key members within the network.

ResultsThere was no statistical difference between network chairpersons and co-ordinatorson either forms of commitment, antecedents or on network performance scores.Table III presents the means, standard deviations and intercorrelations of thevariables in the study. The results support the view that both affective andcontinuance commitment are independent dimensions of the overall networkcommitment construct. As expected, both these forms of commitment were notsignificantly correlated and are differentially related to the posited antecedentvariables. Regression analyses were used to investigate the relationships betweenthe variables, using the two forms of commitment and network performance asdependent variables at the first, second and third set of analyses respectively. Thefirst set of analyses examined relationships between each of the antecedentconditions and affective commitment. The second set of analyses examinedrelationships between each of the antecedent conditions and continuancecommitment. The final set of analyses examined the influence of all antecedentconditions and both forms of commitment on network performance. The results in

MD44,9

1192

MS

DA

lph

a1

23

45

67

89

1011

12

1.C

onfi

den

ce15

.04

3.13

0.78

–2.

Con

flic

tre

solu

tion

14.7

73.

070.

770.

67*

*–

3.F

eed

bac

k13

.95

3.63

0.80

0.56

**

0.41

**

–4.

Mu

tual

gai

n16

.61

3.29

0.79

0.56

**

0.46

**

0.28

*–

5.G

oal

con

gru

ence

15.0

52.

820.

650.

72*

*0.

64*

*0.

44*

*0.

42*

*–

6.In

terd

epen

den

ce15

.82

3.19

0.58

0.26

*0.

200.

41*

*0.

56*

*0.

27*

–7.

Inv

olv

emen

t16

.99

2.66

0.82

0.26

*0.

51*

*0.

37*

*0.

65*

*0.

64*

*0.

38*

*–

8.R

ole

clar

ity

15.5

42.

790.

590.

56*

*0.

40*

*0.

50*

*0.

38*

*0.

59*

*0.

22*

0.51

**

–9.

Sh

ared

val

ues

14.1

93.

260.

700.

80*

*0.

69*

*0.

37*

*0.

45*

*0.

70*

*0.

160.

52*

*0.

44*

*–

10.

Aff

ecti

ve

33.2

14.

410.

710.

47*

*0.

48*

*0.

25*

0.77

**

0.44

**

0.54

**

0.59

**

0.47

**

0.34

**

–11

.C

onti

nu

ance

22.3

25.

470.

682

0.02

20.

042

0.08

0.16

0.05

0.04

0.10

20.

042

0.03

0.05

–12

.N

etw

ork

outc

omes

18.6

74.

34–

0.50

**

0.46

**

0.51

**

0.21

0.61

**

0.17

0.32

**

0.49

**

0.42

**

0.36

**

20.

21–

Notes:

* p,

0.05

,*

* p,

0.01

Table III.Intercorrelations, means,standard deviations andreliabilities of measures

used in the study

Networkcommitment

1193

Table IV show that when regressed against affective commitment, four of the eightposited antecedent conditions have statistically significant positive regressioncoefficients; mutual interdependence, mutual gain, effective conflict resolutionmechanisms and role clarity, accounting for 69 percent of the variance in thisdimension of commitment. However the results show that none of the antecedentvariables included in the study were significantly related to continuancecommitment. Propositions 1, 2, 6 and 7 were therefore supported in that thesefour of the nine posited antecedent conditions were positively associated withaffective commitment.

The final set of analyses examined the relationship between both the positedantecedent conditions and the two dimensions of network commitment on networkperformance. In the first regression analysis all nine antecedent conditions wereentered in the first step to determine if they were related to network performanceoutcomes. The results in Table IV show that both effective feedback and goalcongruence were significantly related to network performance outcomes, togetheraccounting for 48 per cent of the variance in this construct. When both affective andcontinuance commitment were entered in the second step, the value of R 2 changedindicating a significant effect due to both forms of commitment. An examination of thebeta weights demonstrated however that affective commitment was significantlypositively related to network performance while there was a smaller, negativerelationship with continuance commitment. Thus, when statistically controlling for theposited antecedents, both affective and continuance forms of commitment havestatistically significant effects on network performance. Proposition ten was thereforesupported in that affective commitment was positively related to networkperformance. The finding that continuance commitment was negatively related tonetwork performance suggests that the relationship between commitment andperformance within networks is certainly not straightforward.

