Anne Delarue, Geert Van Hootegem, Stephen Procter and · PDF filemechanisms by which...

International Journal of Management Reviews (2008)doi: 10.1111/j.1468-2370.2007.00227.x

International Journal of Management Reviews Volume 10 Issue 2 pp. 127–148 127

© 2007 The AuthorsJournal compilation © 2008 Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden,MA 02148, USA

Blackwell Publishing LtdOxford, UKIJMRInternational Journal of Management Reviews1460-8545© Blackwell Publishing Ltd 2007XXX ORIGINAL ARTICLESxxx 2007Teamworking and organizational performance

Teamworking and organizational performance: A review of survey-based researchAnne Delarue,1 Geert Van Hootegem, Stephen Procter and Mark Burridge

This paper presents a review of recent survey-based research looking at the contributionof teamwork to organizational performance. In particular, it focuses on empirical studiesin which both teamwork and performance are directly measured in a quantitative way.The paper begins by identifying four interrelated dimensions of teamwork effectiveness:attitudinal, behavioural, operational and financial. The first two represent transmissionmechanisms by which organizational performance can be improved. The latter two providedirect measures of organizational outcomes. The review shows that teamworking has apositive impact on all four dimensions of performance. It also reveals that, when teamworkis combined with structural change, performance can be further enhanced. The paperconcludes by highlighting some important research gaps that future studies could address.

Introduction

Teamwork has emerged in recent years as oneof the most important ways in which work isbeing reorganized (Osterman 1994; Watersonet al. 1997). This idea of delegating responsi-bilities to work groups has been diffused undera range of different labels. Human resourcemanagement (HRM), modern sociotechnicaltheory, business process re-engineering andlean production all embrace the core principlesof teamworking (Benders and Van Hootegem1999; De Sitter et al. 1997; Kleinschmidt andPekruhl 1994; Kuipers and Van Amelsvoort

1990) and suggest an important link withorganizational performance (Hammer andChampy 1993; Katzenbach and Smith 1993;Womack et al. 1991).

Various arguments have been advanced toexplain the effectiveness of team-based work.For example, both sociotechnical theory(e.g. De Sitter 1994; Pasmore 1988) and workdesign theory (Hackman and Oldham 1976)have focused on the design of the group’s taskto explain positive results; self-leadership theoryhas identified the supervisory behaviours thathelp self-managing teams achieve success (Manzand Sims 1987); and theories of participative

Teamworking and organizational performance

128 © 2007 The AuthorsJournal compilation © 2008 Blackwell Publishing Ltd

management argue that certain aspects ofthe organizational context contribute to theeffectiveness of teams (e.g. Lawler 1992;Glew et al. 1995).

However, theoretical arguments about theeffectiveness of teams are not enough. Thenext logical step in the cycle of scientificenquiry is the testing of these theories inpractice. Various methodological approachescan be taken to assess the benefits associatedwith teams. Field experiments or intensivecase studies allow the careful monitoring ofthe effects of workplace changes on outcomes,both qualitatively and quantitatively as wellas over considerable periods of time. Suchresearch provides insight and suggests hypo-theses, but it is difficult to generalize onthe basis of its findings (Ichniowski et al. 1996,303). In contrast, survey-based research, ifappropriately conducted, does allow gener-alization to the population at large. Two re-views of the teamworking literature carriedout approximately ten years ago showed thatsome survey-based research was already inexistence (Cohen and Bailey 1997; Guzzo andDickson 1996), but they also indicated thatvery little of this empirical work consideredissues of overall organizational performance.Indeed, some authors have argued that theevidence regarding the impact of teamworkat the level of the workplace is often basedupon anecdotes or descriptive case analyses(Appelbaum et al. 2000, 13; Cohen and Ledford1994, 13–15). However, over the last decade,studies have begun to emerge that attempt toevaluate group performance at different levelsof the organization and to assess the widerbenefits of teamwork.

The purpose of this paper is to carry out acritical examination of this literature examiningthe links between teamworking and perform-ance. Within this, we are particularly interestedin identifying studies that look at teamworkas a managerial strategy for the organizationof work. The review focuses on studies whereboth teamwork and performance have beendirectly captured in a quantitative way. Initially,we identify the various channels through

which teamworking can affect performanceand the effectiveness outcomes that are likelyto result. This theoretical exercise leads tothe development of a framework designedto categorize empirical work on effectivenessresearch and informs a subsequent literaturesearch. The conceptual framework is also usedto provide insight into the themes on whicha considerable body of work agrees. Further,it enables us to identify the specific problemsfaced by teamwork–performance research andto produce recommendations on how suchresearch might be conducted in the future.

The rest of the paper is organized as follows.First, we discuss the definitions of both team-work and performance that will be consideredin this review. Next, we formulate specifichypotheses relating to team effectiveness anddevelop a framework based upon these, whichshows how teams can affect organizationaloutcomes. We then explain the methodologythat has been adopted in this review andprovide background information on the papersidentified. This is followed by a comparisonof the results of the individual studies anda detailed discussion of these based uponhypotheses and framework. The final sectionconcludes the paper with a call for quantitative,survey-based research on the teamworking–performance link to be complemented by morequalitative, case study approaches.

Teamwork and Performance: A Tentative Demarcation of Two Key Notions

Teamworking

Over the years, a number of attempts have beenmade to define teamwork (Hackman 1987;Katzenbach and Smith 1993; Robbins andFinley 1995) and classify teams (Cohen andBailey 1997; Dunphy and Bryant 1996). How-ever, there remains no generally accepteddefinition. At different times and in differentsettings, various terms such as ‘teams’, ‘groups’and ‘work units’ have been used to describe thisform of work organization (Benders and Van

June 2008

© 2007 The Authors 129Journal compilation © 2008 Blackwell Publishing Ltd

Hootegem 1999). These terms have frequentlybeen used in conjunction with adjectivessuch as ‘autonomous’, ‘semi-autonomous’, ‘self-directed’, ‘high-performing’ and ‘self-managed’(Mueller et al. 2000). This all suggests thatworking with a specific definition of teamworkwould be very restrictive and could excludestudies important for this review. We thereforechoose not to adopt a strict definition of‘teamwork’ but to work with a ‘phenomeno-logical bottom-line’ (Schumann et al. 1994).When authors use the words ‘team’ or ‘group’to indicate variables in their research, weconsider this as sufficient for their work tobe considered for inclusion in our review.Whether a particular reality should be con-sidered as ‘teamwork’ is therefore left to theindividual researcher in the field, and it isimportant to bear this in mind in the course ofthis paper (see Van Hootegem 2000, 378).

Performance

It is also difficult to formulate an unambiguousand definitive description of ‘performance’,since this ultimately depends upon theobjectives of the particular organization. Never-theless, a wide range of performance indicatorshave been investigated in organizations, and,for the purposes of this review, we look at theseunder the headings of operational outcomesand financial outcomes. The former wouldinclude productivity (e.g. the number of hoursto assemble a car), the quality of the productor service, innovation and customer satisfaction;the latter, value-added per employee and returnon capital employed. To complicate matters,many of these indicators can be recorded atdifferent levels within an organization. Pro-ductivity, for example, can be measured atdepartment, workplace or company level.

In addition, when one begins to considerthe team-based literature, another set of‘performance’ outcomes come to the fore (Cohenand Bailey 1997; Guzzo and Dickson 1996).A number of these studies are designed to showthe outcomes for individual team members orthe team itself. While some of these measures

– job satisfaction, for example, or absenteeism– may not seem directly relevant to the presentstudy, subsequent discussion will show thatthere are important links with organizationalperformance. We therefore include such studieswithin our remit and categorize these mea-sures under the headings of ‘attitudinal’ and‘behavioural’ outcomes.

