www.elsevier.com/locate/dsw

Journal of Operations Management 24 (2006) 99–123

Could lean production job design be intrinsically motivating?

Contextual, configurational, and levels-of-analysis issues

Suzanne de Treville *, John Antonakis

HEC, University of Lausanne, 616-BFSH-1, 1015 Lausanne, Switzerland

Received 5 July 2003; received in revised form 10 December 2004; accepted 15 April 2005

Available online 20 June 2005

Abstract

Are lean production jobs intrinsically motivating? More than 20 years after the arrival of lean production, this question

remains unresolved. Generally accepted models of job design such as the Job Characteristics Model (JCM, (Hackman, J.R.,

Oldham, G.R. 1976. Motivation through the design of work: test of a theory. Organizational Behavior and Human Performance

16, 250–279.)) cannot explain the occurrence of worker intrinsic motivation in the context of lean production. In this paper, we

extend the JCM to the lean production context to explain the theoretical relationship between job characteristics and

motivational outcomes in lean production. We suggest that a configuration of lean production practices is more important

for worker intrinsic motivation than are independent main effects, and that motivation may be limited by excessive leanness. We

conclude that lean production job design may engender worker intrinsic motivation; however, there are likely to be substantial

differences in intrinsic motivation under differing lean production configurations.

# 2005 Published by Elsevier B.V.

Keywords: Lean production; Intrinsic motivation; Job design; Autonomy

1. Introduction

Are jobs in lean production settings intrinsically

motivating? More than 20 years after the arrival of

lean production in the Western world, this question

remains unresolved. It is also highly controversial.

The extant literature concerning job design and

motivation in lean production is fragmented, with

conflicting claims emanating from operations man-

agement researchers, sociologists, and psychologists.

* Corresponding author. Tel.: +41 21 692 3448.

E-mail address: suzanne.detreville@unil.ch (S.d. Treville).

0272-6963/$ – see front matter # 2005 Published by Elsevier B.V.

doi:10.1016/j.jom.2005.04.001

On one hand, lean production proponents suggest

that workers in lean production settings appear to

display what could be characterized as intrinsically

motivated behavior, appearing to be internally driven

and more productive than in traditional assembly line

settings. This behavior has been linked to improved

manufacturing outcomes and competitiveness (Adler,

1993a; Hayes et al., 1988; Hopp and Spearman, 1996;

Monden, 1983; Schonberger, 1982; Suzaki, 1987;

Womack et al., 1990).

On the other hand, opponents of lean production

argue that it places workers in highly limiting and

alienating conditions; ‘‘motivation’’ in the best case is

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123100

external, with workers simply complying with, and

often resisting, restrictive practices that create

dependent and deskilled workers (e.g., Babson,

1993; Berggren, 1992; Fucini and Fucini, 1990;

Graham, 1995; Kamata, 1982; Milkman, 1997; Post

and Slaughter, 2000; Rinehart et al., 1997).

With few exceptions (e.g., Brown and Mitchell,

1991; Jackson and Mullarkey, 2000), evidence from

both sides is largely anecdotal, and any conclusions

that can be drawn are speculative. Two decades of

discussion have yielded little progress. We attribute

this impasse to the lack of theoretical job-design

models suitable for explaining intrinsic motivation in

the context of lean production. Three key areas need to

be addressed if theory and empirical work are to

advance in this area of motivation: (a) the role of

contextual factors, (b) the configural or synergistic

effects of work practices, and (c) levels of analysis

(i.e., individual and organizational levels) implica-

tions at which effects are evident and their cross-level

consequences. We briefly introduce the three areas

below.

First, to explain possible antecedents of intrinsic

worker motivation in lean production settings, it is

necessary to root job design models in the context in

which motivation occurs (Parker et al., 2001). General

job design intrinsic motivation models traditionally

have been context-free, assuming that intrinsic motiva-

tion can be predicted in any type of context (Blair and

Hunt, 1986; Parker et al., 2001). The lean production

context, however, demonstrates the limits of context-

free models. For example, the Job Characteristics

Model (JCM,Hackman andOldham,1975, 1976, 1980)

– whichwewill use as a platform for developing amore

complete model of work motivation – specifies that

autonomy, defined as freedom concerning work proce-

dures and timing,1 is a sine quanon for the emergence of

intrinsically motivating jobs. Lean production, though,

is characterized by process standardization, with

interdependencies resulting from lean production’s

focus on flow, teamwork, and short cycle times.

Standardized processes reduce worker autonomy

1 As we will explain later, the definition of autonomy inherent in

the JCM is very limited and restricting. Part of reconciling moti-

vated behavior under lean production with job design theory

requires re-incorporating other key aspects of autonomy such as

responsibility and decision-making authority.

almost completely. According to the JCM, lean

production jobs simply cannot be intrinsically motivat-

ing. As we will argue in this paper, however, intrinsic

motivation is theoretically possible in lean production

settings, but the type of explicative model must be

concordant with the contextual forces that act on the

phenomenon observed, including the choice of job

characteristics, outcomes, and moderators.

Second, many motivational models focus on the

effects of individual practices on workers, ignoring the

synergistic effect of various independent practices

operating simultaneously. On the shop floor, however,

workers do not perceive single practices independent

of other practices; workers perceive joint effects, a

whole that is not merely an addition of the component

parts. For example, workers perceiving reduced levels

of autonomymight still be motivated if that perception

is accompanied by other job-design factors that

compensate for, justify, and overcome this apparent

lack of internal motivation. Thus, the impact of a

configuration of practices on workers may be

substantially different from the impact of the bivariate

effects of these practices—the gestalt (whole) effect

may account for more variance in the dependent

measures than the summed (individual) effects of the

parts (as demonstrated by Stajkovic and Luthans,

2003).

Third, current motivational theories are limited by a

lack of multilevel theorizing and testing. Organiza-

tional practices are implemented at the organizational

level of analysis, but the effects of these practices are

hypothesized to impact workers (i.e., at the individual

level). Furthermore, individual-level outcomes (e.g.,

worker performance) are assumed to have impacts at

the organizational level of analysis (e.g., organiza-

tional performance). These cross-level effects are

implicit in most motivation models; however, with

very few exceptions, the theoretical and empirical

consequences of these cross-level processes have been

ignored in the motivation and job design fields over the

last several decades (Klein et al., 1994; Pierce and

Dunham, 1976; Roberts and Glick, 1981; Seibert

et al., 2004).

We believe that questions concerning lean produc-

tion job design and intrinsic motivation are timely and

important to answer. There is a clear need to better

understand the kinds of motivational effects that lean

production practices bring about. Efforts to create

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123 101

alternatives to lean production, while appealing, have

not been successful at convincing large numbers of

firms to replace their lean production practices (Adler

and Cole, 1993; Dankbaar, 1997). Lean production

practices clearly are associated with better organiza-

tional performance (e.g., Flynn et al., 1995; Fujimoto,

1999; McKone et al., 2001; Porter, 1996; Shah and

Ward, 2003; White et al., 1999; Wood et al., 2004),

and lean production is here to stay. Even lean

production critics Rinehart, Huxley, and Robertson

(1997: 2) acknowledge ‘‘If there is one non-debatable

proposition in the early literature, it surely must be the

claim that lean production will be the standard

manufacturing mode of the 21st century.’’

Thus, our major purpose is to develop a theoretical

job-design model that will allow researchers to

grapple more effectively with the intrinsic motiva-

tional impacts of lean production job design. The JCM

is the most widely accepted work design theory of

worker motivation (e.g., Parker et al., 2001; Spector,

2003). In the form proposed by Hackman and Oldham

(1975, 1976, 1980), however, the JCM is not

compatible with the lean production context. Much

of the theory underlying the JCM, however, is useful

for coming to terms with this conundrum. We propose

an extended JCM that we suggest will better explain

intrinsic motivation in the lean production context.

Furthermore, the incompatibility between JCM theory

and lean production practice is sufficiently surprising

to warrant revisiting the JCM’s proposed constructs

and causal relationships (Whetten, 1989).

In the remainder of the paper, we first introduce

lean production as a manufacturing system (see

Section 2). Next, we review the JCM in Section 3,

then extend it to the lean production context (Section

4). We offer a summary and conclusions in Section 5.

2. A review of lean production

2.1. Defining lean production

Lean production is a manufacturing system whose

objective is to streamline the flow of production while

continually seeking to reduce the resources (e.g.,

direct and indirect labor, equipment, materials, space,

etc.) required to produce a given set of items; any slack

in the system is referred to as ‘‘waste’’ (e.g., Womack

et al., 1990). Rather than setting a goal of a specific

level of leanness, lean production is focused on a

continuous improvement process. Each improvement

in flow or reduction in waste leads to new goals

(Monden, 1983; Womack et al., 1990).

Early adopters of lean production changed the

nature of competition in repetitive manufacturing

industries as a result of increased productivity, quality,

and rates of learning. The impact of lean production on

competitiveness was so profound that for several years

after its arrival, the competitive strategy of companies

not using lean production in those industries was

reduced to catching up with early implementers (e.g.,

Adler and Cole, 1993; Hayes et al., 1988; Porter, 1996;

Schonberger, 1982; Womack et al., 1990).

The term ‘‘lean production’’ originated from

researchers at the International Motor Vehicle Project

to distinguish between mass production as carried out

on traditional Western assembly lines and repetitive

manufacturing (on assembly lines or in cells) based on

the Toyota Production System (see Krafcik, 1988). The

Toyota Production System was born out of scarcity in

post-World War II Japan (see Fujimoto, 1999). The

buffers required to maintain a high capacity utilization

given line imbalances, quality problems, alienated

workers with narrow skills, and other sources of

variability were too costly for Toyota. The solution for

Toyota was to operate with minimum inventory buffers

while attempting tomaintain a high capacity utilization:

objectives that are, prima facie, conflicting. To achieve

these objectives simultaneously required the reduction

of variability in all forms.

Quality problems, whether originating in-house or

from external stakeholders (e.g., suppliers), had to be

eliminated. To that end, workers were provided with

expertise and equipment and were encouraged to help

improve the production system. Workers also were

cross-trained so that they could fill in for an absent

colleague or be reassigned as necessary to balance the

line. Demandwas smoothed so that the production line

would be sheltered from its inherent variability; for

example, some cars required more labor time than did

others, and without demand smoothing, a sequence

with several labor-intensive cars together would

require a reduced capacity utilization for a given

performance level.

Most important, Toyota management decided that

high utilization and low inventory (i.e., lean produc-

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123102

tion) would require workers who were committed to

their tasks and to the company, and who could be given

responsibility for quality and for nonmanufacturing

tasks far beyond what was typical in traditional mass

production (Fujimoto, 1999; Monden, 1983). Reduced

buffers imply that production problems would stop the

line unless competent and committed workers were

able to take action to solve such problems immedi-

ately. Alienated workers would be neither willing nor

able to respond to problems that arose, and they would

be unlikely to exert themselves in an effort to produce

the highest quality goods. Also, labor unrest arising

from worker alienation would cause a reduction in

utilization and an increase in inventory.

Therefore, a key element of the Toyota Production

System was a subsystem referred to as ‘‘respect-for-

humanity,’’ which explicitly sought to incorporate

worker suggestions and communicate management

respect for, and appreciation toward, workers (Hopp

and Spearman, 1996; Monden, 1983; Womack et al.,

1990). MacDuffie (1995a: 198) built upon this concept

by stressing the need ‘‘for expansion of work force

skill and conceptual knowledge required for problem-

solving under this approach.’’

The way lean production was conceptualized was a

radical departure from the precepts of traditional

mass-production manufacturing, which becomes

evident when looking at how lean production changes

the ‘‘factory physics’’ (Hopp and Spearman, 1996) of

a given operation. According to Hopp and Spearman,

many of the important performance metrics of a given

operation can be derived by a set of axiomatic

mathematical laws based on the utilization of capacity,

the sizes of buffers, and the various types of

variability: All variability must be buffered by some

combination of capacity and inventory. We suggest

that it is the factory physics of lean production

(reduction of capacity and inventory buffers, requiring

reduction of system variability) that distinguishes it

from other manufacturing systems such as traditional

assembly line or batch manufacturing,2 and we

propose the following definition:

2 Factory physics can also serve to distinguish between lean

production and the manufacturing approach carried out by Volvo’s

Kalmar and Uddevalla factories, which explicitly included capacity

and inventory buffers and accepted variability (Adler and Cole,

1993; Berggren, 1992).

Lean production is an integrated manufacturing

system that is intended to maximize the capacity

utilization and minimize the buffer inventories of a

given operation throughminimizing system variability

(related to arrival rates, processing times, and process

conformance to specifications).