(1) (2)(3)

Network performanceVariable Affective Continuance At step Final

1. Shared values 20.20 20.08 20.13 20.091. Mutual interdependence 0.20 * 20.12 20.05 20.131. Role clarity 0.24 * 20.13 0.12 0.031. Feedback 20.17 20.01 0.27 * 0.31 *

1. Involvement 0.04 0.06 20.19 20.191. Lack of conflict 0.25 * 20.15 0.09 20.011. Mutual gain 0.54 * * 0.33 20.01 20.111. Goal congruence 0.06 0.24 0.51 * * 0.54 * *

1. Confidence in other orgs 0.01 20.14 0.11 0.082. Affective 0.30 *

2. Continuance 20.19 *

Multiple R 0.84 0.29 0.69 0.74R-square 0.69 0.09 0.48 0.54

Notes: *p , 0.05, * *p , 0.01

Table IV.Hierarchical regressionresults

MD44,9

1194

DiscussionThe significance of network commitment for network performanceThe purpose of this paper was to posit and test a theoretical model grounded in theliterature, to explain a number of possible antecedent network conditions associatedwith commitment at the inter-organisational level, or network commitment. In so doingit was suggested that one of the key mechanisms through which improvedperformance in networks can be brought about, is indirectly through increasing thecommitment amongst network partners. In this respect a number of key findings haveemerged from the study demonstrating some degree of support for the construct ofnetwork commitment as a mediating variable influencing network performanceoutcomes, thereby supporting proposition one. However the two different dimensionsof affective and continuance commitment, although both statistically significant, werefound to have opposite relationships with network performance, positive and negativerespectively. The results suggest then, that the link between commitment and networkperformance may vary as a function of the strength of these two dimensions ofnetwork commitment. Previously in the organisational commitment literature, Meyeret al. (1989) found that supervisory ratings of the overall job performance and thepromotability of their employees correlated positively with those employees’ affectivecommitment scores, and negatively with their continuance commitment scores. Mayerand Schoorman (1992) also found that commitment leading to participation wasassociated with continuance commitment whereas by contrast, performance is morestrongly related to value (affective) commitment. Similarly of note here too,continuance network commitment was also found to be negatively and to a lesserdegree related to network performance. Specifically as regards networks, it may wellbe that a continuance form of commitment has a role in influencing network partners tocontinue to participate within a network, and as such has an important role to playoverall in regards enhancing collaboration. However the finding here that it has anegative relationship with network performance has significant implications. Firstly, itdraws attention to important considerations concerning the variegated nature ofnetworks themselves and the extent to which different forms of commitment may havemore salient effects under different contingencies. Certainly the finding here that noneof the antecedent conditions tested was associated with continuance commitment lendssupport to the independent nature of these two commitment dimensions. Of majorinterest however, it may well also be that these different forms of commitment areassociated with different types of collaborative outcomes. Such that for example,continuance commitment may possibly be more closely associated with alternativecollaborative outcomes such as political outcomes for example (Clarke, 2005; Hardyet al., 2003). Secondly, these findings may also shed some light on some of the mixedresults obtained in previous studies examining commitment at the inter-organisationallevel, which have utilised uni-dimensional measures of the construct. Previously forexample, Sarkar et al. (1998) found a negative relationship between the construct ofrelational bonding and termination costs, whereas Morgan and Hunt (1994) found apositive relationship. Termination costs are more likely to be associated withcontinuance commitment. Differences in results may well therefore reflect the failure tosatisfactorily operationalise the multidimensional nature of commitment at theinter-organisational level. Furthermore these findings provide some support for thebenefits of establishing conditions that generate an affective form of commitment

Networkcommitment

1195

within networks as a means through which network performance might be increased.Importantly these findings appear to provide credence to McElroy et al.’s (2001)assertions for recognising the distinct dimensions associated with the commitmentconstruct beyond simply the organisational level.

Antecedent conditions of network commitmentThe findings of this study are also of interest in providing empirical support for someof the antecedent conditions of network commitment posited (demonstrated by solidlines of Figure 1). Importantly the network conditions, mutual interdependence, mutualgain, effective conflict resolution, and role clarity were all found to have statisticallysignificant positive relationships with affective commitment, providing some supportfor stated propositions 2 and 5. The first three of these antecedent conditions supportprevious research examining the antecedents of commitment at theinter-organisational level (Dwyer et al., 1987; Yan and Gray, 1994; Kumar et al.,1995; Holm et al., 1999). Mechanisms for conflict resolution in particular may actuallystrengthen network relationships and lead to both greater commitment as well as trust(Gundlach et al., 1995; Weitz and Jap, 1995). Previously however mixed results havebeen found for the antecedent, role clarity (Lusch and Brown, 1996; Sarkar et al., 1998;Young and Wilkinson, 1989). This may well have been due to the different ways inwhich the commitment construct has been operationalised in these previous studies.

