Description
The shift to world-class manufacturing strategies has necessitated complementary changes in management accounting
systems (MAS). Using survey data obtained from top manufacturing executives at 253 US firms, this study
empirically examines the relationship between the level of just-in-time (JIT) practices implemented by US manufacturing
firms and the performance measures and incentive systems that are incorporated in their MAS
The role of performance measures and incentive systems in
relation to the degree of JIT implementation
Rosemary R. Fullerton
a
, Cheryl S. McWatters
b,
*
a
Utah State University, School of Accountancy, Logan, UT 84322-3540, USA
b
Faculty of Management, McGill University, Samuel Bronfman Building, 1001 Sherbrooke Street West,
Montreal, Canada H3A 1G5
Abstract
The shift to world-class manufacturing strategies has necessitated complementary changes in management account-
ing systems (MAS). Using survey data obtained from top manufacturing executives at 253 US ?rms, this study
empirically examines the relationship between the level of just-in-time (JIT) practices implemented by US manu-
facturing ?rms and the performance measures and incentive systems that are incorporated in their MAS. The statistical
tests provide empirical evidence that the use of non-traditional performance measures such as bottom-up measures,
product quality, and vendor quality, as well as incentive systems of employee empowerment and compensation rewards
for quality production are related to the degree of JIT practices implemented. # 2002 Elsevier Science Ltd. All rights
reserved.
1. Introduction
An appropriate organizational structure, incor-
porating the management accounting system
(MAS), is considered a necessity for the successful
implementation of organizational strategy (Brick-
ley, Smith, & Zimmerman, 2001; Oldham &
Tomkins, 1999). As organizations adapt to tech-
nological change, globalization, and customer
demand, they must ensure that the MAS is
designed congruent with decision-making and
control requirements. Recent studies have empha-
sized that ‘‘trategic priorities should be sup-
ported by appropriate and e?ectively implemented
manufacturing processes and information systems,
including those providing management accounting
information’’ (Chenhall & Lang?eld-Smith, 1998b,
p. 243). Bouwens and Abernethy (2000) address
this issue in the context of customization strate-
gies. To date, relatively little research has exam-
ined what design of the MAS, organizational
structures, and contexts is consistent with the
adoption of lean manufacturing systems, such as
JIT (Selto, Renne, & Young, 1995).
The shift to world-class, integrated manufactur-
ing strategies, including a JIT management philo-
sophy, requires accompanying changes in the
management accounting systems (MAS) that sup-
port their implementation (Milgrom & Roberts,
1995; Safayeni, Purdy, Van Engelen, & Pal, 1991;
Young & Selto, 1993). JIT’s focus on excellence
through continuous improvement requires a deci-
sion-making system that evaluates the changes in
quality, setup times, defects, rework, and
throughput time. The MAS also must provide the
requisite control systems to motivate organiza-
tional members in terms of JIT strategies. Young
0361-3682/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PI I : S0361- 3682( 02) 00012- 0
Accounting, Organizations and Society 27 (2002) 711–735
www.elsevier.com/locate/aos
* Corresponding author.
and Selto (1993) suggest that both management
and researchers must better understand the types
and usages of management controls that are crucial
to measuring the success of manufacturing strate-
gies. In a JIT environment, the control system
should be linked to critical success factors at all
organizational levels, but as Lang?eld-Smith
(1997) and Mia (2000) discuss, the need for
performance measurements as controls is particu-
larly important at the operational level of the
organization.
The purpose of this study is to evaluate empiri-
cally the relationship between the JIT practices
implemented by US manufacturing ?rms and their
respective systems for decision making and con-
trol, speci?cally the performance measures and
incentive systems that these ?rms use. The study
contributes to the management accounting litera-
ture in a number of ways. First, it responds to the
call for further survey research that focuses on the
combination of strategy, management techniques,
and management accounting (Chenhall & Lang-
?eld-Smith, 1998b). Through hierarchical multiple
linear regression (MLR), the current study ?nds a
signi?cant statistical relationship between the
implementation of a JIT strategy and the adoption
of non-traditional performance measurement and
incentive systems within the MAS. Second, rather
than arbitrarily classifying ?rms into JIT or non-
JIT categories, a contribution of this study is its
provision of a comprehensive assessment of JIT
implementation by capturing the degree to which
manufacturing ?rms have implemented 10 basic
practices supporting the JIT philosophy. Research
has found that the more extensive the adoption of
JIT in both breadth and depth, the greater are its
bene?ts (Fullerton & McWatters, 2001; White &
Prybutok, 2001). Third, the research extends
Chenhall and Lang?eld-Smith (1998a), which
examined the in?uence of management accounting
on discrete management techniques used to
implement manufacturing change programs.
While the JIT philosophy is both broad and
ambiguous in nature, this study operationalizes
JIT in terms of the 10 JIT practices classi?ed and
utilized in previous research (Fullerton &
McWatters, 2001; White, Pearson, & Wilson 1999;
White & Prybutok, 2001; White & Ruch, 1990).
Speci?cally, the research examines the manage-
ment accounting control system that supports JIT
implementation. The next section examines the
prior literature related to JIT and the MAS, and
outlines the research hypotheses. Section 3
describes the research method. Sections 4 and 5
present and discuss the empirical results. The ?nal
section summarizes the study, and identi?es lim-
itations and future research directions.
2. Research hypotheses
2.1. The JIT manufacturing environment
JIT is a Japanese-developed manufacturing phi-
losophy emphasizing excellence through the con-
tinuous elimination of waste and improvement in
productivity. According to Schonberger (1987, p.
5), JIT is the ‘‘most important productivity
enhancing management innovation since the turn
of the century.’’ Manufacturing ?rms have been
trying to duplicate this manufacturing system for
over two decades, often under di?erent names, e.g.
total enterprise manufacturing, world-class manu-
facturing, and lean production (White & Prybu-
tok, 2001). JIT continues to be referred to as a
‘‘revolution in world manufacturing,’’ which, with
the help of the Internet, is making dramatic chan-
ges to the traditional production system (Zur-
awski, 2001). In an Industry Week survey, almost
95% of corporate executives identi?ed lean man-
ufacturing (de?ned as manufacturing practices of
preventive maintenance, cellular manufacturing,
focused factory, continuous ?ow, reduced lot
sizes, quick changeover times, and kanban) as
either extremely critical or somewhat critical to
achieving world-class manufacturing status
(Jusko, 1999).
The wasteless production philosophy underlying
JIT is ‘‘continual productivity and quality
improvement in the pursuit of excellence in all
phases of the industrial cycle’’ (White & Ruch,
1990, p. 12). JIT is expected to reduce manu-
facturing costs continuously through better qual-
ity, lower inventory, and shorter lead times.
Achieving these results requires an even produc-
tion ?ow of small lot size incorporating schedule
712 R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735
stability, product quality, short setup times, pre-
ventive maintenance, and e?cient process layout
(Chapman & Carter, 1990; Foster & Horngren,
1987; Hall & Jackson, 1992).
Previous research frequently has classi?ed ?rms
as JIT or non-JIT based on their use of ‘‘con-
tinuous manufacturing’’ or a ‘‘pull system’’ (e.g.
Balakrishnan, Linsmeier, & Venkatachalam, 1996;
Kalagnanam & Lindsay, 1998). However, classi?-
cation as a ‘‘JIT ?rm’’ can range from the imple-
mentation of an inventory management system to
the total integration of JIT practices throughout
the manufacturing system. Confusion remains
over what exactly constitutes JIT (Mia, 2000); yet
the manufacturing practices of ‘‘e?cient material
?ow, improved quality, and increased employment
involvement’’ continue to be sought-after, compe-
titive strengths of world-class manufacturing
?rms. JIT remains the most ‘‘universally accepted
term to describe this broad production system’’
(White & Prybutok, 2001, p. 113). Many ?rms
may practice a majority of JIT practices, as
de?ned in this study, without identifying them-
selves as JIT ?rms.
In order to optimize overall organizational per-
formance, a systems perspective that employs each
element of JIT should be implemented (Suzaki,
1987). ‘‘The potential synergic bene?ts are not
fully realized until all elements of a JIT system are
integrated’’ (White & Prybutok, 2001, p. 114). A
recent survey reported that purchasing managers
‘‘still believe heartily in JIT,’’ but companies need
to more fully implement material ?ow and quality
management activities to reap its full bene?ts
(Milligan, 2000). The current study examines the
degree of JIT implementation by capturing the
extent to which sample ?rms have adopted a
combination of JIT elements. These data allow for
a comprehensive assessment of JIT implementa-
tion as a broad, integrated production system.
Despite the acute awareness of JIT and its pur-
ported bene?ts, its implementation by US ?rms
has been relatively slow and in an ad hoc fashion
(Clode, 1993; Gilbert, 1990; Goyal & Deshmukh,
1992). The implementation lag has been attributed
to a number of factors: the resistance to change,
inadequate understanding of JIT methods, a ten-
dency to implement only its easiest and least costly
elements, an incompatible workforce and work-
place environment, and non-supportive suppliers
(Majchrzak, 1988; Snell & Dean, 1992; Wafa &
Yasin, 1998; White & Prybutok, 2001). An alter-
native explanation for JIT’s limited success in the
USA is the failure of the MAS to provide appro-
priate performance measures and incentives to
support JIT objectives. In their in-depth, cross-
sectional study of a JIT ?rm, Young and Selto
(1993) found that although information related to
critical success factors was well designed and
available, this information was not provided at the
shop-?oor level where it could a?ect operating
decisions.
2.2. The role of the management accounting system
Implementing JIT creates major changes in an
organization’s way of doing business. These chan-
ges should be re?ected in the MAS that provides
the necessary information for improved decision
making and control. This information also should
enhance ?rm productivity by motivating employ-
ees in terms of the organization’s strategic goals
(Sprinkle, 2000). However, the MAS information
often can be misleading in this new environment.
The preoccupation with e?ciency and cost
reduction at the expense of e?ectiveness for-
ces managers to adopt short-term operational
views that detract attention from the crucial
manufacturing strategy and structure and
alienate the work force (Lee, 1987, p. 84).