As suggested by Wacker (2004: 6–7), defining lean

production according to its factory physics follows the

rules of formal conceptual definition development: (a)

following the ‘‘rule of replacement’’, the term ‘‘lean

production’’ can be replaced by ‘‘a manufacturing

system that maximizes capacity utilization and

minimizes buffer inventories through minimizing

system variability’’ without loss of meaning; (b) the

concepts of capacity utilization, buffer inventories,

and variability are clear and unambiguous; (c) it is

parsimonious with respect to terms used. This

definition of lean production is also broad enough

to be inclusive of firms following lean production

principles in spite of implementation differences, as

well as being narrow enough to exclude companies

that do not follow the essential characteristics of lean

production (see McKelvey, 1978).

The demanding factory physics of lean production

are achieved over time through a combination of

synergistic and mutually reinforcing practices, which

have been grouped into several complementary

subsystems including (but not limited to) just-in-time

(JIT) manufacturing, total quality management

(TQM), total preventive maintenance (TPM), Kaizen

(continuous improvement), design for manufacturing

and assembly (DFMA), and supplier management,

with human resources management practices under

the respect-for-workers umbrella serving as the glue to

hold the overall system together (Flynn et al., 1995;

MacDuffie, 1995a; Sakakibara et al., 1997; Shah and

Ward, 2003; Snell and Dean, 1992).

The literature on lean production, JIT, and TQM

reveals a large number of practices that can be

associated with these subsystems (e.g., Davy et al.,

1992; Forza and Choo, 1996; Hall, 1983; Harmon and

Peterson, 1990; McLachlin, 1997; Monden, 1983;

Sakakibara et al., 1997; Schonberger, 1982; Shah and

Ward, 2003; White et al., 1999; Womack et al., 1990),

and a variety of different practices can be used to

achieve a given lean production objective. Spear and

Bowen (1999: 104) noted that ‘‘Toyota does not

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123 103

consider any of the tools or practices – such as kanbans

or andon3 cords, which so many outsiders have

observed and copied – as fundamental to the Toyota

Production System. Toyota uses them merely as

temporary responses to specific problems that will

serve until a better approach is found or conditions

change.’’However, leanproductionpractices are geared

toward achieving one or more elements of lean

production factory physics: leanness through inventory

reduction and increased capacity utilization, and

variability reduction (placing special emphasis on

respect forworkers). In the following section,webriefly

review some of the more common practices, which we

group according to their factory physics dimension.

2.2. Lean production practices

2.2.1. Leanness: reducing inventory and

increasing capacity utilization

Lean production inventory reduction practices are

intended to (a) permit control over the quantity ofwork-

in-process (WIP) inventory in the system, (b) link

production and delivery to some evidence of down-

stream demand, and (c) reduce lot sizes, as batching is

oneof theprimarycausesofhighWIP.WIP iscontrolled

andproduction is linked todownstreamdemandthrough

a ‘‘flow-control’’ mechanism such as kanban (de

Treville, 1987; Monden, 1983; Schonberger, 1982).

Lot size reduction may require reduction of setup times

aswellasaflow-basedlayout(anassemblylineorcell) to

reduce batching due to part transportation (Hall, 1983;

Harmon and Peterson, 1990;Monden, 1983; Schonber-

ger, 1982; Wemmerlov and Johnson, 1997).

Capacity utilization is also increased through the

lean production practice of balancing the production

line, as bottlenecks created by line imbalances limit

the throughput of a given system. Giving workers the

authority to signal the need for extra help by pulling an

andon cord permits the line to operate at a higher

utilization level than would otherwise be possible

(e.g., Womack et al., 1990).

One of the more controversial lean production

practices attempts to increase utilization by providing

3 An andon cord is used by a worker who is having difficulty

completing the task within the cycle, or who has encountered a

production or quality problem, to signal to the team leader a need for

support (Monden, 1983).

employees with less time than needed to accomplish a

given task (e.g., by increasing the speed of the

production line without adding more workers). This

practice follows the assumption made by Ohno and

other senior managers at Toyota that scarcity will result

in a creative tension that will be motivating to workers

(Babson, 1993; Fucini and Fucini, 1990; Imai, 1986;

Monden, 1983; Rinehart et al., 1997; Womack et al.,

1990). There is, therefore, a constant contradiction in

lean production contexts between equipping workers

well enough to get the job done,while leaving resources

scarce enough to encourage search behavior (see Cyert

and March, 1992).

2.2.2. Variability reduction

Lean production variability reduction begins with

standardization and documentation of processes, along

with the requirement that workers perform processes

according to the documents (Adler, 1993a; Adler and

Borys, 1996; Edelson and Bennett, 1998; Fujimoto,

1999; Imai, 1986; Klein, 1991). Lean production and

standard operating procedure (SOP) theory call for the

involvement of workers (usually operating in teams) in

the development of procedures for two reasons: (a) only

the people actually running the process have access to

many key types of knowledge concerning how the

process operates in practice, and (b) it is generally

believed that participation in development of proce-

dureswill giveworkers a sense of ownership, increasing

their willingness to run the process as documented

(Adler, 1993a;Adler andBorys, 1996; de Treville et al.,

2005; Edelson and Bennett, 1998; Fujimoto, 1999;

Imai, 1986; Klein, 1991).

Process standardization and documentation lays a

foundation for statistical process control (SPC), a

second lean production practice dedicated to the

reduction of variability (e.g., Flynn et al., 1995). SPC

is concerned with statistical analysis of process data to

distinguish between random and nonrandom variation

(Edelson andBennett, 1998). For example, process data

can be collected, aggregated, and charted to determine

whether a process is running under statistical control

(i.e., nothing has changed) or whether there is some

factor causing the process variability. Note that in a

situation where a process is not standardized, or

workers do not run the process according to the

documents, it is impossible for a process to run under

statistical control (Edelson and Bennett, 1998).

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123104

Variability also is reduced in lean production

through use of equipment and parts that reduce the

probability of operator error. For example, a machine

can be designed so that it is impossible to insert a part in

the wrong direction, or so that a buzzer sounds if the

machine detects an abnormality (common terms for

such machine design are jidoka or poke-a-yoke,

Fujimoto, 1999;Monden, 1983).Alongwith equipment

(such as andon cords) thatmakes it visually clear that an

error or problem is occurring, lean production also

emphasizes the visual display of quality-related data

(e.g., Hopp and Spearman, 1996; Schonberger, 1982).

Lean production and TQM emphasize the elimina-

tion of variability in incoming raw materials through a

variety of supplier management tools and practices,

ranging from the formation of alliances and asset

specificity (Dyer, 1996), to better exchange of

information with fewer suppliers (Handfield, 1993),

to a simple insistence that parts of consistent quality be

delivered on time (Walleigh, 1986). More generally,

the production line is protected from arrival rate

variability through demand-smoothing practices, so

that the production schedule does not change from day

to day (sometimes even from hour to hour, Monden,

1983).

Under traditional mass production, cycle times

(i.e., the amount of time that a part spends in front of a

given workstation under synchronous production) are

short, typically around 1 min in automobile assembly

(Adler, 1993a; Graham, 1995; Rinehart et al., 1997)

and around 2 min in other assembly operations (see

Harmon and Peterson, 1990: 105 for a discussion of

the relationship between cycle time and efficient

production in assembly production). These cycle

times have remained unchanged in the transition from

traditional mass to lean production.4 Although giving

each worker such a small part of the total task results

in monotonous jobs that are likely to cause repetitive

strain injuries, lean production retains short cycle

times to minimize variability, attempting to compen-

sate for the detrimental impact through job rotation

and involving workers in nonproduction tasks.

Cross-training operators allows teams of workers to

reduce the variability caused by absent or injured team

4 Although short cycle times are the norm in lean production, the

length of the cycle time can be increased for highly skilled workers

or during times of slow demand.

members (Babson, 1993; Imai, 1986; Monden, 1983;

Rinehart et al., 1997; Womack et al., 1990). Improving

the manufacturability of items to be produced also

reduces variability (Womack et al., 1990).

Keeping the plant clean and orderly is a lean

production practice that has been observed to play a

key role in variability reduction. Disorder and dirt

encourage quality problems and hinder problem

solving (Collins and Schmenner, 2003; Davy et al.,

1992; Forza and Choo, 1996; Hayes, 1981).

2.2.3. Respect for workers

Respect for workers can be conceptualized as the

glue that holds the other lean production factory

physics dimensions together. As described previously,

the objectives of respect for workers are to reduce

alienation through expressing respect, recognition,

and appreciation and making the job more interesting

(thereby reducing variability), as well as to make

maximum use of worker knowledge (thereby increas-

ing resource utilization and reducing the need to hold

buffer inventories).

Lean production respect for workers begins, at least

ideally, with a competitive wage (authors such as

Graham (1995) and Milkman (1997) indicate that

workers earn more money in lean production plants

than in other jobs that are locally available), and giving

workers the training and equipment required to allow

them to perform well (Adler, 1993a; Monden, 1983;

Womack et al., 1990). Respect for workers also is

encouraged by the lean production/TQM practice of

grouping workers into teams according to their

production line or cell (Hackman and Wageman,

1995; Kenney and Florida, 1993). Whereas production

tasks under lean production usually are carried out by

individuals (Rinehart et al., 1997), teams of workers

collaborate to attack quality problems and carry out

lateral tasks. Teams take responsibility for quality and

discipline members who do not perform tasks correctly

(Adler, 1993a; Barker, 1993; Graham, 1995; Kenney

and Florida, 1993; Rinehart et al., 1997), and teams

reallocate tasks when a member is injured or absent.

Team membership has been observed in lean produc-

tion implementations to be a source of both support

and stress. Support, in that team members help each

other, and stress, in that team norms, and the social

conformance they can induce (cf. Cialdini and

Goldstein, 2004) could be considered as one of the

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123 105

main extrinsic motivators in lean production (Graham,

1995). This conformance is a source of extrinsic

motivation that creates a ‘‘strong situation’’ (see

Section 4.1.3 on Moderators). Rinehart et al. (1997)

found that workers at CAMI had mixed reactions to

working as part of a team.

Respect for workers calls for the transfer of certain

types of authority and responsibility (including

inspection, trouble-shooting, statistical quality con-

trol, and equipment maintenance) to lower levels of

the organization (Boyer, 1996; Davy et al., 1992;

MacDuffie, 1995a; McLachlin, 1997; Sakakibara

et al., 1997). Other authors, however, note that the

practice of decentralization of authority as discussed

in the lean production literature consists primarily of

the transfer of technical tasks rather than a true shifting

of power (Rinehart et al., 1997).

Finally, respect for workers is demonstrated by

practices such as referring to workers as associates

(e.g., Graham, 1995), emphasizing the relationship

between the worker and the company (Adler, 1993a;

Babson, 1993; Graham, 1995), communication

between management and workers (Monden, 1983),

and even gestures such as giving workers business

cards (Adler, 1993a).

These lean production factory physics have

implications for job design, which as previously

mentioned may have an impact on worker intrinsic

motivation. In the following section, we begin with a

description of intrinsic motivation. Then, we introduce

the JCM, which will serve as a starting point for our

job design model of lean production.

5 Performance improvements arising from motivation (whether

intrinsic or extrinsic) also require competence on the part of the

worker, as well as the necessary resources (Pinder, 1984).6 We thank an anonymous reviewer for pointing out that motivation

should not be included as an outcome measure in the JCM because it

reduces the hypothesized model relations to a tautology (i.e., job

scope! motivating potential of the job! workermotivation). Thus,

outcome measures should include theoretical consequences of the

critical psychological states (i.e., intrinsic motivation), including

performance, satisfaction, turnover, and so forth.

3. The job characteristics model (JCM)

3.1. Work motivation

Work motivation ‘‘is a set of energetic forces that

originate both within as well as beyond an individual’s

being, to initiate work-related behavior, and to

determine its form, direction, intensity, and duration’’

(Pinder, 1984: 8). When a worker ascribes the source

of motivation to outcomes associated with performing

the task then the (work) motivation is referred to as

intrinsic (Brief and Aldag, 1977: 497). In other words,

the individual exerts greater effort because of aspects

of the task itself, rather than because of any (extrinsic)

reward or punishment. Work motivation usually is

comprised of both extrinsic and intrinsic components.

It is generally believed that the intrinsic component

leads to reduced worker alienation and higher

performance outcomes, especially concerning the

quality of work (e.g., Hackman and Lawler, 1971;

Hackman and Oldham, 1976).5

Hackman and Lawler (1971) considered both (a) the

level of what they referred to as intrinsic motivation-

operationalized as worker satisfaction with work

outcomes, which they also referred to as internal

motivation as well as intrinsic motivation and (b) the

focus of that motivation (operationalized as the degree

to which ‘‘employees tend to report feeling internal

pressures to take personal responsibility for their work

and to do high quality work,’’ Hackman and Lawler,

1971: 273). In developing the JCM, Hackman and

Oldham (1975: 162) shifted completely from the term

‘‘intrinsic motivation’’ to use of the term ‘‘internal

motivation,’’which theydefinedas ‘‘thedegree towhich

the employee is self-motivated to performeffectively on

the job—that is, the employee experiences positive

internal feelings when working effectively on the job,

and negative internal feelings when doing poorly.’’

