Journal of Marmgamnl Studies 38:6 September0022-2380

DELIVERING CUSTOMER-ORIENTED BEHAVIOUR THROUGHEMPOWERMENT: AN EMPIRICAL TEST OF HRM ASSUMFI IONS

RlCHARDO

PATRICE ROSENTHAL

's College, London

ABSTRACT

Organizational initiatives to strengthen customer orientalion among front-lineservice workers abound, and have led many commentators to speak of the recon-stitution of service work. The.s<' interventions rest on managers' assumptions aboutwhat engenders the desired customer-oriented behaviours among employees. Weevaluate those assumptions in the context of a major change initiative in asupermarket firm. The logic of the programme mirrors key pret epts in the coti-temporary management literature. These are thai management behaviour, jobdcsijrn and values-based training can produce a sense of empowertnent amongnnployecs, atid that empowerment will generate prosocial customer-orientedbehaviour. Using data from a large scale employee survey, we test the validity ofthose assumptions. Employees who perceived management behaviour in a positiveligiit and who had participated in values-based training were more likely to feelempowered (i.e. to have internalized prosocial service values and to feel a sense ofcompetence and autonomy on the job). Psychologieal empowerment was, in turn,positively related to the customer-oriented behaviour of workers. This study, there-fore, provides support for key assumptions underlyitig HRM theory and practicein services.

ENTRODUCTION

The 'sovereign customer' is seen to demand greater individual attention, flexibil-ity and novelty in the provision of services, as well as goods (du Gay; 1996; Peters,1987; Sturdy, 1998). Service cjuality has thus become a major managementpreoccupation. Increasing numbers of organizations, including those who con-tinue to compete on price, seek also to differentiate themselves on the basis ofsuperior customer service. A key problem for managers is how to ensure appro-priate behaviours on the part of front-line workers: those employees whoactually meet the customer and deliver the service (Bowen and Schneider, 1988;Carlzon, 1987).

Addrf.-i!i for rf.priiils. Ritcardo Ptuxei, The Management Cenu-c. King's tlollcgc, Ldiidon. l.'jt) Siam-lord Street, London ECl 8WA, UK.

Pul»li.slicrs LuJ 20(11. PublislK^if ljy Bliukw<-ll Piil>lish.;rs. lOH Cowk-y lltwcl, Oxiord (>X4 IJF. UKand 350 Main Sireci, NLiklen, M \ (V2I48, USA.

832 R. PECCEI AND P. ROSEN'HIAL

(.)ne route to the provision of high quality senicr is through redoubling thetraditional mechanisms of employee control developed within manufacturingsettings. Managers might opt to simplify jobs and to clarify the 'script': the set ofrules which govern interaction between customers and employees. Supervisionmight be intcn.sifieci, and harsher .sanctions for poor performance imposed. Sucha 'production-line' approach (Levitt, 1972) emphasizing standardization, techno-logical and personal control is pursued by numerous semce organizations, andmay indeed produce cflicient and reliable sendee which meets customers' demands(Bowen and Lawler, 1992). However, it flies in the face of ctirrent definitions ofquality seivice and prescriptions for its achievement. It may produce greater efTi-ciency and reliability, goes the argument, but not the flexibility and sense ofgenuine concern prized by eontemporar)' customers (Bowen and Schneider, 1988;Paiasuraman et al., 1985; Quinn, 1992; Schlesingcr and Heskett, 1991). Scripts,rules and close supervision can 'straightjacket' employees, constraining both theirscope to act in customer interests and their motivation to do so. Hem' then to attainthe responsiveness, eommitment, authenticity and innovation which now charac-terizes the ideal service worker?

In an increasing number of organizations, management's response to this ques-tion takes the form of Vulture change' or 'customer care' programmes targetingthe attittides, orientations and behaviours of front-line stalV. These programmesreflect prescriptions in the human resource management (for a review; see Legge,1995; Sparrow and Marehington, 1998), cpiality management (for a review, seeDean and Bowen, 1994), excellence (Peters, 1987; Peters and Waterman, 1982),and services management (for a review, see Bowen and Schneider, 1988; Redtnanand Matlicws, 199(3) literatures. like the intervTntions, these literatures vary infocus and emphasis (Harris and Ogbonna, 1999). Hovi'cver they share certainassumptions and precepts, fbr exatnpie tlie 'natural' desire of employees to doquality work, and how tliis can be facilitated through appropriate job design,culture and leadership, supportive supervision, hierarchiciil trust and normative-reeducative training.

in this paper, we evaluate the assumptions tliat underpin one such change ini-tiative in a supermarket company. In tlie first section we set out the logic of theintervention. This entails description first, of w/iat. was done (i.e. the managementand human resource practices involved in the programme). Second, and centralto this paper, we address why it was done. Our focus, in other words, is the under-lying logic of the initiative: the a.ssumptions about the set of conditions that wouldfacilitate the customer orientation of front-line stafT. We explore this underlyinglogic and show how it reflects arguments and prescriptions in various strands ofmanagemrnt literature. In the remainder of" the paper, we test these assumptionsempirically, using data from a large scale employee survey.

Evaluation of this kind is important. The partial state of human resource man-agement (HRM) and quality theory is widely discussed (Becker and Gerhart, 1996;Guest, 1987, 1997; Hackman and Wageman, 1995; Hill and Wilkinson, 1995;Knights and McCabe, 1997; Noon, 1992; Patterson et al., 1997). Witliiii the main-stream literature, two issues in particular arc lughlighteci. First, is the need todevelop a fuller theoretical understanding of the mechanisms linking humanresource (HR) practices and management behaviour to key aspects of individualand organizational performance. And second, is the need systematically to testthese theoretical links and develop a sounder empirical basis for identifying best

© Blackwcll Publishent t.td 2001

HRM ASSUMPTIONS 833

practice. A parallel literattire engages in critique of HRM, angtiing theoreticalinconsistencies and (especially in Its 'soft' version) the lack of its application inpractice (Legge, 1995), with writers such as Keenoy (1997; p. 837) speaking of its'overwhelming empirical "refutation"'.

The present study .seeks to contribute to each oi' these strands of the HRM andquality debate. It does so by first modelling and then testing core assumptionslinking HR practices and management behaviours to the customer-orientedbehav iour of front-line workers.

THE INTERVENTION

Shopco is one of the three largest supermarket chains in Britain. Several ^farsago, senior management judged tliat quality of customer sen ice would join, andpe,rhaps stipersede, price as a major competitive issue within mass market foodretailing. Accordingly, they decided to pursue competitive advantage through amore personalized and flexible type of service. Numerous and ongoing customerservice innovations were planned, many of these technological in nature. However,it was asstnned that tiie key to efTective delivery of these inno\ations, atid of day-to-day service, lay with front-line staft It was also assumed that to achieve this newcustomer-orienled ethos, a significajit 'culture change' was required in Shopco. Tothis end, a major initiati\e. Sen ice Excellence, was lauiuhed in 1994.

Through a series ol' in-depth inten'iews with corporate and store managementand supervisors, and analysis of documentary (including training) material, wemapped the aims, objectives and assumptions about appropriate mechanismsunderlying the initiative. In the section below we look at the logic of Sen ice Excel-lence, and show how themes from the literature are reflected in the programme.

Overall Logic of the Initiative

Service Excellence encompassed four main elements: a shifl to a more supportive,participative management style; role modelling by management; job redesign toincrease the discretion for front-line staff; and customer senice training. It wasassumed that together, these changes would create the conditions necessary fordelivery of the new service ethos, namely employees v\h() (1) understood and inter-nalized the values of Sen'ice Excellence, (2) possessed the necessary' skills andabilities, and (3) had the scope to use these competences.

We term this set of conditions f>syc/iologicfil empownment. given its similarity tocertain conceptualizations in the Iitcriiture. Writers such as Tlioniiis and Velthouse(1990) and Spreitzer (1995) define psychological empowerment as increased taskmotivation manifested in employees' cognitions about their work role. These cog-nitions are said to encompa.ss meanuig, or the fit between an individual's valuesand his/her work role; competence, or belief in one's ability to perform the job;self-determination, or a sense of autonomy on the job; and impact, or a sense ofhaving an influence over job outcomes. Together, tiiese perceptions should n^sultin an active rather than passive orientation to a work role (Spreitzer, 1995). Withinthe Shopco context, the first three of these dimensions correspond to under-standing and internalization of values (meaning); possession of the necessaty skillsand abilities (competence); and scope, or discretion, to use these competences(self-de te rmination).

e Blat-kwcll Pubtishrrs Ud 2001

834 R. PEGCEI AND P. ROSENTHAL

Ernpowcrmenl has of" course been conceptualized in a number of difli'rentways, and no (.•on.sen.sus on its status has yet emerged (AVilkinson, 1998). Somewriters focus on 'empowering' work practices such as provision of organizationalinformation to employers, reduction of Imrcaucratic controls and increased taskautonomy (Bowen and Lawler, 1992; Wall and Martin. 1994). These practices areoften hypothesized to increase employee motivation through greater task involve-ment. Another stream of writing adopts a critical perspective on empowerment,treatitig it as a rhetorical device to obscure an increase in management power(Parker and Slaughter, 1993; Sewell and Wilkinson, 1992). In this study, we takean explicilly psychoK)gicai view of emi3owerinent, focusing on individuals' per-ceptions of their work role and modelling and testing certain anteeedents and out-comes of these perceptions (for a parallel approach, see Corsun and Enz, 1999).We will return to conceptions of empowerment in the Discussion section ofthe paper.

We turn below to the four key elements of the Service Excellence initiative.

.New Managerial Attitudes and Behaviour

The rather authoritarian management style traditionally practised witliin Shopcowas judged inimical to the philosophy of Service Excellence. A new managerialethos was 'designed', this to encompass an array of changes in attitudes andbehaviour, towards both customers and employees. These new competences weretransmitted to managers and supervisors throtigh company directives, trainingmaterials and exercises, and through their establishment as explicit elements of theperformance appraisal system in Shopco.