For almost two decades, academics and popular
business authors have proclaimed the obsoles-
cence and ine?ectiveness of current management
accounting and business performance measures in
advanced manufacturing environments (i.e.
Green, Amenkhienan, & Johnson, 1992; Hedin &
Russell, 1992; Hiromoto, 1991; Howell & Soucy,
1987; Johnson & Kaplan, 1987; Mazachek, 1993;
McNair, Lynch, & Cross, 1990; McNair, Mos-
coni, & Norris; 1989; Neely, 1999; Sillince &
Sykes, 1995; Wisner & Fawcett, 1991). Due to
their short-term emphasis, current measures do not
provide an accurate assessment of improvements
that a?ect long-term pro?tability (Hendricks,
R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735 713
1994; Ittner & Larcker, 1998; Johnson & Kaplan,
1989; Kaplan, 1983; Kaplan & Norton, 1996).
‘‘Measures have always had the power to shape a
corporation’s destiny, but the focus on ?nancial
?gures alone limits their utility. Management
accounting of the past forced managers to build
world-class organizations with a truncated set of
chromosomes’’ (Epstein & Birchard, 2000, p. 145).
When systems reward managers and employees
for e?orts counterproductive to JIT, instead of for
e?orts designed to increase quality, eliminate
waste, and reduce throughput time, the wrong
incentives are communicated. A ?rm will imple-
ment JIT more successfully, as well as obtain
information requisite to improving its competitive
position, if its performance measures concentrate
on inventory levels, throughput, lead time, defect
rates, equipment downtime, and employee training
(Wisner & Fawcett, 1991). ‘‘The pursuit of manu-
facturing strategies focused on quality and ?ex-
ibility. . .[has] implications for both the range of
strategic performance measures and the importance
of non-?nancial measures’’ (Lillis, 1999, p. 17).
To operate at peak performance, advanced
manufacturing ?rms must retool their MAS
(Epstein & Birchard, 2000), making waste visible
at all levels and integrating performance measure-
ments with continuous-improvement processes
(Maskell, 2000). The lack of slack and cushion in a
JIT environment renders MAS information on
targets and actual performance more critical than
in non-JIT situations (Mia, 2000). Bene?ts from
JIT implementation, therefore, are enhanced by
complementary changes in a ?rm’s internal
accounting measures (Ahmed, Runc, & Mon-
tagno, 1991; Ansari & Modarress, 1986; Barney,
1986; Bennett & Cooper, 1984; Hendricks, 1994;
Milgrom & Roberts, 1995). The management
accounting profession is responding to the need
for MAS modi?cations that will increase global
competitiveness. Recent changes have been
described as a performance measurement revolu-
tion that seeks to redress the insu?ciency of tra-
ditional performance measures for evaluating
advanced manufacturing techniques (Neely, 1999).
Extensive discussion exists of the association
between increased reliance on non-?nancial per-
formance measures and strategic manufacturing
change (Lillis, 1999). Previous studies have indi-
cated that organizations using more e?cient pro-
duction practices make greater use of non-
traditional information and reward systems
(Abernethy & Lillis, 1995; Banker, Potter, &
Schroeder, 1993a, 1993b; Durden, Hassel, &
Upton, 1999; Ittner & Larcker, 1995, 1998;
Jazayeri & Hopper, 1999; Patell, 1987). World-
class manufacturing systems—advanced manu-
facturing technology, TQM, and JIT— have been
referred to and examined in research studies in
isolation and as synergistic combinations (Flynn,
Sakakibara, & Schroeder, 1995; Patell, 1987; Sim
& Killough, 1998; Snell & Dean, 1994; Swanson &
Lankford, 1998). Ittner and Larcker (1995) looked
exclusively at the relationship between TQM
practices and non-traditional information and
reward systems, and found variates from both
TQM practices and non-traditional accounting
measures that were signi?cantly related. Banker et
al. (1993a, 1993b) found a positive relationship
between the accessibility of non-?nancial infor-
mation on the shop ?oor and the implementation
of both TQM and JIT practices. Sim and Killough
(1998) extended Ittner and Larcker (1995) with a
focus on both TQM and JIT, but did not look at
the direct relationship among performance mea-
sures, JIT, and TQM. Instead, these three mea-
sures were independent variables used to explain
plant-wide customer and quality performance.
Although evidence shows the MAS is expanding
to include more non-?nancial information, the
majority of ?rms still use traditional accounting
criteria much more than non-traditional for both
internal and external performance evaluation (Itt-
ner & Larcker, 1998; Mazachek, 1993). The current
study extends existing research related to world-
class manufacturing practices and their relation-
ship to performance measures by examining the
linkage between the degree of JITimplementation in
US manufacturing ?rms and the use of non-tradi-
tional performance measures and incentive systems.
2.3. Hypothesis 1A
To make decisions in a JIT environment, a ?rm
must measure and report those items that are
a?ected by JIT adoption (i.e. inventory turns,
714 R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735
delivery time, scrap, quality, setup times, and
vendor performance). Young (1992) points out
that without appropriate measures to evaluate and
control the critical measures of success in a JIT
system, its level of performance could be incor-
rectly assessed. According to Foster and Horngren
(1987), JIT ?rms depend less on ?nancial measures
and more on personal observations and non-
?nancial measures. In their 1991 study of Japanese
companies, Daniel and Reitsperger found that
setup times, scrap, and downtime were reported
more frequently to managers supporting zero-
defect strategies than managers supporting more
traditional strategies. Banker et al. (1993b)
concluded that when quality improvement strate-
gies were implemented, non-?nancial information
to workers was more available. Results in a rela-
ted study by Banker et al. (1993a) indicated that
the availability and use of productivity measures
were related to the implementation of JIT and
TQM.
Ittner and Larcker (1995, 1997b) recommended
that the distribution of information encompass all
levels of the organization to overcome the de?-
ciencies of the MAS. Workers need to gather their
own ‘‘bottom-up’’ information using statistical
process control (SPC), Pareto analysis, histo-
grams, and ?ow charts, rather than be dependent
upon ‘‘top-down’’ information that emphasizes
standards and budgets (Johnson, 1992). Ittner and
Larcker (1998) reported more extensive use of
non-?nancial performance measures to supple-
ment traditional accounting-based measures. A
case study of a UK chemical company imple-
menting world-class manufacturing practices
found that non-?nancial measures such as quality,
on-time delivery, inventory levels, and productiv-
ity replaced the previous emphasis on budgets and
?nancial measures (Jazayeri & Hopper, 1999).
However, the speci?c link between the degree of
JIT implementation and the use of non-traditional
performance measures and incentive systems is not
clear. Hypothesis 1A examines this relationship:
H
1A
. Firms implementing a higher degree of JIT
elements such as lean manufacturing practices,
quality improvements, and kanban systems are
more likely to use more non-traditional perfor-
mance measures of quality results, bottom-up
data, benchmarking, waste, and vendor quality.
2.4. Hypothesis 1B
If ?rm incentives are not aligned with organiza-
tional changes, the desired behaviors for new,
integrated manufacturing systems are di?cult to
achieve (Lawler, 1981; Snell & Dean, 1992, 1994).
Compensation incentives for all employees need to
be linked to organizational strategy. Reward sys-
tems also must be congruent with other organiza-
tional systems, with alignment necessary among
the organization’s core values, its processes, its
practices, and its structures (Lawler, 1998, p. 288).
In an advanced manufacturing environment,
reward systems should re?ect critical success
factors of product quality and team-based perfor-
mance (Ittner & Larcker, 1995).
1
Reitsperger
(1986) found that workers in Japanese-managed
corporations outperformed their counterparts in
US- and UK-managed companies, because incen-
tive pay was tied to quality and productivity
measures.
Despite the call for more broadly based strategic
measures, the majority of ?rms rely on traditional
?nancial performance measures as compensation
incentives. Mazachek (1993) demonstrated that
managers considered accounting criteria to be sig-
ni?cantly more important than non-accounting
criteria as indicators of ?rm performance and
evaluators of managerial performance. Ittner and
Larcker’s (1998) review of trends in performance
measurement reiterated this point. Hypothesis 1B
examines the link between compensation rewards
and non-?nancial measures in a JIT environment.
1
The relationship between reward systems and advanced
manufacturing systems is not a clear cut one, as noted by
Lawler, Mohrman, and Ledford (1992, p. 102): ‘‘The total
quality literature has in some cases (Deming, 1986) cautioned
against ‘management by fear’ and especially against the estab-
lishment of individual appraisal systems and standards of per-
formance. The argument is that these fail to take account of the
reality that performance levels are more the product of the
system than of individual performance. As a result, practices
that manage the performance of individuals have not been a
central focus of implementation in total quality management.’’
Lawler et al. include JIT as part of the total quality system.
R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735 715
H
1B
. Firms implementing a higher degree of JIT
elements such as lean manufacturing practices,
quality improvements, and kanban systems are
more likely to tie compensation rewards to non-
?nancial measures.
2.5. Hypothesis 1C
Previous research has emphasized that the suc-
cessful adoption of advanced manufacturing sys-
tems is linked to human resource management
practices (Snell & Dean, 1992). JIT is designed to
show respect for people by using their input in
decision making and broadening their workplace
skills (Billesbach & Hayen, 1994; Golhar, Stamm,
& Smith, 1990; Johnston, 1989; Plenert, 1990;
Schonberger, 1982; Snell & Dean, 1992). Wruck
and Jensen’s (1994) analysis of TQM is pertinent
here. Their study outlines TQM’s association with
employee empowerment in that both TQM and
empowerment require ?rms ‘‘to e?ectively utilize
valuable speci?c knowledge at lower levels of the
organization’’ (p. 258).
Empowerment in decision making, however, has
di?erent implications for a JIT environment.
Although JIT might limit employee discretion in
production-level decisions, it generally increases
responsibility in the areas of operations and qual-
ity control (Snell & Dean, 1992, p. 494). The top-
down nature of JIT implementation a?ects some
aspects of employee empowerment and may con-
tribute to the con?ict between reduced discretion
in JIT environments and the need for responsive
operational decision-making (Klein, 1989, 1991).