A Conceptual Framework of Team Effectiveness

A number of theoretical arguments have beendeveloped to explain why teamworking mightlead to improved organizational performance.Some theories focus on the effort and motiva-tion of individual workers and claim thatthey work harder. Strategic HRM theory,for example, suggests that an appropriatelydesigned HR system, which typically includesteamwork, will have a positive effect on anemployee’s job satisfaction, commitment andmotivation, leading to behavioural changesthat result in improved organizational per-formance (Becker et al. 1997; Dyer and Reeves1995). Similarly, self-leadership theory focuseson participatory decision-making, individualdiscretion and teamwork as important moti-vating factors, and suggests these will lead tomore committed employees who strive forgreater efficiency and effectiveness (Manzand Sims 1980; Sims and Manz 1996). Workdesign theory, however, tends to emphasizeintra-group processes such as job design,task variety and interdependence (Hackmanand Oldham 1980; Wall and Martin 1987),while sociotechnical theory highlights changesin the structure of an organization and itsprocesses as the main mechanism by whichperformance is enhanced (Mueller et al. 2000;Van Hootegem 2000).

It is apparent from this that the teamwork–performance link is related to the more generaldiscussions surrounding HRM and perform-ance, empowerment, self-leadership and soon. However, teamwork research should notbe considered only within these contexts sincea specific teamworking literature has emerged



over the course of time (Campion et al. 1993;Benders and Van Hootegem 1999; Gladstein1994; Salas et al. 2000).

Ichniowski et al. (1996, 300–301) providea summary of the reasons why innovativeworkplaces may be more effective. We use theirframework to address the issue of teamwork,and this forms the basis for the discussionwhich follows and our subsequent classificationof the means by which teamworking could beeffective.

• Working harder: Enhancing employee dis-cretion is often held to have a positive effecton job satisfaction and motivation whichtogether result in employees voluntarilyworking harder. Beyond these effects at theindividual level, group dynamics can alsoplay a role: employees may feel stimulatedby working together towards a common goal.However, there might also be a downsideto working together in that team membersmay start watching each other very closely.Particularly if a system of collective teamreward is implemented, members mayexert strong pressure on each other in anattempt to achieve high levels of teamperformance. A situation where strict normsare imposed and controlled by peersmight even result in what has been calledgroup ‘terror’ or ‘tyranny’ (Barker 1993;Sinclair 1992). In this situation, peer pressurewithin teams could be counterproductivefor the organization. Employees mightexperience higher stress levels, with adetrimental effect on team performance.Teams could also enforce norms that restricteffort to a certain maximum, and this wouldagain have performance implications.This so-called ‘social loafing’ (Karau andWilliams 1993; Shepperd 1993) or ‘shirking’(Auster 1979) is frequently mentioned insocial psychology and organization eco-nomics as an explanation for teamworking’snegative effect on productivity.

• Working more smartly: Innovative workpractices such as teamwork may also leadworkers to work more efficiently. In traditional

working systems, production problems canonly be solved by functional specialists,whereas self-managing teams are capableof solving problems as soon as they occur,thus reducing interruptions to the productionprocess (Salas et al. 2000, 348). In addition,workers often have information that highermanagement lacks, especially on how tomake their own job more efficient (Campionet al. 1993, 826). By encouraging people toexpress their views and to learn from solv-ing problems, the production process can beimproved significantly. Employees may alsoenjoy using their intellectual capacities,finding work more rewarding as a result.

• Organizational changes: It has also beensuggested that the redesign of an organiza-tion along team-based lines can involve therationalization of the production process.Related operations are grouped together,thus allowing more efficient process flowsand a reduction in product/informationhandling (Guzzo and Dickson 1996, 329).The implementation of teamwork simplifiesthe organizational structure and reduces theneed for co-ordination. Decentralizingdecision-making to self-directed teams canthus reduce the number of supervisors andmiddle managers (Ichniowski et al. 1996,301). Such a redesign can therefore leadto significant improvements in terms ofefficiency (e.g. lower costs and throughputtimes).

Figure 1 attempts to summarize these argu-ments. Four major dimensions of performanceare identified, ranging from individual out-comes to workplace measures of performance.These are: (a) attitudinal outcomes such as jobsatisfaction, commitment, trust and involvement;(b) behavioural outcomes, including turnover,absenteeism, extra role behaviour and ‘con-certive’ control (Barker 1993); (c) operationaloutcomes such as productivity, quality of theproduct or service, innovation and flexibility;and (d) financial outcomes such as added value,added value per employee and profitability. Thefigure also indicates the potential moderating

June 2008


role of external factors, allowing for theorganization of work to be influenced by work-place characteristics and by environmentaland institutional factors. Based upon the pre-ceding discussion as well as the work of otherauthors (Becker et al. 1997; Dyer and Reeves1995; Guest 2001), we believe that thesedimensions (a to d) can be differentiated intohierarchical layers, with outcomes in onedimension contributing to outcomes in the next(De Winne and Sels 2003). Changes in work

organization (the introduction of teams) canhave a direct impact on employee behaviour(less absenteeism) and, subsequently, onoperational performance (higher productivity),which, in turn, can contribute to higher levelsof financial performance (higher profits). Onecan therefore speak of a ‘performance chain’(Dyer and Reeves 1995; Guest et al. 2003). Solidarrows are used to represent possible directlinks within this framework, whereas dottedarrows are used to represent indirect links.

Figure 1. A classification model for studies that analyse the link between teamwork and performance.



Figure 1 also helps with the classificationof empirical studies looking at the effective-ness of teamwork, as the arrows all indicaterelationships that could form the focus of aparticular article. Thus the path marked 2 inthe figure (from teamworking to financialoutcomes) can be the focus of a study, but thisdoes not imply that any effects are directones: they would still work through the inter-mediating variables. It would simply be thatthe study in question did not look explicitlyat these intermediate relationships.

Propositions

Given the discussion above, we now formulatethree specific propositions relating to theteamwork–performance link that will beexplicitly examined via our review of theliterature.

(a) Teamwork will be positively linked toorganizational outcomes. The very generalassumption that will guide our review isthat teamwork will have positive effectson organizational performance. As pre-viously indicated, the implementationof teams can increase efficiency andencourage employees to work moresmartly and harder. We should thereforeexpect teamwork to have a significantpositive impact on operational outcomessuch as productivity, quality and flexibility.Since financial outcomes are further downthe ‘performance chain’, they are morelikely to be subject to other influences.Hence, although we propose that team-work will be positively linked to financialoutcomes as well, we should expectthis relationship to be weaker than theone between teamwork and operationaloutcomes.

(b) Any positive link between teamworkingand performance can be explained by theimpact of teamworking on employee atti-tudes and behaviours and/or organiza-tional structures. Figure 1 shows thatthere is no direct link from teamwork to

organizational outcomes, so any effects onperformance would have to work throughthe two different transmission mechanismsidentified. The first of these consists ofworker outcomes: teamwork, it is hypo-thesized, will have a direct impact on theattitudes of the workers (job satisfaction,motivation, commitment), and this willinfluence their behaviours (turnover,absenteeism). These effects are assumedto be positive, but they could be negative,as both accounts emerge from the litera-ture. The second transmission mechanismhas to do with the structural changes thattend to be associated with the implemen-tation of teams. A rationalization of theproduction process and a reduction in theneed for supervision imply that organiza-tions are redesigned to exhibit greatersimplicity and efficiency. Team-basedworking is therefore proposed to be posi-tively associated with structural changessuch as a delayering and decentralization.

(c) Organizational and environmental factorswill moderate the relationship betweenteamwork and organizational perform-ance. Finally, it is important to note that,although in general teamwork is expectedto have a positive impact, its effects willbe moderated by different organizationalfactors (strategy, size, industry) andenvironmental aspects (economic climate,labour market).

Scope and Methods

This paper will review survey-based researchpublished over the last decade that looks atthe links between teamwork and the variousdimensions of performance shown in Figure 1.In this context, articles will only be includedif they contain explicit measures of bothteamwork and one or more of the performanceoutcomes. Furthermore, these articles mustpresent empirical results directly relating to themeasures identified. Hence, HRM articles thatexplicitly identify teamwork and outcome(s)are included, whereas those that make use of

June 2008


composite measures of work practices variables(of which teamworking forms a part) are not.

Initially, an extensive literature search wascarried out of various online research databasesusing phrases such as ‘teamwork’, ‘teams’,‘groups’ and ‘high-performance work practices’in combination with keywords such as ‘per-formance’ and ‘effectiveness’. The snowballmethod was then applied, going through thereference lists of the selected articles andincluding additional studies where appropriate.This approach resulted in the identification ofover 300 articles but, after further investigationbased upon our specific inclusion criteria, thisset of studies was reduced to only 31 that lookempirically at the links between teamworkand performance. The papers in question aremarked with an asterisk in the reference list.These studies all worked with a survey offirms, lines of businesses or plants, making itpossible to test any hypotheses specified.

