JMAR

Volume Eight
1996

Benchmarking and Management

Accounting: A Framework
for Research
Dan EInathan
University of Southern California and Tel Aviv University

Thomas W. Lin
S.Mark Young
University of Southern California
Abstract: Benchmarking, the search for the best practices within and across
, . industries to improve performance, has become a popular management tool.
Increased global awareness and intensified international competition together
with the development of numerous benchmarking clearinghouses will provide
more opportunities for firms to engage in benchmarking. While much has been
written about benchmarking, few attempts have been made to integrate the lit-
erature and formalize a research framework which includes the role of manage-
ment accounting.
The goals of this paper are threefold. First, we integrate the literature and
propose a research framework in which antecedent, contextual and outcome
variables are developed. Second, we discuss the roles that benchmarking plays
within the management accounting function. Finally, we illustrate how research-
ers can apply the research framework when conducting studies to determine
whether benchmarking an activity based cost management system has been
successful and which variables play critical roles in leading to success.

INTRODUCTION
Over the past decade, the competitive business environment has dic-
tated major shifts in corporate strategies, organizational cultures and
designs. To keep pace with these changes, management has turned
aggressively to implementing innovative techniques such as total quality
management, lean manufacturing and reengineering. Management account-
ing in organizations also has taken on a mueh expanded role with the
development of activity based costing and management (Young and Selto
1991). Benchmarking has emerged as a central tool for these Innovations

The authors would like to acknowledge the valuable insights of Rick Anderson (Xerox
Corporation). Marty Russell (Russell Enterprises). Dick SnouJ^er (Hughes Aircraft Com-
pany), and especially Charles M. Conway (Los Angeles County Transportation
Commission). The helpful comments of Shannon Anderson. Jake Birnberg. Dick Chase.
Bill Ferrara. Mahendra Gupta. Oliver Kim, Ken Merchant. Frank Selto. Mike Shields,
and seminar participants at the Fifth Asian-Pacific Conference on International
Issues, Mexico City (November 1993). the First International Conference on
Contemporary Accounting issues, Taipei. Taiwan (January 1994). and the Fourth
Management Accounting Research Conference, San Diego. CA (April 1994) are ap-
preciated.
38 Journal of Management Accounting ResearcK 1996

as organizations attempt to emulate the best practices within and across

Industries (EInathan and Kim 1995). While benchmarking is used widely
in practice, the vast majority of the evidence gathered about the process
and its benefits, thus far. Is anecdotal and not well integrated or under-
stood. In this paper, our goal is to provide more structure for systematic
inquiry by integrating the benchmarking literature from various disciplines
and developing a framework for research.
Background on benchmarking in general and an overview of the
benchmarking process are discussed in the next section of the paper. In
the following section, we develop the framework for benchmarking research
that discusses antecedent, contextual and outcome variables that should
be considered when researching benchmarking. Then, the interaction of
benchmarking and management accounting is discussed. Next, the frame-
work is applied to Illustrate how researchers could study the benchmarking
of an activity based cost management system of one organization by an-
other. The final section of the paper presents a summary and discussion.

BACKGROUND .
Benchmarking Practice
Benchmarking in various forms has been conducted for many years,
but only recently has the proeess been named and systematized. Whtle
benchmarking can be compared to other types of standard-setting pro-
cesses, there are differences which distinguish it from traditional standard
setting. Organizations engage in benchmarking for several reasons: for con-
tinuous improvement of internal operations, to become more externally
competitive, or for organizational survival. Unlike traditional standard set-
ting, benchmarking has a strong external orientation. For instance,
benchmarking is undertaken when an organization believes that others
outside the organization have superior knowledge about processes, tech-
nology, quality or costing methods that go beyond their own current state-
of-the-art. As is often the case, seeking external help is cost-beneficial since
formal, within-firm analysis is not always possible, or is prohibitively
expensive.
Industry surveys are a form of benchmarking which many times in-
volve the collecting and analyzing of marketing data. Usually, though, only
aggregate or average results are available. Industry associations also col-
lect information from member firms, such as market data and wage and
employment level information used for labor negotiations and lobbying
purposes (Kirby 1988). Associations focused on specific organizational func-
tions, such as procurement or facilities management, provide another form
of benchmarking. While associations engage mainly in technical training
and professional certifications of their membership, many publish an an-
nual review of key statistics of their membership and the industry, which
can be used in the early stages of a benchmarking effort.
The Xerox Corporation is credited with advancing benchmarking in the
United States as an important management tool. In the late 1970s, Xerox
lost its leading position in the copier business to IBM, Kodak and several
Japanese firms. The loss of leadership manifested itself in technical areas,
as well as marketing, pricing and customer support. In 1979, Xerox deter-
EInathan. Lin and Young 39