In a JIT environment, strategic priorities need to
be communicated throughout the ?rm, such that
quality improvements support organizational
strategy. Measurement data should be linked to
corporate strategies (Govindarajan & Gupta,
1985; Ittner & Larcker, 1997b; Kaplan & Norton,
1996; Najarian, 1993; Perera, Harrison, & Poole,
1997). Employees should not only be better
informed, but also have the ability to make oper-
ating decisions. Ittner and Larcker (1995, p. 6)
suggested that ‘‘the primary role of MAS in TQM
environments is providing empowered workers
with information for problem solving and con-
tinuous improvement activities.’’ In a JIT envir-
onment, the workers are put in control of
production operations, which requires their invol-
vement in solving production problems (Banker et
al., 1993a). Thus, the MAS must provide infor-
mation that enables e?ective worker empower-
ment. Hypothesis 1C examines the e?ect of JIT
implementation on the role of employees.
H
1C
. Firms implementing a higher degree of JIT
elements such as lean manufacturing practices,
quality improvements, and kanban systems are
more likely to have increased empowerment in
decision making and a clearer understanding of
company strategy.
3. Research method
3.1. Survey instrument
To examine these relationships, a ?ve-page sur-
vey instrument was used to collect speci?c infor-
mation about the manufacturing operations,
product-costing methods, information and incen-
tive systems, JIT practices employed, perceived
results from JIT implementation, and character-
istics of the respondent ?rms. The survey was sent
to executives representing 447 US manufacturing
?rms. Data from the 253 survey responses were
analyzed to determine whether the implementa-
tion of JIT practices is linked to non-traditional
performance measures and incentives in the MAS.
The majority of the questions on the survey
instrument are either categorical or interval Likert
scales. Factor analysis combined the Likert-scaled
questions into independent measures for testing
the research question. The survey instrument was
evaluated in a limited pretest by several business
professors and managers from ?ve manufacturing
?rms for readability, completeness, and clarity.
Appropriate changes were made as per their com-
ments and suggestions.
3.2. Sample ?rms
To select sample JIT ?rms, an extensive literature
search was done to identify all of the US manu-
facturing ?rms thought to be formally practicing
716 R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735
JIT. A potential sample of additional US manu-
facturing ?rms was located on Compaq Dis-
closure. For inclusion in the study, a ?rm was
required to have a primary two-digit SIC code
within the manufacturing ranges of 20 and 39,
have sales between $2 billion and $2 million, and
be included on the COMPUSTAT database.
2
After eliminating the randomly selected ?rms from
this sample because of duplication or inadequate
COMPUSTAT information, manufacturing
executives at 447 ?rms were sent the survey packet.
In contacting the potential respondents, the
purpose of the survey was explained, along with a
request for participation. The objective was to
locate the most senior manufacturing person who
had a broad enough understanding of operations
to complete the questionnaire. Sometimes the
executive would forward the survey to or supply
the name of another individual who could better
answer the questionnaire. A number of the
respondents indicated that they had requested
information from other personnel to complete the
questionnaire. Thus, evidence exists that serious
e?orts were made to answer the survey questions
appropriately. Following a maximum of three
contacts, 254 out of the 447 ?rms surveyed com-
pleted and returned the survey instruments, for an
overall response rate of 56.8%.
3,4
The respondents
had an average of 17 years of management
experience, including 9 years in management with
their current ?rm and various levels of responsi-
bility (see Table 1 for distribution of respondents).
The industry distributions of the self-identi?ed
JIT, non-JIT, and total sample respondent ?rms
are presented in Table 2. The majority (72%) of
the respondent ?rms are from four industries:
chemicals and allied products (SIC-28), industrial
machinery (SIC-35), electronics (SIC-36), and
instrumentation (SIC-38).
5
3.3. Measuring the degree of JIT implementation
Without assuming directional causality, the
degree of JIT practices implemented operates as the
dependent variable to test the linear equation for
the research hypotheses. An objective of this study
is to specify and measure a representative set of
JIT manufacturing practices. Thus, it was neces-
sary to delineate a set of measurable manufactur-
ing practices describing JIT. The struggle to de?ne
JIT stems from an inability to specify a universal
set of elements (White & Ruch, 1990). Di?erent
2
It was determined that surveying one person in companies
with annual sales in excess of $2 billion about the overall prac-
tices in his or her company was problematic. In addition, ?rms
with annual sales of less than $2 million were determined to be
non-representative. However, seven ?rms in excess of $2 billion
were actually sampled, with four responding. These larger ?rms
were included in the sample because they were pre-identi?ed as
JIT ?rms and a contact person (manufacturing executive) was
known.
3
The means of the sales for non-responding and responding
?rms were compared to determine if there was a response bias
related to the size of the ?rms. The mean of the sales from the
responding ?rms is slightly higher at $404 million, than that of
the sales from the non-responding ?rms at $380 million. How-
ever, an ANOVA test shows the di?erences in the means for
responding and non-responding ?rm sales are not statistically
signi?cant. In addition, the means of the industry SIC codes
(represented as dummy variables) for the non-responding and
responding ?rms were compared. An ANOVA test shows no
statistical di?erences in the industry means between the
responding and non-responding ?rms. Thus, there does not
appear to be a response bias related to either ?rm size or
industry.
4
Perhaps one reason for the higher than usual response rate
for this type of research is the respondents’ interest in the
material. Eighty percent of those responding (192) requested a
copy of the research results.
5
The industry distribution for the respondent ?rms is simi-
lar to the total sample industry distribution. Seventy percent of
the ?rms sampled were from these same four industries: che-
micals and allied products, industrial machinery, electronics,
and instrumentation.
Table 1
Distribution of survey respondents
Title Number
VP manufacturing or operations 51
Manufacturing or operations director 43
Manufacturing/production managers 41
Quality assurance managers 30
VP quality 24
Quality director 21
President/CEO 17
Plant manager 17
Miscellaneous or undesignated 10
Total 254
R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735 717
practices deemed important in adopting JIT are
iterated in several studies (Banker et al., 1993a,
1993b; Flynn et al., 1995; Mehra & Inman, 1992;
Moshavi, 1990; Spencer & Guide, 1995; Young,
1992). Moshavi (1990) suggests ?ve essential ele-
ments of JIT: setup time reduction, focus ?ow
processing, containerization (pull system contain-
ers for inventory), parts control (kanban), and
preventive maintenance. Young (1992) discusses
the JIT manufacturing system, kaizen, total qual-
ity control, and JIT purchasing as important
underlying factors of the Japanese manufacturing
system. A literature review through 1990 by White
and Ruch found 16 techniques identi?ed as JIT. A
consensus for 10 of these JIT elements was iterated
by established JIT authors (e.g. Hall, Hay, Mon-
den, Schonberger, Shingo, and Suzaki). These
consensus elements are described in previous
research as encompassing JIT practices and are
used by White et al. (1999; White & Prybutok,
2001) and Fullerton and McWatters (2001) as JIT
indicators. Thus, they were considered broad
enough to represent a comprehensive JIT imple-
mentation for the purposes of this study. The 10
items employed to measure the extent to which a
company has adopted JIT are: focused factory,
group technology, reduced setup times, total
productive maintenance, multi-function employees,
uniform workload, kanban, JIT purchasing, total
quality control, and quality circles.
6
3.3.1. Factors for JIT determinants
Eleven six-point Likert-scaled questions on the
survey instrument measure the extent to which
?rms use JIT.
7
Responses from the 11 JIT-imple-
mentation questions were re?ned with an explora-
tory factor analysis using the principal
components method. Three components of JIT
with eigenvalues greater than 1.0 were extracted
Table 2
Distribution of two-digit SIC codes for sample ?rms
Industry JIT ?rms frequency Non-JIT ?rms frequency Sample frequency Sample per cent
20—Food 1 6 7 2.8
22—Textiles 2 3 5 2.0
25—Furniture and ?xtures 5 1 6 2.4
26—Paper and allied products 1 1 2 0.8
27—Printing/publishing 1 0 1 0.4
28—Chemicals and allied products 4 20 24 9.5
30—Rubber products 3 2 5 2.0
33—Primary metals 3 12 15 5.9
34—Fabricated metals 7 7 14 5.5
35—Industrial machinery 17 24 41 16.2
36—Electronics 24 37 61 24.1
37—Motor vehicles and accessories 6 5 11 4.3
38—Instrumentation 20 35 55 21.7
39—Other manufacturing 1 5 6 2.3
Totals 95 158 253 100.0
6
Further research is necessary to determine which JIT ele-
ments are most important for successful JIT implementation
and how these JIT elements interact. It is not possible to make
these determinations from the data gathered in the current
study. However, Table 6 indicates the degree to which the
individual elements have been implemented by the sample ?rms
and the largest implementation di?erences between JIT and
non-JIT ?rms. In addition, the factor loadings of the individual
JIT measurement variables reported in Table 3 assist in the
understanding of how these individual JIT practices are related.
Finally, each of the three identi?ed JIT constructs in the factor
analysis is signi?cantly correlated with the other two. Although
it is impossible to determine how they interact, it is apparent
that they do.
7
Total quality control is represented by two questions on
the survey: one is related to process quality and the other to
product quality.
718 R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735
from the analysis, representing 63% of the total
variance in the data.
8
The ?rst factor is a manufacturing component
that explains the extent to which companies have
implemented general manufacturing techniques
associated with JIT, such as focused factory,
group technology, uniform work loads, and multi-
function employees. Together these represent ele-
ments of a JIT manufacturing philosophy, although
individual elements of the factor may be adopted by
any high technology manufacturing ?rm.
The second JIT factor is a quality component
that examines the extent to which companies have
implemented procedures for improving product
and process quality. A reason for the association
between TQM and JIT is their common con-
tinuous improvement goals. Successful JIT imple-
mentation requires a high level of quality in
production. Although TQM can be adopted with-
out implementing JIT, it is unlikely that a JIT
manufacturing system can succeed without incor-
porating the underpinning tenets of TQM. Good
quality management and productive maintenance
are keys to JIT survival (Imai, 1998). Quality fre-
quently has been referred to as the cornerstone of
JIT (Banker et al., 1993a; Sim & Killough, 1998;
Swanson & Lankford, 1998; Young et al., 1988).