Supportive rnanagmmit slyk. The management style traditionally manifested inShopco was seen to have resulted in an overly formal and distant relationshipbetween store management and general staff. As an explicit part of Service Excel-lence, this social distance was to be lessened. For example, a company edict wasissued banning the use of last names in interactions among employees, regardlessof their hierarchical level. More generally, Shopco managers, fmm board level tofirst line management, were charged with the development of a more supportive,considerate st>'le. There was to be more frequent interaction between managersand supei-visors and front-line employees. The nature of tbis contact was tocommunicate support and respect for employees (e.g. mistakes were to be dealtwith in a non-coercive manner) and a positive valuing of their contribution to thebtisiness.

This aim mirrors contemporary arguments, based in turn on htiman relationstheoi-y (Likcrt, 1961), that supportive behaviour on the part of management andincreased iniormal contact between management and stall" will have tbe elTect ofreducing 'them and us' attitudes and will, in turn, contribute to the high trustnecessary for flexible behaviour (PfelTcr, 1998). This might, for example, translateinto more upward communication from front-line stafi" about customer desires andideas about improving service. A related asstimption was that supportive and con-siderate treatment of employees by management would be directly reciprocatedby employees in their dealings with customers.

Role modelling in service quality. Consistent with arguments in the literature about theimportance of leadership in generating cultural change (Chin and Bemie, 1985;

C Blackwetl Publishirs Lid 2001

HRM ASSUMPTIONS 835

Schcin, 1992), managers were lo aci as role models in rcj ard to customer service.This mcaiu die active involvement of management in the training exercise (seebelow). It also meant frequently being seen at the sharp end, for example, workingon the tills or stocking shelves along wilh general staiV. Phially, it often ineaiu grant-ing accommodations to customers that would have been surprising in ibe 'oldregime', and then disseminating stories alxiut tbese instances around the ronipiiny.This can be seen as a mechanism by which stories, legends and myths are gener-ated as a way of s^inboiizing and redefining the organization and tlie service itonens customers (Keenoy and Antliony. 1992).

Management behaviour was, in short, to aet as a visible demonstration toemployees of senior level commitment to Ser\'ice Excellence. This aspeet of theinitiative reflects the assumption, consistent with prescriptions in the HRM andquality literatures that management commitment is crucial to the success a\' inter-\'entions such as this (Hill and Wilkin.son, 1995; O;iklancl, 19913). It increases thelikelihood that needed resouires will be made available, for example. Kurtlier, andbnportanily, it is assumed that management's commitment in the ejvs of shopfioorstaff is a necessary condition to employees' own acceptance of cultural change(Schein, 1992).

Job RedesignJob redesign constituted the third main element of the programme. Front-lineworkers were to be given increased discretion in their dealings with customers, intwo senses. First, they were allowed greater autonomy in terms of actions they couldtake on behalf of customers without reference to supen. isors. These included grant-ing exchanges or credit for unsiitisfactory products; taking the customer's word inevent of a disputed price; granting accommodations, writing letters and oiTeringgifts in the event of customer dissatisfaction; and removing from the selling floorany product tluit did not in their view meet qtiality standards.

Second, front-line workers were given wider latitude in how they interacted withcustomers. Employees were no longer bound by a company 'script' mandatingforms of words and requisite smiles. Rather they t ould converse wilh customersin ways the employee deemed appropriate, lor example, to call a customer 'love'if that is the natural form of address; or to be chatty and sociable, or more formal,depending upon the indi\idual customer's proclivities.

A key rationale for increased discretion in front-line jobs was to introduce scopefor continuous improvement on the pan of individual emplo^x'es. It was a.ssumed,in other words, that employers would use their increased autonomy for the benefitof custoniers. This increased discretion was to take place witliin a climate <A' hightrust evidenced by supportive supemsion and management {see above). It wasespoused as vital that the inevitable mistakes made by stafT should not be censuredand that correct decisions should be met with ready praise.

This shift towards 'self-management' is consistent with prescriptions in the ser-vices management literature. The nature of .ser\'iccs (e.g. the unpredictabilityarising from the participation of custcjmers) has led many writers to argue thatreliance on rigid scripts and direct controls is inappropriate to quality strategies inthis .sector (Ashfortli and Himiphrey.. 1993; Bowen and Schneider. 19H8; Peters,1987; Schneider, 1990a). This perspective suggests tliat self-management andlatitude for self-expre.ssion are necessary conditions both for employee willingnessand capacity for quality serv ice.

(D Blackwell Publishers Lid 2001

836 R. PECCEI AND P. ROSEPn'HAL

The wider literature on job design (for reviews, see Parker and Wall, 1998; Walland Martin, 1994) provides some empirical evidence from manufacturing settingof a positive relationship between job autonomy and outcomes such as motivation(Hacknian and Oldham, 1980); performance (Wall et al., 1992); and more proac-tive role orientations (Parker et al., 1997). Few studies have been performed inservice settings; however there is eWdence that perceived autonomy is positivelyassociated with a sense of authenticity in the work role (Wharton, 1993).

StaJJ Training in Service Excellence

Training constituted the fourth main element of the inten,'ention. A set of moduleswas developed to communicate the new service values to employees, to gain theircommilmenl to the programme and t<} instil the necessary confidence and inter-personal skills. It was stressed to managers that the training sessions should bebased on group discussions, and that these should be unstructured, participativeand fun.

Qualitative interxnews and analysis of documentary material revealed the basic'messages' presented to stafF via the training ses.sions:

• The route to better service and their own sati.sfaction was not rules or acompany 'script', hut freedom: the freedom to 'be themselves' wilh customers;to respond ui ways appropriate to themselves and to the customer.

• That they could and should become 'excellent', through developing their ownself-confidence and drawing on tlieir existing strengths (e.g. bringing usefulstyles of interacting in their personal or home life into their dealings withcustomers).

• That ever\'one could win: customers and the company, but also employeesthrough the satisfaction of having done their job well and through personalbenefits of improved self-coniidence and social skills.

The training programme can be described as a normative-reeducative approachto change (Chin and Benne, 1985), offering staff a new way of looking at them-selves and the world. The content of this new view is redolent of the themes andlanguages of excellence and enterprise culture (du (iay, 1996; Sturdy, 1998). Rules,scripts and supervisors are to give way to enterprising individuals who develop anduse their OWTI personalities and make their own decisions about how best to servecustomers.

De.nred Outcomes of Service ExcellenceP.sychological empowerment. It was assumed that the above changes would foster theright conditions for the piogramine's success. As noted above, we term these con-ditions psvchological empowerment. Fii'st, employees would come to understand andinternalize the principles of Service Excellence, and therefore be clear aboutwhat was expected in regard to treatment of custoniers. Second, they would feelthey had the necessary skills and abilities to do their job well and to cope withunexpected problems. Finally, they would perceive tliey had tbe autonomy to(decide how to use their abilities for the benefit of customers. These primary mech-anisms would, it was assumed, deliver tbe ciistomci-orientcd beliaviour sought hymanagement.

© Bliickwcll Publishers Ud 2001

HRM ASSUMPTIONS 837

Custorner-oriented behaviour. The tiltimate aim of Ser\'ice Excellence was changes inemployees' behaviour towards customers, the general nature of which lia.s alreadybeen characterized as more personalized, flexible and receptive to individual cus-tomer demands. An essential point is that the wellspring of this service is notunpersonal rules and close supeivision, but employees' own commitnient lo cus-tomers. The creation of this commitment, and the customer-oriented behaviourit would spur, was the objective of the programme, and it was assumed lliat successhere would realize the company\s aim of enhanced service quality.

RESEARCH MODELS

The section above has oudined Senice Excellence and its underlying assumptions.In the seclion to (bllow, we evaluate empirically the soundness of those assump-tions. It is worth noting that we are not evaluating the impact of the programmeper 5ffJ'' Our aim is not, for example, to a-ace increases in customer-orientedbehaviour prior to and post the intenentiou. The focus of our evaluation Is, rather,the underlying logic of the progi^anime. This logic, simply stated, is that certainmanagement behaviours and human resource practices in the work context (asexperienced by employees) engender customer orientation. If these assumptionsare soiuid, then employees within the Shopco system who more strongly perceivethese behaviours and practices to be operating (e.g. who view their supervisor assupportive, or who feel they have autonomy in their job) should be more likely toengage in customer-oriented behaviour.

Against this background we now formally model the main assumptionsand h)iiotheses underlying the programme prior to testing them against theShopco data.

In line with the arguments of the programme outlined abov'c, the presentanalysis focuses on three m<iin sets of variables. The core dependent variablein the analysis is the extent to which staff engage in behaviour designed to satisfycustomers. More generally, the core dependent variable can usefully be thoughtof as a form of prosocial organizational behaviour (PSOB) (Brief and Motovvidlo,! 986) which is specifically directed towards customers. This type of PSOB includesdirect forms of helping beha\iour towards customers, as well as more indirectexpressions of prosocial behaviour in the form, for example, of suggestionsfor improvement in customer service. We refer to this t^pe of PSOB as cuslomer-oriented behaviour (COBEH), which we define as the extent to which employeesengage in continuous improvement and exert effort on the job on behalf ofcustomers (Pcccei and Rosenthal, 1997). COBEH is the key outcome variable inthe analysis and is hypothesized to be affected by two main sets of antecedentsrelating to the level of psychological empowerment experienced by employeeson the job, and to perceived management behaviours and HR practices in theorganization.