Figure 1.The relationship betweennetwork commitment, itsantecedents and networkperformance

MD44,9

1196

This may also explain the failure to find any significant relationships with theantecedent conditions of shared values, goal congruence and effective performancefeedback here, whilst positive relationships have previously been found in a study bySarkar et al. (1998). Yet it should be noted that all the antecedent conditions positedhere were found to be significantly correlated with affective commitment (Table III).This may suggest that some of the antecedent conditions tested may well influenceaffective commitment through their effects on a further intervening variable. Mostnotably trust. A previous study by Zeldin and Jonnson (2000) for example, also failed todemonstrate a relationship between commitment and shared values at theinter-organisational level, yet they did find a significant positive relationshipbetween shared values and trust. Supporting much of the literature in the area, theseauthors also found that trust was positively associated with commitment at theinter-organisational level (De Ruyter et al., 2001; Kwon and Suh, 2005; Morgan andHunt, 1994; Wong and Sohal, 2002). The results here also failed to demonstrate arelationship between confidence in network members and commitment. Confidencehowever has figured prominently in the way in which trust has often beenconceptualised (Moorman et al., 1993; Rotter, 1967). In some instances then despite thesignificant positive correlations found between these posited antecedent conditionsand commitment, they may well influence affective commitment although indirectlythrough their capacity for generating trust.

Theoretically these three posited antecedent conditions variables for which nosignificant relationships were found were suggested to enhance commitment throughtheir capacity to increase compatibility between network partners and reduce levels ofuncertainty. Yet it can also be argued that all three of these conditions can becategorised typically as control mechanisms, with the latter two processes (goalcongruence and performance feedback) classified as types of formal control whileshared values an example of social control (Eisenhardt, 1985; Das and Teng, 1998).Elsewhere Cullen et al. (1995) in their study examining commitment in joint ventures,found that formal control was not related to commitment. The results here then wouldseem to indicate that any relationship between these latter three specific conditions andcommitment is not as straightforward as perhaps suggested by the theoreticalconsiderations discussed earlier in this paper. It may also be the case that other factorsmay well determine their salience in fostering network commitment, not least the typeof network under consideration (Das and Teng, 1998) and importantly the particulardimension of network commitment being examined. In relation to shared values forexample, it may well be that this is more closely associated with the dimension ofnormative network commitment rather than affective commitment as was posited here.Previously O’Reilly and Chatman (1986) have found a positive relationship betweenshared values and a measure of organisational commitment that corresponded closelyto a normative dimension. At the inter-organisational level too then, shared values maybe more closely associated with a normative form of network commitment. Problemsencountered with establishing a reliable measure of normative commitment in thisstudy however meant that this was not able to be tested here. It does howeverunderscore the need for future research that examines the differing dimensions ofcommitment and their potential antecedents at the inter-organisational level.

Nevertheless the results do support a clear role for formal control mechanismswithin networks (goal congruence and performance feedback) as having a direct

Networkcommitment

1197

influence on network performance. This would appear to support much of the literaturerelating to supporting collaboration between organisations where a number of authorshave stressed interventions that seek to increase compatibility between collaboratingpartners’ goals and institute evaluative processes that attempt to provide data on howthe network for example may be performing in order to guide ameliorative action(Clarke, 2005; Cummings and Worley, 1997; Gray, 1999). The failure to find a positiverelationship between the network condition, involvement and decision making andnetwork commitment was surprising given the theoretical rational outlined. This couldsuggest that the items contained in the measure used may not have sufficiently tappedthis variable appropriately. The items relating to Involvement in decision-making usedhere may simply have captured participation without any real sense of power sharing(Conger and Kanungo, 1988). Yet it may be this latter aspect that is the far moresignificant factor in relation to fostering commitment (Emerson, 1962). Similarly themeasurement of confidence in network members may have also been far too vague tocapture the specific confidence locus potentially associated with affective commitment.Das and Teng (1998) for example suggest that it is specifically confidence in theco-operation of network members that fosters network relational attributes rather thanmerely confidence in network organisations per se. Future research should explore theextent to which measures that better cover the domain of interest associated with theseantecedent conditions produce similar or differing results.