A central assumption of this type of literaturereview is that the sample of studies identifiedfairly represents all work done in the field,published and unpublished. As a field matures,journals may tend to accept only articles thatreport strong effects or statistically significantfindings. If this is the case, there could beupward ‘publication bias’ in the magnitude ofreported effects (Rust et al. 1990). This biascould exist in the present paper, although theidentification of only 31 relevant articles sug-gests that this area of study is far from mature.

Table 1 sets out the most importantcharacteristics of the studies we are reviewing.As can be seen, most of the studies surveyedthe managers of the establishments (16), witha smaller number focusing on employees (8)or a combination of both types of survey (5).In addition, two studies focused exclusivelyon existing company records. The surveys canbe further classified if we take a look at theresearch population. First, there are the majorsurveys that cover private sector (industry-wide)or both private and public sector (economy-wide) establishments (16), which involvestratified random samples of large- and medium-sized workplaces. Second, there are surveys that

concentrate on the effect of work organizationon performance in specific industries (7). Thisfocus should ensure that results are com-parable across observations since outcomes,work practices and control variables can bemore precisely measured. The approach alsoenriches the quantitative analyses with theauthor’s detailed knowledge of the industry’shistory, technology, industrial relations andproduct market (Ichniowski et al. 1996). How-ever, concentration on a single industry makesit difficult to generalize the results to othersettings. A number of studies narrow theirfocus further and look only at different plantsof one company or even different productionlines within one firm (7). Finally, there is onesurvey that examines the link between workorganization and performance in more thanone country.

The last column of Table 1 indicateswhether the study has a longitudinal or across-sectional design. It can be seen thateight studies analyse multiple periods of data.In non-experimental studies, cross-sectionaldata make it very difficult to determine causalityand to rule out the possibility that omittedvariables have an influence on the result(Leamer 1983). The advantage of panel data isthat they provide information on changes,which is necessary for estimating fixed-effectsmodels. However, there can be a ‘time lag’problem, whereby the reorganization of worktakes some time and a plant may still be intransition when the outcomes are measured.In many cases, the implementation phase ofthe change can be accompanied by a declinein performance due to the costs of associatedmeasures (such as extra training) or becauseof resistance from the employees. This meansthat it is not clear precisely when the perform-ance measurement should be carried out.

It is beyond the scope of this paper toreview in detail all the different variables thatwere analysed in each study, but we still wishto take a closer look at the way in which theindependent and dependent variables wereoperationalized. These findings are summarizedin Tables 2 and 3, respectively.

Teamw

orkin

g an

d o

rgan

ization

al perfo

rman

ce

134©

2007 The Authors

Journal compilation ©

2008 Blackwell Publishing Ltd

Table 1. Methodology

Type of methodology used Research population CountryLongitudinal/cross-sectional

1 Bacon and Blyton 2000 Survey of workers (union representatives) Iron and steel industry Britain Cross-sectional2 Banker et al. 1996 Survey of company records Electromechanical assembly plant US Longitudinal3 Batt 1999 Survey of employees and company records Telecommunications company US Cross-sectional4 Batt 2001 Survey of employees and company records Telecommunications company US Cross-sectional5 Batt 2004 Survey of employees Telecommunications company US Cross-sectional6 Batt and Appelbaum 1995 Survey of employees Telecommunications and apparel companies US Cross-sectional7 Benders et al. 1999 Survey of managers Economy-wide 10 EU countries Cross-sectional8 Black and Lynch 2001 Survey of managers Industry-wide US Longitudinal9 Boning et al. 2001 Survey of managers and employees Steel mills US Longitudinal10 Cappelli and Neumark 2001 Survey of managers Industry-wide US Longitudinal11 Cohen et al. 1996 Survey of managers and employees Telephone company US Cross-sectional12 Cooke 1994 Survey of managers Industry-wide US Cross-sectional13 Delarue et al. 2004 Survey of managers Economy-wide Belgium Cross-sectional14 DeVaro 2006 Survey of managers Economy-wide Britain Cross-sectional15 Dunlop and Weil 1996 Survey of managers Apparel industry US Cross-sectional16 Elmuti 1997 Survey of managers Industry-wide US Cross-sectional17 Glassop 2002 Survey of managers Economy-wide Australia Cross-sectional18 Godard 2001 Survey of employees Economy-wide Canada Cross-sectional19 Harley 2001 Survey of managers and employees Economy-wide Britain Cross-sectional20 Hamilton et al. 2003 Survey of company records Textile manufacturer US Longitudinal21 Ichniowski et al. 1997 Survey of managers and company records Steel finishing lines US Longitudinal22 Mathieu et al. 2006 Survey of employees Office equipment company Canada Cross-sectional23 McNabb and Whitfield 1997 Survey of managers Economy-wide Britain Cross-sectional24 Osterman 2000 Survey of managers Industry-wide US Longitudinal25 Paul and Anantharaman 2003 Survey of managers and employees Software companies India Cross-sectional26 Power and Waddell 2004 Survey of managers Economy-wide Australia Cross-sectional27 Procter and Burridge 2004 Survey of managers Economy-wide Britain Cross-sectional28 Stewart and Barrick 2000 Survey of employees Industry-wide US Cross-sectional29 Tata and Prasad 2004 Survey of managers and employees Industry-wide US Cross-sectional30 West et al. 2002 Survey of managers Acute hospitals Britain Cross-sectional31 Zwick 2004 Survey of managers Industry-wide Germany Longitudinal

Jun

e2008

© 2007 The A

uthors135

Journal compilation ©

2008 Blackwell Publishing Ltd

Table 2. Types of independent variables studied

Label for teamsDefinition/description of teamwork

Features of teams

Organizational and/or external factors

1 Bacon and Blyton 2000 Teams – high road and low road teams Description pp.1436–1438 Yes No2 Banker et al. 1996 High-performance work teams Description pp.873–876 No Yes3 Batt 1999 Self-managed teams Description p.542 Yes Yes4 Batt 2001 Self-managed teams Description pp.4–5 No Yes5 Batt 2004 Self-managed teams Description pp.186–187 No Yes6 Batt and Appelbaum 1995 Self-managed teams Description pp.358 and 361 Yes Yes7 Benders et al. 1999 Team-based work Definition p.13 Yes No8 Black and Lynch 2001 Self-managed teams – No Yes9 Boning et al. 2001 Problem-solving teams – No Yes10 Cappelli and Neumark 2001 Self-managed or autonomous teams Definition p.748 No No11 Cohen et al. 1996 Self-managing work teams Definition p.653 Yes No12 Cooke 1994 Work teams – No Yes13 Delarue et al. 2004 Teams – lean/sociotechnical teams Definition p.17 No Yes14 DeVaro 2006 Team production – Yes Yes15 Dunlop and Weil 1996 Modular production groups Description p.338 No Yes16 Elmuti 1997 Self-managed teams Definition p.185 No No17 Glassop 2002 Self-managing work groups Definition p.227 No No18 Godard 2001 Teams Definition p.786 Yes Yes19 Harley 2001 Autonomous teams Description p.729 No Yes20 Hamilton et al. 2003 Modular production teams Description pp.475–476 No No21 Ichniowski et al. 1997 Team-based work organization Definition p.294 No Yes22 Mathieu et al. 2006 Empowered teams Description p.101 Yes Yes23 McNabb and Whitfield 1997 Teamworking – No Yes24 Osterman 2000 Self-managed teams – No Yes25 Paul and Anantharaman 2003 Teamwork Description p.1251 No Yes26 Power and Waddell 2004 Self-managed work teams Description pp.249–250 No No27 Procter and Burridge 2004 Teamwork – semi-autonomous teams Description pp.15–17 Yes Yes28 Stewart and Barrick 2000 Work teams Description p.140 Yes No29 Tata and Prasad 2004 Self-managed teams Description p.254 Yes No30 West et al. 2002 Teams – No Yes31 Zwick 2004 Shop-floor employee participation – teamwork Description p.720 No Yes



Definition and Measurement of Teamwork

As indicated previously, divergent workforms may all be called ‘teamwork’. Furtherinformation on structural features is thereforerequired in order to determine the type ofgroup work identified. In most of the articles,teamwork was based on the delegation ofresponsibilities to employees and decentralizeddecision-making. While some of the studiesin our review are based on sociotechnical orwork design theory, thus emphasizing team-work and task design as the direct predictorsof performance, others take strategic HRMtheory as their starting point, defining teamworkor employee participation as one item in a list

of possible ‘innovative work practices’ thatcan be introduced to enhance the ability of thehuman resources in organizations to contributeto organizational performance (Doorewaardet al. 2002).