The third JIT factor identi?ed is one of uniquely
JIT practices that describe the extent to which
companies have implemented JIT purchasing and
kanban. The likelihood is low that companies who
are not fully committed to a JIT program would
adopt these practices. A description of the speci?c
survey questions that support these factors is
found in Appendix B. For results of the factor
analysis for JIT elements, refer to Table 3.
3.4. Independent variables
Nine constructs were selected to examine the
non-traditional performance measures and incen-
tive systems of manufacturing ?rms. Four of these
constructs, which represent performance measures
for evaluating manufacturing productivity,
measure hypothesis 1A: bottom-up data gathering
techniques; benchmarking for products, services,
and processes; frequency of measurements and
reports on quality; and manufacturing performance
measures. The ?rst three of these constructs were
de?ned in Ittner and Larcker’s 1995 TQM study.
The last construct is similar to one examined by
Durden et al. (1999) in examining the use of non-
?nancial manufacturing performance indicators in
a JIT environment. Three constructs related to
Table 3
Factor analysis (VARIMAX rotation) and factor loadings for JIT variables
a
Factor 1
JITMANUF
Factor 2
JITQLTY
Factor 3
JITUNIQUE
Cronbach’s alpha 0.831 0.946 0.684
Focused factory 0.740
Group technology 0.770
Reduced setup times 0.706
Productive maintenance 0.668
Multi-function employees 0.501
Uniform work load 0.731
Product quality improvement 0.917
Process quality improvement 0.902
Kanban system 0.820
JIT purchasing 0.825
a
All loadings in excess of 0.300 are shown. n=253.
8
All of the 11 elements loaded greater than 0.50 onto one of
the three constructs except for number 11, asking about the use
of ‘‘quality circles’’. It was evident from initial observations of
the survey responses that only a few ?rms (both JIT and non-
JIT) used quality circles. Thus, this question was eliminated
from further testing representing JIT.
R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735 719
performance incentives through compensation are
examined in hypothesis 1B: compensation ties to
non-?nancial performance; compensation ties to
quality and team performance; and compensation
ties to traditional pro?tability measures. The last
two constructs of the research analysis testing
hypothesis 1C are: communication of the strategic
plan to middle managers, ?rst-line supervisors, and
non-management personnel; and empowerment of
employees in decision making.
3.4.1. Factors for performance measures and
incentive systems
Thirty-nine items from the survey instrument
were evaluated to measure the nine performance-
measure and incentive-system constructs. To
reduce and summarize the collected data, these
survey items were subjected to a factor analysis.
Using the principal components method, the fac-
tor analysis revealed ten distinct factors with
eigenvalues greater than 1.0, which accounted for
73% of the total variance in the data.
9
The VAR-
IMAX rotation resulted in the following factors:
QLTYREV: The frequency with which quality
issues are measured and reported
to management strata.
COMPQLTY: The importance of quality and
teamwork in determining
compensation.
BOTTOM: The use of bottom-up data
gathering techniques such as
Pareto analysis, histograms,
and cause-and-e?ect diagrams to
evaluate operations.
COMPBDGT: The importance of adherence to
budget items in determining
compensation.
BENCH: The use of benchmarking to
evaluate operations.
PERFWASTE: The use of performance measures
related to waste and ine?ciency
in evaluating the manufacturing
system
STRPLAN: The extent to which employees
understand the ?rm’s strategic
plan.
PERFVEND: The use of performance measures
related to timeliness and vendor
performance in evaluating the
manufacturing system.
COMPNF: The use of non-?nancial measures
to determine compensation.
EMPOWER: The extent to which line managers
and non-management personnel
are empowered to make decisions.
A description of the speci?c survey questions that
support these factors is found in Appendix A. For
the results from the factor analysis, refer to Table 4.
3.5. Control variables
Four control variables (covariates) are included
in the regression testing. Firm size (SIZE) a?ects
most aspects of a ?rm’s strategy and success;
therefore a ?rm’s net sales are used to control for
?rm size. The net sales for each sample ?rm are
obtained from COMPUSTAT data. Whether a
?rm follows a more innovative strategy can a?ect
its willingness to make changes. Innovative ?rms
are more risky and generally more pro?table
(Capon, Farle, & Hoenig, 1988). Innovation
(INNOV) is measured by a ?rm’s response on the
survey instrument as to whether it is a leader or a
follower in product technology, product design,
and process design (Ittner & Larcker, 1995). The
industry in which a ?rm operates often a?ects its
competitive behavior and performance measures.
Thus, the industry for each ?rm, as identi?ed on
COMPUSTAT, is tested by the use of the ?rm’s
two-digit SIC code (SIC). Organizational structure
can in?uence a ?rm’s ability to be ?exible and
9
Seven factors loaded as expected (QLTYREV, BOTTOM,
BENCH, COMPQLTY, COMPNF, STRPLAN, and
EMPOWER). The construct for evaluating compensation
rewards from traditional ?nancial measures loaded onto two
factors. One was for compensation incentives related to var-
iances and budgets, which created the COMPBDGT variable.
The other factor was for compensation rewards related to tra-
ditional pro?tability measures. This factor made no signi?cant
contributions to any of the regression tests; therefore, it was
eliminated from the ?nal analyses. The expected single con-
struct measuring manufacturing performance loaded onto two
factors as per the factor analysis: PERFWASTE and
PERFVEND.
720 R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735
Table 4
Factor analysis (VARIMAX rotation) and factor loadings for performance measures and incentive systems variables
a
Factor 1
QLTYREV
Factor 2
COMPQLTY
Factor 3
BOTTOM
Factor 4
COMPBDGT
Factor 5
BENCH
Factor 6
PERFWASTE
Factor 7
STRPLAN
Factor 8
PERFVEND
Factor 9
COMPNF
Factor 10
EMPOWER
Cronbach’s alpha 0.920 0.909 0.873 0.862 0.907 0.775 0.828 0.783 0.869 0.809
TM reviews quality results 0.719
TM reviews quality consequences 0.793
MM reviews quality results 0.836
MM reviews quality consequences 0.857
LS reviews quality results 0.819
LS reviews quality consequences 0.828
MM compensation—quality 0.773
MM compensation—throughput 0.703
MM compensation—teamwork 0.734
LS compensation—quality 0.854
LS compensation—throughput 0.827
LS compensation—teamwork 0.783
Use cause-and-e?ect diagrams 0.729
Use histograms 0.791
Use ?owcharting 0.658
Use Pareto analysis 0.778
Use scatter diagrams 0.748
Use SPC charts 0.652
MM compensation—variances 0.788
MM compensation—budget 0.704
LS compensation—variances 0.794
LS compensation—budget 0.780
Benchmarking of operations 0.844
Benchmarking of products 0.867
Benchmarking of delivery systems 0.836
Performance measures—downtime 0.691
Performance measures—scrap 0.721
Performance measures—rework 0.689
Performance measures—setups 0.685
MM understand strategic plan 0.791
LS understand strategic plan 0.793
NM understand strategic plan 0.776
Performance measures—on-time 0.514
Vendor performance—quality 0.866
Vendor performance—on-time 0.879
MM compensation—non-?nancial 0.867
LS compensation—non-?nancial 0.873
LS empowerment 0.713
NM empowerment 0.741
a
All loadings in excess of 0.300 are shown. n=253. TM=top management; MM=middle management; LS=line supervisors; NM=non-management.
R
.
R
.
F
u
l
l
e
r
t
o
n
,
C
.
S
.
M
c
W
a
t
t
e
r
s
/
A
c
c
o
u
n
t
i
n
g
,
O
r
g
a
n
i
z
a
t
i
o
n
s
a
n
d
S
o
c
i
e
t
y
2
7
(
2
0
0
2
)
7
1
1
–
7
3
5
7
2
1
make major operational changes. If a ?rm is
highly centralized, the employees will be much less
involved in decision making and organizational
changes than if it is more decentralized. Kalaga-
nam and Lindsay (1998) demonstrated how adapt-
ing more organic (decentralized) organizational
structures led to greater bene?ts from JIT adop-
tion. The organizational structure (STRUCTR) of
a ?rm is identi?ed on the questionnaire.
3.5.1. Factors for control variables
The six survey questions related to ?rm innova-
tion and organizational structure were reduced
and summarized using factor analysis. These six
variables converged into two anticipated distinct
factors with eigenvalues in excess of 1.0, account-
ing for 66% of the total variance in the data. The
VARIMAX rotation resulted in the following
control variables:
STRUCTR: The extent of centralization or
decentralization of a ?rm’s
organizational structure.
INNOV: The extent to which the ?rm considers
itself a leader in product and process
design and product technology.
A detailed description of the speci?c questions
that support these control variables is found in
Appendix B. Refer to Table 5 for the rotated fac-
tor solution.
3.6. Construct validity and reliability analysis
The factor solutions for the de?ned constructs
support the construct validity of the survey instru-
ment. Convergent validity is demonstrated by each
factor having multiple-question loadings in excess
of 0.5. In addition, discriminant validity is sup-
ported, since none of the questions in the factor
analyses have loadings in excess of 0.3 on more
than one factor.
10
Cronbach’s alpha is used as the
coe?cient of reliability for testing the internal
consistency of the constructs validated by the fac-
tor analysis. The alpha coe?cients for all of the
constructs are in excess of 0.7.
11
(The alpha coef-
?cients are included in Tables 3–5.) Overall, these
tests support the validity of the measures repre-
senting the constructs used in this study.
4. Research results
4.1. Descriptive statistics
One objective of this study is to capture the
degree to which the sample ?rms have imple-
mented JIT practices. On the survey instrument,
the respondents were asked to provide the degree
to which they were using 10 individual aspects of
JIT (scaled from 1 to 6). Respondents also were
asked to indicate whether their ?rm had formally
implemented JIT. Descriptive statistics depicting
the means for each of the individual elements,
along with the three JIT factors, and the total
combination of the JIT elements are shown on
Table 6. The data are presented in terms of the
total sample, the JIT sample ?rms, and the non-
JIT sample ?rms.