Finally a number of limitations associated with the study should be recognised inconsidering the findings obtained. Firstly, the selection of both the chairperson andco-ordinator from each of the networks to provide data on the antecedent conditionsand commitment measures may have introduced some bias into the measures collected.These individuals were selected due to their in-depth knowledge of the nature of thenetwork and network conditions and because arguably their commitment is deemed tobe the most significant in determining network performance outcomes. However theseparticular antecedent conditions may well have very different degrees of significancefor performing these roles compared to other members of the network. It thereforecannot be discounted that the results obtained may well have been due to network databeing collected from only these two key members of the network. Previously Nummela(2003) has suggested that as far as possible, measures of commitment at theinter-organisational level should be obtained from all network members to increase thevalidity of the measures obtained. In this study there was generally a high level ofagreement between the scores obtained on variables for chairpersons and co-ordinatorsfrom the same network, suggesting some degree of support for the consistencyobtained in the measures used. Nonetheless far more research needs to be undertakento determine the extent to which data from significant members of networks,particularly relating to commitment, genuinely reflects that of network members morewidely. Secondly, the cross-sectional nature of the research only allows us to identifyassociations, so longitudinal studies are necessary in order to gain more concrete dataregarding the direction of any significant relationships found. Such studies would alsoprovide greater insights into the differing developmental trajectories of the antecedentconditions identified here and how these then impact on the development of networkcommitment. Indeed we need to understand far more on how these differing antecedentconditions interact with one another and whether there is a sequential element to theireffectiveness. The nature of this network may well also constrain the extent to which

MD44,9

1198

these findings are necessarily characteristic of other networks more widely.Organisations and agencies operating within the public sector and within healthcarein particular are subject to a range of particular legal and political regulatoryframeworks that may well influence both the significance of particular antecedentconditions, their influence on differing types of commitment as well as the potentialimpacts of these types of commitment. Despite these limitations this would seem to bethe first study to appear in the literature that has sought to examine relationshipsbetween network conditions and different dimensions of the commitment construct theresults of which has highlighted the differing roles that types of commitment may playwithin networks. As our understanding of the differing outcomes of collaborationassociated with networks increases, this should enable us to consider more carefullyhow different forms of commitment might be implicated in producing differingcollaborative outcomes. This knowledge should then lead to us developing far bettermanagement interventions in order to improve both network relationships and theoutcomes desired by both network members and where relevant those of policymakers.

ConclusionsThis paper has argued that commitment may play a significant role specificallyassociated with determining performance outcomes in networks. In so doing atheoretical model was put forward drawn from the literature to suggest possibleantecedent conditions associated with commitment at the inter-organisational level.Importantly a number of these posited relationships were found to be significantlyassociated with affective commitment. However affective and continuancecommitment at the inter-organisational level were found respectively to have apositive and negative relationship with network performance. The findings suggestthen that we need to discover far more how different forms of commitment influencecollaborative outcomes associated with different forms of networks. Importantly inthis respect the nature of different networks may well influence the salience ofparticular variables in terms of the genesis of commitment at the inter-organisationallevel. Further research is therefore required to fully identify those different antecedentconditions that may foster commitment at the inter-organisational level. Of mostsignificance despite the study’s limitations, the findings provide an initial indicationthat utilising differing dimensions associated with the commitment construct may bean important step forward if we are to begin to build a comprehensive theory tounderpin the commitment construct and its impact at the inter-organisational level.Further research should therefore seek to confirm the significance of those networkconditions found to be associated with affective commitment here, as well as identifyothers that are associated with other dimensions of network commitment. This shouldthen assist network facilitators and managers at the operational level in consideringwhere energies should be directed for developing key antecedent conditions in order toimprove network performance.

References

Allen, N.J. and Meyer, J.P. (1990), “The measurement and antecedents of affective, continuanceand normative commitment to the organization”, Journal of Occupational Psychology,Vol. 63, pp. 1-18.

Networkcommitment

1199

Anderson, J.C. (1995), “Relationships in business markets: exchange episodes, value creation andtheir empirical assessment”, Journal of the Academy of Marketing Science, Vol. 23 No. 4,pp. 346-50.

Badaracco, J.L. (1991), The Knowledge Link: How Firms Compete through Strategic Alliances,Harvard Business School Press, Boston, MA.

Baruch, Y. and Winkelmann-Gleed, A. (2002), “Multiple commitments: a conceptual frameworkand empirical investigation in a community health service trust”, British Journal ofManagement, Vol. 13, pp. 337-57.

Boddy, D., Macbeth, D. and Wagner, B. (2000), “Implementing collaboration betweenorganizations: an empirical study of supply chain partnering”, Journal of ManagementStudies, Vol. 37 No. 7, pp. 1003-17.