Clearly, there are various ways in which these sets of antecedents might belinked to COBEH. Here we identify and test three alternative explanatory modelswhich we label the simple additive, the full mediation, and the partial mediationmodels, respectively. The three models are .shown schematically in figure I. Westart by outlining the full mediation model since, of the three models, this is the

© Biackwell Publishers Ltd 2001

838 R. PECCEI AND P. ROSENTHAL

1 (B) Partial mediation model

PartiripaiioTi in SKIrainuii;

Siipijuniw .111 cl cusiiimrv

Suppttnivc and cusionicioriirnird siipcrvwioii

^ -

^ ""-"-v.. — ' — — — — ^

^^^^^y^ ^-^-"^^

Intnrnalization of SE

Job awlotiomy

*

l[b) Pull medlarion model

l(c) Simple additive model

Job compeifiic

Siipportiw and custoiiif roriented siipcn'uioi)

Job autonnm\

ritrtirr 1. .Mtcrnative explanatory models of COBEH

—

Rirtkipalion in SEtraining

Siip]«irtivc and (iisioiiuT

Suj.)|>iirtivv ami iiisioinrroricntfd !iii]«-T\ision

Intcrnali/ation (if SH

Job compete lit f

Job autonomy

COHMI

one that most closely captures and reflects the key assumptions undcrlyiug the SEinitiative. We then briefiy outline the alternative models.

hull Afcdiation ModelAs shown in figure l(b), the full mediation model assumes that there is no directrelationship between the set of perceived management behavioiirs and HR prac-

<>Blaclnvrll Publishers t.id '2001

HRM ASSUMPTIONS 839

tices and employees' level of customer-oriented behaviour. Rather, thi.s modelhypothesizes that the impatit of the management beha\iour and HR practice vari-ables on COBEH is purely indirect, being completely mediated by the level of psy-chological empowerment of employees. Thus, in line witli the main assumptionsof the programme, the model hypothesizes COBEH to be directly alletted byemployee empowerment which, in turn, is assumed to be influenced by the set ofperceived management behaviours and HR practices in tlie organization.

The notion of empowerment is captured l)y means of three variables in thepresent analysis, namely, internalizalion of SE,job competence, and job autonomy. Thesevarialjles, as noted, closely parallel the 'meaning', 'competence', and 'self-drtermination' components of psychological empowerment ideniilied by Spreitzer(1995) as part of her four-dimensional conceptualization of this construct. Specifi-cally, internalization of SE refers to the extent to which employees feel they havea good understanding of what Service Excellence is about, think SE is importantfor the future of their store, and are committed to it. Employees who exhibit astrong sense of internalization of SE are expected to place a stronger value onproviding high quality service to customers and to have a greater understandingof their role requirements in relation to customers. As such, they are expected toexhibit higher levels of motiv ation to engage in customer-oriented job behaviour(Hackman and Oldliiun, 1980; Porier and Lavvlei; 1968).

Job competence relates to individuals' feelings of personal mastery and self-efBcacy linked to the performance of their work roles (Bandura, 1989). Tt refers tothe extent to which indiv iduals perceive tliat they have the necessaiy training, skillsand competence to do their job well and cope with any unexpected problems intheir work. Job autcinomy refers to the extent to which employees feel they are ableto make their i vvn decisions m their job. As such, it is closely linked lo the degreeof control and self-determination which individuals feel tliey are able to exerciseat work (Deci et al., 1989; Wall ct al., 1992) as reflected, for example, in theiropportunity to decide about the pace and methods of work (Hackman andOldham, 1980; Spector, 1986).

More generally, the present notion of empowerment could be said to parallelstandard models of individual job performance proposed in the organizationalpsychology and behaviour literattire (e.g. Campbell and Pritchard, 1976). Thesemodels view performance as a function of three main factors, namely individuals'motivation, (ompetence and opportunity t<j use iheir skills at work. The internal-ization, competence and autonomy dimensions of empowerment identified abovemay be thought of as tapping employees' willingness, capacity and opportunity toengage in COBEH, respectively, and, therefore, as closely corresponding to thethree main components of standard models of job performance. Hence, they rep-resent imj)ortant preconditions of customer-oriented behaviour, and in line withthe assumptions of the programme, are each hypothesized to be positively relatedto COBEH.

Employee empowennent is, in turn, hypothesized to be affected by the set ofmanagement behaviour and HR practice variables on t]ie left hand side of themodel. These variables direct attention to the remaining set of programmeassumptions, namely those relating to managerial role modelling, supy^ortiveleadership styles and customer service traming as means of empowering stalT andmaking them more customer oriented. (Note that the assumptions relating tojob design have already been captured indirectly in the rnodel through the job

e Blatkwcil Publisli<-rs Lid 2(HI1

840 R. PECCEI AND R ROSENTHAL

autonomy variable.) More specifically, role ninde/ling was initially captured by meansof two variables in the analysis, perceived mariagemenl commitment to lustotner .service andperceived supervisor commitmml to customer service. These variables refer to the extent towiiich employees perceive store management and immediate supervisors to beconimitted to customer service and to set an example of good SCIAice in tlieir dailyjob. Supportive leadership was also initially measured by means of two variables,namely perceived management concern and perceived mperiyisory support. T h e fu-st of thesevariables refers to the extent to which employees perceive that they are treatedfairly and sympatlietically by store management, while the second refers to theextent to whi( h they perceive iheir immediate supervisor to be supportive and par-ticipative in his or her behav iour.

As we explain more fuUy below, initial analysis of tJiese various constructsrev-ealed that employees in Shopco did not make a clear distinction between theperceived supportiveness and tiie perceived customer (orientation of their imme-diate supenisors. Nor did they make a clear distinction between the supponive-ness and customer orientation of store management. /Vs a result, we decided tocombine the perceiv ed role modelling and leadership style variables into two com-posite constructs tapping individuals' overall evaluation of their immediate super-visor and of store management, respectively. The.se composite constructs are theones included in the models in figure 1 under the labels of supportive and customerorimUd supervision and .supportive and cu.stomer oriented management. Finally, the assump-tions relating to SE tiainiug arc captured in the models by means of a single vari-able mca.suring whether individuals had participated in any of the SE trainingsessions at their store.

In line with the assumptions of the programme, the management behaviourand HR practice variables are hypothesized to be related to employee empower-ment. Although the precise links involved with the diflerent dimensions of empower-ment are not spelled out in detail in the programme, diflerent strands of researchand literature suggest that the relevant links in the model are likely to be positive.More specifically, the literature on organizational training and socialization (.'Mienand Meyer, 1990; Cascio, 1998; Thurley and Wirdenius, 1973; Van Maancn andSchein, 1979) suggests that employees' internalization of prosociiil customerservice values and norms, as well as their acquisition of relev'ant customer serviceskills, IS likely to be strongly influenced by two main factors; (I) the extent of theirinvolvement in relevant planned processes of" organizational training and educa-don; and (2) the strength of the organ ization iU climate for customer service. Thefirst factor directs attention to formal processes of customer service socializationwithin the organization, wliile the second focuses on more informal processesbased on individuals' day-to-day experiences and interactions.

Central to formal processes of customer service socialization within contempo-rary organizations, as we have seen, are .systematic programmes of customer t areand service training targeted at both new recruits and existing staff (Norniann,1991; Schneider, 1990a). These programmes are concerned as much with devel-oping a stnmg sei"vice orientation among front-line stafl, as they are with impart-ing relevant job skills to employees and acquainting tliem with new service qualitystandards, tools and techniques (Roberts and Corcoran-Nantes, 1993; Rosenthalet al., 1997). The success of these programmes is clearly likely to vary dej>endingon a whole range of factors, including the nature of the programme itself, its dura-tion, and the combination of training and socialization mechanisms employed

e BlackwrU Pubtishers Ud 20(11

HRM ASSimPTIONS 841

(Schein, 1992;Wanous, 1980). Howevct; other thins^beint;; equal, systematic train-ing in senace quality and exposure to customer service values and norms can beexpected to facilitate the development of both a strong service orientation and astrong sense of job competence amongst employees.

Participation in formal training and education programmes is, however, onlyone means through which employees may come to acquire relevant job skills andassimilate prosocial customer service values and norms. Equally, if not moreimportant, is the ongoing process of learning and socialization which occursthrough day-to-day experiences in the organization. Particularly important in thisrespect is the organizational climate for customer ser\ i( e (Schneider, 1990b). Thaiis lo say, ihc extent to which employees perceive customer senice to be a majorfocus of concern within the organization (Peccei and Rosenthal, 2000).

Central to this arc their perceptions of the extent to which management andsupemsors place a high priority on customer-oriented service iuid beha\'e in waysthat are consistent with espoused customer service values and norms. The strongerthe perceived cust(jnier service climate, tlie more systematic tlie situational cueing,validation and reinforcement of prosocial customer service values and norms islikely to be. Such 'strong' situations (Mischel, 1977) should facilitate the develop-ment and assimilation by employees of prosocial customer service standards andskills b) signalling to them what is \iilued in the system and by focusing their atten-tion on widely shared attitudes and behaviours (O'Reilly and Chatman, 1996;Peccei and Rosenthal, 2000). Hence, we expect both ihe perceived mana,t!;etnentand supervisory behaviour variables in tlae model to contiibute to tlie developmentof a strong service orientation, as well as to a strong sense of job competence,amongst employees.

The management and supervisory behaviour variables in the model can also beexpected to have an impact on the third and last dimension of empowerment,namch- employees' perceived degree of autonomy on the job. Clearly, the degreeof control individuals are able to exercise over various aspects of tlieir work is likelyto depend, above all, on the way in which jobs are designed in the first place(Hackman and Oldham, 1980; Parker and Wall, 1998). leadership styles, however,can also make a difference. In particulai; as human relations writers (e.g. Likert,1961) have traditionally emphasized, and as is also suggested by some of the morerecent leadership literature (e.g. Conger and Kanungo. 1987; Evans, 1970),'command and contror forms of supei-vision are likely to consuain tiie exercise ofjob autonomy by subordinates. Lf:tadership styles that are perceived to be more sup-portive and participative, on the other hand, are likely to enhance subordinates'scope for decision-making and facilitate the development of a stronger sense of jobautonomy. Hence, we expect tlie two management and superv' isory variables in themodel to have a positive effect on joh autonomy. We do not, howex'er, necessarilyexpect job autonomy to be direcdy influenced by the training participation variable.