The third column of Table 2 gives an over-view of the labels of the independent variablesused in the 31 studies. It is striking that onlya minority of the studies give a clear, explicitdefinition of teamwork. A number of authorsonly give a (sometimes rather vague or implicit)description of what they associate with a‘team’. In the studies from the HRM tradition,teamwork is usually only one componentof a series of measures designed to improveemployee involvement and participation,with little specification of the type of work

Table 3. Types of outcomes measured

Number of measures

Operational outcomes

Financial outcomes

Behavioural or attitudinal outcomes

Structural changes

Objective / subjective measures

1 Bacon and Blyton 2000 More than 3 X X X Subjective2 Banker et al. 1996 2 X Objective3 Batt 1999 2 X Both4 Batt 2001 More than 3 X X X Objective5 Batt 2004 3 X Subjective6 Batt and Appelbaum 1995 3 X X Subjective7 Benders et al. 1999 More than 3 X X X X Subjective8 Black and Lynch 2001 1 X Objective9 Boning et al. 2001 1 X Objective10 Cappelli and Neumark 2001 3 X X Objective11 Cohen et al. 1996 More than 3 X X X Subjective12 Cooke 1994 1 X Objective13 Delarue et al. 2004 More than 3 X X Both14 DeVaro 2006 1 X Subjective15 Dunlop and Weil 1996 2 X X Objective16 Elmuti 1997 More than 3 X X X Subjective17 Glassop 2002 More than 3 X X X X Subjective18 Godard 2001 More than 3 X Subjective19 Harley 2001 More than 3 X Subjective20 Hamilton et al. 2003 2 X Objective21 Ichniowski et al. 1997 2 X Objective22 Mathieu et al. 2006 2 X Both23 McNabb and Whitfield 1997 1 X Subjective24 Osterman 2000 1 X Objective25 Paul and Anantharaman 2003 More than 3 X X Both26 Power and Waddell 2004 1 X Subjective27 Procter and Burridge 2004 3 X X Subjective28 Stewart and Barrick 2000 1 X Subjective29 Tata and Prasad 2004 1 X X Subjective30 West et al. 2002 More than 3 X Objective31 Zwick 2004 1 X X Objective

June 2008


organization that underpins them. Twoexamples from studies that do give a definitionof teamwork also show that such definitionsare not always the same:

A self-managing work group is a group ofinterdependent individuals that have acceptedresponsibility for a group task and share thisresponsibility by monitoring and controlling thecontributions of its members. (Glassop 2002, 227)

Self-managed or autonomous teams give theirmembers authority over decisions that in othercontexts are made by supervisors, such as how toperform their tasks or, in more advanced situationswhich tasks to perform. (Cappelli and Neumark2001, 748)

These definitions do have some similarities,but the differences in emphasis suggest thatthere is no one-on-one relation between whatis called a ‘team’ and the organizational workform that the term is designed to represent.Teamwork is not a unitary phenomenon; thesemantic notion covers different ontologicalrealities.

A number of authors have made pleas fora more differentiated model of team design.Some criteria for classification that have beensuggested are the degree of self-management,the nature of the team membership, the teamstructure (size, breadth of job definition ofthe team members, degree of autonomy) andcomposition, the type of tasks performed andthe technology and equipment used (Dunphyand Bryant,1996, 678; Cohen and Bailey1997, 248). Taking into account these teamcharacteristics, a typology of teamwork couldbe developed, making it possible to re-expressthe initial research question, ‘Does teamworkincrease organizational performance?’ as ‘Whichtype of teams generates which organizationaloutcomes?’ However, our review shows that,to measure the presence of teamwork, moststudies are satisfied with a yes/no response ora percentage of employees working in a teamsystem. In Table 2, we indicate the studiesthat do include some features of teams in theiranalysis. It should be noted that only one study

makes reference to team type (Ichniowskiet al. 1997) and only two to team size (Bankeret al. 1996; Elmuti 1997), with none of thesehaving an explicit measure in their analysis.Instead, ‘features’ usually refers to aspects ofautonomy.

Definition and Measurement of Performance

Various performance outcomes are measuredin the 31 studies identified, with 15 of thearticles employing three or more measures ofeffectiveness. Ten studies focus only on one(operational or financial) outcome. In the othersix, two indicators of organizational perform-ance are examined. Operational outcomes areincluded in almost every study, as an attemptto measure the direct link between teamworkand organizational performance. Few studiestry to make the picture more complete byfocusing on transmission mechanisms. Only tenstudies analysed worker outcomes, the mostcommon being absenteeism, turnover and jobsatisfaction. Even more exceptional are studiesthat pay attention to structural changes.

We can make a distinction between, onthe one hand, studies that featured objectivemeasures of performance often taken fromproduction records and, on the other, those thatrely on perceptions of performance, derivedfrom interviews or from the responses ofemployees or managers to survey questions.Examples of objective outcomes are monthlydefect rates (Banker et al. 1996), quality mea-sured as the percentage of tons produced thatmeet specific quality standards (Ichniowskiet al. 1997) and weekly productivity measuredas the hours required to produce a batch ofgarments (Hamilton et al. 2003). Subjectiveinformation on performance outcomes isprovided either by one of the parties involved(for example, when a respondent is asked tocompare the quality of their product or servicewith other establishments in the same industry)(Procter and Burridge 2004) or as work teameffectiveness rated by the team members andteam managers (Cohen et al. 1996).



There may appear to be advantages inhaving objective measures, but the accountancyprotocols on which financial indicators arebased have come under critical scrutiny inthe wake of a number of corporate scandals(Guest et al. 2003, 294). Similarly, subjectivemeasures have been subject to some criticism(Starbuck and Mezias 1996), but they mayhave the advantage of being both more up todate (Hart and Banbury 1994) and more com-parable across industries (Covin et al. 1994).The last column of Table 3 shows a rough balancebetween objective and subjective measures,with a limited number of studies making useof both.

Teams and Performance: What Can We Learn from Empirical Studies?

Over the last decade, considerable progresshas been made in understanding team effec-tiveness. The results from our set of studiesreveal a number of significant findings whichwill be classified according to the relation-ships shown in Figure 1. The numbers belowcorrespond to the numbers shown in this figure.Table 4 shows specific results related to each ofthe papers identified. All the articles presentstatistical analysis related to their data. Samplesizes vary substantially, but this type of analysisallows for such differences. Furthermore, almost

Table 4. Results

Behavioural or attitudinal outcomes

Structural changes

Operational outcomes

Financial outcomes

1 Bacon and Blyton 2000 + + +2 Banker et al. 1996 +3 Batt 1999 +4 Batt 2001 + + +5 Batt 2004 +6 Batt and Appelbaum 1995 + +7 Benders et al. 1999 + + + +8 Black and Lynch 2001 09 Boning et al. 2001 +10 Cappelli and Neumark 2001 0 011 Cohen et al. 1996 + + +12 Cooke 1994 +13 Delarue et al. 2004 + 014 DeVaro 2006 +15 Dunlop and Weil 1996 + +16 Elmuti 1997 + + +17 Glassop 2002 + + + 018 Godard 2001 +19 Harley 2001 020 Hamilton et al. 2003 +21 Ichniowski et al. 1997 022 Mathieu et al. 2006 +23 McNabb and Whitfield 1997 024 Osterman 2000 +25 Paul and Anantharaman 2003 + +26 Power and Waddell 2004 027 Procter and Burridge 2004 + +28 Stewart and Barrick 2000 +29 Tata and Prasad 2004 + +30 West et al. 2002 +31 Zwick 2004 + +

Notes: + indicates a statistically significant impact and/or performance enhancing effect. 0 indicates no statistically significant effect.