It is important to note that ‘‘internal motivation’’ is

not actually motivation per se, but instead results from

the intrinsic motivation captured by the critical

psychological states described below (Spreitzer,

1995; Thomas and Velthouse, 1990).6

3.2. Job characteristics

Hackman and Lawler (1971) proposed that a

substantial portion of the variation in worker outcomes

(internal ‘‘motivation’’; i.e., feeling good about doing

well, job satisfaction, performance, and absenteeism)

could be explained by the characteristics or specific

attributes constituting the job and how workers

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123106

8 The multiplicative MPS has (seldom-heeded) implications for

testing. Given the scale dependency of the multiplicative term (the

cross product of the three psychological states), it is essential to first

partial out the main and lower-level interaction effects from the

multiplicative term in the regression equation (Aiken and West,

1991; Cohen and Cohen, 1983; Evans, 1991; Jaccard et al., 1990).

Evans (1991: 7) went as far as to state that anything short of this

procedure would yield results that would be ‘‘profoundly and fatally

flawed’’ (see also Ferris and Gilmore, 1985) and that the extant

literature was ‘‘cluttered with suspect results that continue to be

cited approvingly by subsequent authors and by studies that con-

perceived these attributes. Hackman and Oldham

(1975, 1976, 1980) developed this theory into the

JCM, which quickly became the most generally

accepted theory of job-design motivation and remains

inwide use today (Jex, 2002; Spector, 2003).According

to the JCM, objective changes to a given job are

expected to change how the worker perceives the job

along five core job dimensions: skill variety, task

identity, task significance, autonomy, and feedback.

3.3. Critical psychological states

The presence of the above five core job dimensions

is theorized to determine the extent to which workers

will experience meaningfulness (caused by skill

variety, task identity, and task significance), respon-

sibility (caused by autonomy), and knowledge of the

results of their work (caused by feedback). These three

critical psychological states theoretically mediate the

relations between worker perceptions of job char-

acteristics and various outcome measures (internal

motivation, high quality work performance, satisfac-

tion, and low absenteeism and turnover Hackman and

Oldham, 1975, 1976).7

Thomas and Velthouse (1990) and Spreitzer (1995)

treated the critical psychological states – to which they

added ‘‘competence’’ – as dimensions of intrinsic

motivation, which they equated with empowerment

(empowerment and intrinsic motivation were also

treated as equivalent by Conger and Kanungo, 1988).

3.4. The motivating potential score (MPS)

Integral to the JCM theory is a summary index that

serves to translate the measures of job characteristics

into a single number, which then serves as an estimate

of the (intrinsic) motivating potential of a given job.

The MPS is calculated as follows:

MPS

¼ ðskill varietyþ task identityþ task significanceÞ3

� autonomy� feedback:

7 The 1980 version of the JCM proposed high internal work

motivation, satisfaction (growth and general), and high work effec-

tiveness as outcomes, with absenteeism eliminated due to lack of

empirical support (Hackman and Oldham, 1980: 90, 93).

Hackman and Lawler (1971) originally proposed

that a high level of each of the states was a necessary

condition for a job to be maximally motivating. They

proposed a configural grouping (i.e., a specific kind of

interaction confined to predetermined levels of the

interacting independent variables). This configural

effect was tested by grouping together subjects who

perceived high levels of each of the three critical

states. Hackman and Lawler compared these subjects

on various dependent outcomes with subjects who

perceived varying levels of the critical states (either all

low, or various combinations). Using this approach,

they found significant differences between groupings.

The ‘‘high-high-high’’ group scored significantly

higher than the ‘‘low-low-low’’ or mixed groups on

various outcome measures. Hackman and Lawler

(1971: 277) concluded that this approach produced

more robust results than a main-effects correlational

model.

Following Hackman and Lawler (1971), Hackman

and Oldham (1975) proposed that each of the three

critical psychological states was needed for motiva-

tion to occur and suggested that a multiplicative effect

best represented this proposition. There is, however, a

major incongruence between how this ‘‘synergistic’’

effect was originally conceived by Hackman and

Lawler and how it was later tested by Hackman and

Oldham (and the extant literature). A multiplicative

effect tests a full interaction model (as in moderated

multiple regression, see Aiken and West, 1991) rather

than the specific interaction effect originally implied

by JCM theory.8

tinue to use the suspect methods because ’they have always been

done that way’ . . . pleas to conduct multiplicative tests correctly

have been ignored by the reviewers and editors of the very journals

that published the original critiques of the inappropriate techniques’’

(Evans, 1991: 13).

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123 107

Along similar lines, Meyer et al. (1993) specifically

suggested that work motivation models – including the

JCM – should be tested configurally, because their

attributes may be related in a reciprocal and nonlinear

fashion, contrary to traditional models that examine

linear relationships and unidirectional causation.

3.5. Moderators

JCM theory proposes that its causal relationships are

moderated by individuals’ Growth Need Strength

(GNS, Hackman and Oldham, 1975, 1976, 1980).

Hackman and Lawler (1971) based their inclusion of

individual differences in general, andGNS inparticular,

on the expectancy theory of motivation (e.g., Lewin,

1938; Vroom, 1964) and on the motivational implica-

tions of higher-order needs (Alderfer, 1967, 1969;

Maslow, 1943, 1954). In other words, individuals high

in GNS are theorized to benefit more than low-GNS

individuals from enriched jobs because they would be

Fig. 1. The impact of lean production job chara

more satisfied in enriching conditions. By implication,

low-GNS individuals might be overwhelmed by

conditions that are too challenging and would require

them to use skills that were beyond their capabilities.

The validity of these claims is explored later.

4. A JCM model for lean production: context,

configurations, and levels

In this section, we extend the JCM to the lean

production context, considering issues related to

context, configurations, and levels. Our extended

model is shown in Fig. 1. Apart from the direct effects

of lean production systems on organizational perfor-

mance, we suggest that lean production systems have

an effect on individuals’ motivation. As depicted, our

model offers a testable and parsimonious representa-

tion of how organizational-level factors (i.e., the lean

production system) affect individual-level outcomes

cteristics on worker intrinsic motivation.

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123108

(e.g., intrinsic motivation, work performance), which

in turn can be linked to organizational performance.

These causal relations are moderated by the degree of

ideal lean implementation such that when lean

production is implemented in an ideal way, the job

characteristics configuration will engender motivation

whereas when lean production is implemented less-

than-ideally, outcomes are negative. The model,

propositions depicting its causal relations, and the

contextual factors that bound it are presented in Fig. 1.

4.1. Context

In building a job design model specific to the

context of lean production, we are responding to

recent calls to introduce contextual factors into

organizational research (Johns, 2001; Rousseau and

Fried, 2001). Context affects theory building and

testing. Because context constrains the variability that

can be observed, it has implications for the type of

phenomenon that emerges (Rousseau and Fried,

2001). Dubin’s (1976) criteria for theory building

explicitly stated that theories should specify their

boundaries, that is, the conditions (i.e., contexts) under

which the theory is expected to hold. The effect of

context on individual organizational behavior has not

been adequately considered as a boundary condition

of theories (George and Jones, 1997). Parker et al.

(2001) argued that work motivation theories need to be

adapted to explain motivation in different contexts

because the contextual factors can accentuate or

attenuate the explanatory power of independent

variables. As regards the JCM, Roberts and Glick

(1981) specifically noted that contextual factors that

limit how the theory operates have not been

considered (see also Griffin et al., 1981).

4.1.1. The core job characteristics in the lean

production context

Consider the implementation of lean production on

a traditional mass-production assembly line. What is

the expected main effect of the changed factory

physics (reduced inventory, increased capacity utili-

zation, and reduced variability) on worker perceptions

of job characteristics as specified by the JCM? We

begin our evaluation of the JCM in the lean production

context with an evaluation of these main effects in

order to highlight both consistencies and inconsis-

tencies. In this subsection we deal only with the core

job characteristics in the JCM.

4.1.1.1. Skill variety. Does lean production cause

workers to have more variety in their jobs, and to use

more of their skills and talents? Lean production

proponents argue that cross training and job rotation

(resulting in ‘‘multiskilling’’), problem-solving, and

participation (e.g., in the development of standard

operating procedures) cause lean production jobs to

embody more skill variety than found in typical

assembly line work (Adler, 1993a,b; Barker, 1993;

Jackson and Mullarkey, 2000; MacDuffie, 1995a;

Womack et al., 1990). Snell et al. (2000) found that

integrated manufacturing – including JIT and TQM in

addition to advanced manufacturing technology – was

positively related to investment in training and

development of workforce skill in a study of 74

plants. Lean production critics counter that multis-

killing is actually ‘‘multi-tasking,’’ especially given

the short cycle times common to lean production (e.g.,

Rinehart et al., 1997). Moreover, front-line workers

have limited opportunities to engage in problem-

solving and participation (Fucini and Fucini, 1990;

Graham, 1995; Hackman and Wageman, 1995;

Rinehart et al., 1997), and training has little to do

with development of technical skills (e.g., Graham,

1995; Parker and Slaughter, 1995; Rinehart et al.,

1997).

Overall, then, worker perceptions of skill variety

should be positively related to their involvement in

problem solving and participation in activities such as

development of standard operating procedures, and

should have a positive but small relationship with

cross training and job rotation. It has been observed in

practice, however, that as leanness increases over a

critical threshold (i.e., leanness becomes excessive),

worker participation, training, and job rotation

decrease (e.g., Adler et al., 1997; Graham, 1995;

Rinehart et al., 1997).

Proposition 1. Lean production implementation will

result in an increase in skill variety in situations where

workers participate in problem solving, receive train-

ing, and – to a lesser extent – rotate jobs.

Proposition 2. The relationship between lean produc-

tion implementation and skill variety is moderated by

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123 109

excessive leanness, such that when leanness is exces-

sive, the relationship between lean production imple-

mentation and skill variety goes from positive to

negative.

4.1.1.2. Task identity. As conceptualized by the JCM,

task identity has two components: does the worker

complete a whole unit of work, and is the job designed

to permit the worker to see how the individual

contribution relates to the whole product? Consider

the wording of a survey item designed by Hackman

and Oldham (1980: 298): ‘‘The job is a moderate-sized

‘chunk’ of the overall piece of work; the person’s own

contribution can be seen in the final outcome.’’ In the

lean production context, short cycle times make a

positive response to the first part of this question

unlikely. It is possible, however, that implementation

of a flow-based layout, cross-training, and inventory

reduction (as inventory obscures the flow of the

product, reducing the link between various stages)

may result in a modest increase in the score awarded to

the second part of the question (see Adler, 1993a).

Proposition 3. Lean production implementation is

positively related to task identity in that the worker

is better able to see how his or her task contributes to

the whole product.

4.1.1.3. Task significance. In developing the JCM,

Hackman and Oldham (1976: 257; 1980: 75) used the

example of assembling aircraft brakes to emphasize

that tasks that appeared trivial or to have no inherent

value might be perceived as significant by workers for

a variety of reasons. Do the objective job changes that

occur under lean production cause the worker to

perceive tasks as being more significant in ways that

are not captured by the dimensions of skill variety and

task identity?

Task significance as described in the JCM is

independent of job design: significance comes from

the product or from identifying with the social

agenda of the firm. Although the job characteristic of

task significance may be useful in explaining why a

job such as aircraft brake system assembly might

lead to a higher level of experienced meaningfulness

than a job assembling a less essential product, we

propose that in the present case, the inherent

meaning of the task does not aid in evaluating the

impact of lean production job design on the

motivating potential of a given job. Consideration

should therefore be given to removal of this job

characteristic in developing a model for the lean

production context in the interest of parsimony.

Proposition 4. Lean production implementation is

not related to task significance.

4.1.1.4. Feedback. Workers under lean production

receive substantially more feedback from the process

itself than under traditional assembly line manufac-

turing. As employees carry out work activities, they

receive direct and clear information concerning

performance effectiveness. That feedback resulting

from inventory reduction engenders performance

improvement is a tenet of lean production and JIT

theory (e.g., Schonberger, 1982; Schultz et al., 1999;

Suri and de Treville, 1986). Feedback is also

encouraged through a flow-based layout, which makes

it easier for a downstream worker to communicate

demand and defect information upstream.

Efforts to reduce variability are also likely to

increase process feedback. Consider, for example, the

implementation of jidoka and poke-a-yoke devices

that make it immediately clear to the worker whether

the piece or fixture is being inserted correctly. Under

the TQM umbrella, process information is widely

disseminated to workers (MacDuffie, 1997).

Proposition 5. Lean production implementation is

positively related to feedback, resulting from both

the process and co-workers.