Alternative Explanatory ModelsIn the full mediation model outlined above, the three empowerment variablesare treated as intervening variables mediating the impact of the managementbehaviour and HR practice variables on COBEH. This does not necessarily meanthough that the latter variables, in addition to having an indirect effect on COBEHthrough their impact on the empowerment variables, may not also have a directeffect on customer-oriented behaviour. This would imply only partial, rather than

© Biackwclt Publishers Ltd 2001

842 R. PECCEI AND P. ROSEiNTHAL

complete, mediation vvitli respect to the intervening empowerment variables(Barron and Kenny, 1986), and suggest that, apart from psychological empower-ment, there may also be otJier mechanisms linking the set of perceived manage-ment hehaviours and HR practices to COBEH.

The programme assumptions are not specific on this point. Theoretical andempirical work on prosocial and organizational citizenship behaviour suggests,however, that a partial mediation tnodel of the t^pe depicted in figure l(a) maywell represent a plausible alternative to the complete mediation model outliiiedabove.'' Central to this new model is the idea that, because of their potentialimpact on employee job .sat:isfaction. commitment, and positive alfect, the man-agement behaviour and HR practice variables are likely to have a dijx'ct positiveeffect on COBEH. This direct efiect is assumed to be above and beyond theindirect efiect tiiat those variables are hypothesized to have on COBEH throughtheir impact on the intei^vening empowerment variables.

Finally a second alternative to the full mediation model discussed above, is thesimple additive mode! depicted in figure l(c). Although this model incorporatesseveral of the hypotheses from the previous two models, its distinguishing featureis the absence of mediation effects. In other words it does not explicitly hypothe-size any links between the set of management behaviour and HR practice vari-ables and the set of empowerment variables. Thus., the predictor variables in thismodel are all assumed to be at tlie same level of analysis and hypothesized to havea direct additive effect on COBEH.

In summary, in this section v\e modelled the main assumptions underlying theSE initiative in Shopco. f his includes the full mediation mode! outlined above andtwo plausible alternative models. The three models are tested below against theShopco data.

METHODS

Dafn Sample and Methods

The three models were tested using data fn:)m a large-scale attitude survey cover-ing approximately 2100 staif employed in seven Shopco stores. The stores includedin the survey em])Ioyed between 90 and 450 employees. The survey was conducteda few months after the launch of the SE initiative, and was part of a widerevaluation of the SE programme witiiin Shopco (see Rosenthal et al., 1997). Theresearch instrument used in the survey was a self-completion questioiuiaire whichwe distributed to all stall' in each of the participating stores.

A 35 per cent response rate was obtained for the survey, providing a total sampleof 711 respondents, including .54 supervisors and 663 general staff (e.g. tilloperators, stock controllers, shelf fillers). The sample was representative of thestore populations in terms of a number of key demographic characteristics includ-ing age, sex, part-time/full-time status and tenure in the organization.

MeasuresThe individual questionnaire items used to construct the relevant scales are shownin table I. Responses to all items were scored on five-point Likert scales measur-ing respondents' degree of agreement or disagreement with the item in question(1 = strongly disagree; 5 — strongly agree).

O Blackwell Piiblisht-rs Ud 2001

0.860.840.78

0.070.120.28

0.040.170.11

3.4938.8

0.070.130-27

0.800.790.G4

0.150.120.48

1.5617.3

0.090.100.09

0.230.120.07

0.870.860.53

1.06II.H

HRM ASSUMPTIONS 843

Table I. Factor analysis results

Factor / Factor 2 Factor 3

(a) Intgrnalization of SE,job competence, and job autonomy items

InternalKation of SEI aiTi strongly cummittcd lo SESE is important for llie iulurf of my storeI feel I have a good understanding of what SE is about

Job rompdrnci'1 knou' how lo deal with most pn)l)lfms in my jobI have had enough training to do my job wellI am always comfortable dealing with customers

Job autonomyI ran make my own decisions in carr^ong out my job1 have the freedom to de( ide what I do on my jobI can use my personal judgt-mcnt in carrying out my job

Ejgenvabie% variance explainedK M O = 0.79: BTS = 0.00000

(b) Role modtUi}^ and leadershi/r style itenvi

Supervisory supportMy ininifdiati- boss .support.'i me in getting my Job doneMy immediate boss praises me when I do a good jobMy iiratiediiite boss encourages me to speak up when I

disagree wilh a decision

Supervisor commitment to customer serviceMy immediate boss sets a personal example of good

customer .service in his/her daily jobMy immediate boss puts a lot of emphasis on giving

good service to customersMy immediate boss is genuinely committed to SE

Xianagemmt commitment to customer serviceThe senior management team in my store sets a personal

example of giwd customer service in its daily activitiesManagement in this store is genuinely committed to SE

Atanagemmf concernThe senior management team in my store is sincere in its

attempts to meet the employees' point of view1 fi'cl quite confident ihat management in this store will

alwa>'s tr\ to treat me fairlyThe senior management team in my store would be quite

prepared to gain advantage by deceiving employees(Re\'ersed)

Eigenvtilue% \'ariance explainedKMO = 0.89: BTS = 0.(10000

0.810.810.79

0.74

0.70

0.70

0.22

0.27

0.24

0.36

0.02

5.4549.6

0.250.080.07

0.33

0.37

0.43

0.7H

0.73

0.76

0.72

0.63

1.4813.6

© Blackwell Publishers Ltd

844 R. PECCEI AND P. ROSENTHAL

Dependent variable. Customer-oriented behaviour was measured tising the six-itemGOBEH scale developed by Peccei and Rosenthal (1997). This is designed to taprespondents' (seif-ieported) propensity to engage in continuous improvement andexert eflbrt on behalf of customers. Sample items included 'I put a lot of elTortinto my job to xry to satisfy customers" and 'I often make suggestions about howto improve customer service in my department'. The full scale showing the threeitems designed to tap continuous improvement and the three effort items are repro-duced in the Appendix. Factor analysis (all factor analyses used principal compo-nents with varimax rotation) of the six items yielded a single factor which, whenscaled into an overall measure of ClOBKH, exhibited an acceptable level ol' inter-nal reliability {coefficient alpha = 0.80). Half the sample rated tJiemselves relativelylow on this measure, scoring below four on the five-point scale.

Intervening empowerment variables, hterna/ization of SE and job competence were eachmeasured with three questionnaire items developed specifically for this study. Jobautonomy was measured with three items adapted from Price et al. (1992). Whenfactor analysed together these nine items loaded on three separate factors corre-sponding to the three hypothesized dimensions of empowerment and were scaledaccordingly. As can be seen from table I(a), however, the last autonomy item alsoloaded hea\'ily on the job competence factor and was therefore dropped from theconstmction of the filial job autonomy scale.

The results of the factor analysis indicate that the three empowermentmeasures represent separate constnicts. As might be expected, however, thesethree variables are moderately interrelated. The average intercorrelation betweenUiem is 0.3S. In the following analysis we will treat them as separate variablesrepresenting distinct, but related, dimensions of p.sychological empowerment onthe job.

Alariagemeiit behaviour and HR practice variables. Perceived martagenwnt and perceived .super-visor commitment to customer service were measured with two and three items respec-tively, all developed specifically for this study. Perceii'ed management concern wasmeasured using an adapted three-item version of the 'faith in management' com-ponent of Cook and Wall's (1980) interpersonal trust a( work scale. Perceived super-visory .support was measured with a three-item scale adapted from Price et ai. (1992)(see table l(b)). Finally, participation in SE training was measured by asking respon-dents whether they had taken part in any of the formal SE training sessions heldat the stores (no = 0; yes = 1). Overall, nearly three quarters of respondents hadundergone SE training by the time of the survey.

As can be seen from table I(b), when factor analysed together the 11 rolemodelling and leadership items yielded oniy two rather than four factors. The firstfactor included all the items relating to respondents* assessment of their immedi-ate supervisor, whiie the second compri.sed all the items relating to management.,\s a result, we decided to recombine the items into just two composite variables.The first of" these composite variables, labelled supportive and customer orietjted .super-vision, was constructed Ijy scaling the six items from factor one in table I(b). Theother composite variable u.sed in the analysis was labelled supportive and customeroriented management and was constructed by scaling the five items which loaded ontothe second factor in the table.

O BlacUcIl Publishers Ltd 2001

HRM ASSUMPTIONS 845

Control variables. We used a number of control variables in the analysis. These weredesigned lo capture basic structural and demographic factors that might have animpact on COBEH as well as on the predictor variables in the model. Tbe con-trol measures used included a series of store dtimtny variables, hierarchical hvel(1 = general stall; 2 = super\'isors), age^ organizational tenure, sex (0 = male; I = lemale),work status (0 = part-time; 1 = full time), and frequency of cmtomer contact on the jobmeasured on a five point scale.

Table II presents means, standard deviations, correlations and reliabilities(where appropriate) for all the variables used in the analysis, including the controlmeastires. As can be seen, all multiple-item scales exhibited acceptable le\'els ofinternal reliability.

Analysis Procedures

The fltil mediation, partial mediation and simple additive models presented infigure 1 were tested and compared using hierarchical regression analysis. As weexplain below, the procedures used lo test the ftill mediation model also seivcd asthe Ijasis for evaluating and comparing the three models. Central to tlie fullmediation model is the assumption that ihe impael of the various managementbehaviour and HR practice (MB&HR) variables on COBEH will he tomplctelymediated by the three empowerment variables of internalization of SE, job com-petence, and job autonomy. Following Barron and Kenny (1986) we tested formediation in three steps. First, we regressed each of the three empowerment \'ari-ables on tbe set of MB&HR variables. Second, we regressed COBEH on ihe setof MB&HR variables. And third, we regressed COBEH on tbe set of MB&HRvariables and on the three empowerment variables together, controlling in eachcase for the background structural and demographic factors. In all cases the set ofcontrol variables were entered first into the ecjuations, followed, where relevant,by the set of empowerment and MB&HR variables respectively

For complete mediation to he operating the following four conditions must hold.First, the MB&RH variables must significantly affect the empowerment variablesin the firsi sel of regressions so that when they are added to ihe control variables,tliey significantly add to the level of explained variation in the empowermentvariables. Second, the MB&HR variables must signifieanUy afTcct COBEH in ihesecond equation so thai when they are added to the control variables, they signifi-cantly add to the le\el of explained \ariation in COBFJH. 1 hii'd, the empower-ment variables must significantly afTect COBEH in the tliird equation so that whenthey are added to the control variables, they significantly add to the lo\el ofexplained variation iii ClOBEH. /Vnd fourth, tlie impact of the MB&HR \ariableson COBEH must no longer be significant in the third equation so that addingthem to the set of (ontrol antl empowerment variables in the analysis does notsignificantly add to the level of explained variation in COBEH.