June 2008


all the articles have been deemed to be ofpublishable standard by independent referees.Hence, we attempt to draw conclusions on thebasis of the evidence presented in this set of31 papers.

Teamwork and Operational Outcomes: (1)

A positive relationship between teamworkand operational performance is found in anumber of studies. In their study of a textilemanufacturer, Hamilton et al. (2003) found thatteam-sewing increased productivity by approxi-mately 18%. According to Cohen et al. (1996),a form of work organization incorporating teamsand strong employee involvement had asignificant impact on both quality and effici-ency. Reducing the number of managementlayers, working with flexible job descriptionsand the introduction of teamworking were allpositively associated with different operationaloutcomes in the study of Bacon and Blyton(2000). West et al. (2002) focused on HR effec-tiveness in acute hospitals using health out-comes as operational measures, and found asignificant relationship between teamworkingand reduced patient mortality. An improve-ment in both quality and labour productivitywas also recorded by Banker et al. (1996), Batt(1999), Batt (2001), Benders et al. (1999),Elmuti (1997), Mathieu et al. (2006), Paul andAnantharaman (2003), Procter and Burridge(2004), Stewart and Barrick (2000) and Tataand Prasad (2004). Positive productivity effectswere found in studies relating to the Australianeconomy (Glassop 2002), US steel mills (Boninget al. 2001) and the US apparel industry(Dunlop and Weil 1996). Finally, Batt andApplebaum (1995) report improved quality.

However, little or no statistically significantevidence of beneficial productivity effects wasfound in five studies. A study that linked thework organization of Flemish firms to differ-ent performance indicators (Delarue et al. 2004)found no relationship between the applicationof teamwork and labour productivity, althoughpositive correlations with innovation andpersonnel fit were registered. Ichniowski et al.

(1997) found that bundles of HRM practiceshad a significant impact on performance,but teamwork did not when considered inisolation. Similarly, Power and Waddell (2004)found that, although empowerment had asignificant effect, this was not true of theirmeasure of self-managed work teams. BothBlack and Lynch (2001) and Cappelli andNeumark (2001) found little evidence of posi-tive productivity effects in their industry-widestudies, although it should be noted that thislack of significance applied to a number ofthe independent variables used in their studies(including bundles of work practices).

With no studies indicating a negative effectand 18 out of the 23 that measured operationaloutcomes finding beneficial effects, we cansay that, in general, teamwork is likely to havea positive impact on operational performance.This confirms our first proposition – aninteresting, if not entirely unexpected, result.Although the research settings are differentand the studies measure teamwork and organ-izational outcomes in different ways, a degreeof consistency emerges in their findings.

Teamwork and Financial Outcomes: (2)

We might expect a weaker link with financialoutcomes than with operational outcomes,since factors other than teamwork can affectthe former. Nevertheless, the overall pattern ofrelationships between teamwork and financialmeasures is consistent with the assumption thatteamwork is an effective practice in increas-ing the competitiveness of firms. Cooke (1994)found that teamwork had a significant impacton value added per employee. Similarly, Zwick(2004) found that the introduction of shopfloor participation, in which teamwork formeda substantial part, significantly increasedeconomic value added in German firms.Teamwork also proved a positive predictor offinancial performance in the studies of Batt(2001), Benders et al. (1999), Cohen et al. (1996),DeVaro (2006), Dunlop and Weil (1996),Elmuti (1997), Paul and Anantharaman (2003)and Procter and Burridge (2004).



Only three studies in our sample found nosignificant association between teamworkand financial outcomes. Glassop (2002) triedto measure the effect of self-managing workgroups on profitability, but the results wereinconclusive at the 5% significance level.Similarly, McNabb and Whitfield (1997)found the impact of teamworking on relativefinancial performance to be significant only atthe 10% level. Cappelli and Neumark (2001)found their various measures of teamwork hadno effect on revenue per worker. Their care-ful conclusion is that high-performance workpractices tend to raise both productivity andworker compensation, with the net effect onoverall profitability being unclear.

These results give support to the propositionthat teamwork has positive effects on financialoutcomes. The three insignificant results canperhaps be explained by the fact that financialoutcomes work through intermediate variablesrather than there being any direct effect.

Teamwork and Worker Outcomes: (3) and (4)

Attitudinal or behavioural outcomes at theemployee level were measured in ten studies.Godard’s (2001) survey of Canadian employeesfocused exclusively on this type of outcome.He examined whether work practices associatedwith the high-performance model have mainlynegative or positive implications for workers.He found that team-based work has strongand statistically significant positive correlationswith belongingness, task involvement, jobsatisfaction, empowerment, commitment andcitizenship behaviour. However, the resultsfor team autonomy and responsibility weregenerally insignificant. Batt (2004), Batt andAppelbaum (1995) and Elmuti (1997) foundthat the job design characteristics associatedwith self-managed teams significantly improvejob satisfaction and organizational commitment.A similar result was found by Bacon and Blyton(2000), who showed that workers under ‘high-road’ teamworking scored more positively onmotivation, interest in their job and enjoyment

of the job. Benders et al. (1999), Cohen et al.(1996) and Delarue et al. (2004) all foundthat the large-scale use of group work leadsto lower levels of absenteeism, whereas thestudy by Glassop (2002) found that firms withteams had lower levels of employee turnover.

This provides a summary of reported posi-tive employee outcomes, but it is also possiblethat teamwork might cause negative employeeexperiences. A team-based work organizationcan intensify ‘concertive’ control (Barker 1993)and workload. A few of the studies explicitlylooked at this issue, but no such results emerged.Harley (2001) reported no significant dif-ferences between team members and non-team members in terms of their reportedlevels of stress, satisfaction and commitment.Overall, therefore, these results suggestthat the impact of teamwork on employeeattitudes (e.g. motivation, commitment) andbehaviour (e.g. turnover, absenteeism) is largelypositive.

Teamwork and Structural Changes: (5)

Only a few of the studies in our review oper-ationalized structural changes as an outcomeassociated with the introduction of teamwork.However, such changes can imply a reductionin the complexity of the production processand in the need for co-ordination. In theirEuropean study, Benders et al. (1999) recordeda reduction of the throughput time as one ofthe positive effects of team-based work. BothDunlop and Weil (1996) and Hamilton et al.(2003) report an impact on production leadtime, although in both studies this variablewas considered as part of ‘operational per-formance’. Another aspect of the organizationthat can be influenced by teamwork is hier-archical structure. Decentralizing decision-making to the lowest level in organizationscan mean that there is less need for managersand supervisors. Bacon and Blyton (2000)indicate that a reduction in the number ofmanagement layers was an important reasonfor the introduction of teamwork. Tata andPrasad (2004) present results that show that team

June 2008


effectiveness is increased when teamwork iscombined with a decentralized structure; andZwick (2004) shows that a combination ofteamwork and a flatter organizational structurehas an additional positive effect upon pro-fitability. Glassop (2002) found that firms thatwork with self-managing work groups havea broader span of control, thus being lesshierarchical. Similarly, Batt (2001) found thatthe execution of supervisory tasks by workersincreased with the existence of self-managedteams. Finally, Osterman (2000) reports thatteamwork is positively and significantly relatedto establishment layoffs, with these layoffsfalling disproportionately upon managers.These results suggest that, when teamwork isassociated with structural change, performancecan be further enhanced, but there is insufficientevidence to conclude that this is a widespreadphenomenon.

The Link Between Teamwork and Performance, Controlling for Organizational and Environmental Factors: (6)

The majority of the studies used regressiontechniques to estimate the impact of teamworkon performance, and included in their modelssuch organizational characteristics as size ofthe establishment, age of the plant, productdiversity and features of the workforce ascontrols. Apart from these typical controlvariables, some authors suggested that workpractices such as teamwork would be moreeffective when ‘bundled’ with supporting man-agement practices. The reasoning behind thisis that workers cannot make good decisionswithout sufficient information and training,and they are thus unlikely to make suggestionsfor improvement if they feel this will costthem their jobs or reduce their pay (Ichniowskiet al. 1996, 302).