The ANOVA comparison of the means between
the JIT and non-JIT ?rms, found on Table 6,
consistently shows highly signi?cant di?erences
(p
The shift to world-class manufacturing strategies has necessitated complementary changes in management accounting
systems (MAS). Using survey data obtained from top manufacturing executives at 253 US firms, this study
empirically examines the relationship between the level of just-in-time (JIT) practices implemented by US manufacturing
firms and the performance measures and incentive systems that are incorporated in their MAS
The role of performance measures and incentive systems in
relation to the degree of JIT implementation
Rosemary R. Fullerton
a
, Cheryl S. McWatters
b,
*
a
Utah State University, School of Accountancy, Logan, UT 84322-3540, USA
b
Faculty of Management, McGill University, Samuel Bronfman Building, 1001 Sherbrooke Street West,
Montreal, Canada H3A 1G5
Abstract
The shift to world-class manufacturing strategies has necessitated complementary changes in management account-
ing systems (MAS). Using survey data obtained from top manufacturing executives at 253 US ?rms, this study
empirically examines the relationship between the level of just-in-time (JIT) practices implemented by US manu-
facturing ?rms and the performance measures and incentive systems that are incorporated in their MAS. The statistical
tests provide empirical evidence that the use of non-traditional performance measures such as bottom-up measures,
product quality, and vendor quality, as well as incentive systems of employee empowerment and compensation rewards
for quality production are related to the degree of JIT practices implemented. # 2002 Elsevier Science Ltd. All rights
reserved.
1. Introduction
An appropriate organizational structure, incor-
porating the management accounting system
(MAS), is considered a necessity for the successful
implementation of organizational strategy (Brick-
ley, Smith, & Zimmerman, 2001; Oldham &
Tomkins, 1999). As organizations adapt to tech-
nological change, globalization, and customer
demand, they must ensure that the MAS is
designed congruent with decision-making and
control requirements. Recent studies have empha-
sized that ‘‘
ported by appropriate and e?ectively implemented
manufacturing processes and information systems,
including those providing management accounting
information’’ (Chenhall & Lang?eld-Smith, 1998b,
p. 243). Bouwens and Abernethy (2000) address
this issue in the context of customization strate-
gies. To date, relatively little research has exam-
ined what design of the MAS, organizational
structures, and contexts is consistent with the
adoption of lean manufacturing systems, such as
JIT (Selto, Renne, & Young, 1995).
The shift to world-class, integrated manufactur-
ing strategies, including a JIT management philo-
sophy, requires accompanying changes in the
management accounting systems (MAS) that sup-
port their implementation (Milgrom & Roberts,
1995; Safayeni, Purdy, Van Engelen, & Pal, 1991;
Young & Selto, 1993). JIT’s focus on excellence
through continuous improvement requires a deci-
sion-making system that evaluates the changes in
quality, setup times, defects, rework, and
throughput time. The MAS also must provide the
requisite control systems to motivate organiza-
tional members in terms of JIT strategies. Young
0361-3682/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PI I : S0361- 3682( 02) 00012- 0
Accounting, Organizations and Society 27 (2002) 711–735
www.elsevier.com/locate/aos
* Corresponding author.
and Selto (1993) suggest that both management
and researchers must better understand the types
and usages of management controls that are crucial
to measuring the success of manufacturing strate-
gies. In a JIT environment, the control system
should be linked to critical success factors at all
organizational levels, but as Lang?eld-Smith
(1997) and Mia (2000) discuss, the need for
performance measurements as controls is particu-
larly important at the operational level of the
organization.
The purpose of this study is to evaluate empiri-
cally the relationship between the JIT practices
implemented by US manufacturing ?rms and their
respective systems for decision making and con-
trol, speci?cally the performance measures and
incentive systems that these ?rms use. The study
contributes to the management accounting litera-
ture in a number of ways. First, it responds to the
call for further survey research that focuses on the
combination of strategy, management techniques,
and management accounting (Chenhall & Lang-
?eld-Smith, 1998b). Through hierarchical multiple
linear regression (MLR), the current study ?nds a
signi?cant statistical relationship between the
implementation of a JIT strategy and the adoption
of non-traditional performance measurement and
incentive systems within the MAS. Second, rather
than arbitrarily classifying ?rms into JIT or non-
JIT categories, a contribution of this study is its
provision of a comprehensive assessment of JIT
implementation by capturing the degree to which
manufacturing ?rms have implemented 10 basic
practices supporting the JIT philosophy. Research
has found that the more extensive the adoption of
JIT in both breadth and depth, the greater are its
bene?ts (Fullerton & McWatters, 2001; White &
Prybutok, 2001). Third, the research extends
Chenhall and Lang?eld-Smith (1998a), which
examined the in?uence of management accounting
on discrete management techniques used to
implement manufacturing change programs.
While the JIT philosophy is both broad and
ambiguous in nature, this study operationalizes
JIT in terms of the 10 JIT practices classi?ed and
utilized in previous research (Fullerton &
McWatters, 2001; White, Pearson, & Wilson 1999;
White & Prybutok, 2001; White & Ruch, 1990).
Speci?cally, the research examines the manage-
ment accounting control system that supports JIT
implementation. The next section examines the
prior literature related to JIT and the MAS, and
outlines the research hypotheses. Section 3
describes the research method. Sections 4 and 5
present and discuss the empirical results. The ?nal
section summarizes the study, and identi?es lim-
itations and future research directions.
2. Research hypotheses
2.1. The JIT manufacturing environment
JIT is a Japanese-developed manufacturing phi-
losophy emphasizing excellence through the con-
tinuous elimination of waste and improvement in
productivity. According to Schonberger (1987, p.
5), JIT is the ‘‘most important productivity
enhancing management innovation since the turn
of the century.’’ Manufacturing ?rms have been
trying to duplicate this manufacturing system for
over two decades, often under di?erent names, e.g.
total enterprise manufacturing, world-class manu-
facturing, and lean production (White & Prybu-
tok, 2001). JIT continues to be referred to as a
‘‘revolution in world manufacturing,’’ which, with
the help of the Internet, is making dramatic chan-
ges to the traditional production system (Zur-
awski, 2001). In an Industry Week survey, almost
95% of corporate executives identi?ed lean man-
ufacturing (de?ned as manufacturing practices of
preventive maintenance, cellular manufacturing,
focused factory, continuous ?ow, reduced lot
sizes, quick changeover times, and kanban) as
either extremely critical or somewhat critical to
achieving world-class manufacturing status
(Jusko, 1999).
The wasteless production philosophy underlying
JIT is ‘‘continual productivity and quality
improvement in the pursuit of excellence in all
phases of the industrial cycle’’ (White & Ruch,
1990, p. 12). JIT is expected to reduce manu-
facturing costs continuously through better qual-
ity, lower inventory, and shorter lead times.
Achieving these results requires an even produc-
tion ?ow of small lot size incorporating schedule
712 R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735
stability, product quality, short setup times, pre-
ventive maintenance, and e?cient process layout
(Chapman & Carter, 1990; Foster & Horngren,
1987; Hall & Jackson, 1992).
Previous research frequently has classi?ed ?rms
as JIT or non-JIT based on their use of ‘‘con-
tinuous manufacturing’’ or a ‘‘pull system’’ (e.g.
Balakrishnan, Linsmeier, & Venkatachalam, 1996;
Kalagnanam & Lindsay, 1998). However, classi?-
cation as a ‘‘JIT ?rm’’ can range from the imple-
mentation of an inventory management system to
the total integration of JIT practices throughout
the manufacturing system. Confusion remains
over what exactly constitutes JIT (Mia, 2000); yet
the manufacturing practices of ‘‘e?cient material
?ow, improved quality, and increased employment
involvement’’ continue to be sought-after, compe-
titive strengths of world-class manufacturing
?rms. JIT remains the most ‘‘universally accepted
term to describe this broad production system’’
(White & Prybutok, 2001, p. 113). Many ?rms
may practice a majority of JIT practices, as
de?ned in this study, without identifying them-
selves as JIT ?rms.
In order to optimize overall organizational per-
formance, a systems perspective that employs each
element of JIT should be implemented (Suzaki,
1987). ‘‘The potential synergic bene?ts are not
fully realized until all elements of a JIT system are
integrated’’ (White & Prybutok, 2001, p. 114). A
recent survey reported that purchasing managers
‘‘still believe heartily in JIT,’’ but companies need
to more fully implement material ?ow and quality
management activities to reap its full bene?ts
(Milligan, 2000). The current study examines the
degree of JIT implementation by capturing the
extent to which sample ?rms have adopted a
combination of JIT elements. These data allow for
a comprehensive assessment of JIT implementa-
tion as a broad, integrated production system.
Despite the acute awareness of JIT and its pur-
ported bene?ts, its implementation by US ?rms
has been relatively slow and in an ad hoc fashion
(Clode, 1993; Gilbert, 1990; Goyal & Deshmukh,
1992). The implementation lag has been attributed
to a number of factors: the resistance to change,
inadequate understanding of JIT methods, a ten-
dency to implement only its easiest and least costly
elements, an incompatible workforce and work-
place environment, and non-supportive suppliers
(Majchrzak, 1988; Snell & Dean, 1992; Wafa &
Yasin, 1998; White & Prybutok, 2001). An alter-
native explanation for JIT’s limited success in the
USA is the failure of the MAS to provide appro-
priate performance measures and incentives to
support JIT objectives. In their in-depth, cross-
sectional study of a JIT ?rm, Young and Selto
(1993) found that although information related to
critical success factors was well designed and
available, this information was not provided at the
shop-?oor level where it could a?ect operating
decisions.
2.2. The role of the management accounting system
Implementing JIT creates major changes in an
organization’s way of doing business. These chan-
ges should be re?ected in the MAS that provides
the necessary information for improved decision
making and control. This information also should
enhance ?rm productivity by motivating employ-
ees in terms of the organization’s strategic goals
(Sprinkle, 2000). However, the MAS information
often can be misleading in this new environment.
The preoccupation with e?ciency and cost
reduction at the expense of e?ectiveness for-
ces managers to adopt short-term operational
views that detract attention from the crucial
manufacturing strategy and structure and
alienate the work force (Lee, 1987, p. 84).
For almost two decades, academics and popular
business authors have proclaimed the obsoles-
cence and ine?ectiveness of current management
accounting and business performance measures in
advanced manufacturing environments (i.e.