Provided that the signs on individual eoeflicients are in die predicted direction,empirical confirmation of all four of the abo\c conditions provides support forthe full mediation model. By implication, it also pro\ades pnma facie evidenceagainst the other two models since it indicates that the management behaviourand HR practice variables do not have a direct effect on COBEH. It would alsosuggest that the empowerment variables are significantly linked to the MB&HRvariables. This is contrary to the assumptions of the simple additive model and

© Blackwell Publishers Ud 2001

846 R. PEGCEI AND P. ROSENTHAL

Table II. Means, standard df\,iations, reliabilities and correlations \'aHabIcs

1 Store 22 Store- 33 Store 44 Stort- a5 Store 66 Store 77 Hierarthical level8 Organizational tciiurr9 Agt-

10 Sex11 VVt)rk status\i t'.iistrjnier comatt13 Paniiipaiion in SE

training11 Supporlivc and

custnnn'r-fjririucdnian:ii!i'ni(-iii

15 Siip|»orti\r andt-ustnnier-iiricnled

16 lni<'rnali/:ition of SE1 7 Job rntnjK'Icncr18 Job aiiloniniiy19 ('iisnimor-iirienu'd

beliavi(>ur|t:OBKH)

Metut

0.19O.iO0.200.120.070.181.074.97

32.830.750.294.340.74

3.42

3.M

4,234.183.473.85)

sn

0.390.300.400.320.250.390.264.00

11.600.430.461.240.44

0.78

0.81

0.650.580.890.54

/

-0 .16-0 .24-0 .18-0 ,13-0 .23-0 .02

0.210.190.11

-0.110.080.13

0.18

0,02

0.190.09O.(M(t.OR

-0.16-0 .12-0 .09-0 .16

0.010.100.060.07

-fJ.O2-0.00

0.10

0.04

0.04

0.090.03

-0.070.04

-o.ia-0.14-6.24

0.00-0.06-0.12-0.04-0.02-O.OI

0.20

-0.19

-0.06

-fl.O50.000.01

-0.02

• /

-0.10-O.IH-0.0.5-0.00-0.09-0.04-0.03-0.12-0.30

0.13

O.UG

-O.Ofi-0 .07-41.03- ^ . 0 4

.'1

-0.13O.OI

-0.13-0.04

0.020.0.50.02

-0.04

-0.00

-0.03

-0.05-0.03-0.01

(1.01

0.03-0.02

0.040.03

-^).03-0.04-0.20

-O.IO

O.(K>

-O.IO-0.04

0.03-O.IO

7

0.240,04

-0.160,31

-0.040.09

0.00

0.06

0.070.000.20O.lll

Cratii>arh'!i alphas npiirl ^ r r>0 ,07 [>< 0.(15.l-brr>0.1I p<O.OI.lwr>0.i:ip<0,OOI.

on ihc in I I tnr llir ini[llipli--ilriii si".

would thrrefore further undermine the validity of this model compared to the fullmediation one.'''

RESULTS

The results of the regression analyses used lo test the various hypotheses andcompare the three models are shown in table III. Tlie first three eciuationsexamined the impact of the pereeivcd management behaviours and HR practiceson the three empowerment variables of internalizaiion of SE, job competenceand job auionomy. resjx'ctivcly. Equation 4 examined the impact of the manage-ment behaviour and HR practice variables on COBEH, while equation 5 exam-ined the direct impact on ClOBEH of both Lhe empowerment and themanagement behaviour and HR practice variables together.

Before examining the results for the main variables in the study, it is worth notingthat only three of the control \ariables in the final regression analysis (equation 5)had a significant impact on COBEH. These included employees' age and fre-quency of contract with customers, which were both positively related to COBEH,and organizational tenure, whicii was negatively associated with custfjmer-orientedbehaviour. Age and frecjuency of customer contact also emerged as .significant pre-dictors of some of the empowerment variables (see equations I to 3). None of theother control variables, including any of the store dummies, had an impact on

O Blarkwtll Publishers Ltd 2001

HRM ASST rMPTIONS 847

10 tt 12 i:i t-i /.j Ki n iH 19

0.420.05 0.240.22 0.02 -0.35O.(M -0.02 0.04 -0.060.28 O.Il 0.07 0.12 0.28

-0.09 0.10 0.05 -fl. 17 O.(H -0.07 fO.R2|

-0.04 O.OH -O.OI - 0 . 0 8 0 .08 0 .02 0.56 [0.89]

0.090.040.080.06

0.20O.OI0.090.25

0.070.02

-0.13-0.01

-0.03-0,02

0.200.02

0.190.260.070.23

0.220.19O.O.T0.17

0.41O.!80.200.27

0.370.290.300.30

fo.aoi0.380.240.58

[0.69]

0.40[0.76]0.30 [0.80]

COBEH, althougli some of them (r.g. hierarchical level, sex, and work status) weresignifitanLly related to reported levels of job autonomy in our sample.

Turning to the main variables in the analysis, the results in table III provideclear support for all four of the test conditions for mediation ouliincd ahove. First,equations 1 to 'S show dial the management behaviour and HR practice variablestogether were able to explain a significant amount of extra variance, over andabove that exjilained by the control variables, in all three of the empo\Aermentmeasures (for internalization of SE: AR" = 0.17; AF = 54.38; p < 0.000; for jobcompetence: AR^ = 0.08; AF = 20.54; p < 0.000; for job autonomy: AR- = 0.07;AF = 19.68; p < 0.000). Second, equation 4 shows ihai when tlic ('mpowcrnicntvariables were noi included in the anaiysis, the MB&HR \ariablc.s logcthcrexplained a significant amount of extra variance in COBKH, over and above tliatexplained by the control \'ariablcs (AR^ = 0.08: AF = 22.15; p < 0.000). 'Fliird, thefinal regression (equation 5), which int hided both the empowermeni and MB&HRvariables, shows that the empowerment variables were able to explain a significantamount of extra variance in GOBKH, over and above that explained b\- thecontrol variables (AR-' = 0.28; AF = 108.05; p < 0.000). And lourtli, equation 5also shows, however, that once the empowerment variables were added to theanalysis, the MB&HR variables were no longer able to explain a significantamount of extra variance in COBEH (AR' = 0.00; AF = 0.61; p > 0.05).

Taken together these results provide strong support for the full mediation modelrather than for either one of tbe other two alternative explanatory models of

© BlackwL-lI Publislicrs U d 2001

848 R. PECCEI .\NY) V. ROSENTHAL

Table [II. Miiltipic n'gn-ssion rrsnlts

Independent mmbte.\

Controt vartahle:>

Store 2Store 3Store 4Store 5Store 6Ston- 7Hierarchical levelOrganizational tenureAgeSexWork statusCustomer contact

.MB and HR varinhles

IWticipaiion in SE trainingSupportive and customer

orienied managementSupportive and customer

oriciiU'd supcrvi.sion

Empowermeni iviriabte.i

Ini ernai ication of SEJob competenceJob autonomy

Total FAdj. R'(N)

Control variables; AR^AF

Empowemient variables: AR~AF

MBit HR viirial)l<'s: AR-'AF

Dependent vnriat>k

Ihiemalization

ofSE

0.070.04

-O.OI-0.03-0.03-0.05

0.03-0.03

0.1.3**0.020.020.12***

0.15***0.27***

0,20***

0.30***18.57***0.28***

(671)

0.12***7.74***

0.17***54.33***

s

2Jobcompettnce

0.01-0.00-0.02-0.05-0.03-0.03-0.02

0.04-0.03

O.OI-0.02

0.19***

O.U*0.02

0.26***

0.1.1***7.96***0.13***

(671)

0.07***4.42***

0.08***20.54***

3Jabaulnmmy

0.06-0.05

0.060.01

-0 ,0 !0.070.13**O.OI0.09*

-0.08*0.21***0.06

-0.020.12*

0.20***

0.19***10.26***0.1 7***

(671)

0.12***7.29***

0,07***19.68***

4COBEH

-0 .03-O.OI-0 .02-0.02-O.OI-O.IO*

0.07-0.08*

0.27***-0.04

0.030.20***

0,08*0.12**

0.20***

0.23***12,98***0.21***

(671)

0.15***9.74***

0.08***22.15***

COBEH

-0,07-0 .02-0.02

0.000.00

-0 .080.05

-0 .08*0.21***

-0.05O.OI0.11***

0.00-O.OI

0.05

0.43***0.17***0.09***

0.43***27.72***

0.42***(671)

0,15***9,74***

0,28***108,25***

0.000.61

Figuo-s in llir lop pan of itic tabk- are sEandaixlized liclii toclIicicnLs.• p < 0.03, ••p < O.Ul. ***p < 0,(101.

COBEH. tn partieular, they indicate that while the management behaviour andHR practice variables did not have a direct impact on COBFH. they had asigniikant indirect effect through their impact on the three employee empower-ment variables which were, in turn, fbund to have a significant positive effect oncustomer-oriented behaviour. Of the three empowerment variables, the inter-nalization of SE had by far the strongest impact on COBEH (see equation f)).'''

In summary, the results of our analysis provide strong .support for the lullmediation model and for the main assumptions underlying the SE initiative. They

Ptibtuhers Ud 200 i

HRM ASSUMPTIONS 849

suggest that participation in SE training, as well as both perceived managementand supervisory behaviour to\Aards customers and subordinates, had a significantpositive impact on employees' sense of psychological empowerment at work andthat ihis, in turn, had a strong positi\'e effect on their behaviour towards customers.Speciliciilly; otii' results indicate a clear tendency for employees who had internal-ized the message and values of Service Excellence, who felt they had the skills ajidcompetence to peiform well on the job. and who felt the)' had the freedom to maketheir own decisions at work, to be significantly more likely lo engage in proactiveforms of customer service behaviour.