In this context, reward has been seen as animportant variable moderating the teamwork–performance link. Gain-sharing, profit-sharingand other programmes that provide employeeswith a share in improved performance are

typically seen as strongly complementary tothe introduction of teamwork. ‘Compensation’in one form or another is included in tenstudies in our review, but the evidence is farfrom clear cut. Boning et al. (2001), Cappelliand Neumark (2001) and Hamilton et al. (2003)all show that teamwork is associated withgreater levels of worker compensation. Tataand Prasad (2004) indicate that team rewardhas a positive and significant impact oneffectiveness. Similarly, Ichniowski et al. (1997)argue that it is important to analyse a firm’swork policies as a part of a coherent incentivesystem. The results of their study in the steelindustry show that systems of HRM practices,including an incentive-pay plan, determineproductivity and quality, while marginalchanges in individual work practices havelittle effect. Cooke (1994) shows that teamworkand group-based incentives yield substantialgains in firm-level performance (measured asvalue added net of labour costs per employee),although other results in the same papersuggest little difference between bundles andindividual practices. In contrast, both Batt andApplebaum (1995) and Osterman (2000) findthat teamwork is associated with lower paylevels. The former also find that compensationhas no effect on performance, with this resultbeing supported by other work undertakenby one of the authors (Batt 1999, 2001).

Another organizational characteristic thatcomes to the fore in some studies is strategy.For example, the hypothesis of Dunphy andBryant (1996) that low-intensity teamworkingimpacts principally on cost, whereas semi-autonomous teamworking impacts on value,was tested by Procter and Burridge (2004).There appears to be some tentative evidencethat, although cost-focused firms achievehigher levels of financial performance whenthey adopt semi-autonomous teamworking, itis teamworking per se that has an impact onperformance in value-focused establishments.Strategy is also a moderating variable in theresearch of Delarue et al. (2004).

An institutional factor that often appears(mostly in the US) is the existence or level of



unionization. Cooke (1994) examined whetherunion representation positively or negativelyinfluences the effectiveness of employeeparticipation programmes and group-basedincentives. He found that unionized firms, onaverage, provide a better environment for tap-ping the benefits of employee participationprogrammes than do non-union firms. Unionaffiliation was also included in the analysesundertaken by Batt and Appelbaum (1995),Black and Lynch (2001), Godard (2001), Harley(2001) and Ichniowski et al. (1997).

Discussion and Suggestions for Future Research

Main Findings

The results of this literature review generallypoint in one direction: adopting team structurescan yield positive outcomes for organizations.In terms of our specific propositions, we cansay the following:

Proposition A. Teamwork will be positively linkedto organizational outcomes: Confirmed. Mostof the studies that analysed the relationshipbetween teamwork and operational or financialoutcomes found a positive link.

Proposition B. Any positive link between team-working and performance can be explained bythe impact of teamworking on employee attitudesand behaviours and/or organizational structure:Confirmed. Almost all the studies that analysedthe link with behavioural and attitudinaloutcomes at the level of the employees foundthat worker outcomes improved. The resultswith regard to structural change also suggest apositive impact.

Proposition C. Organizational and environ-mental factors will moderate the relationshipbetween teamwork and organizational per-formance: Confirmed. The studies in the reviewoffer support to the idea that contextual factorsact to moderate the teamwork–performancelink.

Methodological Issues

Before assessing the implications of thesefindings, a word of caution is necessary withregard to terms such as ‘effects’ or ‘impact’.We can look at this in the context of the recentmove in the HRM–performance literaturetowards specific consideration of issues ofcausality. The statistically significant relation-ships between HRM practices and companyperformance which are presented in the litera-ture evoke difficult ‘chicken or egg’ questions(Sels 2002, 6). It is unclear what comes first:do HRM practices cause good performance ordoes good performance lead to the implemen-tation of certain HRM practices? The resultsof a study by Guest et al. (2003), for example, onthe effect of HRM on corporate performancein the UK, supported the view that profitabilitycreates scope for more HRM. Thus, while theresults reveal a positive association betweenHRM and profitability, they do not supportthe causality assumption that HRM leads tothe higher profitability.

However, it can be argued that this causalitydilemma is less pertinent here, as the analogywith teamwork research only holds to a certainextent. Why would a high-performing organi-zation invest their profits in a redesign ofthe production process when they are self-evidently being well-served by the existing one?Moreover, the implementation of an alterna-tive work system such as teamwork would becostly in terms of both money and time. Inthis context, it therefore seems implausible thatcausality would run in the opposite direction.

Future Research

This final section attempts to contribute to ongo-ing research looking at the team–performancelink by highlighting some important gaps thatfuture studies could address.

Embeddedness in a ‘division of labour’approach. A distinction should be made bet-ween, on the one hand, decisions in organiz-ations that lead to workplace change and, on

June 2008


the other, decisions that have an impact onemployment relations. In management terms,this difference can be translated into opera-tions management versus HR management.Teamwork should not only be considered asan aspect of the HR system, nor simply as aredesign that can be undertaken in a vacuum,but rather as a dimension of the division oflabour as well. The way in which jobs aredesigned is strongly embedded in choicesmade at higher levels of the organization inareas such as production organization, workorganization and technological choice. Few ofthe studies in the review focus on this broadercontext. However, decisions which are madeat these levels exclude some options con-cerning the design of teams. The decision ofan organization to work with a line-orientedproduction system, for example, will stronglydetermine the type of teamwork it can adopt.

Structural features of teams as inputs to a teamtypology. A second important issue is that alot of the existing research examines the effectof the implementation of teamwork per se onperformance, without taking into accountthe structural features of the teams, such astask design or the degree of autonomy andself-leadership. We recommend that these‘independent variables’ should be elaboratedupon in future research and that more com-plete descriptions of the technology, task andproducts/services associated with the teamsshould be provided.

It seems appropriate to classify teams inan autonomy–interdependence matrix. At oneextreme of this matrix would be a group ofworkers, which may be called a team, buthave no job enrichment or reciprocal inter-dependence because preparing, supportingand regulating tasks remain concentrated atthe management level. At the other extremewould be a self-leading team whose membershave a large range of responsibilities, a highdegree of interdependence and identify withteam targets. The team leader’s role in thiscase would be to facilitate the team’s self-managing capacity. Between these two extremes,

there could be various intermediate forms,some with low autonomy and high inter-dependence and others with high autonomyand low interdependence.

Team development as a necessary complementto team design. Most studies still do notaddress how teams change over time. It isimportant to recognize the association betweenteam evolution, team structure and perform-ance improvements. Teams in the first stage oftheir development process can exhibit differ-ent structural features from mature teams.Ancona et al. (2001) make a call to sharpen thetemporal perspective on conducting organiza-tional research. The constitution of a team–performance link has to be seen as a process intime, and hence an indicator of team maturity isindispensable. Only three studies in our reviewinclude such an indicator in their model ofteam effectiveness (Banker et al. 1996; Glassop2002; Hamilton et al. 2003).

Teamwork as only one element of a bundle ofpractices. Teamwork should also be considerednot in isolation, but as one item in a systemwhich, if appropriately configured, can reapbenefits through interactive and mutually rein-forcing effects. Such systems could includeemployee involvement schemes, training anddevelopment programmes, and forms of profitsharing (Ramsay et al. 2000, 502). The inter-action between teamwork and these otherelements should be further examined to giveus a better understanding of how teamworkcontributes to organizational performance.

Need for a contingency approach. There is alsoa need to take into account in more detailthe organization and industry in which teamsare embedded and the objectives that under-pin operations. Management strategy, marketposition, competitiveness and the broad eco-nomic situation may be important moderatingconditions in explaining how teams canpositively influence organizational productivity.In addition, internal contingencies must be con-sidered. It is not the management–employment



relationship (HR policy) or the structure ofthe division of labour (work system) alonethat influences the performance of an organi-zation, but the interaction between the two.

Most of the studies we have reviewed hereremain ‘black box’ research. The most importantconclusion of this review is that, although thereis a positive association between teamworking

Figure 2. The relationship between teamwork and performance: refinement of the model.

June 2008


and performance, the nature of this link hasnot been definitively established. Therefore,we make a plea for more complex modelbuilding to address the issue of how teamscontribute to organizational performance. Afirst step is a more differentiated model ofteam design, team development and teamperformance measurement. There is someevidence to suggest that the conventional linkbetween teamworking and performance(through job design and motivation) does work,but there needs to be a lot more exploration ofthe linkage that operates through structuralchanges.