Green, Amenkhienan, & Johnson, 1992; Hedin &
Russell, 1992; Hiromoto, 1991; Howell & Soucy,
1987; Johnson & Kaplan, 1987; Mazachek, 1993;
McNair, Lynch, & Cross, 1990; McNair, Mos-
coni, & Norris; 1989; Neely, 1999; Sillince &
Sykes, 1995; Wisner & Fawcett, 1991). Due to
their short-term emphasis, current measures do not
provide an accurate assessment of improvements
that a?ect long-term pro?tability (Hendricks,
R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735 713
1994; Ittner & Larcker, 1998; Johnson & Kaplan,
1989; Kaplan, 1983; Kaplan & Norton, 1996).
‘‘Measures have always had the power to shape a
corporation’s destiny, but the focus on ?nancial
?gures alone limits their utility. Management
accounting of the past forced managers to build
world-class organizations with a truncated set of
chromosomes’’ (Epstein & Birchard, 2000, p. 145).
When systems reward managers and employees
for e?orts counterproductive to JIT, instead of for
e?orts designed to increase quality, eliminate
waste, and reduce throughput time, the wrong
incentives are communicated. A ?rm will imple-
ment JIT more successfully, as well as obtain
information requisite to improving its competitive
position, if its performance measures concentrate
on inventory levels, throughput, lead time, defect
rates, equipment downtime, and employee training
(Wisner & Fawcett, 1991). ‘‘The pursuit of manu-
facturing strategies focused on quality and ?ex-
ibility. . .[has] implications for both the range of
strategic performance measures and the importance
of non-?nancial measures’’ (Lillis, 1999, p. 17).
To operate at peak performance, advanced
manufacturing ?rms must retool their MAS
(Epstein & Birchard, 2000), making waste visible
at all levels and integrating performance measure-
ments with continuous-improvement processes
(Maskell, 2000). The lack of slack and cushion in a
JIT environment renders MAS information on
targets and actual performance more critical than
in non-JIT situations (Mia, 2000). Bene?ts from
JIT implementation, therefore, are enhanced by
complementary changes in a ?rm’s internal
accounting measures (Ahmed, Runc, & Mon-
tagno, 1991; Ansari & Modarress, 1986; Barney,
1986; Bennett & Cooper, 1984; Hendricks, 1994;
Milgrom & Roberts, 1995). The management
accounting profession is responding to the need
for MAS modi?cations that will increase global
competitiveness. Recent changes have been
described as a performance measurement revolu-
tion that seeks to redress the insu?ciency of tra-
ditional performance measures for evaluating
advanced manufacturing techniques (Neely, 1999).
Extensive discussion exists of the association
between increased reliance on non-?nancial per-
formance measures and strategic manufacturing
change (Lillis, 1999). Previous studies have indi-
cated that organizations using more e?cient pro-
duction practices make greater use of non-
traditional information and reward systems
(Abernethy & Lillis, 1995; Banker, Potter, &
Schroeder, 1993a, 1993b; Durden, Hassel, &
Upton, 1999; Ittner & Larcker, 1995, 1998;
Jazayeri & Hopper, 1999; Patell, 1987). World-
class manufacturing systems—advanced manu-
facturing technology, TQM, and JIT— have been
referred to and examined in research studies in
isolation and as synergistic combinations (Flynn,
Sakakibara, & Schroeder, 1995; Patell, 1987; Sim
& Killough, 1998; Snell & Dean, 1994; Swanson &
Lankford, 1998). Ittner and Larcker (1995) looked
exclusively at the relationship between TQM
practices and non-traditional information and
reward systems, and found variates from both
TQM practices and non-traditional accounting
measures that were signi?cantly related. Banker et
al. (1993a, 1993b) found a positive relationship
between the accessibility of non-?nancial infor-
mation on the shop ?oor and the implementation
of both TQM and JIT practices. Sim and Killough
(1998) extended Ittner and Larcker (1995) with a
focus on both TQM and JIT, but did not look at
the direct relationship among performance mea-
sures, JIT, and TQM. Instead, these three mea-
sures were independent variables used to explain
plant-wide customer and quality performance.
Although evidence shows the MAS is expanding
to include more non-?nancial information, the
majority of ?rms still use traditional accounting
criteria much more than non-traditional for both
internal and external performance evaluation (Itt-
ner & Larcker, 1998; Mazachek, 1993). The current
study extends existing research related to world-
class manufacturing practices and their relation-
ship to performance measures by examining the
linkage between the degree of JITimplementation in
US manufacturing ?rms and the use of non-tradi-
tional performance measures and incentive systems.
2.3. Hypothesis 1A
To make decisions in a JIT environment, a ?rm
must measure and report those items that are
a?ected by JIT adoption (i.e. inventory turns,
714 R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735
delivery time, scrap, quality, setup times, and
vendor performance). Young (1992) points out
that without appropriate measures to evaluate and
control the critical measures of success in a JIT
system, its level of performance could be incor-
rectly assessed. According to Foster and Horngren
(1987), JIT ?rms depend less on ?nancial measures
and more on personal observations and non-
?nancial measures. In their 1991 study of Japanese
companies, Daniel and Reitsperger found that
setup times, scrap, and downtime were reported
more frequently to managers supporting zero-
defect strategies than managers supporting more
traditional strategies. Banker et al. (1993b)
concluded that when quality improvement strate-
gies were implemented, non-?nancial information
to workers was more available. Results in a rela-
ted study by Banker et al. (1993a) indicated that
the availability and use of productivity measures
were related to the implementation of JIT and
TQM.
Ittner and Larcker (1995, 1997b) recommended
that the distribution of information encompass all
levels of the organization to overcome the de?-
ciencies of the MAS. Workers need to gather their
own ‘‘bottom-up’’ information using statistical
process control (SPC), Pareto analysis, histo-
grams, and ?ow charts, rather than be dependent
upon ‘‘top-down’’ information that emphasizes
standards and budgets (Johnson, 1992). Ittner and
Larcker (1998) reported more extensive use of
non-?nancial performance measures to supple-
ment traditional accounting-based measures. A
case study of a UK chemical company imple-
menting world-class manufacturing practices
found that non-?nancial measures such as quality,
on-time delivery, inventory levels, and productiv-
ity replaced the previous emphasis on budgets and
?nancial measures (Jazayeri & Hopper, 1999).
However, the speci?c link between the degree of
JIT implementation and the use of non-traditional
performance measures and incentive systems is not
clear. Hypothesis 1A examines this relationship:
H
1A
. Firms implementing a higher degree of JIT
elements such as lean manufacturing practices,
quality improvements, and kanban systems are
more likely to use more non-traditional perfor-
mance measures of quality results, bottom-up
data, benchmarking, waste, and vendor quality.
2.4. Hypothesis 1B
If ?rm incentives are not aligned with organiza-
tional changes, the desired behaviors for new,
integrated manufacturing systems are di?cult to
achieve (Lawler, 1981; Snell & Dean, 1992, 1994).
Compensation incentives for all employees need to
be linked to organizational strategy. Reward sys-
tems also must be congruent with other organiza-
tional systems, with alignment necessary among
the organization’s core values, its processes, its
practices, and its structures (Lawler, 1998, p. 288).
In an advanced manufacturing environment,
reward systems should re?ect critical success
factors of product quality and team-based perfor-
mance (Ittner & Larcker, 1995).
1
Reitsperger
(1986) found that workers in Japanese-managed
corporations outperformed their counterparts in
US- and UK-managed companies, because incen-
tive pay was tied to quality and productivity
measures.
Despite the call for more broadly based strategic
measures, the majority of ?rms rely on traditional
?nancial performance measures as compensation
incentives. Mazachek (1993) demonstrated that
managers considered accounting criteria to be sig-
ni?cantly more important than non-accounting
criteria as indicators of ?rm performance and
evaluators of managerial performance. Ittner and
Larcker’s (1998) review of trends in performance
measurement reiterated this point. Hypothesis 1B
examines the link between compensation rewards
and non-?nancial measures in a JIT environment.
1
The relationship between reward systems and advanced
manufacturing systems is not a clear cut one, as noted by
Lawler, Mohrman, and Ledford (1992, p. 102): ‘‘The total
quality literature has in some cases (Deming, 1986) cautioned
against ‘management by fear’ and especially against the estab-
lishment of individual appraisal systems and standards of per-
formance. The argument is that these fail to take account of the
reality that performance levels are more the product of the
system than of individual performance. As a result, practices
that manage the performance of individuals have not been a
central focus of implementation in total quality management.’’
Lawler et al. include JIT as part of the total quality system.
R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735 715
H
1B
. Firms implementing a higher degree of JIT
elements such as lean manufacturing practices,
quality improvements, and kanban systems are
more likely to tie compensation rewards to non-
?nancial measures.
2.5. Hypothesis 1C
Previous research has emphasized that the suc-
cessful adoption of advanced manufacturing sys-
tems is linked to human resource management
practices (Snell & Dean, 1992). JIT is designed to
show respect for people by using their input in
decision making and broadening their workplace
skills (Billesbach & Hayen, 1994; Golhar, Stamm,
& Smith, 1990; Johnston, 1989; Plenert, 1990;
Schonberger, 1982; Snell & Dean, 1992). Wruck
and Jensen’s (1994) analysis of TQM is pertinent
here. Their study outlines TQM’s association with
employee empowerment in that both TQM and
empowerment require ?rms ‘‘to e?ectively utilize
valuable speci?c knowledge at lower levels of the
organization’’ (p. 258).
Empowerment in decision making, however, has
di?erent implications for a JIT environment.
Although JIT might limit employee discretion in
production-level decisions, it generally increases
responsibility in the areas of operations and qual-
ity control (Snell & Dean, 1992, p. 494). The top-
down nature of JIT implementation a?ects some
aspects of employee empowerment and may con-
tribute to the con?ict between reduced discretion
in JIT environments and the need for responsive
operational decision-making (Klein, 1989, 1991).