Brown, S., Flynn, M. and Wistow, G. (1992), Back to the Future: Joint Work for People withLearning Disabilities, Nuffield Institute for Health, London.

Bucklin, L.P. and Sengupta, S. (1993), “Organizing successful co-marketing alliances”, Journal ofMarketing, Vol. 57 No. 2, pp. 32-46.

Caldwell, D.F., Chatman, J.A. and O’Reilly, C.A. (1990), “Building organizational commitment:a multifirm study”, Journal of Occupational Psychology, Vol. 63, pp. 245-61.

Churchill, G.A. (1979), “A paradigm for developing better measures of marketing constructs”,Journal of Marketing Research, Vol. 16 No. 2, pp. 64-73.

Clarke, N. (2005), “Transorganization development for network building”, Journal of AppliedBehavioral Science, Vol. 41 No. 1, pp. 30-46.

Conger, J.A. and Kanungo, R.N. (1988), “The empowerment process: integrating theory andpractice”, Academy of Management Review, Vol. 13, pp. 471-82.

Cullen, J.B., Johnson, J.L. and Sakano, T. (1995), “Japanese and local partner commitment to IJVs:psychological consequences of outcomes and investments in the IJV relationship”, Journalof International Business Studies, Vol. 26, pp. 91-115.

Cummings, T.G. and Worley, C.G. (1997), Organization Development and Change, 6th ed., SouthWestern College Publishing, Cinninati, OH.

Daniel, H.Z., Hempel, D.J. and Srinivasan, N. (2002), “A model of value assessment incollaborative R&D programs”, Industrial Marketing Management, Vol. 31, pp. 653-64.

Das, T.K. and Teng, B.S. (1998), “Between trust and control: developing confidence in partnercooperation in alliances”, Academy of Management Review, Vol. 23 No. 3, pp. 491-512.

De Ruyter, K., Moorman, L. and Lemmink, J. (2001), “Antecedents of commitment and trust incustomer-supplier relationships in high technology markets”, Industrial MarketingManagement, Vol. 30, pp. 271-86.

Drug Strategy Directorate (2004), National Drug Strategy Performance Management FrameworkResource Pack, Home Office Drug Strategy Directorate, London.

Dwyer, R.F., Schurr, P.H. and Sejo, O. (1987), “Developing buyer-seller relationships”, Journal ofMarketing, Vol. 51, pp. 11-27.

Dyer, J.H. and Singh, H. (1998), “The relational view: cooperative strategy and sources ofinterorganizational competitive advantage”, Academy of Management Review, Vol. 23,pp. 660-79.

Eisenhardt, K.M. (1985), “Control: organizational and economic approaches”, ManagementScience, Vol. 31, pp. 134-49.

Emerson, R.M. (1962), “Power-dependence relations”, American Sociological Review, Vol. 27,pp. 31-41.

MD44,9

1200

Frazier, G.L., Spekman, R.E. and O’Neal, C.R. (1988), “Just-in-time exchange relationships inindustrial markets”, Journal of Marketing, Vol. 52, pp. 52-67.

Gray, B. (1999), “The evolution of collaborative research in the last decade: toward a dynamictheory”, in Schruijer, S. (Ed.), Multi-Organizational Partnerships and Co-operative Strategy,Dutch University Press, Tilberg, pp. 9-16.

Gundlach, G.T., Achrol, R.S. and Mentzer, J.T. (1995), “The structure of commitment inexchange”, Journal of Marketing, Vol. 59 No. 1, pp. 78-92.

Hagedorn, J., Link, A.N. and Vonortas, N.S. (2000), “Research partnerships”, Research Policy,Vol. 29, pp. 567-86.

Halinen, A. (1997), Relationship Marketing in Professional Services: A Study of Agency-ClientDynamics in the Advertising Sector, Routledge, London.

Hardy, C., Phillips, N. and Lawrence, T.B. (2003), “Resources, knowledge and influence:the organizational effects of interorganizational collaboration”, Journal of ManagementStudies, Vol. 40 No. 2, pp. 321-46.

Heide, J.B. and John, G. (1998), “Do norms matter in marketing relationships?”, Journal ofMarketing, Vol. 56, pp. 32-44.

Himmelman, A. (1992), Communities Working Collaboratively for a Change, The HimmelmanConsulting Group, Minneapolis, MN.

Hiscock, J. and Pearson, M. (1999), “Looking inwards, looking outwards: dismantling the ‘BerlinWall’ between health and social services?”, Social Policy and Administration, Vol. 33 No. 2,pp. 150-63.