The results of this review as well as thediscussion concerning the direction of futureresearch suggest that our initial classificationmodel should be extended to incorporate anumber of additional elements. This revisedframework is shown in Figure 2. The greyboxes indicate items which, to date, have beenlargely ignored in the literature, but whichwe consider important for future research inthis area. Overall, the model can be seen as ablueprint for future research on teams.

Conclusion

There are long-standing theoretical argumentsto support the belief that the adoption of for-mal team structures and the use of employeeparticipation programmes can provide manybenefits for organizations. More recently,a number of researchers have attempted tomeasure the link to performance directly andcomprehensively, sometimes using complexeconometric models. Our review of a sampleof these studies shows that the balance of theevidence supports the proposition that team-work improves organizational performance.

However, we should take into account thelimits of survey research. Even though addi-tional control measures are often includedin the models, there will always be omittedvariables, and quantitative studies will neverexplain all the variation we observe. In viewof this reductionism, the question arises ofwhether a survey design is the most appropriate

method to test such extended models. This isespecially so because measures at the organi-zational, team and individual level are allrequired in order to grasp the performanceproblem fully. Following Ichniowski et al.(1996, 330), we recommend that any quantita-tive analysis should be complemented bydetailed case studies that can observe hard-to-quantify data and shed light on crucial detailsof how the implementation of teamwork leadsto success. It is, in other words, important toget inside the black box and to establish howand why people behave and perform as they do.

Note

1 Corresponding author. E-mail: [email protected]

References

Ancona, D.G., Goodman, P.S., Lawrence, B.S. andTushman, M.L. (2001). Time: a new research lens.Academy of Management Review, 26, 645–663.

Appelbaum, E., Bailey, T., Berg, P. and Kalleberg,A.L. (2000). Manufacturing Advantage. Why High-Performance Work Systems Pay Off. New York:Cornell University Press.

Auster, R.D. (1979). Shirking in the theory of thefirm. Southern Economic Journal, 45, 867–874.

*Bacon, N. and Blyton, P. (2000). High road and lowroad teamworking: perceptions of managementrationales and organizational and human resourceoutcomes. Human Relations, 53, 1425–1458.

*Banker, R.D., Field, J.M., Schroeder, R.G. andSinha, K.K. (1996). Impact of work teams on man-ufacturing performance: a longitudinal field study.Academy of Management Journal, 39, 867–890.

Barker, J.R. (1993). Tightening the iron cage: concer-tive control in self-managing teams. AdministrativeScience Quarterly, 38, 408–437.

*Batt, R. (1999). Work organization, technology, andperformance in customer service and sales. Indus-trial and Labour Relations Review, 52, 539–564.

*Batt, R. (2001). The economics of teams amongtechnicians. British Journal of Industrial Relations,39, 1–24.

*Batt, R. (2004). Who benefits from teams? Compar-ing workers, supervisors, and managers. IndustrialRelations, 43, 183–212.



*Batt, R. and Appelbaum, E. (1995). Worker parti-cipation in diverse settings: does the form affectthe outcome and if so, who benefits? British Journalof Industrial relations, 33, 353–378.

Becker, B.E., Huselid, M.A., Pickus, P.S. and Spratt,M.F. (1997). HR as a source of shareholder value:research and recommendations. Human ResourceManagement, 36(1), 39–47.

*Benders, J., Huijgen, F., Pekruhl, U. and O’Kelly,K.P. (1999). Useful but Unused Group Work inEurope. Findings from the EPOC Survey. Dublin:European Foundation for the Improvement ofLiving and Working Conditions.

Benders, J. and Van Hootegem, G. (1999). Teams andtheir context: moving the team discussion beyondexisting dichotomies. Journal of Management Studies,26, 609–628.

*Black, S.E. and Lynch, L.M. (2001). How to com-pete: the impact of workplace practices and inform-ation technology on productivity. Review of Economicsand Statistics, 83, 434–445.

*Boning, B., Ichniowski, C. and Shaw, K. (2001).Opportunity Counts: Teams and the Effectivenessof Production Incentives. NBER Working PaperNo. 8306. Cambridge, MA: National Bureau ofEconomic Research.

Campion, M., Medsker, G. and Higgs, A. (1993).Relations between work group characteristics andeffectiveness: implications for designing effectivework groups. Personnel Psychology, 46, 823–850.

*Cappelli, P. and Neumark, D. (2001). Do ‘high-performance’ work practices improve establishment-level outcomes? Industrial and Labour RelationsReview, 54, 737–775.

Cohen, S.G. and Bailey, D.E. (1997). What makesteams work: group effectiveness research fromthe shop floor to the executive suite. Journal ofManagement, 23, 239–290.

Cohen, S.G. and Ledford, G.E. (1994). The effective-ness of self-managing teams: a quasi-experiment.Human Relations, 47, 13–43.

*Cohen, S.G., Ledford, G.E. and Spreitzer, G.M.(1996). A predictive model of self-managing workteam effectiveness. Human Relations, 49, 643–676.

*Cooke, W.N. (1994). Employee participation pro-grams, group-based incentives, and company perform-ance: a union-nonunion comparison. Industrial andLabour Relations Review, 47, 594–609.

Covin, J., Slevin, D. and Schultz, R. (1994). Implement-ing strategic missions: effective strategic, structuraland tactical choices. Journal of Management Studies,31, 481–505.

*Delarue, A., Van Hootegem, G., Huys, R. and Gryp, S.(2004). Dossier: Werkt teamwerk? De PASO-resultaten rond arbeidsorganisatie doorgelicht.Leuven: Hoger Instituut voor de Arbeid, DepartementTEW, Departement Sociologie (KU Leuven).

De Sitter, L.U. (1994). Synergetisch produceren. Humanresources mobilisation in de productie: een inleidingin de structuurbouw. Assen: Van Gorcum.

De Sitter, L.U., Den Hertog, J.F. and Dankbaar, B.(1997). From complex organizations with simplejobs to simple organizations with complex jobs.Human Relations, 50, 497–534.

*DeVaro, J. (2006). Teams, autonomy, and the financialperformance of firms. Industrial Relations, 45,217–269.

De Winne, S. and Sels, L. (2003). HRM and perform-ance, research without theory? A literature review.Research Report, Department of Applied Economics,0333, 44.

Doorewaard, H., Huys, R. and Van Hootegem, G.(2002). Team responsibility structure and teamperformance. Personnel Review, 31, 356–370.

*Dunlop, J.T. and Weil, D. (1996). Diffusion andperformance of modular production in the U.S.apparel industry. Industrial Relations, 35, 334–355.

Dunphy, D. and Bryant, B. (1996). Teams: panaceas orprescriptions for improved performance? HumanRelations, 49, 677–699.

Dyer, L. and Reeves, T. (1995). Human resourcestrategies and firm performance: what do weknow and where do we need to go? InternationalJournal of Human Resource Management, 6, 656–670.

*Elmuti, D. (1997). The perceived impact ofteambased management systems on organizationaleffectiveness. Team Performance Management,3(3), 179–192.

Gladstein, D. (1984). Groups in context: a modelof task group effectiveness. Administrative ScienceQuarterly, 29, 499–517.

*Glassop, L.I. (2002). The organizational benefits ofteams. Human Relations, 55, 225–249.

Glew, D.J., O’Leary-Kelly, A.M., Griffin, R.W. and VanFleet, D.D. (1995). Participation in organizations:a preview of the issues and proposed frameworkfor future analysis. Journal of Management, 21,395–421.

*Godard, J. (2001). High performance and the trans-formation of work? The implications of alternativework practices for the experience and outcomes ofwork. Industrial and Labour Relations Review, 54,776–805.

June 2008


Guest, D.E. (2001). Human resource management:when research confronts theory. InternationalJournal of Human Resource Management, 12,1092–1106.

Guest, D.E., Michie, J., Conway, N. and Sheehan, M.(2003). Human resource management and cor-porate performance in the UK. British Journal ofIndustrial Relations, 41, 291–314.

Guzzo, R.A. and Dickson, M.W. (1996). Teams inorganizations: recent research on performance andeffectiveness. Annual Review of Psychology, 47,307–338.