4.1.1.5. Autonomy. Implementation of lean produc-

tion results in a massive reduction in autonomy,

defined by Hackman and Oldham (1980: 79) as ‘‘The

degree to which the job provides substantial freedom,

independence, and discretion to the individual in

scheduling the work and in determining the proce-

dures to be used in carrying it out.’’ Individuals under

lean production have almost no freedom, indepen-

dence, or discretion in either how the work is

scheduled (due to inventory reduction, short cycle

times, and a flow-based layout), or the procedures to

be used (due to process standardization).

What is more, employee involvement activities

may reduce such freedom even more. Researchers

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123110

10 Philosopher Immanuel Kant saw autonomy as self-government

stemming from morality, with morality arising from knowledge

have noted that process improvements made by

workers under traditional assembly line manufactur-

ing often resulted in workers having a few seconds of

extra break time, whereas those extra seconds now are

used to benefit the company instead of the worker (see

Adler, 1993b; Fucini and Fucini, 1990; Graham, 1995;

Post and Slaughter, 2000; Rinehart et al., 1997; Sewell

and Wilkinson, 1992).

JCM theory, therefore, would predict an absence of

the critical psychological state of experienced

responsibility and hence low levels of the outcome

variables. The assumption, however, might not be

tenable given the restricted definition used and failure

to take into account how ‘‘autonomy’’ might interact

with other contextual factors/variables. For example,

Langfred (2004) demonstrated that contexts providing

workers with low individual autonomy (concerning

procedures) could actually engender high perfor-

mance. The result depended on the trust and

monitoring levels in a team. Thus, in the case of

autonomy – in lean production contexts in particular –

one size might not fit all. Other researchers (Friedman,

1977; Jurgens, 1986; Sayer, 1986) have suggested that

it is more appropriate to speak of responsible

autonomy, which Wood (1989) makes clear is a

fundamentally different construct from autonomy as

traditionally used in the job design literature. This

conceptualization of autonomy seems to be applicable

to the context of lean production, in which workers are

expected to have higher levels of actual responsibility

and decision-making authority than under traditional

assembly line manufacturing (e.g., Davy et al., 1992;

Flynn et al., 1999; Jackson et al., 1993).9 Klein (1989,

1991) and Hackman and Wageman (1995) observed

that worker responsibility and participation could

engender ‘‘autonomy’’ even in the absence of freedom

concerning procedures and timing. Mullarkey, Jack-

son, and Parker (1995: 69) assessed individual

autonomy using three measures: timing control,

method control, and boundary control (a measure of

vertical integration tapping) ‘‘the extent to which

operators are involved in a variety of activities that are

9 Promises of responsibility and authority made in the context of

lean production are not always kept, (see, for example, Fucini and

Fucini, 1990; Graham, 1995; Rinehart et al., 1997). These authors

note an apparent reduction in worker motivation and commitment

when workers realize that the promised responsibility and authority

are not materializing.

associated with traditional supervisory or first-line

management activities.’’ Jackson and Mullarkey

(2000) found that, compared with workers using

conventional manufacturing practices, choice auton-

omy regarding timing was rated as significantly lower

by lean production workers. However, they found that

lean production workers experienced a higher level of

production responsibility (see also Mullarkey et al.,

1995).

One can conclude from the lean production

context, therefore, that autonomy is not a single

construct, but instead can be viewed as comprising two

distinct constructs: (a) choice—that is, freedom

concerning procedures and timing, corresponding to

the JCM definition of autonomy (Hackman and

Lawler, 1971; Hackman and Oldham, 1975, 1976,

1980), and (b) responsible autonomy—an increase in

accountability arising from decentralization of author-

ity, power sharing, and participation in decision

making.

Responsible autonomy is less related towhether the

worker can operate without constraints and more

related to the degree to which the worker plays an

active role in setting the rules by which he or she is

bound and whether these rules are congruent with the

worker’s reasoning. Aworker, for example, who plays

an active role in development of standard operating

procedures, who is held personally accountable for

results, and who chooses to operate under the

constraints inherent in standardization because of

the expected performance improvement and fulfill-

ment of work-related needs is acting in a morally

autonomous way despite the lack of choice (de

Treville et al., 2005).10

Again, however, as leanness becomes excessive,

workers no longer have the time to participate, and

management no longer has the time to communicate

with workers or give the training required for effective

power sharing.

and self-discipline (Schneewind, 1998). A later philosopher stated,

‘‘The fundamental attribute of moral agency is autonomy—that is,

self-legislation or the capacity of a will to give laws to itself’’

(Dodson, 1997: 94). Following these concepts, we suggest that the

construct of responsible autonomy is more consistent with philo-

sophical views of autonomy than is the JCM conceptualization of

autonomy.

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123 111

Proposition 6. Lean production implementation is

negatively related to choice.

Proposition 7. Lean production implementation may

be positively related to responsible autonomy in situa-

tions where workers actively participate in developing

procedures, and responsibility and decision-making

authority are transferred from higher levels of the

organization.

Proposition 8. The relationship between lean produc-

tion implementation and responsible autonomy is

moderated by excessive leanness, such that as leanness

becomes excessive, the relationship between lean

production implementation and responsible autonomy

goes from positive to negative.

11 The multiplicative MPS correlated .08 with performance. This

result, however, is meaningless from a statistical perspective

because of the failure to partial out the main and lower-level

interaction effects from the multiplicative term in the regression

equation. The additive index of the five job characteristics correlated

.27 with performance; however, this result must be interpreted with

caution because adding independent dimensions into an index biases

the correlation with other variables, especially if the measures are

not highly intercorrelated or measure different constructs, as is the

case with the JCM (Paunonen and Gardner, 1991; Perloff and

Persons, 1988).

4.1.2. The critical psychological states: adding

experienced self-efficacy

In reconsidering the set of job characteristics

appropriate for predicting worker performance, we are

following the suggestion of Taber and Taylor (1990:

492), who stated, ‘‘Even though the [Job Diagnostic

Survey, used to instrumentalize the JCM theory,] has

been used in hundreds of studies, it is not too late to

ask the fundamental question: ‘Why these job

characteristics?’’’ Do the three critical psychological

states proposed by Hackman and Oldham (1975, 1976,

1980) account for the job design-related intrinsic

motivation in general work contexts (and in the lean

production context in particular)? If yes, then the

critical psychological states should be at least

moderately or strongly related to motivational out-

comes, especially work performance. The empirical

results, though, have been disappointing.

The most extensive meta-analysis of the JCM,

conducted by Fried and Ferris (1987), generally

supported the directions of relations suggested by the

theory (see also Loher et al., 1985), as well as the

mediatory hypothesis regarding the critical psycho-

logical states (see also Renn and Vandenberg, 1995).

However, Fried and Ferris’s results indicated no

relationship between job characteristics and perfor-

mance. Specifically, the meta-analytic correlations

(90% credibility values) between the five job

characteristics and rated (objective) performance that

Fried and Ferris reported ranged from .00 (task

significance and autonomy) to .13 (task identity). The

analogous correlations between the three critical

psychological states and performance were .03

(knowledge of results) and .00 (experienced mean-

ingfulness and responsibility).11

Along similar lines, Kelly (1992) reviewed 31

studies and found that job redesign did lead to higher

worker satisfaction, but not to increased performance,

leading Kelly to suggest a ‘‘twin-track’’ model of job

redesign incorporating elements related to goal setting

and work methods improvement, given that the

determinants of worker performance and worker

satisfaction might not be the same.

Therefore, in order to be useful in the context of

analyzing lean production, the critical psychological

states (and core job characteristics) should be

expanded to include affective dispositions that result

in increased performance, such as experienced

competence or self-efficacy, which Bandura (1986:

391) defined as ‘‘people’s judgments of their

capabilities to organize and execute courses of action

required to attain designated types of performances . . .concerned not with the skills one has but with

judgments of what one can do with whatever skills one

possesses.’’ Apart from being an important determi-

nant of motivation (Bandura, 1977), self-efficacy,

whether task-specific or generalized, is strongly linked

to job performance and job satisfaction (Judge and

Bono, 2001; Stajkovic and Luthans, 1998), and it may

shield individuals from physical and psychological

job-related strain (Jex and Bliese, 1999).

More generally, building on theorizing by Thomas

and Velthouse (1990), Renn and Vandenberg (1995)

suggested that worker feelings of competence (i.e.,

self-efficacy) should be included in the JCM as a

mediatory critical psychological state (see also

Conger and Kanungo, 1988). Spreitzer (1995) model,

which was an extension of Thomas and Velthouse’s

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123112

model, also included competence as a mediatory

variable.

More important, task-related self-efficacy correlates

more strongly with performance (i.e., .53) in situations

where task complexity is low – reminiscent of lean

production settings – thanwhen task complexity is high

(i.e., .24, Stajkovic and Luthans, 1998). In fact,

Stajkovic and Luthans specifically mentioned that

task-related self-efficacy of workers could be improved

by providing accurate descriptions of work tasks and

training in using the appropriate means in executing the

tasks (e.g., use of standard operating procedures) to

render complex processes simpler. Again, such

suggestions are highly concordant with what is done

in lean production settings (seeGist andMitchell, 1992,

for further discussion on how organizational-level

factors could be linked to individual-level efficacy).

For these reasons, and given that the lean production

philosophy has as a main tenet that job design increases

the competence level of the worker (Boyer, 1996;

MacDuffie, 1995a), we included task-related self-

efficacy as a fourth critical psychological state.

Proposition 9. The relationship between worker per-

ceptions of job characteristics under lean production

and worker outcomes is mediated by the critical

psychological state of task-related self-efficacy.

4.1.2.1. Work facilitation as a core job dimen-

sion. The concept of work facilitation – originally

discussed in the leadership literature – refers to actions

centered on removing obstacles that inhibit worker

performance and on the provision of resources that are

instrumental for the achievement of worker goals (see

Antonakis and House, 2002; Bowers and Seashore,

1966). This job dimension is theoretically linked to

self-efficacy (e.g., Conger and Kanungo, 1988) and is

a foundational practice of lean production.

One of the primary differences between lean and

traditional assembly line production is the expectation

that workers should have the resources, equipment, and

training required to do their job well (Monden, 1983;

Womack et al., 1990). Practices that reduce variability –

such as designing the product for manufacturability,

supplier management, leveling the production sche-

dule, and designing equipment to reduce the chance of

worker error – also serve to make the worker’s job

easier, and his or her performance higher.

Standard operating procedures – which reduce

worker choice by definition – also result in work

facilitation. Imai (1986: xxiv), one of the main lean

production authors focusing on continuous improve-

ment or kaizen, definedSOPs as ‘‘a set of policies, rules,

directives, and procedures. . . . for all major operations,

which serves as guidelines that enable all employees to

perform their jobs successfully.’’ All major lean

production authors (Adler, 1993a, b; Monden, 1983;

Suzaki, 1987; Womack et al., 1990) have insisted that

SOPs are the primary tool that permits workers to take

on true responsibility. Adler (1993a; see also Adler and

Borys, 1996) suggested that access to accurate SOPs

increased worker self-efficacy beliefs.

Furthermore, there is a relationship between

standardization – requiring that employees perform

their tasks according to SOPs – and participation.

SOPs are the mechanism by which employee ideas are

transformed into company practice, substantially

increasing worker perceptions of participation (de

Treville et al., 2005; Edelson and Bennett, 1998).

It has been observed in practice, however, that

workers’ sense of having the resources needed to

perform their tasks effectively decreases as leanness

becomes excessive. There is no longer time to invest in

SOP development, and some workers even stop

following SOPs as they search for (often inappropri-

ate) shortcuts to complete their task more quickly

(Barker, 1993; Graham, 1995; Rinehart et al., 1997).

More important, excessive leanness results in an

substantial increase in worker injuries (Adler et al.,

1997; Babson, 1993; Fucini and Fucini, 1990;

Graham, 1995; Kamata, 1982; Parker and Slaughter,

1995; Rinehart et al., 1997): workers who know that

they are likely to become injured on the job – or who

feel pressured to work although injured – are unlikely

to feel that they have the resources necessary to

perform their tasks well.

Proposition 10. Lean production implementation is

positively related to work facilitation.

Proposition 11. The relationship between lean pro-

duction implementation and work facilitation is mod-

erated by excessive leanness, such that the relationship

between lean production implementation and work

facilitation changes from positive to negative as lean-

ness becomes excessive.

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123 113

Before considering how the lean production job

characteristics interact to affect the critical psycho-

logical states that make up intrinsic motivation, we

must address how the lean production context affects

the choice of moderators.

4.1.3. Moderators

4.1.3.1. Growth need strength (GNS). The JCM

theory suggests that enriching jobs for low-GNS

workers might lead to lower values for the critical

psychological states and outcome measures. Unfortu-

nately, the majority of studies testing this moderating

hypothesis have not used the correct statistical proce-

dures, having failed to test a full moderated regression

model (Evans, 1991). Also, it is well established that

dichotomizing groups based on interval-level GNS

scores (as has often been done) is incorrect because it

results in a loss of information and statistical power

(Aiken and West, 1991; Cohen and Cohen, 1983).