DISCUSSION

A niUTiber of theoretical and policy implications emerge from ihc present research.At a theoretical and empirical level, the study contributes to the growing literatureon empowerment, its nature, conseqtiences and antecedents within a servicecontext (Bowen and Uiwler, 1992; Spreitzer, 1995, 1996; Wilkinson, 1998). Despitethe emphasis placed on empowerment in the quality, services management andcritical literatures (Bowen and Lawler, 1992; Hill and Wilkinson. 1995; Wilkinsonet al., 1992), the concept itself remains under-researched and little is known of itsantecedents and outcomes in organizational setting.

One implication of the sttidy concerns the need for a clear conceptualizationof empowerment. As noted earlier, no consensus on tlie concept or its opera-tionalization has yet emerged in the literature. The theoretically grounded, multi-dimensional approach taken in this study may provide a useful way forward. AsWilkinson (1998, p. 51) notes, 'employees interpret, eviiluate and (re)act towardsmanagerial initiatives, and . . . serve to audit in their own way the viability ofmanagerial initiatives'. The concept of psychological empowermeni as adaptedfrom Spreitzer (1995) oilers a way to capture these miployee interpretations atid tliusmove away from objectivist approaches to the concept, whether of a mainstreamor critical nature. For example, operationalizing empowerment through the per-ceptions of workers provides a complement to critical approaches restitig on a prioriconceptions of rhetoric masking increzised management control. Taking system-atic account of employee experience offers a way of exploring the extent lo wliichany 'rhetoric* has been accepted or rejected by workers, surely oi' interest in criti-cal analyses of contemporary management (Sturdy, 1998).

Our empirical findings pertaining to empowerment also carr\' implications forboth mainstream and critical reseaich. As stated above, all three dimensions of theconstruct were related to COBEH. This suggests that conceptions of empower-ment shotild not be redticed to notions of autonomy or self-determination. Workerdiscretion, in other word.s, may be a necessaiy but insufTicieni dri\ er ibr customer-oriented behaviour. Also required are a sense of personal competence, and mostimportantly according to our results, an agreement with organizational goiils atid\'alues about the importance of customer-oriented sersice. This more complexconception of empowerment as a multidimensional subjective state directs atten-tion beyond is.sues of job design to the management of meaning and culture withinorganizations.

More broadly, the finding.? from our study provide support for key argumentsand assumptions underlying mainstream HRM theory and practice. In particular

0 RIackwrll Publishers Ud 2001

850 R. PECCEI AND P. ROSKNTIIAL

they suggest thai htiman resource management sysiems, as percei\'eH and experi-enced by indi\iduais at the workplace, can significantly affect employees' pro-social beliaviours towards customers. Specifically, otir findings stiggest that theadojjtion by organizations of the set of management bchaviotirs and HR prac-tices examined here may indeed help to enhance customer-oriented beliaviouramongst front-lijie workers. These include first, the systematic provision of cus-tomer care training that emphasizes values as well as skills. Second, the demon-stration by management and supervisors of commitment to customer service aridsupport of subordinates in dieir day-to-day work behaviour. And third, the designof jobs so as to give front-line employees greater control over their work decisions.These practices should positively aifert the psychological empf)werment associatedwith customer-oriented beha\'iour.

This is not to suggest that the contingency arguments about empowermentadvanced by Bowen and I.^wler (1992) are misplaced. Under certain conditions(e.g. a strategy of C<JS1 lniuimization, customers who prioritize speed of service,etc.), a more standardized and tightly controlled production-line approach(Levitt. 1972) to service delivery- may indeed be ver\' effecti\e and result in highleveLs of customer salisfaclion. Our study is important, however, in that it showsthat even within the relatively routinized and low skill context of supermarketservice, the adoption of less Tayloristic practices which enhance employee em-powertnent may have a significant positive impact on work behaviour oi Iroiu-liiieemployees.

At a more general Ie\'el. the study contributes to an understanding of the mecha-nisms through which HR practices and percei\ed managetnetit conduct mayaffect individtial work behaviottr. Opening this 'black box' is crucial to ihe broadermainstream research agenda on the impact of HRM on organizational perfor-mance (Becker and Cicrhart. 1996: Becker and Huselid, 1998: (iuest, 1997).Specifically^ our study suggests that employees" psychological empowermeni islikely to be a core mechanism linking pereeivo<l tiumagement beh;t\ iours and HRpractices to COBl'iH. This is not to suggest that factors such as organizationalcommitment, positive aflect, or job satisfaction may not also ha\'e an importanteffect on customer-oriented behaviour. However important these other mecha-nisms might be thtitigh. our research suggests that a sense of empowermeni islikely to form a central part of any theoretical model designed to explain theimpact of HRM on this key aspect of work behaviour in a ser\-ice context.

CONCtAiSlONS

Despite the growing emphasis placed by organizations on initiatives de.signed lostrengthen customer orientation among front-line service workers, few systematicattempts have been made lo examine and to validate the assumptions underlyingsuch programmes. Our ptirpose was to address this gap through analysis of amajor customer care initiali\e in one of the largest supermarket companies in theUK, with a view to contributing to the wider debate in iho literature about theimpact of HRM on employee work behaviour and performance in a servicecontext.

To this end, we first reconstructed the main asstunptions underlying ihe pro-gramme. The.se are thai supportive leadership, management role modelling, job

C BlackwTU Publishers Ud 2001

HRM ASSUMPTIONS 851

redesign, and customer care training ean produce a sense of psychologicalempowerment among employees, and that empowerment will, in turn, enhancecustomer-oriented behaviour. These assumptions were captured in tliree alterna-tive models and tested using data from a large scale representative stiivey ofShopco employees. The results provided unambigiious support fbr the full medi-ation model of customer-oriented behaviour.

Specifically., two main findings emerged from the study. The first is the p{jsitiveassociation between all three dimensions of psychological empowerment and cus-tomer orientation. Autonomy, internalization of service values and a sense of com-petence were all related to customer orientation, with internalization emerging asthe strongest driver of COBEH. The .second is that all the perceived managementbehaviours and HR practices examined were linked to COBEH, but only indirectly.,thn)ugh their impact on empowerment. Specifically, customer care trainingenhanced both the felt competence of employees and their internalization of cus-tomer service values. Supervisors who were perceived to be supportive and cus-tomer oriented contributed to all three dimensions of employee empowerment,while perceived role modelling and supportive leadership by management con-tributed to the internalization and autonomy dimen.sions of empowerment, buthad no significant impact on employees' fell job competence.

Taken as a whole therelbre, the results of our research provide clear support forthe idea that the adoption by organizations of progressive human resource prac-tices and management behaviotus of the type examined here, ran indeed have apositive impact on the psychological empowerment experienced by employees, andthat this sense of empowerment can, in ttirn, significantly contribute to enhanc-ing the customer-oriented behaviour of the front-line workers in\olvcd.

Nevertheless, because of limitations in the present study, the results and con-ditsions otitlined above are necessarily tentative in nature. Further research isclearly required to test their robu.stness. First, future research should address thegeneralizability of our findings. ;\lthough based on a reasonably large and repre-sentative sample from seven stores, the present study is still limited to a singleorganization. The second issue concerns the cross-sectional nature of our data. Afull exploration of the catisiil links in the tnodcls of COBEH presented in die studyrequires the use of longitudinal data designed to capture changes in the para-meters of relevant variables over time. Such longitudinal research is all the moreimportant given the complex causal rcladonships that are Ukely to obtain betweensome of the main variables in the analysis.''''

The third Lssue concerns the use of self-report measures of customer-orientedbehaviour, as well as of other key antecedent variables in the analysis. As notedby Howard (1994), the use of self-reports may be less of a problem than is some-times a.ssumed in the literature. Furthermore, the self-report COBEH scale usedin the present study exhibited satisfactory psychometric properties. Nevertlieless,future studies shotild explore the tise of non-self-report measures of customer-oriented behaviour involving, for example, supervisor)' or customer ratings ofemployees' behaviour. This would help to minimize possible problems of common-method variance linked to the use of self-reports for both the dependent and theindependent variables in the analysis (PodsakofT and Organ, 1986; Spector, 1994).It would also serve to extend the analysis to tlie end users of the service. Thiswould enable researchers to explore the extent to which prosocial servicebehaviours actually result in customer satisfaction in different service contexts.

© Blackwtll Publishers Lid 2001

852 R. PECCEI AND P. ROSENTHAL

On a more substantive Ie\'el, future research might u.seftilly extend the presentanalysis in several ways. First, it can explore possible interactions between someof the main variables included in the present study. Second, it can explore addi-tional aspects of HRM systems (e.g. payment stnuiures, tcamworking) that mightinfluence the customer sei"vice behaviour of front-line staff, either directly, or indi-rectly through their impact on t!ie different dimensions of empowerment. Third,it can examine the relative influence on COBEH of employee empowerment andof the other tiiain mechanisms outlined in the study, including organizational com-mitment, positive job alfecl, and job salisfiu tion. And fourth, flittire research mightconsider the structure of empowerment it.self and examine possible causal linksbetween the different dimensions of this construct.

Finally, a nutnber of issues raised by this study are approjiriately accessed viaqualitative mctliods. One such example concerns the concept of empowerment,in particular otir finding of the relative importance of internalization of organi-zatioiiiil values and gt)als as a predictor (jf customer sen ice beha\ iour. This findingresonates with the emphasis now placed in both mainstream and critical literatureon the management of meaning and subjectivity' in organizations. In so doing, itraises fundamental questions about value intrrnalization among workers and howil relates to issues of identity, indoctrination and discourse (Rosenthal et al., 1997;Sturdy, 1998).

