1、OrganizingFlexibility:TheFlexibleFirmina New CenturyArne L.KallebergAbstractResearch on organizational flexibility should examine the linkages betweennumerical and functional flexibility.Unfortunately,studies of each type offlexibility generally neglect the other.Moreover,the most popular conception
2、of the interplay between these two forms of flexibility the coreperipherymodel is incomplete in important ways.I discuss evidence and limitationsof the coreperiphery model of the flexible firm,and outline some promisingattempts to conceptualize how organizations may combine functional andnumerical f
3、lexibility.I focus mainly on the USA and the UK,though I alsoreview evidence and issues involved in cross-national differences in organiz-ational flexibility.1.IntroductionSocial and economic changes in the past quarter-century have underscoredthe need for organizations to have greater flexibility i
4、n their employmentsystems.Observers inallindustrial countriesregularly emphasizethe import-ance of human resource management practices that enable organizationsto adapt quickly to rapid developments in technology,greater diversity inlabour markets,growing international and price competition in produ
5、ctmarkets,andcorporatefinancialrestructuringincapitalmarkets.Apopularexpression of this concern has been the idea of the flexible firm,whichdenotes the kinds of organizational forms that enable employers to obtainthe flexibility they need.Researchers have emphasized two distinct strategies of flexib
6、le labourutilization:enhancing employees ability to perform a variety of jobs andparticipate in decision-making,and reducing costs by limiting workersinvolvement in the organization.These two strategies have been referred toArne L.Kalleberg is in the Department of Sociology,University of North Carol
7、ina at ChapelHill.British Journal of Industrial Relations39:4 December 2001 00071080 pp.479504#Blackwell Publishers Ltd/London School of Economics 2001.Published by Blackwell Publishers Ltd,108 Cowley Road,Oxford OX4 1JF,UK and 350 Main Street,Malden,MA 02148,USA.variouslyas:functionalv.numericalfle
8、xibility1(Atkinson1984;Smith1997;Hunter et al.1993),internal v.external flexibility(Cappelli and Neumark2001),clan v.market(Ouchi 1980),dynamic v.static flexibility(Colcloughand Tolbert 1992;Deyo 1997),and organization-focused v.job-focusedemployment relations(Tsui et al.1995).Recent research on org
9、anizational flexibility has proceeded along tworelativelydistinct streams,each focusingonone of these twoflexible labourutilization strategies.The majority of studies emphasize the correlates ofhigh performance work systems that are believed to enhance functional orinternal flexibility;the other gro
10、up examines processes of externalizationdesigned to reduce costs and provide organizations with numerical flexi-bility.Unfortunately,relatively few studies have examined the interplaybetween functional and numerical flexibility.Consequently,we know com-paratively little about the relative costs and
11、benefits associated with organ-izations pursuing one as opposed to the other form of flexibility or acombination of the two.In particular,insufficient attention has been paidto whether the benefits to employers of high-performance work systems areshared with their employees in a win-win or mutual ga
12、ins enterprise,orwhether there are disadvantages to one or more parties to attaining func-tional flexibility.Those studies that have examined the interrelations between functionaland numerical flexibility in organizations have tended to conceptualize thisrelationshipprimarily in termsof a relatively
13、simplecoreperipherymodel(e.g.Atkinson 1984;Mangum et al.1985:599601;Osterman 1988:8589;OlmstedandSmith1989;Handy1990:87115).Thedebateaboutthiscore-periphery model has focused mainly on whether it is an accuraterepresentation of employers labour utilization strategies.This relativelynarrow concern ha
14、s hindered the development of more refined conceptual-izations of the varieties of ways in which the two flexible forms of workorganization may be related and of the patterns of conflict and accommo-dation created by employers attempts to balance the advantages and dis-advantages associated with the
15、m.In this paper I argue that advancing our knowledge of organizationalflexibility depends on our ability to develop models of the relationshipbetween functional and numerical flexibility,and to specify the conditionsunder which an organization will establish various combinations of stand-ard(i.e.reg
16、ular,full-time)and non-standard(e.g.part-time,temporary andcontract work)employment relationships.Accomplishing these goals willfacilitate the examination of trade-offs associated with the various labourutilization strategies and their correlates and consequences for organiz-ational,worker and indus
17、trial relations outcomes.Ifirstcriticallyevaluateresearchoneachofthetwoformsofflexibilityintermsoftheir failure toconsidertheother.Ithenreview briefly some oftheevidence onthe coreperiphery model of the flexible firm,and outline somepromising attempts to conceptualize employment systems that overcom