Results generally indicate that GNS does not

moderate the theory. Tiegs et al. (1992) did not find

GNS to moderate the hypothesized relations in the

theory, even through they had a very large sample

(n = 6405) and thus enough power to detect an

interaction effect. Rentsch and Steel (1998) – who

conceptualized GNS in terms of competence and need

for achievement – did not find a moderating effect.

Ganster (1980) – who also conceptualized GNS in

terms of need for achievement – did find a moderating

effect; however, it was opposite to that hypothesized

by the theory (i.e., individuals low in need for

achievement working in high-scope conditions were

more satisfied than individuals with high need for

achievement working in high-scope conditions).

The results of meta-analyses claiming to support the

moderating role of GNS are highly questionable. As

shown by Kanetkar et al. (1995), when integrating

independent studies that have tested the apparent

moderation effects of the GNS, correlation coefficients

cannot be synthesized because they confound the form

of the moderating relation. Unstandardized regression

coefficients should be used instead. Thus, the results of

meta-analytic studies that found support for the GNS

moderator hypothesis cannot be interpreted given their

use of correlation coefficients (e.g., Fried and Ferris,

1987; Loher et al., 1985; Spector, 1985).

Moving to the lean production context, is the role of

GNS as moderator expected to increase or decrease?

Lean production can be viewed as a ‘‘strong’’

situation; that is, there are rigid behavioral and

performance norms and scripts (i.e., extrinsic motiva-

tional levers) that may inhibit the predictive power of

individual differences such as GNS (see Bouffard and

Price, 1974; Davis-Blake and Pfeffer, 1989; Kenrick

and Funder, 1988; Mischel, 1977; Rousseau and Fried,

2001). Furthermore, from a statistical perspective,

uniform (i.e., strong) contexts – like those constituting

lean production – would result in range restriction in

terms of the variation that can be observed in a variety

of variables, including individual differences (see

Schneider, 1987).

The situational strength of lean production is

increased by the employee selection process, which

can be expected to eliminate applicants with

excessively low or high GNS (see Graham, 1995

for an extensive description of typical lean production

selection and hiring procedures), and by the fact that it

is common for firms to invest heavily in how workers

interpret lean production (Adler, 1993a; Fucini and

Fucini, 1990; Graham, 1995; Parker and Slaughter,

1995; Rinehart et al., 1997).

Therefore, given the lack of reliable empirical

evidence to support the moderating effect of GNS

combined with the situational strength of lean

production, we propose:

Proposition 12. The relationship between worker

perceptions of core job characteristics and the critical

psychological states is not moderated by GNS.

4.1.3.2. Excessive leanness. The objective of lean

production is to improve flow while continuously

reducing the resources required for production.

Production areas frequently are understaffed. It is

typical in lean production to equip workers, yet make

slightly fewer resources available than are required, to

encourage creativity and problem-solving behavior

(Monden, 1983; Womack et al., 1990). Lean produc-

tion proponents claim that limiting resources creates

challenging goals, which increases worker motivation.

This is consistent with a human relations view of

motivation, suggesting that staffing should be slightly

less than the amount required so that employees are

exposed to opportunities to utilize not only their

motoric, but also their cognitive and conative, skills

(Argyris, 1964).

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123114

Lean production opponents disagree with the above

reasoning, noting that insufficient resources limit

workers’ capacity to engage in problem-solving

activities, as well as claiming that the resulting

production pressure causes stress, control, and

compliance. As noted, excessive leanness has been

observed to result in increased injuries, resistance in

the form of collective andon pulls or worker activism,

and severe limitations on worker freedom (e.g., Fucini

and Fucini, 1990; Graham, 1995; Rinehart et al.,

1997). At an extreme level, excessive leanness results

in complete worker alienation (e.g., Kamata, 1982).

Behavioral economists would refer to such a

continual reduction of resources as a reduction in

organizational slack (Cyert and March, 1992; Simon,

1976)—slack referring to extra resourcesmaintained in

the organization that support flexibility and creativity.

The existence of organizational slack has an apparently

contradictory effect on motivation and creativity.

Whereas slack is required for creativity, its presence

serves as a relaxing agent in the process of searching for

new ideas, encouraging individuals to reduce the effort

that they put into the search process (Bourgeois, 1981;

Simon, 1976). This contradictory effect is observed in

lean production. The continual reduction of resources

and understaffing causes intensive search behavior, but

the excessive scarcity of resources implies that workers

do not have the time or energy to invest in creative

problem solving (Lawson, 2001).

The concept of double-loop learning (Argyris,

1964) is useful in explaining part of the apparent

contradiction. Lean organizations are forced to be very

efficient in searching for faster and leaner ways to

perform a given task (i.e., single-loop learning), but

they have little capacity for double-loop learning (i.e.,

examining the governing variables of a system). Much

of the learning required of workers under lean

production is expected to be single loop, which is

appropriate for repetitive and simple situations.

However, double-loop learning is required for the

organization to be able to adapt to a changing

environment. Thus, excessive leanness would be

expected to reduce the workers’ ability to truly

participate in the development of the organization over

time, as well as the organization’s adaptive ability.

Lack of organizational slack – in restricting workers to

single-loop learning – might therefore be expected to

limit intrinsic motivation.

Beyond the problem of leanness eventually

restricting double-loop learning, anecdotal evidence

suggests that leanness often becomes excessive under

lean production. That is, so much slack has been

removed from the system that workers are put under

excessive pressure, are no longer able to do their job

correctly, cannot afford to rotate jobs, cannot

participate, and begin to experience a major increase

in injuries (Adler et al., 1997; Landsbergis et al.,

1996). Rinehart et al. (1997) even claimed that lean

production will always result in excessive leanness.

As mentioned previously, lean production theory

emphasizes respect for workers (i.e., in an attempt to

limit worker alienation). The impact of respect for

workers is consistent with theories regarding worker

perceptions of justice and how equity is distributed in

the workplace, and how these perceptions induce

motivation in the workplace (Ambrose, 1999; Cohen-

Charash and Spector, 2001; Colquitt et al., 2001).

In extending the JCM to cover the context of lean

production, therefore, we have proposed that the

relationships between lean production implementation

and skill variety, responsible autonomy, and work

facilitation are moderated by excessive leanness.

Indicators of excessive leanness that can be observed

include increased injury rates, decreased suggestion

rates per employee, decreased job rotation, failure to

follow standard operating procedures, and general

worker discontent. Lean production implementations

where management has slowed the development of

leanness have been observed in practice (e.g.,

Mullarkey et al., 1995), suggesting that metrics

designed to detect excessive leanness should be

tracked regularly.

4.2. Configurations

We would like to explore the idea that lean produc-

tion practices (and, by extension, the job characteristics

and critical psychological states) should be studied

from a configural or holistic perspective. This

perspective argues that manufacturing processes (or

job characteristics) can be synergistic (i.e., mutually

reinforcing). The configural model implies that the unit

of analysis is a set or bundle of practices that cause

motivation in workers and hence organizational perfor-

mance, and that independent main effects account for

less of the variation in the dependent variables—ormay

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123 115

even lead to a completely different prediction (see

Meyer et al., 1993 for a discussion on configural

theory). Boyer et al. (2000) suggested that theory

development in the field of operations management has

been hindered by insufficient configural theorizing.

As stated by Miller and Friesen (1978: 921) ‘‘there

are limitations to looking at one bivariate relationship at

a time and ignoring its context. A more holistic

approach is proposed to get at the most common

complexes of attributes and relationships which

comprise function and deteriorative organizational

and strategy making states.’’ Also, Foti (in press) noted

that taking a variable approach (i.e., main effects/

bivariate or multiplicative/interaction, as opposed to a

holistic/gestalt) (a) ignores how different patterns of

variables might be evident in the unit of observation

(suggesting that combinations of variables might be

more important than levels of individual variables) and

(b) focuses on interactive effects of multiple variables

andnot configurations of variables (i.e., types of units of

analysis). Finally, single variables in isolation might

have less importance on outcomes than when inves-

tigated in the contexts of other variables.

Configurations are theoretically determined based

on certain distinguishing factors (such as those used by

Krafcik, 1988 to contrast lean production with

Fordism and Craft production systems). It is useful

to consider a few examples in which configurations

have been used effectively. Jackson and Mullarkey

(2000) compared various dependent outcomes as a

function of configurational setting (i.e., lean produc-

tion versus traditional manufacturing). Devaraj et al.

(2001) grouped plants according to theoretically

derived typologies (i.e., bundles) to test the effect

of the grouping factor on performance measures (see

also Reeves et al., 2003). The configuration argument

implies sampling factories-as-configurations, with the

configuration hypothesis contrasted with the main

effects hypothesis by testing whether configurations

predict variation in dependent outcomes better than do

independent main effects.

Use of a configural approach has both cognitive/

motivational and statistical implications, as is dis-

cussed below.

4.2.1. Cognitive/motivational implications

The literature involving social and nonsocial

cognition strongly suggests that individuals store,

process, and recall information in a configural or

schematic form (Cantor and Mischel, 1977; Fiske,

1995; for an extensive review see Fiske and Taylor,

1991). That is, individuals categorize social or

nonsocial objects based on groupings of attributes,

which guide information processing (see also Meyer

et al., 1993). Each part that constitutes the whole is

given meaning in relation to the other parts; each part

is not evaluated in isolation from the whole (Meyer

et al., 1993).

A very simple example to demonstrate the holistic

or gestalt perspective of cognition is found in the early

work by Asch (1946). In this classic study, Asch found

that a trait description of an individual (e.g.,

intelligence) was interpreted differently depending

on which other traits were used to describe the

individual (see also Fiske, 1995). For example,

imagine an individual who is intelligent, skillful,

industrious, cold, determined, practical, and cautious.

One would naturally assume this person to be shrewd

and ruthless. Now, imagine the following individual:

intelligent, skillful, industrious, warm, determined,

practical, and cautious. The meaning of intelligence

changes substantially when replacing the word ‘‘cold’’

with ‘‘warm.’’ The person is now wise and humane. In

an algebraic or additive model, each descriptor is

weighted according to how it is independently

evaluated by the perceiver. Using an additive approach

to information processing, the trait ‘‘intelligence’’ has

the same value in both lists of traits, leading to a rather

simplistic and unrealistic linear account of human

information processing.

Similarly, we suggest that the way in which

workers interpret the value of job characteristics will

depend on the configuration of practices they perceive.

That is, workers do not determine whether they like

their job or are motivated by simply summing their

isolated evaluations of individual practices. Rather,

workers make a complex and holistic evaluation by

giving each job characteristic meaning from the other

practices with which it occurs. For example, having

low choice autonomy might not be seen as demotivat-

ing to workers, especially when they have high

responsible autonomy, feel efficacious, use a variety of

skills, identify with their task, and receive appropriate

feedback.

An example of the practical importance of

configural theorizing is evident in the work of

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123116

Stajkovic and Luthans (2003), who argued and found

that various motivational dimensions (i.e., use of

monetary incentives, feedback, and social recogni-

tion) induced more potent motivational effects when

used in combination. They coded experiments

according to the type of motivational approach used

(i.e., the unit of analysis was the approach or

approaches used, e.g., each of the three approaches

used in isolation, three, two-way combinations and

one, three-way combination)—thus testing the main

and configural effects (i.e., a specific interaction).

Using meta-analytic methodology, Stajkovic and

Luthans found that the combinatorial model of the

three motivation practices together – that is, the

configural effect of high levels of each of the three

practices occurring simultaneously – produced sig-

nificantly larger effects than any one of the approaches

by itself.

Supporting the ‘‘configuration-in-context’’ argu-

ment, MacDuffie (1995a) and Adler (1993a,b; see also

Adler and Borys, 1996) have suggested that choice

autonomy could undermine the positive effects of

factors such as responsibility and feedback, thereby

reducing intrinsic motivation and commitment even

though the independent main effect of choice on

intrinsic motivation is generally expected to be

positive. That is, the positive impact of a high

feedback-responsibility – low choice configuration on

intrinsic motivation might be greater than a low

feedback-responsibility – high choice configuration

(Adler, 1991; Barker, 1993; Graham, 1995).

4.2.2. Statistical implications

From a statistical perspective, if the interactive

(i.e., moderation) effects of variables working

synergistically cannot be predicted linearly, the form

of the regression function might be difficult to model

and detect (Jaccard et al., 1990). In these situations,

theory should guide the way in which the interaction

hypothesis is tested. In this case, we refer to an

interaction of the job characteristics as being a specific

combination (i.e., a configuration) of independent

variables that are grouped together following theory.