Holm, D.B., Eriksson, K. and Johanson, J. (1999), “Creating value through mutual commitment tobusiness network relationships”, Strategic Management Journal, Vol. 20, pp. 467-86.

Huxham, C. (1996), Creating Collaborative Advantage, Sage, London.

Huxham, C. and Vangen, S. (2000a), “Ambiguity, complexity and dynamics in the membership ofcollaboration”, Human Relations, Vol. 43 No. 6, pp. 771-806.

Huxham, C. and Vangen, S. (2000b), “Leadership in the shaping and implementation ofcollaboration agendas: how things happen in a (not quite) joined-up world”, Academy ofManagement Journal, Vol. 43 No. 6, pp. 1159-75.

Kim, K. and Frazier, G.L. (1997), “Measurement of distributor commitment in industrial channelsof distribution”, Journal of Business Research, Vol. 40, pp. 139-54.

Kumar, N., Scheer, L.K. and Steenkamp, J. (1995), “The effects of perceived interdependence ondealer attitudes”, Journal of Marketing Research, Vol. 32, pp. 348-56.

Kwon, I.G. and Suh, T. (2005), “Trust, commitment and relationships in supply chainmanagement: a path analysis”, Supply Chain Management: An International Journal,Vol. 10 No. 1, pp. 26-33.

Lasker, R.D., Weiss, E.S. and Miller, R. (2001), “Partnership synergy: a practical framework forstudying and strengthening the collaborative advantage”, The Millbank Quarterly, Vol. 79No. 2, pp. 179-205.

Lawler, E.J. and Yoon, J. (1996), “Commitment in exchange relations: test of a theory of relationalcohesion”, American Sociological Review, Vol. 61 No. 1, pp. 89-108.

Lewis, J.D. (1990), Partnerships for Profit: Structuring and Managing Strategic Alliances, FreePress, New York, NY.

Locke, E.A. and Latham, G.P. (1990), A Theory of Goal Setting and Task Performance,Prentice-Hall, Englewood-Cliffs, NJ.

Networkcommitment

1201

Loxley, A. (1997), Collaboration in Health and Welfare – Working with Difference, Jessica KingPublishers, London.

Lusch, R.F. and Brown, J.R. (1996), “Interdependency, contracting and relational behaviour inmarketing channels”, Journal of Marketing, Vol. 60, pp. 19-38.

McElroy, J.C., Morrow, P.C. and Laczniak, R.N. (2001), “External organizational commitment”,Human Resource Management Review, Vol. 11, pp. 237-56.

Madhok, A. (1995), “Revisiting multinational firms’ tolerance for joint ventures: a trust-basedapproach”, Journal of International Business Studies, Vol. 26 No. 1, pp. 117-38.

Mankin, D. and Cohen, S. (2003), Complex Collaborations in the New Global Economy,Jossey-Bass, San Francisco, CA.

Mathieu, J.E. and Zajac, D.M. (1990), “A review and meta-analysis of the antecedents, correlatesand consequences of organizational commitment”, Psychological Bulletin, Vol. 108 No. 2,pp. 171-94.

Mayer, R.C. and Schoorman, F.D. (1992), “Predicting participation and production outcomesthrough a two-dimensional model of organizational commitment”, Academy ofManagement Journal, Vol. 35, pp. 671-84.

Medcof, J. (1997), “Why do many alliances end in divorce?”, Long Range Planning, Vol. 30,pp. 718-32.

Meyer, J.P. and Allen, N.J. (1988), “Links between work experience and organizationalcommitment during the first year of employment: a longitudinal analysis”, Journal ofOccupational Psychology, Vol. 61, pp. 195-209.

Meyer, J.P. and Allen, N.J. (1997), Commitment in the Workplace, Sage Publications, ThousandOaks, CA.

Meyer, J.P. and Herscovitch, L. (2001), “Commitment in the workplace: toward a general model”,Human Resource Management Review, Vol. 11, pp. 299-326.

Meyer, J.P., Allen, N.J. and Smith, C.A. (1993), “Commitment to organizations and occupations:extension and test of a three-component conceptualisation”, Journal of Applied Psychology,Vol. 78, pp. 538-51.

Meyer, J.P., Paunonen, S.V., Gellaty, I.R., Goffin, R.D. and Jackson, D.N. (1989), “Organizationalcommitment and job performance: it’s the nature of the commitment that counts”, Journalof Applied Psychology, Vol. 74, pp. 152-6.