18、esome of its simplistic assumptions about how organizations may combine480 British Journal of Industrial Relations#Blackwell Publishers Ltd/London School of Economics 2001.features of both functional and numerical flexibility.I conclude with adiscussion of some key elements of a research agenda for
19、the new centurythat is needed to understand better employers labour utilization strategies.Myfocusismainlyonorganizationalflexibilityintherelativelyunregulatedeconomies of the United States and United Kingdom,though I also reviewevidence from other countries and discuss explicitly some issues involv
20、ed incross-national differences in organizational flexibility.2.Studying functional and numerical flexibility:unidimensionalapproachesI first provide a brief overview of the two main research traditions onorganizational flexibility,as a prelude to discussing the interrelationsbetween them.Functional
21、 Flexibility and High-Performance Work OrganizationsA great deal of attention has been paid in recent years to forms of workorganization and human resource management practices that are designedto provide employees with skills,incentives,information and decision-making responsibility that improve bu
22、siness performance and facilitateinnovation.The pervasiveness of these studies has led some to suggest thata focus on such new work and human resource management practicesconstitutesanewparadigmthatisreplacingunionsandcollectivebargain-ing as the core innovative force in industrial relations researc
23、h(Godardand Delaney 2000).These new practices have been called,variously,high-performance work organizations(Appelbaum et al.2000;Osterman 2000),transformed work organizations(Osterman 1994),flexible or alternativeworkplace practices(Gittleman et al.1998),employee involvement systems(Cotton 1993),fl
24、exible production systems(MacDuffie 1995),progressivehuman resource management practices(Delaney and Huselid 1996),high-commitment systems(Walton 1985)and high-involvement management(Lawler 1988;Wood 1999).Common to these labels is the notion that the hierarchical system ofcontrol characteristic of
25、Taylorist or Fordist forms of production has beenreplaced by a type of work organization that empowers workers to par-ticipate in decision-making,enables them to work in teams and enhancestheir commitment to the organization by,among other things,linking theircompensation to the organizations perfor
26、mance.These mechanisms en-hance an organizations functional flexibility by providing employees withmultiple skills so that they can be redeployed relatively quickly from onetasktoanother(seethereviewsbyAppelbaumandBatt1994;OECD 1999;Wood 1999).Moreover,studies of high-performance work organizationsg
27、enerally focus on work done by regular,full-time workers such as those inOrganizing Flexibility 481#Blackwell Publishers Ltd/London School of Economics 2001.the organizations core occupation(e.g.Osterman 1994,2000).For theseworkers,high-performance work organizations are generally assumed toresult i
28、n mutual gains for both the firms and their workers(especially thosein the core of the organization)(Kochan and Osterman 1994;Appelbaumet al.2000).By emphasizing the work done by regular,full-time workers,studies ofhigh-performance work organizations have tended to neglect flexible staff-ing strateg
29、ies involving non-standard forms of employment.As a result,these studies generally do not consider explicitly whether a necessary con-dition for securing the benefits of stable,functionally flexible employmentfor a group of highly committed workers is a demand buffer of externalflexibilityprovidedby
30、agroupof temporaryandothernon-standardworkers,who often have dead-end,insecure and low paid jobs and relatively littlecommitment to the organization(Atkinson 1984:31;Goldthorpe 1984:32935;Mangumetal.1985:609;Osterman1988:85;Kyotani1999:1824).Rather than a winwin scenario,then,the use of high-perform
31、ance workpractices may result in disadvantages to some,which are rooted in theincreased segmentation and polarization between advantaged,regular full-time(durables)and disadvantaged,non-standard labour force members(disposables)(Hyman 1988:557;Vallas 1999:935).The emphasis of studies of high-perform
32、ance work organizations onmanagerial outcomes,a harmony of interests and mutual gains has sup-pressed attention to conflicting interests and diverted attention from socialissues related to workers labour market rights and opportunities(Godardand Delaney 2000:492).Restricting the analysis primarily t
33、o regular,full-time committed workers has obscured the fact that achieving functionalflexibility may not necessarily require the sole use of such workers,assuggested by Smiths(2001:1720)study of how temporary workers canhelp the organization obtain functional flexibility and by Dragos(1996)discussio