This specific combination is different from testing

general interaction effects – as in multiplicative effects

in moderated multiple regression analysis—across

various levels of independent variables. Power to

detect moderation/interaction effects is also reduced in

this case because of the joint residual variance

distribution, which is dramatically reduced as com-

pared to a traditional experimental grouping procedure

as demonstrated, for example, by simulation studies

by McClelland and Judd (1993). In other words, in

moderated multiple regression, the number of groups

analyzed equals the number of subjects (Sharma et al.,

1981), which thus does not test a specific type of

moderation effect, but instead a general moderation

effect across subjects.

One way to test whether configurational models

differentially predict outcomes is to divide observa-

tions into what could be termed causally homogenous

groups to test whether parameter estimates or

structural relations vary among groups (Antonakis

et al., 2003; Baron and Kenny, 1986; James et al.,

1982; McKelvey, 1978; Mulaik and James, 1995). The

statistical tests employed could include subgroup

analysis using ANOVA models (Sharma et al., 1981),

or more complicated procedures using structural

equation modeling, to test how structural (and

measurement) relations vary among groups (see

Joreskog, 1971)—that is, whether the means of

dependent variables or causal paths between variables

vary significantly as a function of the grouping factor

or configuration.

Proposition 13. The variance predicted in worker

outcomes will be significantly larger when using the

ideal configural grouping model than when using an

additive or full multiplicative model.

What groups emerge in the lean production context?

Wehave proposed that lean production always results in

feedback and task identity. High values of these job

characteristics, with all other job characteristics rated

low, would form a minimum case of lean production.

Lean production at its best would consist of high values

of all job characteristics except for choice. An

intermediate configuration has been observed, in which

workers have high levels of skill variety, task identity,

feedback, andwork facilitation, but have low values for

responsible autonomy and choice. These three config-

urations are listed in Fig. 1. Obviously, the different

configurations would result in quite different gestalts

and, hence, different levels of intrinsicmotivation. Lean

production implementations have beenobserved to start

off with an ideal configuration, only to deteriorate as

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123 117

leanness increases (e.g., Adler et al., 1997; Fucini and

Fucini, 1990; Graham, 1995; Rinehart et al., 1997).

Proposition 14. A lean production implementation

resulting in an ideal configuration of job character-

istics will be positively related to worker intrinsic

motivation.

Proposition 15. A lean production implementation

that results in a poor, or non-ideal configuration of job

characteristics will be negatively related to worker

intrinsic motivation.

Proposition 16. A lean production implementation

that results in a configuration of job characteristics that

is less than the ideal but better than the non-ideal case

will have a small or null relationship with worker

intrinsic motivation.

4.3. Levels of analysis

Implicit in the previous discussion is that organiza-

tion-level factors (i.e., lean production practices)

affect individual- or group-level factors (e.g., worker

perceptions of job characteristics and intrinsic

motivation of workers) and, by implication, other

organizational-level factors (e.g., profit) in a largely

homogenous cross-level manner. In this section, we

investigate the consequences and implications of

assuming cross-level effects.

Cross-level effects, in which a phenomenon at a

higher level of analysis (e.g., organizational level)

influences phenomena at lower levels of analysis (e.g.,

individual level) can be thought of as contextual (Johns,

2001) or configural effects (Meyer et al., 1993).

Irrespective of how they are conceptualized, cross-

level relationships act as boundary conditions to

theories, specifying the domains in which the predic-

tions of the theory hold. For example, if a motivation

model predicts that variation in organization-level

characteristics affects variation in person-level vari-

ables, an implicit – but often untested – claim that is

made is the existence of a cross-level effect that impacts

workers in a largely similar way (see Dansereau et al.,

1984; Klein et al., 1994; Rousseau and Fried, 2001).

Levels effects are often implicit in theories, but they are

rarely tested, and work design theories are not an

exception (Mossholder and Bedeian, 1983).

Assumptions regarding levels can be tested with

appropriate statistical methods (e.g., within-and-

between analysis or WABA, see Dansereau et al.,

1984; or hierarchical linear modeling or HLM, see

Hoffman, 1997) that seek to determine whether a

higher-level factor (e.g., a lean production configura-

tion) homogenously impacts lower-level factors (e.g.,

critical psychological states or motivation). If the

researcher has sampled various production configura-

tions (i.e., factories grouped according to the

configuration they display in their practices) and

personal level variables nested in those groupings,

then it is easy to test whether:

1. T

here is statistically meaningful between-group

variation in outcome variables (i.e., that motivation

in the different groupings exhibits differential

variation);

2. T

he variation in outcome variables within groups is

statistically homogenous;

3. G

roup-level effects in predictions (i.e., the relation

between critical psychological states and motiva-

tion) operate at the group level.

If the above conditions are satisfied, then one can

make the claim that a meaningful cross-level effect

may be prevalent, suggesting that changes in practices

do have a meaningful effect on outcome variables at

the group – or, when concerning profits, at the factory

– level of analysis (Mossholder and Bedeian, 1983).

In the only published example that we could locate

that tested the levels effects of the JCM, Coleman

(1996) found that the theory operated at the individual

level of analysis (in white-collar type jobs, which

would be more associated with a weak situation than

that of a factory setting). This result suggests that

individual perceptions may be idiosyncratic and

depend on how workers interpret environmental

events. In such cases, the impact of a group-level

intervention – such as job design – would not be

expected to affect worker intrinsic motivation homo-

genously unless management also intervenes to

influence worker interpretive systems.

Thus, individuals’ interpretive systems – and

whether they are idiosyncratic or homogenous – play

an important role in transforming data from interven-

tions and external events into perceptions that affect

motivational states. Interpretive systems refer to

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123118

causal attributions and inferences made by workers,

howworkers evaluate their performance, and how they

envision future events (see Thomas and Velthouse,

1990). This perspective complements the literature on

social cognition regarding the effect of expectations

on information processing (see Fiske and Taylor,

1991; Gilbert andMalone, 1995), as well as that on the

motivational effects of leaders on followers (Antona-

kis and Atwater, 2002; Shamir et al., 1993). In the lean

production case, a cross-level effect may be more

likely in situations where management invests in how

workers interpret key elements of lean production.

Along similar lines, the level of social identification

of employees with the company, plant, or work team

should be considered in evaluating levels effects. Social

identification – which plays a key role in lean

production (e.g., Graham, 1995; Kenney and Florida,

1993; MacDuffie, 1995b) – would be expected to

increase support for and commitment to a group, and it

might encourage ‘‘internalization of, and adherence to,

group values and norms’’ (Ashforth and Mael, 1989:

26), again making a cross-level effect more likely.12

Proposition 17. Implementation of lean production is

more likely to exhibit a cross-level effect on workers

in contexts where management invests in worker

interpretive systems, and where worker social identi-

fication with the company, plant, or work team is high.

5. Summary and conclusions

Is lean production job design theoretically able to

cause intrinsic motivation, such that workers increase

their effort because of the task itself, or does the

worker behavior observed under lean production

result only from increased management control? Must

managers hoping for worker commitment avoid lean

production because of its standardization and inter-

dependence, or are there paths to intrinsic work

motivation within the lean production context? More

than 25 years after the arrival of lean production, these

questions have remained unanswered.

The philosophy behind the JCM is a reasonable

starting point for such an investigation, given the

12 Lean producers have been observed to use celebrations and

rituals to promote a sense of social identity and loyalty, with varying

degrees of success (Graham, 1995).

assumption that much of the impact of lean production

implementation on motivation occurs through changes

in job design. Extending the JCM to the lean

production context, we have been able to link the

factory physics of lean production to job character-

istics perceived by lean production workers. The lean

production context suggests division of autonomy into

choice and responsible autonomy, eliminating a

primary conflict between the JCM and lean production

theory. Given the emphasis in lean production theory

on training and equipping workers, we added the job

characteristic of work facilitation. As lean production

factory physics do not change the significance of the

task, we eliminated the job characteristic of task

significance from our extended model.

The JCM was developed to provide a theoretical

structure to explore the relationship between job

enrichment and intrinsic motivation, which was then

expected to result in various work-related outcomes.

Our objective in the lean production context has been

different: Job design arising from lean production’s

factory physics results in work performance, appar-

ently mediated by intrinsic motivation in some cases.

Given the tenuous link between the JCM critical

psychological states and performance, we added

experienced self-efficacy as a fourth critical psycho-

logical state both because it is a dimension of intrinsic

motivation arising from job design and because it is

positively related to performance.

We proposed that the lean production job character-

istics operate configurally to cause worker intrinsic

motivation. The lean production literature suggested

three natural configurations: (a) feedback and task

identity (occurring in all lean production implementa-

tions); (b) feedback, task identity, skill variety, and

work facilitation; (c) feedback, task identity, skill

variety, work facilitation, and responsible autonomy.

The job characteristic of choice is low in all lean

production implementations. The lean production

configuration emerging is moderated by excessive

leanness, with skill variety, work facilitation, and

responsible autonomy unlikely to be present in

conditions where leanness is excessive. The addition

of responsible autonomy, in particular, substantially

increases the intrinsic motivation emerging from a lean

production implementation. The different configura-

tions explain someof thevariancebetween the idealistic

worker outcomes observed by authors such as Adler

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123 119

(1993a) and the bleak despair described by Kamata

(1982).

Lean production represents a group-level interven-

tion that is proposed to impact individual workers

homogenously. This cross-level effect needs to be

tested, and may be more likely to hold in strong

situations where management invests in molding

worker perceptions of lean production, and where

social identity is strong. In some lean production

implementations, weaker situational strength may

result in worker reactions that are idiosyncratic rather

than homogenous. In such cases, group-level inter-

vention will not result in intrinsic motivation. Thus,

the roles of leaders might become more salient by

framing events and motivating workers on an

individual level in a customized way (Antonakis

et al., 2004).

It has not been our intention to argue that lean

production turns assembly line work into a great job.

Our goal is more pragmatic. After decades of job

enrichment, assembly line work remains common,

especially in competitive markets, and assembly line

work is the only option available for many workers.

Lean production appears to have made such work

more tolerable – even motivating – for many assembly

line workers (Adler, 1993a,b). Under what circum-

stances would this be true? If choice autonomy is not

an option, what are the alternatives? More counter-

intuitively, could authors like Adler (e.g., 1996) and

MacDuffie (1995a) be correct in suggesting that

workers actually are able to embrace lack of choice

and still be intrinsically motivated?

The extensions to the JCM required to cover the

lean production context have implications for work

motivation theory development in the field of

operations management in general (both manufactur-

ing and services). The long-held (and probably

untenable) assumption concerning the criticality of

choice autonomy to work motivation has resulted in

neglect of factors such as work facilitation and self-

efficacy belief. Just as leadership research is currently

exploring how instrumental leadership (i.e., engender-

ing follower work facilitation) can result in perfor-

mance outcomes that are competitive with

transformational leadership (based on leader char-

isma, Antonakis and House, 2004), restoring work

facilitation to its right place might usher in a new era

of work motivation theory in operations management.

The motivational effects of ‘‘ideal’’ and ‘‘too’’ lean

implementations are, theoretically, different. Recog-

nizing the two-edged nature of leanness can help to

avoid the motivational – and ethical – pitfalls inherent

in the search for greater efficiency. We hope with this

model to have made a first step toward their resolution.

Acknowledgements

The authors would like to gratefully acknowledge

the helpful comments of Paul Adler, Michael

Cusumano, Jan Klein, Ulrich Jurgens, Cathie Ramus,

and Francis Yammarino, as well as the anonymous

reviewers, on earlier versions of this manuscript. We

take full responsibility for all remaining errors and

omissions.

References

Adler, P.S., 1991. Workers and flexible manufacturing systems:

three installations compared. Journal of Organizational Beha-

vior 12, 447–460.

Adler, P.S., 1993a. The new ‘‘learning bureaucracy’’: New United

Motor Manufacturing, Inc. In: Staw, B.M., Cummings, L.L.

(Eds.), Research in Organizational Behavior, 15. JAI Press,

Greenwich, CT, pp. 111–194.

Adler, P.S., 1993b. Time-and-motion regained. Harvard Business

Review 97–108.

Adler, P.S., Borys, B., 1996. Two types of bureaucracy: enabling and

coercive. Administrative Science Quarterly 41, 61–89.

Adler, P.S., Cole, R.E., 1993. Designed for learning: a tale of two

auto plants. Sloan Management Review, Spring 85–94.

Adler, P.S., Goldoftas, B., Levine, D.I., 1997. Ergonomics,

employee involvement, and the Toyota production system: a

case study of NUMMI’s 1993 model introduction. Industrial and

Labor Relations Review 50 (3), 416–437.

Aiken, L.S., West, S.G., 1991. Multiple Regression: Testing and

Interpreting Interactions. Newbury. Sage Publications, Park,

CA.

Alderfer, C.P., 1967. An organizational syndrome. Administrative

Science Quarterly 12, 440–460.

Alderfer, C.P., 1969. An empirical test of a new theory of human

needs. Organizational Behavior and Human Performance 4,

142–175.

Ambrose, M.L., 1999. Old friends, new faces: Motivation research

in the 1990’s. Yearly Review of Management of the Journal of

Management: 1–74.