Milgrom, P. and Roberts, J. (1992), Economics, Organization and Management, Prentice-Hall,Englewood-Cliffs, NJ.

Mobley, W.H. (1982), Employee Turnover: Causes, Consequences and Control, Addison-Wesley,Reading, MA.

Moorman, C., Deshpande, R. and Zaltman, G. (1993), “Factors affecting trust in market researchrelationships”, Journal of Marketing, Vol. 57, January, pp. 81-101.

Morgan, R.M. and Hunt, S.D. (1994), “The commitment-trust theory of relationship marketing”,Journal of Marketing, Vol. 58 No. 3, pp. 20-38.

Morris, J.H. and Koch, J.L. (1979), “Impacts of role perception on organizational commitment, jobinvolvement and psychosomatic illness among three vocational groupings”, Journal ofVocational Behavior, Vol. 14, pp. 88-101.

Mowday, R.T. (1999), “Reflections on the study and relevance of organizational commitment”,Human Resource Management Review, Vol. 8 No. 4, pp. 387-401.

Mowday, R.T., Porter, L.W. and Steers, R.M. (1982), Organizational Linkage: The Psychology ofCommitment, Absenteeism and Turnover, Academic Press, New York, NY.

MD44,9

1202

Mulroy, E.A. and Shay, S. (1998), “Motivation and reward in nonprofit interorganizationalcollaboration in low-income neighborhoods”, Administration in Social Work, Vol. 22 No. 4,pp. 1-17.

Nohria, N. and Eccles, R.G. (Eds) (1998), Network and Organizations: Structure, Form and Action,Harvard Business School, Boston, MA.

Nummela, N. (2003), “Looking through a prism – multiple perspectives to commitment tointernational R&D collaboration”, Journal of High Technology Management Research,Vol. 14, pp. 135-48.

Ohmae, K. (1989), “The global logic of strategic alliances”, Harvard Business Review,March-April, pp. 143-54.

O’Reilly, C.A. and Chatman, J. (1986), “Organizational commitment and psychologicalattachment: the effects of compliance, identification and internalization on prosicialbehaviour”, Journal of Applied Psychology, Vol. 71, pp. 492-9.

Ovretveit, J. (1993), Coordinating Community Care – Multidisciplinary Teams and CareManagement, OUP, Buckingham.

Parker, H. (2000), “Interfirm collaboration and the new product development process”, IndustrialManagement & Data Systems, Vol. 100, pp. 255-60.

Pennings, I.M. (1970), “Work value systems of white collar workers”, Administrative ScienceQuarterly, Vol. 15, pp. 397-405.

Purdue, D. and Razzique, K. (1999), “Connections and expectations: partnerships, communityleaders and the problem of succession”, in Schruijer, S. (Ed.), Multi-OrganizationalPartnerships and Co-operative Strategy, Dutch University Press, Tilberg.

Putterhill, M.S. and Rohrer, T.S. (1995), “A causal model of employee commitment in amanufacturing setting”, International Journal of Manpower, Vol. 16 Nos 5/6, pp. 56-69.

Rahim, M.A. (1983), “Measurement of organizational conflict”, Journal of Group Psychology,Vol. 109, pp. 189-99.

Reichers, A.E. (1985), “A review and reconceptualisation of organizational commitment”,Academy of Management Review, Vol. 19, pp. 465-76.

Ring, P.S. and Van de Ven, A. (1992), “Structuring cooperative relationships betweenorganizations”, Strategic Management Journal, Vol. 13, pp. 483-98.

Ring, P.S. and Van de Ven, A. (1994), “Developmental processes of co-operativeinterorganizational relationships”, Academy of Management Review, Vol. 19, pp. 90-118.

Rizzo, J.R., House, R.J. and Lirtzman, S.I. (1970), “Role conflict and ambiguity in complexorganizations”, Administrative Science Quarterly, Vol. 15, pp. 150-63.

Root, F.R. (1988), “Some taxonomies of international cooperative arrangements”, inContractor, F.J. and Lorange, P. (Eds), Cooperative Strategies in International Business,Lexington Books, Lexington, MA, pp. 69-80.

Rotter, J.B. (1967), “A new scale for the measurement of interpersonal trust”, Journal ofPersonality, Vol. 35 No. 4, pp. 651-65.

Sarkar, M.B., Aulakh, P.S. and Cavusgil, S.T. (1998), “The strategic role of relational bonding ininterorganizational collaborations: an empirical study of the global construction industry”,Journal of International Management, Vol. 4 No. 2, pp. 85-107.