Hackman, J.R. (1987). The design of work teams.In Lorsch, J.W. (ed.), Handbook of OrganizationalBehaviour. Englewood Cliffs, NJ: Prentice Hall.

Hackman, J.R. and Oldham, G.R. (1976). Motivationthrough the design of work: test of a theory. Organ-izational Behaviour and Human Performance, 16,250–279.

Hackman, J.R. and Oldham, G.R. (1980). WorkRedesign. Reading, MA: Addison-Wesley.

*Harley, B. (2001). Team membership and theexperience of work in Britain: an analysis of theWERS98 data. Work, Employment and Society, 15,721–742.

Hammer, M. and Champy, J. (1993). Reengineeringthe Corporation: A Manifesto for Business Revolution.New York: HarperCollins.

*Hamilton, B.H., Nickerson, J.A. and Owan, H.(2003). Team incentives and worker heterogeneity:an empirical analysis of the impact of teams onproductivity and participation. Journal of PoliticalEconomy, 111, 465–497.

Hart, S. and Banbury, C. (1994). How strategy-making processes can make a difference. StrategicManagement Journal, 15, 251–269.

Ichniowski, C., Kochan, T.A., Levine, D., Olson, C.and Strauss, G. (1996). What works at work: over-view and assessment. Industrial Relations, 35,299–333.

*Ichniowski, C., Shaw, K. and Prennushi, G. (1997).The effects of human resource managementpractices on productivity: a study of steel finishinglines. American Economic Review, 87, 293–313.

Karau, S.J. and William, K.D. (1993). Social loafing:a meta-analytic review and theoretical integration.Journal of Personality and Social Psychology, 65,681–706.

Katzenbach, J. and Smith, D. (1993). The Wisdom ofTeams: Creating the High Performance Organiza-tion. Cambridge, MA: Harvard Business SchoolPress.

Kleinschmidt, M. and Pekruhl, U. (1994). Kooperation,Partizipation und Autonomie: Gruppenarbeit indeutschen Betrieben. Arbeit, 4, 150–172.

Kuipers, H. and Van Amelsvoort, P. (1990). Slagvaardigorganiseren, inleiding in de sociotechniek alsintegrale ontwerpleer. Deventer: Kluwer.

Lawler, E.E. (1992). The Ultimate Advantage: Creatingthe High Involvement Organization. San Francisco,CA: Jossey-Bass.

Leamer, E.E. (1983). Let’s take the con out of econo-metrics. American Economic Review, 73, 31–43.

Manz, C. and Sims, H. (1980). Self management as asubstitute for leadership: a social learning theoryperspective. Academy of Management Review, 5,361–367.

Manz, C.C. and Sims, H.P. (1987). Leading workersto lead themselves: the external leadership of self-managing work teams. Administrative ScienceQuarterly, 32, 106–128.

*Mathieu, J.E., Gilson, L.L. and Ruddy, T.M. (2006).Empowerment and team effectiveness: an empiricaltest of an integrated model. Journal of AppliedPsychology, 91, 97–108.

*McNabb, R. and Whitfield, K. (1997). Unions,flexibility, team working and financial performance.Organization Studies, 18, 821–838.

Mueller, F., Procter, S. and Buchanan, D. (2000).Teamworking in its context(s): antecedents, natureand dimensions. Human Relations, 53, 1387–1424.

Osterman, P. (1994). How common is workplacetransformation and who adopts it? Industrial andLabour Relations Review, 47, 173–188.

*Osterman, P. (2000). Work reorganization in an eraof restructuring: trends in diffusion and effects onemployee welfare. Industrial and Labour RelationsReview, 53, 179–196.

Pasmore, W.A. (1988). Designing Effective Organiza-tions: the Sociotechnical Systems Perspective. NewYork: Wiley.

*Paul, A.K. and Anantharaman, R.N. (2003). Impactof people management practices on organizationalperformance: analysis of a causal model. Inter-national Journal of Human Resource Management,14, 1246–1266.

*Power, J. and Waddell, D. (2004). The link betweenself-managed work teams and learning organisa-tions using performance indicators. The LearningOrganization, 11, 244–259.

*Procter, S. and Burridge, M. (2004). Extent, intensityand context: teamworking and performance inthe 1998 UK Workplace Employee Relations



Survey (WERS 98). IIRA HRM Study Group WorkingPapers in Human Resource Management, No. 12.

Ramsay, H., Scholarios, D. and Harley, B. (2000).Employees and high-performance work systems:testing inside the black box. British Journal ofIndustrial Relations, 38, 501–531.

Robbins, H. and Finley, M. (1995). Why Teams Don’tWork. Princeton, NJ: Peterson’s/Pacesetter Books.

Rust, R.T., Lehmann, D.R. and Farley, J.U. (1990).Estimating publication bias in meta-analysis.Journal of Marketing Research, 27, 220–227.

Salas, E., Burke, C.S. and Cannon-Bowers, J.A. (2000).Teamwork: emerging principles. InternationalJournal of Management Reviews, 2, 339–356.

Schumann, M., Baethge-Kinsky, V., Kuhlmann, M.,Kurz, C. and Neumann, U. (1994). TrendreportRationalisierung – Automobilindustrie, Werkzeug-maschinenbau, Chemische Industrie. Berlin: EditionSigma.

Sels, L. (2002). Strategisch Management van HumanResources. Maakt het een verschil? Inaugurale lesBelgische Francqui-leerstoel 2002–2003. UniversiteitAntwerpen, Faculteit Toegepaste EconomischeWetenschappen.

Shepperd, J.A. (1993). Productivity loss in perform-ance groups: a motivation analysis. PsychologicalBulletin, 113, 67–81.

Sims, H. and Manz, C. (1996). Company of Heroes:Unleashing the Power of Self-leadership. New York:Wiley.

Sinclair, A. (1992). The tyranny of a team ideology.Organization Studies, 13, 611–626.

Starbuck, W. and Mezias, J. (1996). Opening Pandora’sbox: studying the accuracy of managers’ perceptions.Journal of Organizational Behaviour, 17, 99–117.

*Stewart, G.L. and Barrick, M.R. (2000). Teamstructure and performance: assessing the mediatingrole of intrateam process and the moderating roleof task type. Academy of Management Journal, 43,135–148.

*Tata, J. and Prasad, S. (2004). Team self-management,organizational structure, and judgements of teameffectiveness. Journal of Managerial Issues, 16,248–265.

Van Hootegem, G. (2000). De draaglijke traagheidvan het management. Leuven: Acco.

Wall, T. and Martin, R. (1987). Job and work design.In Cooper, C. and Robertson, I. (eds), InternationalReview of Industrial and Organisational Psycho-logy. Chichester: Wiley.

Waterson, P., Clegg, C., Bolden, R., Pepper, K., Warr,P. and Wall, T. (1997). The Use and Effectivenessof Modern Manufacturing Practices in the UnitedKingdom. Sheffield: Institute of Work Psychology.

*West, M.A., Borrill, C., Dawson, J., Scully, J., Carter,M., Anelay, S., Patterson, M. and Waring, J.(2002). The link between the management ofemployees and patient mortality in acute hospitals.International Journal of Human Resource Manage-ment, 13, 1299–1310.

Womack, J.P., Jones, D.T. and Roos, D. (1991). TheMachine that Changed the World. The Story ofLean Production. How Japan’s Secret Weapon inthe Global Auto Wars Will Revolutionize WesternIndustry. New York: Harper Perennial.

*Zwick, T. (2004). Employee participation andproductivity. Labour Economics, 11, 715–740.

Anne Delarue and Geert Van Hootegem arefrom the Catholic University of Louvain, Centrefor Sociological Research, Section Work andOrganisation, E. Van Evenstraat 2B, 3000Leuven, Belgium. Stephen Procter is from theUniversity of Newcastle, Business School,Armstrong Building, Newcastle Upon TyneNE1 7RU, UK. Mark Burridge is from theSchool of Management, University of Leicester,Leicester LE1 7RH, UK.

Anne Delarue, Geert Van Hootegem, Stephen Procter and · PDF filemechanisms by which...

Documents

Transcript of Anne Delarue, Geert Van Hootegem, Stephen Procter and · PDF filemechanisms by which...