In conclusion, this study represents a first attempt formally to model and testassumptions about the link between perceived management behaviours and HRpractices, employee empowerment, and the ctistomer orientation of front-lineserxice workers. More research is needed to t onfirm the present findings and toextend and develop the analysis furtlier. However, the study provides strong initialsupport for some core assumptions underlying HRM theory and practice and, inso doing, provides an important insight into the mechanisms whereby humanresource management systems may be related to employee behaviour and perfor-mance in a seivice context.

NOTES

111 Fcir an evaluation of the impact of the intervention, sec Rosenthal ct al. {1997). Find-ings reported in Rosc-nthal t:t al. (1977) .sugge.st ihai key elemcnt.s of the intirveniiondid in practice ocriir. For example, employees reported in interviews iliat ihey were infact given more autonomy. This pertained to decisions that the>' could take withoutreference to supervisors, and to their freedom of expression with customers. Overall,qualitative and quantitative data revealed a strong endorsement of the ititenrtuionon tile part of front-line stafl.

[2] A detailed di.scussi(>n of this literature i.s beyond the s(ope of ihe present analysis (lora review, see Ixe, 1999). Here, it is suflieient to note t vo main points. First, that theavailable empirical evidence strongly suggests thai positive affect, as well as job satis-faction and organizational commitment, have a significant positive impact on \ariousforms of proseHial behavitnir at work (George. 1991; Moorman et al., l'J93; Organ,1988: O'Reilly and Chatman, 1986: OrRan and Konovsky, 1989; Organ and Ryan,1995; Weiss and Cropanzano, 1996). And scroncl, that supportive leadership and, toa lesser extent, the systematic provision of training to employees, have consistentlyheen found to he sigiiificaTU picdinors of employee job .satisfaction, organizationalcommimient and positive job afleft in the organizational hteratun* (Ix'e, 1999; IJncolnand Kal!clM!rg, 1990; Meyer and Allen, 1997; Speclor, 1997).

<0 Bbrlcwrll Publuhrr Ud 2001

HRM ASSUMPTIONS 833

[3] Spt'cifically, the partial mediation model is supported, above the other two models,oiily if (a) conditions one. two and three alx)\f hold, itiu/iln iliv iinpai:i ol" the MB&HRvariables on COBKH remains sig^iifirani in the ihird equation. Sijiiilarly, tlie simpleadditive model is supported, abo\e die other two modeLs, only if (a) condition oneabove is found not to hold, mid (b) the impact of both the MB&HR and ihi* em-powerment variable's on CIOBEH i.s significani in ilie ihird ecjuation.

[4] In term.s of indi\ idual links, ihe result.s in tal)le III provide no support for hypothesespositing a direct litik between each of the three management behaviour and HR prac-tice variables and COBLH. None of the beta etK-Hicients Ibr these \'ariables. in fact,achieve significance in equation li. The results do. htnvo'cr. provide siipimn for all butone of the hypotheses that ibrm part of the full merliation model. This is thr [)ntpose{Isigiiilicant positive rc-lationship between the percei\cd supportiveness and customerorientation of management and employees' sense of job competence. Thus, contrarylo expectations, the (oelficient for this variable in equation 2 failed lo achieve signifi-cance. Finally, as expected, parti( i|)aiion in SK training wiis not found to have a sig-nificant impact on job autonomy (sec equation 3).

[5] Some of these links, in faci, are likely to be non-recursive in nature and involvecomplex patterns of mutual causation over time One such set of relationships, forexample, is that Ijetween supportixe and customer-oriented sujjervision, job autonomy,•And CIOBEH. In thr full mediation model, the peireived nature of su[>er\ision Lstreated a.s an antecedent of autonomy which, in lurn, is treated as an antecedent ofctistomcr-orienled behaviour. Here though, causation may also br operating in theopposite direction. Employees who are more helpful lo customers, for instance, maywell be viewed more positively l>y customer-orieiUed supervisors who, a.s a result, mayalso be more supporti\e towards them and trust them enough to gi\e them greaterautonomy on the job. Future research might useliilly explore these more complex pat-terns of mutual causation among key variables using weU-fcH:used longitudinal designs.

APPENDIX: SCALE ITEMS

Customer-oriented Behaviour (COBEH)

ContintwtLS improvement items(1) I am always working to itnprove the ser\'ic(* T give to customers.(2) I have specific ideas about how to improx'e tbe .ser\''ice I give to customers.(3) I often make su^estions about how to improve customer service In my

department.

Effort items(4) I put a lot of effort into my job to try to satisfy customers.(5) No matter how I feel, I always put myself out for every customer ! serve.(6) I often go out of my way to help customers.

REFERENCES

ALLEN, N . and MEYER, J. (1990). 'Organizational s(x;iaUzation tactics: a bngitudinal analy-sis of links to newcomers' commitment and role orientation'. Academv of ManagementJournal, 33, IH7~5S.

AsHPORTH, B. and HtiMPimHv, R. (1993). 'Emotional labor in service roles: the influenceof identity'. AcMdemy of ManagemenI Review, 18, 88 115.

O Blackwell PublLshrni Ud 2001

854 R. PECGEI AND P. ROSENTHAL

A. A. (1989). 'Human agency in social cognitive theory'. American Psycholoffsl, 44,117.5 84.

BARRON, R . and KKNNY, 1). (1986). 'The moderator-mediator variable distinction in socialps>Thological research: conceptual, strategic and statistical considerations". J"'""""/«//'''''-ionality and Social Psycliology. 5 1 , 117!i-B2.

BECKKR, B. F.. and CIRHART. H. (1996). 'Tlie itnpaci of human resource management onorganizational peiibrniancc: progress and prospects'. Academy of Mariaganenl Review. 39 .779-801.

BECKER, B. E . and HiisEtjn. M. A. (1998), 'High performance work systems and firm per-formance: a s)iithe.sis of reseanrh and managerial implications*. Research in Ihsomiel andHuman Resource Manageninit. 16. .~)3-10I.

BOWEN, D . and LAWLKR, K. (1991^). 'The empowerment of service workers: what, why. howand when'. Sloan Management RemeuK Spring, 31 -9.

BOWEN, D . and ScttNKitJER. B. (1988). "Scnnce marketing and management: implicationsfor organizational behaxior". Research in Organizatumal Ikhaviour, 10, 43 80.

BRIEF, A. and MorowttiLO, S. (1986). 'Prosocial organizational behaviors'. Academy ofManagement Remew. 11, 710-25.

CAMrBKU,-J. and pRrrcibvRO. R. (1976). 'Motivation theory in industrial and organiza-tional psychology'. In Dunnette. M. (Ed.), Handbook of Industrial and OrganizationalPsychoiogv. Chicago: Rand McNalK; 63- 130.

CARLZON, J. (1987). Moments of Truth. Cambridge, MA: Ballinger.CASI^IO, W. (1998). Applied Psychology in Human Rrsaurce Mariagmnit. Ne\v York: Prentice-Hal I.CHIN, R . and BKNNK. K . (198.5). 'Genera! strategies for efiecting change iti human systems'.

In Beiinis, VV., Bonne. K. and (Jhin. R. (Ecl.s), The Plannirtg of Change. New York: Holt,Rcinehari and Winston, 22-45.

CONGER, J. and K.\NI'NGO, R . (1987). 'Ibwards a behavioral theory of charismaticleadership in organizalional settings'. Academy of Managemait Review, 12, 637 47.

COOK, J. and VVALI.. T . (1980). 'New work attitude measures of trust, organizationalcommitment and pei-sonal need non-fulfilment'. Journal of Occupational Psyc/u}log)\ 5 3 ,39 52.

CoRSt^, D. and ENZ, C . (1999). 'Predicting psychological empowerment among serviceworkers: t!ie clXvvi of support-I)asrd relationships'. Hitman Relations, 52 . 20.5 24.

I)i:.\N.J. and B{AVKN, D . (1994). "Management ilieoiy and total quality: improving reseairhand practice through theory development'. Academy of Management Review., 19, 392 418 .

DKCI. E . . CIOXNFJ.. J . and RYAN., R . (1989). 'Self-determination in a work organization'.

Juuriial of Applied Ps^rhoiugr, 74, 580 90.DU CAY. P. (1996). Constunption and Identity at Work. l/)iuion: Sage.EvAWi, M. (1970). 'The elfects of supen i.sor Inrhavior on the patli-goal relation.ship'.

Organisational liehavior and Human Performance, 5, 277-98.CEORGK.,J. M . (1991). 'State or trait: effects of positive mood on proscx'ial behaxiors at

\\'urk\ Journal of Applied Psychology, 7 5 . 107-16 .

GuEsr, D, (1987). 'Hvunan resource management and industrial relations'. Journal ofManagemmt Stutiifs. 24. .503-21.

GtiEST, D, (1997). 'Human resoune management and performance: a review and researchagenda'. The Internationa! Journal of Human Re.\ourtr Management. 8, 263- 77.

HACKMAN. J. and C)IJ>HAM, (i. (1980). Work Rede.'.ign. Reading. MA: Addison-Wesley.HACKMA.\.J. and WAGEMAN, R . (1995). 'Total quality management: empirical, conceptual

and prac tical issues'. Administrative Science Quartnly, 40 . 309 20.HARRIS. L . and OGBONNA, E . (1999). 'Developing a market-oriented culture: a critical

evaluation". Jnurridl of Management Studies. 36, 1 I 7 96.Hn.i., S. and WH^KINSON, \ . (1995). 'In search ttf TQM". Employee Relations, 17, 9-26.HOWARD, G . S. (1994). 'Why do people say nasty things about self-reports?'. Jowrnfl/^' Orga-

nizaiionat Behavior. 15, ;W9 404.

e BlackwfU Publishers Ud 2(H)1

HRM ASSUMPTIONS 853

KP-ENOY, T. (1997). '"HRMism" and ihc languages of re-representation'. Journal ofManagement Studies, 34, 825-41.

KEEKOY, T. and ANTHONY, P. (1992). "HRM: meaning, mfiaphor and morality'. InBlyton, P. and 'I'urnbuU, P. (Kds), Reassessing Human Re.source Management. l^)nclon: Sage,233-55.