Sayeed, O.B. (2001), Organizational Commitment and Conflict: Studies in Healthy OrganizationalProcesses, Sage Publications, New Delhi.

Scott, C.M. and Thurston, W.E. (2004), “The influence of social context on partnerships inCanadian health systems”, Gender, Work and Organization, Vol. 11 No. 5, pp. 481-505.

Networkcommitment

1203

Sink, D.W. (1991), “Transorganization development in urban policy coalitions”, HumanRelations, Vol. 44, pp. 11-21.

Sullivan, H., Barners, M. and Matka, E. (2002), “Building collaborative capacity through theoriesof change”, Evaluation, Vol. 8 No. 2, pp. 205-26.

Walter, A. and Ritter, T. (2003), “The influence of adaptations, trust, and commitment onvalue-creating functions of customer relations”, Journal of Business & IndustrialMarketing, Vol. 18 Nos 4/5, pp. 353-65.

Webb, A. (1991), “Co-ordination: a problem in public sector management”, Policy and Politics,Vol. 19, pp. 229-41.

Weitz, B.A. and Jap, S.D. (1995), “Relationship marketing and distribution channels”, Journal ofthe Academy of Marketing Science, Vol. 4, pp. 305-20.

Wong, A. and Sohal, A. (2002), “An examination of the relationship between trust, commitmentand relationship quality”, International Journal of Retail & Distribution Management,Vol. 30 No. 1, pp. 34-50.

Yan, A. and Gray, B. (1994), “Bargaining power, management control and performance in UnitedStates-China joint ventures: a comparative study”, Academy of Management Journal,Vol. 37 No. 6, pp. 1478-517.

Young, L.C. and Wilkinson, I.F. (1989), “The role of trust and cooperation in marketing channels:a preliminary study”, European Journal of Marketing, Vol. 23 No. 2, pp. 109-22.

Zeldin, M. and Jonnson, P. (2000), “An examination of the main factors affectingtrust/commitment in supplier-dealer relationships: an empirical study of the Swedishwood industry”, The TQM Magazine, Vol. 12 No. 4, pp. 245-65.

AppendixShared values

5. There is compatibility between the values of the network and the values of my organisation.24. If the values of this network and my organisation were different, my organisation would not

be part of this network.25. The different organisational cultures or values of network members often get in the way of

the network getting things done w.

Goal congruence14. The goals of my employing organisation and the network are similar.21. There is a clear vision of what the network is hoping to achieve.28. There are conflicting ideas about what the network should be trying to achieve w.

Effective conflict resolution15. Organisations in the network work hard to try to resolve problems or difficulties that arise.39. Disagreements or disputes often characterise the relations organisations in this network.27. There is compatibility between how network members consider difficulties should be

addressed.

Involvement6. I play a large part in making decisions in this network.

30. I am able to influence the agenda in this network.26. Network members are receptive to my suggestions about how things could be done.

Confidence4. I have confidence in the work of other organisations in the network.

MD44,9

1204

22. I consider members from other organisations participating in this network to be competent.37. I am unsure of the motives of other members participating in the network.

Role clarity1. I am clear about my role in this network.

18. There is often confusion amongst network members about who is responsible for doingwhat.

12. I am aware of the roles and responsibilities of each of the individual organisations involvedin the network.

Mutual interdependence10. My organisation could achieve its objectives without the involvement of the other

organisations in this network w.13. My organisation is dependent on the other organisations in this network in order to achieve

its goals?31. Participating in this network is essential for improving my own organisation’s services.

Mutual gainParticipating in this network brings benefits to all the organisations involved.

9. My organisation gains more from being part of this network than if it were it not.33. The benefits of participating in this network outweigh the costs involved.

Feedback11. I regularly receive feedback on how well the network is achieving its goals.40. Information is readily available on how well the network is actually functioning.32. The network has developed systems and procedures for monitoring its performance.

Network performance outcomes1. To support the implementation of drug education in the context of personal, social and

health education in all schools, the youth service, further education, the community and withparents based on evidence of good practice.

2. To ensure that prevention/education plans targeting young people address local problems ina holistic way involving schools, parents, the youth service, further education, thecommunity and statutory agencies through targeted interventions.

3. To have established a maximum waiting time for admission into a drug treatment serviceand to be monitoring agencies’ performance.

Corresponding authorNicholas Clarke can be contacted at: n.r.clarke@soton.ac.uk

Networkcommitment

1205

To purchase reprints of this article please e-mail: reprints@emeraldinsight.comOr visit our web site for further details: www.emeraldinsight.com/reprints