Antonakis, J., Atwater, L., 2002. Leader distance: a review and a

proposed theory. Leadership Quarterly 13, 673–704.

Antonakis, J., Avolio, B.J., Sivasubramaniam, N., 2003. Context and

leadership: an examination of the nine-factor full-range leader-

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123120

ship theory using the Multifactor Leadership Questionnaire

(MLQ Form 5X). Leadership Quarterly 14, 261–295.

Antonakis, J., House, R.J., 2002. An analysis of the full-

range leadership theory: The way forward. In: Yammarino,

F.J., Avolio, B.J. (Eds.), Transformational and Charismatic

Leadership: The Road Ahead. JAI Press, Amsterdam, pp.

3–34.

Antonakis, J., House, R.J. 2004. On instrumental leadership: Beyond

transactions and transformations. Paper presented at the Gallup

Leadership Institute Conference, University of Nebraska.

Antonakis, J., Schriesheim, C.A., Donovan, J.A., Gopalakrishna-

Pillai, K., Pellegrini, E., Rossomme, J.L., 2004. Methods for

studying leadership. In: Antonakis, J., Cianciolo, A.T., Stern-

berg, R.J. (Eds.), The Nature of Leadership. Sage, Thousand

Oaks.

Argyris, C., 1964. Integrating the Individual and the Organization.

John Wiley and Sons, New York.

Asch, S.E., 1946. Forming impressions of personality. Journal of

Abnormal and Social Psychology 41, 258–290.

Ashforth, B.E., Mael, F., 1989. Social identity theory and the

organization. Academy of Management Review 14 (1), 20–39.

Babson, S., 1993. Lean or mean: The MIT model and lean produc-

tion at Mazda. Labor Studies Journal 18 (2), 3–25.

Bandura, A., 1977. Social Learning Theory. Prentice-Hall, Engle-

wood Cliffs, NJ.

Bandura, A., 1986. Social Foundations of Thought and Action: A

Social Cognitive Theory. Prentice-Hall, Englewood Cliffs, NJ.

Barker, J.R., 1993. Tightening the iron cage: concertive control in

self-managing teams. Administrative Science Quarterly 38,

408–437.

Baron, R.M., Kenny, D.A., 1986. The moderator-mediator variable

distinction in social psychological research: conceptual, strate-

gic, and statistical considerations. Journal of Personality and

Social Psychology 51, 1173–1182.

Berggren, C., 1992. Alternatives to lean production: work organiza-

tion in the Swedish auto industry. ILR Press, Ithaca, NY.

Blair, J.D., Hunt, J.G., 1986. Getting inside the head of the manage-

ment researcher one more time: Context-free and context-spe-

cific orientations in research. Yearly Review of Management of

the Journal of Management 12 (2), 147–166.

Bouffard, D.L., Price, R.H., 1974. Behavioral appropriateness and

situational constraints as dimensions of social behavior. Journal

of Personality and Social Psychology 30, 579–586.

Bourgeois, L.J., 1981. On the measurement of organizational slack.

Academy of Management Review 6 (1), 29–39.

Bowers, D.G., Seashore, S.E., 1966. Predicting organizational

effectiveness with a four-factor theory of leadership. Adminis-

trative Science Quarterly 11, 238–263.

Boyer, K.K., 1996. An assessment of managerial commitment to

lean production. International Journal of Operations and Pro-

duction Management 16 (9), 48–59.

Boyer, K.K., Bozarth, C., Mc Dermott, C.M., 2000. Configurations

in operations: an emerging area of study. Journal of Operations

Management 18, 601–604.

Brief, A.P., Aldag, R.J., 1977. The intrinsic-extrinsic dichotomy:

toward conceptual clarity. Academy of Management Review 2,

496–500.

Brown, K.A., Mitchell, T.R., 1991. A comparison of just-in-time and

batch manufacturing: the role of performance obstacles. Acad-

emy of Management Journal 34 (4), 906–917.

Cantor, N., Mischel, W., 1977. Traits as prototypes: Effects on

recognition memory. Journal of Personality and Social Psychol-

ogy 35, 38–48.

Cialdini, R.B., Goldstein, N.J., 2004. Social influence: Compli-

ance and conformity. Annual Review of Psychology 55, 591–

621.

Cohen, J., Cohen, P., 1983. Applied Multiple Regression/Correla-

tion Analysis for the Behavioral Sciences.. Hillsdale, NJ, Lawr-

ence Erlbaum Associates.

Cohen-Charash, Y., Spector, P.E., 2001. The role of justice in

organizations: a meta-analysis. Organizational Behavior and

Human Decision Processes 86 (2), 278–321.

Coleman, D.F., 1996.In: Job characteristics and work alienation at

multiple levels of analysis, Paper presented at the Proceedings

of the Administrative Sciences Association of Canada (Organi-

zational Behavior Division).

Collins, R.S., Schmenner, R.W. 2003. Explaining variable plant

performance. Paper presented at the EUROMA/POMS Joint

International Conferece, Como.

Colquitt, J.A., Conlon, D.E., Wesson, M.J., Porter, C.O.L.H., Ng,

K.Y., 2001. Justice at the millenium: a meta-analytic review of

25 years of organizational justice research. Journal of Applied

Psychology 86 (3), 425–445.

Conger, J.A., Kanungo, R.N., 1988. The empowerment process:

integrating theory and practice. Academy of Management

Review 13, 471–481.

Cyert, R.M., March, J.G., 1992. A Behavioral Theory of the Firm,

2nd ed. Blackwell, Oxford.

Dankbaar, B., 1997. Lean production: Denial, confirmation or

extension of sociotechnical systems design? Human Relations

50 (5), 567–583.

Dansereau, F., Alutto, J.A., Yamarino, F.J., 1984. Theory Testing in

Organizational Behavior: The Varient Approach. Prentice-Hall,

Englewood Cliffs, NJ.

Davis-Blake, A., Pfeffer, J., 1989. Just a mirage: the search for

dispositional effects in organizational research. Academy of

Management Review 14, 385–400.

Davy, J.A., Gritzmacher, K., Merritt, N.J., White, R.E., 1992. A

derivation of the underlying constructs of just-in-time man-

agement systems. Academy of Management Journal 5 (3),

653–670.

de Treville, S. 1987. Disruption, Learning, and System Improvement

in Just-in-Time Manufacturing: dissertation Abstracts Interna-

tional, (University Microfilms No. 8717014).

de Treville, S., Antonakis, J., Edelson, N.M. 2005. Can standard

operating procedures be motivating? Reconciling process varia-

bility issues and behavioral outcomes. Total Quality Manage-

ment and Business Processes 16 (2), 231–241.

Devaraj, S., Hollingworth, D.G., Schroeder, R.G., 2001. Generic

manufacturing strategies: an empirical test of two configura-

tional typologies. Journal of Operations Management 19, 427–

452.

Dodson, K.E., 1997. Autonomy and authority in Kant’s Rechtslehre.

Political Theory 25 (1), 93–111.

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123 121

Dubin, R., 1976. Theory building in applied areas. In: Dunnette,

M.D. (Ed.), Handbook of Industrial and Organizational Psychol-

ogy. Rand McNally, Chicago, pp. 17–40.

Dyer, J.H., 1996. Does governance matter? Keiretsu alliances and

asset specificity as sources of Japanese competitive advantage.

Organization Science 7 (6), 649–666.

Edelson, N.M., Bennett, C.L., 1998. Process Discipline: How to

Maximize Profitability and Quality Through Manufacturing

Consistency. Quality Resources, New York.

Evans, M.G., 1991. The problem of analyzing multiplicative com-

posites. American Psychologist 46, 6–15.

Ferris, G.R., Gilmore, D.C., 1985. A methodological note on job

complexity indexes. Journal of Applied Psychology 70, 225–

227.

Fiske, S.P., 1995. Social cognition. In: Tesser, A. (Ed.), Advanced

Social Psychology. Boston, McGraw-Hill, pp. 149–193.

Fiske, S.T., Taylor, S.E., 1991. Social Cognition, 2nd ed. New York,

McGraw-Hill.

Flynn, B.B., Flynn, J.E., Schroeder, R.G., 1999. World class man-

ufacturing: an investigation of Hayes and Wheelwright’s foun-

dation. Journal of Operations Management 249–269.

Flynn, B.B., Sakakibara, S., Schroeder, R.G., 1995. Relationship

between JIT and TQM: practices and performance. Academy of

Management Journal 38 (5), 1325–1360.

Forza, C., Choo, A.S., 1996. Work organization in lean production

and traditional plants: What are the differences? International

Journal of Operations and Production Management 16 (2), 42–

62.

Foti, R.J., in press. Patterns and variables: Evidence from leadership

emergence and effectiveness. Journal of Applied Psychology.

Fried, Y., Ferris, G.R., 1987. The validity of the job characteristics

model: a review and meta-analysis. Personnel Psychology 40,

214–249.

Friedman, A., 1977. Responsible autonomy versus direct control

over the labour process. Capital and Class 1, 43–57.

Fucini, J.J., Fucini, S., 1990. Working for the Japanese. The Free

Press, New York.

Fujimoto, T., 1999. The Evolution of a Manufacturing System at

Toyota. New York, Oxford University Press.

Ganster, D.C., 1980. Individual differences and task design: a

laboratory experiment. Organizational Behavior and Human

Decision Processes 26, 131–148.

George, J.M., Jones, G.R., 1997. Organizational spontaneity in

context. Human Performance 10, 153–170.

Gilbert, D.T., Malone, P.S., 1995. The correspondence bias. Psy-

chological Bulletin 117, 21–38.

Gist, M.E., Mitchell, T.R., 1992. Self -efficacy: a theoretical analysis

of its determinants and malleability. Academy of Management

Review 17 (2), 183–211.

Graham, L., 1995. On the Line at Subaru-Isuzu: The JapaneseModel

and the American Worker. ILR Press, Ithaca, NY.

Griffin, R.W., Welsh, A., Moorhead, G., 1981. Perceived task

characteristics and employee performance: a literature review.

Organizational Behavior and HumanDecision Processes 35, 73–

93.

Hackman, J.R., Lawler, E.E., 1971. Employee reactions to job

characteristics. Journal of Applied Psychology 55 (3), 259–286.

Hackman, J.R., Oldham, G.R., 1975. Development of the Job

Diagnostic Survey. Journal of Applied Psychology 60 (2),

159–170.

Hackman, J.R., Oldham, G.R., 1976. Motivation through the design

of work: test of a theory. Organizational Behavior and Human

Performance 16, 250–279.

Hackman, J.R., Oldham, G.R., 1980. Work Redesign. Addison-

Wesley, Reading, MA.

Hackman, J.R., Wageman, R., 1995. Total quality management:

empirical, conceptual, and practical issues. Administrative

Science Quarterly 40 (2), 309–342.

Hall, R.W. 1983. Zero Inventories. Homewood, Il.: Dow Jones-

Irwin.

Handfield, R., 1993. A resource dependence perspective of Just-in-

Time purchasing. Journal of Operations Management 11 (3),

289–311.

Harmon, R.L., Peterson, L.D., 1990. Reinventing the Factory. The

Free Press, New York.

Hayes, R.H., 1981. Why Japanese factories work. Harvard Business

Review 59 (4), 57–65.

Hayes, R.H., Wheelwright, S.C., Clark, K.B., 1988. Dynamic

Manufacturing: Creating the Learning Organization. New York,

Free Press.

Hoffman, D.A., 1997. An overview of the logic and rationale of

hierarchical linear models. Journal of Management 23, 723–744.

Hopp, W.J., Spearman, M.L., 1996. Factory Physics, 1st ed. Irwin,

Chicago.

Imai, M., 1986. Kaizen. Random House, New York.

Jaccard, J., Turrisi, R., Wan, C.K., 1990. Interaction Effects in

Multiple Regression. Sage Publications, Newbury Park, CA.

Jackson, P.R., Mullarkey, S., 2000. Lean production teams and

health in garment manufacture. Journal of Occupational Health

Psychology 5, 231–245.

Jackson, P.R., Wall, T.D., Martin, R., Davids, K., 1993. New

measures of job control, cognitive demand and production

responsibility. Journal of Applied Psychology 78, 753–762.

James, L.R., Mulaik, S.A., Brett, J.M., 1982. Causal Analysis:

Assumptions, Models, and Data. Sage Publications, Beverly

Hills.

Jex, S.M., 2002. Organizational Psychology. Wiley, New York.

Jex, S.M., Bliese, P.D., 1999. Efficacy beliefs as a moderator of the

impact of work-related stressors. Journal of Applied Psychology

84, 349–361.

Johns, G., 2001. In praise of context. Journal of Organizational

Behavior 22, 31–42.

Joreskog, K.G., 1971. Simultaneous factor analysis in several

populations. Psychometrika 36, 409–426.