34、n of management by fear in the low-job-security environmentcharacteristic of the disposable workplace.Numerical Flexibility and ExternalizationA second strand of research on organizational flexibility focuses onemployers attempts to obtain numerical flexibility and otherwise reducecosts by using wor
35、kers who are not their regular,full-time employees.2Forms of externalized labour have been called,variously,flexible staffingarrangements(Christensen 1989;Houseman 2001),market-mediated workarrangements(Abraham1990;AbrahamandTaylor1996),contingentwork(Polivka and Nardone 1989;Blank 1998;Barker and C
36、hristensen 1998),and non-standard work arrangements(Kalleberg et al.1997;Felstead andJewson1999;Cousins1999;Yeandle1999).Agrowingliteraturehassoughttodocumentandexplainthegrowthoforganizationsuseofflexiblestaffingarrangements(e.g.Pfeffer and Baron 1988:28492;Davis-Blake and Uzzi1993:199206;Houseman
37、2001;Kalleberg et al.2001).482 British Journal of Industrial Relations#Blackwell Publishers Ltd/London School of Economics 2001.An organizations externalized work-force includes several kinds of non-standard employment relations.Organizations can limit the duration ofemploymentthroughtheuseofshort-t
38、ermtemporaryworkerswho arehiredfor finite periods on an as-needed basis.They can also obtain numericalflexibility and often reduce costs by externalizing administrative controlthrough the use of temporary help agency or contract workers.Theseworkersareconsideredtobeemployeesofthetemporaryagencyorcon
39、tractcompany,not the client organization.They include both highly skilled(e.g.consultants and independent professionals)and low skilled(e.g.clericals,foodservice)workers.Additionally,organizationscanalsolimitthedurationof employment by using part-time workers,though part-timers are oftenrelativelype
40、rmanentmembersoftheorganizationandsomaybeconsideredas part of its core work-force.Data on trends in organizations use of various types of non-standardwork arrangements are sparse,making it difficult to adjudicate debatesabouttheextenttowhichtheirusereflectsafundamentaltransformationinemploymentrelat
41、ions.IntheUnitedStates,therehasbeenarapidgrowthintemporary help agency employment since the early 1970s,though it stillconstitutes a small proportion of the labour force(e.g.Segal and Sullivan1997:11819).Somewriters(e.g.Cappelli1999)havearguedthatthesenon-standard work arrangements represent a new d
42、eal between employers andemployees in which market mechanisms have replaced hierarchical employ-ment systems and wherein notions of loyalty and the social contract,especially for highly skilled workers,are obsolete.Others have maintainedthat these changesreflect more minor changesinthe allocationof
43、risk fromemployers to employees in light of changing economic conditions(Jacoby1999).Similarly,Britishstudieshaveraisedquestionsaboutwhethertheuseof flexible staffing arrangements is a new phenomenon.Marginson(1991:3542)found a great deal of continuity in the use of non-standard work inmany large UK
44、 companies in the 1980s(see also Casey 1991:1948),whileAtkinson(1987:901)arguedthatsubcontractingincreasedinGreatBritainin the early 1980s(though Pollert,1988:28991,concluded that there waslittle increase in subcontracting in the private sector).Studies of numerically flexible staffing arrangements
45、have generallyneglected to examine the impacts of these arrangements on an organiz-ations regular workers,leaving us with an incomplete view of the dis-advantages associated with using non-standard work.Human resourcemanagement strategies that try simultaneously to minimize labour costsvia contingen
46、t workers and to increase commitment and motivation ofcore workers may not be successful,since managements recruitment oftemporary employees may give rise to conflict and tension betweentemporaryandpermanentworkers(Geary1992:25961,267).Nonetheless,using temporaries may enhance the internal job mobil
47、ity of relatively highskilled,regular workers by reducing the number of such permanent em-ployeeswhoarecompetingforpromotions,asshownbyBarnettandMiner(1992:2724).Organizing Flexibility 483#Blackwell Publishers Ltd/London School of Economics 2001.3.Combining functional and numerical flexibility:the c
48、oreperipherymodelA third group of studies of organizational flexibility have considered theinterrelations between functionaland numerical flexibility,and have soughtto explain how organizations are able to obtain simultaneously these seem-ingly contradictory forms of flexibility.Three decades ago,Do
49、eringer andPiore(1971:173)observed that employers in the USA could use sub-contracting and temporary work to shift demand into the secondary sector,thereby providing manning flexibility in the primary enterprise.Writershave used various labels to refer to firms labour utilization strategies thatcomb
50、ine numerical and functional flexibility,including:coreperipherymodel(Atkinson 1984;Osterman 1988;Harrison 1994;Drago 1998),core/ring configuration(e.g.Olmsted and Smith 1989),shamrock organization(Handy1990:87115),two-tierorganization(ChristenseninDoeringeretal.1991),and the attachmentdetachment mo