KNIGHTS, D . and M(;C'ABE, D. (1997). "'How would you measure something like that?":quality' in a retail bank'. Journal of Management Studies, 34. 371-88.

LEE, H-J. (1999). /MFf ctive Stales al Work and Prosocial Organization id BehavTOur: A CaseStudy of Healtli Care Workers in the NHS'. Unpublished PhD Thesis, Ixindon Schoolof Eeonomirs and PolilieaJ Seience, University of London.

LEGOR. K . (1993). Ihtman Resource Alanagement: Rlutoric.s and Rmlities. Basingstokc:Macmiilan.

LEvrrr, T. (1972). 'Production line approach lo services'. Harvard Business Review, 50,802-10.

LIKERT, R . (1961). .New Patterns of Management. New York; McGraw-Hill.Li.NcoLN. J. and KAi.LF.BERt:, A. (1990). Culture. Control and Commitjnent. Cambridge:

Cambridge University Press.MEVER,J. P. and ALLEN, N.J. (1997). Commitment in the liwkpkce: Theory, Research, and Appli-

cation. London: Sage.MISCHEL, \V. (1977). 'The interaction of person and situation'. In Magnus.son, D. and

Kndler, N. S. (Eds). Personality at the Crossroads: Current Issues in Interactional Psychology.

HULsdale, NJ: Eribaum, 33:i-.i2.MOORMAN, R . H. , NrEHOFF, B. R and ORGAN, D. W (1993). 'Treating employees fairly and

organizational ciiizenship beha\ior: sorting tlie efiects of job satisfariion. organizaiionalcommitnient. and jjroredural justice'. Emplover Responsibilities and Rights "Journal. 6,209-25.

NOON, M . (1992). 'HRM: a map. a model or a theor\?\ In Blyion, P and lurnbull. P.(Eds), Reassessing Human Resource Management. Ijindon: Sage, 16-32.

NoRMANN, R. (1991). Sendee Management. New York: John WUey & Sons.O'REnj.Y, C. A. and CIIAITVLVN, J. A. (1986). 'Organizaiional commitment and psycho-

logicai altachment: The effecls of compliance, idenlification. and inlernaiization onprosocial behavior'. Journat of Applied Psyfliohgy\ 71, 492 9.

O'REILI.Y, C . and CiiArMAN, J. (1996). 'Culture as social control: corporations, cults andcommitment". In Staw; B. and Cummings, L. L. (Eds), Research in OrganizationalBehavior^ 18, 157- 200.

OAKLAND, J. (1993). Total Quality Management. Oxford: Butler^vorth-Heincmann.ORGAN. D. W (1988). A re-statement of the satisfaction-performance hypothesis'. Journal

of Management, 14, 547-57.ORG.\N. D. W. and K(JNOVSKV, M . (1989). 'Cogniti\e versus afTective determinants of orga-

nizational citizenship behavior'. Journal of Applied Psychology, 74. 157-64.ORGAN, D. W. and RYAN, K . (1995). 'A meta-analytic review of attitudinal and disposi-

tional predictors of organizational citizenship behavior'. I^omet Psvcliology, 48,775-802.

PAt4.\stiRAMAN, A., ZEITHAML, V. and BERRY. A. (1985). \\ roncepiual model of servicequality and its implications Ibr future research'. Journal of Marketing, 49, 41-50.

PARKER, M . and SI^UGHTER,J. (1993). 'Should Lhe labour movement buy TQ%1?\ Journalof (hgiinization Change Management, 6, 43 56.

PARKER. S. K.. and WALL. T I). (1998). JwA and Hork De.iign. Thousand Oaks, CA: Sage.PARKER, S., WALL, T. and JACKSON, P. (1997). "That 's not my job": developing flexible

employee work orientations'. Academy of Management Journal, 40. 899 929.P.A.1TERSON, M., WEST, M. , IJWVTON, R . and NiCKEii,,J. (1997). impact of people man-

agement practices on business pirrfbrmance'. Issues in People Management, 22. l.rf)ndon:Institute of Personnel and De\'eiopment.

© Blarkwrll PubiLshrn Ud 2001

856 R. PECCEI AND P. ROSKNTHAL

PECCEI, R . and ROSENTHAL, R (1997). 'The antecedent.s of employee commitment to cus-tomer service: evidence from a UK service context'. International Journal oJ Human ResourceManagement. 8, 66 86.

PFXCF.L R . and ROSENTHAL, R (2000). 'Front-line responses to customer care programmes:a theoretical atid empirical analysis'. International 'journal of Human Resource Management^11,562-90.

Pir;n.RS, T (1987). Thriving on Chaos. New York: Alfred A. Knopf.PtrrERs, T and WATERMAN, R . (1982). In Search of Excellence. New York: Harper and

Row.Pn->TER, J. (1998). The Hitman Equation. Boston, MA: Harvard Busines.s School Press.PonsAKOFF. R and ORGAN. D . (1986). 'Self-reports in organizational research: problems

and p'cos\>e(:vi. Journal of Management, 12, 531 44.PORTER, L. and LAWIXR, E . (1968). Managerial Attitudes and Performance. Homewood: Dorsey

Press.PRICE, J., MUEUXR, C . and CLTUUVA.N, D. (1992). Study of Teacher Turnover in the Chicago Ihiblic

.Schools. Mimeo, Univer.sity of Iowa.N. J. (1992). IntfUigent Enterprise. New York: The Free Press.

N, T. and MAIHEWS, B. (1996). 'Service quality and human resource management:a review and researcli agenda". Personnel RCVIFU; 27. .')7 77.

RoBER'is, K. and C <>Rt:oR \N-NAVi-ES. Y. (1995). 'TOM, lhe new training and industrialrelations'. In Wilkinson, A. and Wllmott, H. (Eds), Making Quality Critical: Plerspectives onOrganisational Change. Ixindon: Routlt;dge.

RosF.NTHAL. P., Hu.t,. S. and PECCEI, R . (1997). 'Checking out service: evaluating excel-Iciue, HRM and I'QN-I in retailing'. Work. Employment and Society, 11 . 481 503.

SCHEIN. E . (1992). Organizational Culture and Uadenhip. San Francisco: Jossey-Bass.ScHLESL\GER, L. and HESKETT,J. (1991). 'Breaking the cycle of faiJure in services'. Shan

Management Rei'iew. .Spring. 17-28.SI:HNEIDF-R, B. (1990a). 'Creating .service oriented organizations: simultaneous and sequen-

tial models for change'. In Bowen, D. E., Chase. R. B. and Cummings. T. G. & .\sso-ciates (Eds), Service Alanagement Effectiveness: Halanchig Strategy, Organization and Hum/inResources, Operations, and Marketing. San Francisco: Jossey-Bass.

SCHNEIDER, B. (1990b). 'The climate for sen ice: an application of the climate construct'.In Schneider. B. (I'-d.J, Organizational Climate and Culture. San Fran(isio: Jossey-Bass,383 412.

SEWIXL. G . and WHXINSON, B . (1992). 'Empowerment or emascttlation? Shopfloor sur-veillance in a total qualily oi^anisation'. In Blyton, P and Turnlmll. P (IJcis). ReassessingHuman Resource .Managetnent. Sage: London. 97 115.

SPARROW, P. and MARCHINGTON, M . (Eds). (1998). Humati Resowtce Managanent: The MewAgenda. London: Financial Times.

SPECTOR, P. (1986). 'Perceived control by employees: a meta-anaJysis of studies concern-ing autonomy and participation at work". Human Relations. 39, 1005 16.

SPECTOR, R (1994). 'Using s(4f-rcport questionnaires in OB researcli: a comment on theuse of a controversial method'. Journal of Organizational Behavior, 15, 385 92.

SPECTOR, R {\997). Job Satisfaction. London: Sage.SPREIT/ER. G. (1995). 'Psychological em|M)\\erment in the workplace: dimensions, mea.s-

urement and validation'. Academy of Managnnent Journal. 38, 1442-65.SPREIT7-ER, G. (1996). 'Social structural characteristics of psychological empowerment'.

Academy of Management Journal, 39, 483-504.STURDY, A. (1998). 'Customer care in a consumer .s<Kiety: .smiling and .sometimes meaning

it?". Organization, 5, 27-53.THOMAS. K . and Via.niotisii, B. (1990). 'Cognitive elements of empowerment'. Acadenty of

ManagemerU Review, 15, 666-81.TIllIRLE^^ K. and WiRDtNius, H. (1973). Supervision: A Reappraisal. Ixindon: Heinemann.

C BUrtnvrll Publishers Ud 2001

HRM ASSUMPTIONS 857

VAN M. \ANE. \ . J . and S<^HI-,IN. E . (1979). 'Toward a iheory of organizational socialization'.In Cummings, L. and Staw; B. M. (Eds), Research in Organizational liehavior, 1, 209 64.

WALL, T. and NIAJ<TIN, R . (1994). 'Job and work design'. In Cooper, C. and Robertson, I.(Eds), A v Review.^ in Managerial Psychology. Chichester: Wiley and Sons, 158 88.

WAU.., T , JACKSON, P and DAvnos, K. (1992). 'Operator work design and robotics systemspedbrmance: a serendipiious field ^ixidy'. Journal of .ipplied Psychology, 77, 353-62.

WANOUS, J. (1980). Organizational Entry. Reading, M J \ : Addison-Wesley.WEISS, H . and CROPANZANO, R . (1996). Affective events theory'. In Cummings, L. L. and

Staw, B. M. (Eds), Research in Organizational Behavior, IB. I 74.WHARTON, A. (1993). 'Tlie afTeciKe conscqut'n<cs of seivice work: managing emotions on

the job'. Hork and Occupations. 20. 205-32.WtiJUNSON, A. (1998). 'Empowerment: theory and practice'. Personnel Review, 27, 40-56.WiLKLNsoN, A., MARCHINGION, M . , GOODMAN, J. and ACKERS, P. (1992). Total quality

management and employee involvement'. Human Resource Management Journal, 2, 1 -20.

Blackwcit Publishers Ud 2001