Judge, T.A., Bono, J.E., 2001. Relationship of core self-evaluation

traits – self-esteem, generalized self-efficacy, locus of control,

and emotional stability – with job satisfaction and job perfor-

mance: a meta-analysis. Journal of Applied Psychology 86, 80–

92.

Jurgens, U., 1986. Kontrollierte Autonomie—arbeitspolitische Stra-

tegien in der internationalen Automobilindustrie. ASI-News

(10).

Kamata, S., 1982. Japan in the Passing Lane: An Insider’s Account

of Life in a Japanese Auto Factory. Pantheon Books, New York.

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123122

Kanetkar, V., Evans, M.G., Everell, S.A., Irving, D., Millman, Z.,

1995. The effect of scale changes on meta-analysis of multi-

plicative and main effects models. Educational and Psycholo-

gical Measurement 55, 206–224.

Kelly, J., 1992. Does job re-design theory explain job re-design

outcomes? Human Relations 45, 753–774.

Kenney, M., Florida, R., 1993. Beyond Mass Production. Oxford

University Press, New York.

Kenrick, D.T., Funder, D.C., 1988. Profiting from controversy:

lessons from the person-situation debate. American Psycholo-

gist 43 (1), 23–34.

Klein, J.A., 1989. The human costs of manufacturing reform.

Harvard Business Review 67 (2), 60–64.

Klein, J.A., 1991. A reexamination of autonomy in light of new

manufacturing practices. Human Relations 44 (1), 21–38.

Klein, K.J., Dansereau, F., Hall, R.J., 1994. Levels issues in theory

development, data collection, and analysis. Academy of Man-

agement Review 19, 195–229.

Krafcik, J.F., 1988. Triumph of the lean production system. Sloan

Management Review 30 (1), 41–52.

Landsbergis, P.A., Cahill, J., Schnall, P., 1996. New systems of work

organization: Impacts on job characteristics and health. Job

Stress Network 1–10.

Langfred, C.W., 2004. Too much of a good thing? Negative effects

of high trust and individual autonomy in self-managing teams.

Academy of Management Journal 47, 385–399.

Lawson, M.B., 2001. In praise of slack: time is of the essence.

Academy of Management Executive 15 (3).

Lewin, K., 1938. The Conceptual Representation of the Measure-

ment of Psychological Forces. Durham, Duke University

Press.

Loher, B.T., Noe, R.A., Moeller, N.L., Fitzgerald, M.P., 1985. A

meta-analysis of the relation of job characteristics to job satis-

faction. Journal of Applied Psychology 70, 280–289.

MacDuffie, J.P., 1995a. Human resource bundles and manufacturing

performance: Organizational logic and flexible production sys-

tems in the world auto industry. Industrial and Labor Relations

Review 48, 197–221.

MacDuffie, J.P., 1995b. Human resource bundles and manufacturing

performance: organizational logic and flexible production sys-

tems in the world auto industry. Industrial and Labor Relations

Review 48 (2), 197–221.

MacDuffie, J.P., 1997. The road to ‘‘root cause’’: Shop-floor pro-

blem solving at three auto assembly plants. Management

Science 43 (4), 479–502.

Maslow, A.H., 1943. A theory of human motivation. Psychological

Review 50, 370–396.

Maslow, A.H., 1954. Motivation and Personality. New York, Harper.

McClelland, G.H., Judd, C.M., 1993. Statistical differences of

detecting interactions and moderator effects. Psychological

Bulletin 114, 376–390.

McKelvey, B., 1978. Organizational systemics: Taxonomic lessons

from biology. Management Science 24, 1428–1440.

McKone, K.E., Cua, K.O., Schroeder, R.G., 2001. The impact

of total productive maintenance practices on manufacturing

performance. Journal of Operations Management 19, 39–

58.

McLachlin, R., 1997. Management initiatives and just-in-time man-

ufacturing. Journal of Operations Management 15, 271–292.

Meyer, A.D., Hinings, C.R., Tsui, A.S., 1993. Configurational

approaches to organizational analysis. Academy ofManagement

Journal 36 (6), 1175–1195.

Milkman, R., 1997. Farewell to the Factory: Auto Workers in the

Late Twentieth Century. Berkeley, University of California

Press.

Miller, D., Friesen, P.H., 1978. Archetypes of strategy formulation.

Management Science 24, 921–933.

Mischel, W., 1977. The interaction of person and situation. In:

Magnusson, D., Endler, D. (Eds.), Personality at the Crossroads:

Current Issues in Interactional Psychology. Lawrence Erlbaum

associates, Publishers, Hillsdale, NJ, pp. 333–352.

Monden, Y., 1983. Toyota Production System. Industrial Engineer-

ing and Management Press, Institute of Industrial Engineers,

Norcross, GA.

Mossholder, K.W., Bedeian, A.G., 1983. Cross-level inference and

organizational research: perspectives on interpretation and

application. Academy of Management Review 8, 547–558.

Mulaik, S.A., James, L.R., 1995. Objectivity and reasoning in science

and structural equationmodeling. In: Hoyle, R.H. (Ed.),Structural

Equation Modeling: Concepts, Issues, and Applications. Sage

Publications, Thousand Oaks, CA, pp. 118–137.

Mullarkey, S., Jackson, P.R., Parker, S.K., 1995. Employee reactions

to JIT manufacturing practices: a two-phase investigation. Inter-

national Journal of Operations and Production Management 15

(11), 62–79.

Parker, M., Slaughter, J., 1995. Unions and management by stress.

In: Babson, S. (Ed.), Lean Work: Empowerment and Exploita-

tion in the Global Auto Industry. Wayne State University Press,

Detroit, pp. 41–53.

Parker, S.K., Wall, T.D., Cordery, J.L., 2001. Future work design

research and practice: towards an elaborated model of work

design. Journal of Occupational and Organizational Psychology

74, 413–440.

Paunonen, S.V., Gardner, R.C., 1991. Biases resulting from the use

of aggregated variables in psychology. Psychological Bulletin

109, 520–523.

Perloff, J.M., Persons, J.B., 1988. Biases resulting from the use of

indexes: an application to attributional style and depression.

Psychological Bulletin 103, 95–104.

Pierce, J.L., Dunham, R.B., 1976. Task design: a literature review.

Academy of Management Review 1, 83–97.

Pinder, C.C., 1984. Work Motivation: Theory, Issues, and Applica-

tions. Scott, Foresman and Company, Glenview, IL.

Porter, M.E., 1996. What is strategy? Harvard Business Review 74

(6), 61–78.

Post, C., Slaughter, J., 2000. Lean Production: Why Work is Worse

than Ever and What’s the Alternative? , http://solidarity.org/

leanproduction.html.

Reeves, T.C., Duncan, W.J., Ginter, P.M., 2003. Strategic config-

urations in health services organizations. Journal of Business

Research 56, 31–43.

Renn, R.W., Vandenberg, R.J., 1995. The critical psychological

states: an underrepresented component in job characteristics

research. Journal of Management 21, 279–303.

S. de Treville, J. Antonakis / Journal of Operations Management 24 (2006) 99–123 123

Rentsch, J.R., Steel, R.P., 1998. Testing the durability of job

characteristics as predictors of absenteeism over a six-year

period. Personnel Psychology 51, 165–190.

Rinehart, J., Huxley, C., Robertson, D., 1997. Just Another Car

Factory? Cornell University Press, Ithaca, NY.

Roberts, K.H., Glick, W., 1981. The job characteristics approach to

task design: a critical review. Journal of Applied Psychology 66

(2), 193–217.

Rousseau, D.M., Fried, Y., 2001. Location, location, location:

contextualizing organizational research. Journal of Operational

Behavior 22, 1–13.

Sakakibara, S., Flynn, B.B., Morris, W.T., Schroeder, R.G., 1997.

The impact of just-in-time manufacturing and its infrastructure

on manufacturing performance. Management Science 43 (9),

1246–1259.

Sayer, A., 1986. New developments in manufacturing: the just-in-

time system. Capital and Class 10, 43–71.

Schneewind, J.B., 1998. The Invention of Autonomy: A History of

Modern Moral Philosophy. Cambridge University Press, Cam-

bridge.

Schneider, B., 1987. The people make the place. Personnel Psy-

chology 40, 437–453.

Schonberger, R.J., 1982. Japanese Manufacturing Techniques: Nine

Hidden Lessons in Simplicity. The Free Press, New York.

Schultz, K.L., Juran, D.C., Boudreau, J.W., 1999. The effects of low

inventory on the development of productivity norms. Manage-

ment Science 45 (12), 1664–1678.

Seibert, S.E., Silver, S.R., Randolph, W.A., 2004. Taking empower-

ment to the next level: a multiple-level model of empowerment,

performance, and satisfaction. Academy ofManagement Journal

47, 332–349.

Sewell, G., Wilkinson, B., 1992. Someone to watch over me:

Surveillance, discipline and the just-in-time process. Sociology

26 (2), 271–289.

Shah, R., Ward, P.T., 2003. Lean manufacturing: context, practice

bundles, and performance. Journal of Operations Management

21, 129–149.

Shamir, B., House, R.J., Arthur, M.B., 1993. The motivational

effects of charismatic leadership: a self-concept based theory.

Organization Science 4, 577–594.

Sharma, S., Durand, R.M., Gur-Arie, O., 1981. Identification and

analysis of moderator variables. Journal of Marketing Research

18, 291–300.

Simon, H.A., 1976. Administrative Behavior: A Study of Decision-

Making Processes in Adminstrative Organizations, 3rd ed. The

Free Press, New York.

Snell, S.A., Dean, J.W.J., 1992. Integrated Manufacturing and

human resource management: a human capital perspective.

Academy Management Journal 35 (3), 467–504.

Snell, S.A., Lepak, D.P., Dean, J.W., Youndt, M.A., 2000. Selection

and training for integrated manufacturing: the moderating

effects of job characteristics. Journal of Management Studies

37 (3), 445–466.

Spear, S.J., Bowen, H.K., 1999. Decoding the DNA of the Toyota

Production System. Harvard Business Review 97–106.

Spector, 2003. Industrial and Organizational Psychology: Research

and Practice. John Wiley, New York.

Spector, P.E., 1985. Higher-order need strength as a moderator of the

job scope-employee outcome relationship: a meta-analysis.

Journal of Occupational Psychology 58, 119–127.

Spreitzer, G.M., 1995. Psychological empowerment in the work-

place: dimensions, measurement, and validation. Academy of

Management Journal 38 (5), 1442–1465.

Stajkovic, A.D., Luthans, F., 1998. Self-efficacy and work-related

performance: a meta-analysis. Psychological Bulletin 124 (2),

240–261.

Stajkovic, A.D., Luthans, F., 2003. Behavioral management and task

performance in organizations: conceptual background, meta-

analysis, and test of alternative models. Personnel Psychology

56, 155–195.

Suri, R., de Treville, S., 1986. Getting from just-in-case to just-in-

time: insights from a simple model. Journal of Operations

Management 6 (3), 295–304.

Suzaki, K., 1987. The New Manufacturing Challenge: Techniques

for Continuous Improvement. The Free Press, New York.

Taber, T.D., Taylor, E., 1990. A review and evaluation of the

psychometic properties of the job diagnostic survey. Personnel

Psychology 43, 467–500.

Thomas, K.W., Velthouse, B.A., 1990. Cognitive elements of

empowerment: an ‘‘interpretive’’ model of intrinsic task motiva-

tion. Academy of Management Review 15, 666–681.

Tiegs, R.B., Tetrick, L.E., Fried, Y., 1992. Growth need strength and

context satisfactions as moderators of the relations of the job

characteristics model. Personnel Psychology 43, 467–500.

Vroom, V.H., 1964. Work and motivation. Wiley, New York.

Wacker, J.G., 2004. A theory of formal conceptual definitions:

Developing theory-building measurement instruments. Journal

of Operations Management 22 (6), 629–650.

Walleigh, R.C., 1986. What’s your excuse for not using JIT?

Harvard Business Review 64 (2), 38–54.

Wemmerlov, U., Johnson, D.J., 1997. Cellular manufacturing at 46

user plants: implementation experiences and performance

improvements. International Journal of Production Research

35 (1), 29–49.

Whetten, D.A., 1989. What constitutes a theoretical contribution?

Academy of Management Review 14 (4), 490–495.

White, R.E., Pearson, J.N., Wilson, J.R., 1999. JIT manufacturing: a

survey of implementations in small and large U.S. manufac-

turers. Management Science 45 (1), 1–15.

Womack, J.P., Jones, D.T., Roos, D., 1990. The Machine that

Changed the World. Harper Collins Publishers, New York.

Wood, S., 1989. The Japanese management model: tacit skills in

shop floor participation. Work and Occupations 16 (4), 446–460.

Wood, S.J., Stride, C.B., Wall, T.D., Clegg, C.W., 2004. Revisiting

the use and effectiveness of modern management practices.

Human Factors and Ergonomics 14 (4), 413–430.