In a JIT environment, strategic priorities need to
be communicated throughout the ?rm, such that
quality improvements support organizational
strategy. Measurement data should be linked to
corporate strategies (Govindarajan & Gupta,
1985; Ittner & Larcker, 1997b; Kaplan & Norton,
1996; Najarian, 1993; Perera, Harrison, & Poole,
1997). Employees should not only be better
informed, but also have the ability to make oper-
ating decisions. Ittner and Larcker (1995, p. 6)
suggested that ‘‘the primary role of MAS in TQM
environments is providing empowered workers
with information for problem solving and con-
tinuous improvement activities.’’ In a JIT envir-
onment, the workers are put in control of
production operations, which requires their invol-
vement in solving production problems (Banker et
al., 1993a). Thus, the MAS must provide infor-
mation that enables e?ective worker empower-
ment. Hypothesis 1C examines the e?ect of JIT
implementation on the role of employees.
H
1C
. Firms implementing a higher degree of JIT
elements such as lean manufacturing practices,
quality improvements, and kanban systems are
more likely to have increased empowerment in
decision making and a clearer understanding of
company strategy.
3. Research method
3.1. Survey instrument
To examine these relationships, a ?ve-page sur-
vey instrument was used to collect speci?c infor-
mation about the manufacturing operations,
product-costing methods, information and incen-
tive systems, JIT practices employed, perceived
results from JIT implementation, and character-
istics of the respondent ?rms. The survey was sent
to executives representing 447 US manufacturing
?rms. Data from the 253 survey responses were
analyzed to determine whether the implementa-
tion of JIT practices is linked to non-traditional
performance measures and incentives in the MAS.
The majority of the questions on the survey
instrument are either categorical or interval Likert
scales. Factor analysis combined the Likert-scaled
questions into independent measures for testing
the research question. The survey instrument was
evaluated in a limited pretest by several business
professors and managers from ?ve manufacturing
?rms for readability, completeness, and clarity.
Appropriate changes were made as per their com-
ments and suggestions.
3.2. Sample ?rms
To select sample JIT ?rms, an extensive literature
search was done to identify all of the US manu-
facturing ?rms thought to be formally practicing
716 R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735
JIT. A potential sample of additional US manu-
facturing ?rms was located on Compaq Dis-
closure. For inclusion in the study, a ?rm was
required to have a primary two-digit SIC code
within the manufacturing ranges of 20 and 39,
have sales between $2 billion and $2 million, and
be included on the COMPUSTAT database.
2
After eliminating the randomly selected ?rms from
this sample because of duplication or inadequate
COMPUSTAT information, manufacturing
executives at 447 ?rms were sent the survey packet.
In contacting the potential respondents, the
purpose of the survey was explained, along with a
request for participation. The objective was to
locate the most senior manufacturing person who
had a broad enough understanding of operations
to complete the questionnaire. Sometimes the
executive would forward the survey to or supply
the name of another individual who could better
answer the questionnaire. A number of the
respondents indicated that they had requested
information from other personnel to complete the
questionnaire. Thus, evidence exists that serious
e?orts were made to answer the survey questions
appropriately. Following a maximum of three
contacts, 254 out of the 447 ?rms surveyed com-
pleted and returned the survey instruments, for an
overall response rate of 56.8%.
3,4
The respondents
had an average of 17 years of management
experience, including 9 years in management with
their current ?rm and various levels of responsi-
bility (see Table 1 for distribution of respondents).
The industry distributions of the self-identi?ed
JIT, non-JIT, and total sample respondent ?rms
are presented in Table 2. The majority (72%) of
the respondent ?rms are from four industries:
chemicals and allied products (SIC-28), industrial
machinery (SIC-35), electronics (SIC-36), and
instrumentation (SIC-38).
5
3.3. Measuring the degree of JIT implementation
Without assuming directional causality, the
degree of JIT practices implemented operates as the
dependent variable to test the linear equation for
the research hypotheses. An objective of this study
is to specify and measure a representative set of
JIT manufacturing practices. Thus, it was neces-
sary to delineate a set of measurable manufactur-
ing practices describing JIT. The struggle to de?ne
JIT stems from an inability to specify a universal
set of elements (White & Ruch, 1990). Di?erent
2
It was determined that surveying one person in companies
with annual sales in excess of $2 billion about the overall prac-
tices in his or her company was problematic. In addition, ?rms
with annual sales of less than $2 million were determined to be
non-representative. However, seven ?rms in excess of $2 billion
were actually sampled, with four responding. These larger ?rms
were included in the sample because they were pre-identi?ed as
JIT ?rms and a contact person (manufacturing executive) was
known.
3
The means of the sales for non-responding and responding
?rms were compared to determine if there was a response bias
related to the size of the ?rms. The mean of the sales from the
responding ?rms is slightly higher at $404 million, than that of
the sales from the non-responding ?rms at $380 million. How-
ever, an ANOVA test shows the di?erences in the means for
responding and non-responding ?rm sales are not statistically
signi?cant. In addition, the means of the industry SIC codes
(represented as dummy variables) for the non-responding and
responding ?rms were compared. An ANOVA test shows no
statistical di?erences in the industry means between the
responding and non-responding ?rms. Thus, there does not
appear to be a response bias related to either ?rm size or
industry.
4
Perhaps one reason for the higher than usual response rate
for this type of research is the respondents’ interest in the
material. Eighty percent of those responding (192) requested a
copy of the research results.
5
The industry distribution for the respondent ?rms is simi-
lar to the total sample industry distribution. Seventy percent of
the ?rms sampled were from these same four industries: che-
micals and allied products, industrial machinery, electronics,
and instrumentation.
Table 1
Distribution of survey respondents
Title Number
VP manufacturing or operations 51
Manufacturing or operations director 43
Manufacturing/production managers 41
Quality assurance managers 30
VP quality 24
Quality director 21
President/CEO 17
Plant manager 17
Miscellaneous or undesignated 10
Total 254
R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735 717
practices deemed important in adopting JIT are
iterated in several studies (Banker et al., 1993a,
1993b; Flynn et al., 1995; Mehra & Inman, 1992;
Moshavi, 1990; Spencer & Guide, 1995; Young,
1992). Moshavi (1990) suggests ?ve essential ele-
ments of JIT: setup time reduction, focus ?ow
processing, containerization (pull system contain-
ers for inventory), parts control (kanban), and
preventive maintenance. Young (1992) discusses
the JIT manufacturing system, kaizen, total qual-
ity control, and JIT purchasing as important
underlying factors of the Japanese manufacturing
system. A literature review through 1990 by White
and Ruch found 16 techniques identi?ed as JIT. A
consensus for 10 of these JIT elements was iterated
by established JIT authors (e.g. Hall, Hay, Mon-
den, Schonberger, Shingo, and Suzaki). These
consensus elements are described in previous
research as encompassing JIT practices and are
used by White et al. (1999; White & Prybutok,
2001) and Fullerton and McWatters (2001) as JIT
indicators. Thus, they were considered broad
enough to represent a comprehensive JIT imple-
mentation for the purposes of this study. The 10
items employed to measure the extent to which a
company has adopted JIT are: focused factory,
group technology, reduced setup times, total
productive maintenance, multi-function employees,
uniform workload, kanban, JIT purchasing, total
quality control, and quality circles.
6
3.3.1. Factors for JIT determinants
Eleven six-point Likert-scaled questions on the
survey instrument measure the extent to which
?rms use JIT.
7
Responses from the 11 JIT-imple-
mentation questions were re?ned with an explora-
tory factor analysis using the principal
components method. Three components of JIT
with eigenvalues greater than 1.0 were extracted
Table 2
Distribution of two-digit SIC codes for sample ?rms
Industry JIT ?rms frequency Non-JIT ?rms frequency Sample frequency Sample per cent
20—Food 1 6 7 2.8
22—Textiles 2 3 5 2.0
25—Furniture and ?xtures 5 1 6 2.4
26—Paper and allied products 1 1 2 0.8
27—Printing/publishing 1 0 1 0.4
28—Chemicals and allied products 4 20 24 9.5
30—Rubber products 3 2 5 2.0
33—Primary metals 3 12 15 5.9
34—Fabricated metals 7 7 14 5.5
35—Industrial machinery 17 24 41 16.2
36—Electronics 24 37 61 24.1
37—Motor vehicles and accessories 6 5 11 4.3
38—Instrumentation 20 35 55 21.7
39—Other manufacturing 1 5 6 2.3
Totals 95 158 253 100.0
6
Further research is necessary to determine which JIT ele-
ments are most important for successful JIT implementation
and how these JIT elements interact. It is not possible to make
these determinations from the data gathered in the current
study. However, Table 6 indicates the degree to which the
individual elements have been implemented by the sample ?rms
and the largest implementation di?erences between JIT and
non-JIT ?rms. In addition, the factor loadings of the individual
JIT measurement variables reported in Table 3 assist in the
understanding of how these individual JIT practices are related.
Finally, each of the three identi?ed JIT constructs in the factor
analysis is signi?cantly correlated with the other two. Although
it is impossible to determine how they interact, it is apparent
that they do.
7
Total quality control is represented by two questions on
the survey: one is related to process quality and the other to
product quality.
718 R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735
from the analysis, representing 63% of the total
variance in the data.
8
The ?rst factor is a manufacturing component
that explains the extent to which companies have
implemented general manufacturing techniques
associated with JIT, such as focused factory,
group technology, uniform work loads, and multi-
function employees. Together these represent ele-
ments of a JIT manufacturing philosophy, although
individual elements of the factor may be adopted by
any high technology manufacturing ?rm.
The second JIT factor is a quality component
that examines the extent to which companies have
implemented procedures for improving product
and process quality. A reason for the association
between TQM and JIT is their common con-
tinuous improvement goals. Successful JIT imple-
mentation requires a high level of quality in
production. Although TQM can be adopted with-
out implementing JIT, it is unlikely that a JIT
manufacturing system can succeed without incor-
porating the underpinning tenets of TQM. Good
quality management and productive maintenance
are keys to JIT survival (Imai, 1998). Quality fre-
quently has been referred to as the cornerstone of
JIT (Banker et al., 1993a; Sim & Killough, 1998;
Swanson & Lankford, 1998; Young et al., 1988).
The third JIT factor identi?ed is one of uniquely
JIT practices that describe the extent to which
companies have implemented JIT purchasing and
kanban. The likelihood is low that companies who
are not fully committed to a JIT program would
adopt these practices. A description of the speci?c
survey questions that support these factors is
found in Appendix B. For results of the factor
analysis for JIT elements, refer to Table 3.
3.4. Independent variables
Nine constructs were selected to examine the
non-traditional performance measures and incen-
tive systems of manufacturing ?rms. Four of these
constructs, which represent performance measures
for evaluating manufacturing productivity,
measure hypothesis 1A: bottom-up data gathering
techniques; benchmarking for products, services,
and processes; frequency of measurements and
reports on quality; and manufacturing performance
measures. The ?rst three of these constructs were
de?ned in Ittner and Larcker’s 1995 TQM study.
The last construct is similar to one examined by
Durden et al. (1999) in examining the use of non-
?nancial manufacturing performance indicators in
a JIT environment. Three constructs related to
Table 3
Factor analysis (VARIMAX rotation) and factor loadings for JIT variables
a
Factor 1
JITMANUF
Factor 2
JITQLTY
Factor 3
JITUNIQUE
Cronbach’s alpha 0.831 0.946 0.684
Focused factory 0.740
Group technology 0.770
Reduced setup times 0.706
Productive maintenance 0.668
Multi-function employees 0.501
Uniform work load 0.731
Product quality improvement 0.917
Process quality improvement 0.902
Kanban system 0.820
JIT purchasing 0.825
a
All loadings in excess of 0.300 are shown. n=253.
8
All of the 11 elements loaded greater than 0.50 onto one of
the three constructs except for number 11, asking about the use
of ‘‘quality circles’’. It was evident from initial observations of
the survey responses that only a few ?rms (both JIT and non-
JIT) used quality circles. Thus, this question was eliminated
from further testing representing JIT.
R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735 719
performance incentives through compensation are
examined in hypothesis 1B: compensation ties to
non-?nancial performance; compensation ties to
quality and team performance; and compensation
ties to traditional pro?tability measures. The last
two constructs of the research analysis testing
hypothesis 1C are: communication of the strategic
plan to middle managers, ?rst-line supervisors, and
non-management personnel; and empowerment of
employees in decision making.
3.4.1. Factors for performance measures and
incentive systems
Thirty-nine items from the survey instrument
were evaluated to measure the nine performance-
measure and incentive-system constructs. To
reduce and summarize the collected data, these
survey items were subjected to a factor analysis.
Using the principal components method, the fac-
tor analysis revealed ten distinct factors with
eigenvalues greater than 1.0, which accounted for
73% of the total variance in the data.
9
The VAR-
IMAX rotation resulted in the following factors:
QLTYREV: The frequency with which quality
issues are measured and reported
to management strata.
COMPQLTY: The importance of quality and
teamwork in determining
compensation.
BOTTOM: The use of bottom-up data
gathering techniques such as
Pareto analysis, histograms,
and cause-and-e?ect diagrams to
evaluate operations.
COMPBDGT: The importance of adherence to
budget items in determining
compensation.
BENCH: The use of benchmarking to
evaluate operations.
PERFWASTE: The use of performance measures
related to waste and ine?ciency
in evaluating the manufacturing
system
STRPLAN: The extent to which employees
understand the ?rm’s strategic
plan.
PERFVEND: The use of performance measures
related to timeliness and vendor
performance in evaluating the
manufacturing system.
COMPNF: The use of non-?nancial measures
to determine compensation.
EMPOWER: The extent to which line managers
and non-management personnel
are empowered to make decisions.
A description of the speci?c survey questions that
support these factors is found in Appendix A. For
the results from the factor analysis, refer to Table 4.
3.5. Control variables
Four control variables (covariates) are included
in the regression testing. Firm size (SIZE) a?ects
most aspects of a ?rm’s strategy and success;
therefore a ?rm’s net sales are used to control for
?rm size. The net sales for each sample ?rm are
obtained from COMPUSTAT data. Whether a
?rm follows a more innovative strategy can a?ect
its willingness to make changes. Innovative ?rms
are more risky and generally more pro?table
(Capon, Farle, & Hoenig, 1988). Innovation
(INNOV) is measured by a ?rm’s response on the
survey instrument as to whether it is a leader or a
follower in product technology, product design,
and process design (Ittner & Larcker, 1995). The
industry in which a ?rm operates often a?ects its
competitive behavior and performance measures.
Thus, the industry for each ?rm, as identi?ed on
COMPUSTAT, is tested by the use of the ?rm’s
two-digit SIC code (SIC). Organizational structure
can in?uence a ?rm’s ability to be ?exible and
9
Seven factors loaded as expected (QLTYREV, BOTTOM,
BENCH, COMPQLTY, COMPNF, STRPLAN, and
EMPOWER). The construct for evaluating compensation
rewards from traditional ?nancial measures loaded onto two
factors. One was for compensation incentives related to var-
iances and budgets, which created the COMPBDGT variable.
The other factor was for compensation rewards related to tra-
ditional pro?tability measures. This factor made no signi?cant
contributions to any of the regression tests; therefore, it was
eliminated from the ?nal analyses. The expected single con-
struct measuring manufacturing performance loaded onto two
factors as per the factor analysis: PERFWASTE and
PERFVEND.
720 R.R. Fullerton, C.S. McWatters / Accounting, Organizations and Society 27 (2002) 711–735
Table 4
Factor analysis (VARIMAX rotation) and factor loadings for performance measures and incentive systems variables
a
Factor 1
QLTYREV
Factor 2
COMPQLTY
Factor 3
BOTTOM
Factor 4
COMPBDGT
Factor 5
BENCH
Factor 6
PERFWASTE
Factor 7
STRPLAN
Factor 8
PERFVEND
Factor 9
COMPNF
Factor 10
EMPOWER
Cronbach’s alpha 0.920 0.909 0.873 0.862 0.907 0.775 0.828 0.783 0.869 0.809
TM reviews quality results 0.719
TM reviews quality consequences 0.793
MM reviews quality results 0.836
MM reviews quality consequences 0.857
LS reviews quality results 0.819
LS reviews quality consequences 0.828
MM compensation—quality 0.773
MM compensation—throughput 0.703
MM compensation—teamwork 0.734
LS compensation—quality 0.854
LS compensation—throughput 0.827
LS compensation—teamwork 0.783
Use cause-and-e?ect diagrams 0.729
Use histograms 0.791
Use ?owcharting 0.658
Use Pareto analysis 0.778
Use scatter diagrams 0.748
Use SPC charts 0.652
MM compensation—variances 0.788
MM compensation—budget 0.704
LS compensation—variances 0.794
LS compensation—budget 0.780
Benchmarking of operations 0.844
Benchmarking of products 0.867
Benchmarking of delivery systems 0.836
Performance measures—downtime 0.691
Performance measures—scrap 0.721
Performance measures—rework 0.689
Performance measures—setups 0.685
MM understand strategic plan 0.791
LS understand strategic plan 0.793
NM understand strategic plan 0.776
Performance measures—on-time 0.514
Vendor performance—quality 0.866
Vendor performance—on-time 0.879
MM compensation—non-?nancial 0.867
LS compensation—non-?nancial 0.873
LS empowerment 0.713
NM empowerment 0.741
a
All loadings in excess of 0.300 are shown. n=253. TM=top management; MM=middle management; LS=line supervisors; NM=non-management.
R
.
R
.
F
u
l
l
e
r
t
o
n
,
C
.
S
.
M
c
W
a
t
t
e
r
s
/
A
c
c
o
u
n
t
i
n
g
,
O
r
g
a
n
i
z
a
t
i
o
n
s
a
n
d
S
o
c
i
e
t
y
2
7
(
2
0
0
2
)
7
1
1
–
7
3
5
7
2
1
make major operational changes. If a ?rm is
highly centralized, the employees will be much less
involved in decision making and organizational
changes than if it is more decentralized. Kalaga-
nam and Lindsay (1998) demonstrated how adapt-
ing more organic (decentralized) organizational
structures led to greater bene?ts from JIT adop-
tion. The organizational structure (STRUCTR) of
a ?rm is identi?ed on the questionnaire.
3.5.1. Factors for control variables
The six survey questions related to ?rm innova-
tion and organizational structure were reduced
and summarized using factor analysis. These six
variables converged into two anticipated distinct
factors with eigenvalues in excess of 1.0, account-
ing for 66% of the total variance in the data. The
VARIMAX rotation resulted in the following
control variables:
STRUCTR: The extent of centralization or
decentralization of a ?rm’s
organizational structure.
INNOV: The extent to which the ?rm considers
itself a leader in product and process
design and product technology.
A detailed description of the speci?c questions
that support these control variables is found in
Appendix B. Refer to Table 5 for the rotated fac-
tor solution.
3.6. Construct validity and reliability analysis
The factor solutions for the de?ned constructs
support the construct validity of the survey instru-
ment. Convergent validity is demonstrated by each
factor having multiple-question loadings in excess
of 0.5. In addition, discriminant validity is sup-
ported, since none of the questions in the factor
analyses have loadings in excess of 0.3 on more
than one factor.
10
Cronbach’s alpha is used as the
coe?cient of reliability for testing the internal
consistency of the constructs validated by the fac-
tor analysis. The alpha coe?cients for all of the
constructs are in excess of 0.7.
11
(The alpha coef-
?cients are included in Tables 3–5.) Overall, these
tests support the validity of the measures repre-
senting the constructs used in this study.
4. Research results
4.1. Descriptive statistics
One objective of this study is to capture the
degree to which the sample ?rms have imple-
mented JIT practices. On the survey instrument,
the respondents were asked to provide the degree
to which they were using 10 individual aspects of
JIT (scaled from 1 to 6). Respondents also were
asked to indicate whether their ?rm had formally
implemented JIT. Descriptive statistics depicting
the means for each of the individual elements,
along with the three JIT factors, and the total
combination of the JIT elements are shown on
Table 6. The data are presented in terms of the
total sample, the JIT sample ?rms, and the non-
JIT sample ?rms.
The ANOVA comparison of the means between
the JIT and non-JIT ?rms, found on Table 6,
consistently shows highly signi?cant di?erences
(p