mined that it needed to study its manufacturing cost structure and pricing
strategy. Xerox's benchmarking effort found that competitors were pricing
and selling machines at its manufacturing cost. The results of that study
led Xerox to expand benchmarking into other areas. Since then, Xerox has
trained scores of employees in developing and implementing benchmarking
for all the functional areas of its business units throughout the world, and
has assisted in training employees from many other firms in performing
similar activities (see Camp 1989; McCamus 1991 and Kearns and Nadler
1992 for more description). Cooperative benchmarking emerged when Xerox
recognized that it was more efficient to collaborate with the benchmarked
companies. Additionally, it was easier to create benchmarking partner-
ships with companies who were not direct competitors.
Benchmarking gained more momentum with its inclusion as a compo-
nent of the Malcolm Baldrige National Quality Award, Numerous organiza-
tions and private consultants now offer their services in setting up
benchmarking programs and several major institutions have started
benchmarking initiatives to pool their resources to act as information clear-
inghouses. Organizations such as the Consortium for Advanced Manufac-
turing International (CAM-I) have been pioneers in bringing large numbers
of firms together to benchmark a variety of technical functions including
aspects of management accounting systems. For instance, it was CAM-I
that independently developed their own cost management system (CMS)
through a consortium of firms, government organizations and universities
(Berliner and Brimson 1988).
In 1990, The Strategic Planning Institute (SPI) in Cambridge, Massa-
chusetts established the SPI Council on Benchmarking with about 50 mem-
bers (Kharbanda 1993), and the American Productivity and Quality Center
in Houston established the International Benchmarking Clearinghouse (IBC)
in 1992 with more than 199 members (Main 1992). The IBC is offering
benchmarking training and library resources on numerous functional ar-
eas and products, including the controllership function. In 1993, the Insti-
tute of Management Accountants (IMA) established a Continuous Improve-
ment Center (CIC) which set up a benchmark database for financial man-
agement functions to provide services to member firms by assisting them
in identifying "best practices" and in improving business processes (IMA
1993).
Internationally, Japanese firms have engaged in informal benchmarking
for many years, and the Canadian Federal Department of Industry, Sci-
ence and Technology established the Interfirm Comparison Program in the
early 1970s to facilitate benchmarking (Rivest 1991), Other countries also
have adopted benchmarking as a prerequisite for quality certification such
as Japan (the Deming award), Europe (the ISO 9000 Standard), Hong Kong
(the Hong Kong Benchmarking Clearinghouse) and Canada (the Award for
Business Excellence) (SMAC 1993). In 1992, the American Quality Foun-
dation and Ernst & Young conducted a survey of 580 firms worldwide in
four industries—computers, autos, hospitals and banks—and found that
31 percent of the U.S. firms regularly benehmarked their products and
services with only seven percent saying that they did not engage in
benchmarking (Main 1992).
40 Journal of Management Accounting ResearcK 1996

The Benchmarking Process

The process of benchmarking has been modeled in a variety of forms
by practitioners—companies and consultants. While these process models
may be slightly different from each other, they all have common elements.
Spendolini (1992) conducted a systematic comparison of 24 models, and
identified a five-stage generic benchmarking model.
In the first stage, the organization must decide on which activities and
functions will be benchmarked. Next, a benchmarking team is coalesced.
In the third stage, benchmarking partners are identified. In stage four, an
extensive analysis of the organization and its activities is undertaken and
data on inputs, outputs, flows and costs are identified and analyzed.* A
"benchmarking (or performance) gap" is determined by comparing the
organization's own performance with the "best" performance that emerges
from the data, and the organization establishes functional goals—new prac-
tices to be adopted, the scope of the change, cost targets, and the time-
table for implementation. In the final stage, the organization takes action.
Benchmarking may provide evidence of successful alternative practices
or approaches in other organizations, but the adopting organization still
has to develop a "benchmarking culture" and overcome hesitation and re-
sistance from within to implement a new approach or process. Observa-
tions by many practitioners and academics suggest that gaining the sup-
port of top and mid die-management for change is a critical step towards
successful implementation. Often, the organization has to undergo a sig-
nificant cultural change, such as being willing to disclose what was once
considered proprietary information regarding products and activities to their
benchmarking partners, who may include current or potential competi-
tors. The next section identifies the variables that organizations should
consider to manage benchmarking successfully. These variables provide a
framework from which research on benchmarking can proceed.

A FRAMEWORK FOR BENCHMARKING RESEARCH

As mentioned earlier, benchmarking is a process by which an organi-
zation targets key areas for improvement, studies the best practices of
others, and implements processes and systems to enhance its own
performance. The process of continuously comparing the organization's
performance on critical aspects of operations against the best-in-industry
(direct competitors) or best-in-class (companies recognized for superior per-
formance in certain functions), determines which activities and costs should
be targeted for improvement. When applied properly, benchmarking can
spark new ideas and approaches which may come from outside industries.
The research framework developed via an integration of literature (see fig-
ure 1), consists of three sets of variables that organizations desiring to
benchmark (hereafter benchmarkors) others (hereafter benchmarkees)
should consider before embarking on a benchmarking project. This frame-

' When the process is cooperative, data collection requires several additional steps;
benchmarking partners should be identified and approached, information sharing
agreements should be worked out. and data comparability issues have to be resolved.
EInathan and Kim (1995) analyze how much information is gathered by individual firms
about potential partners before deciding whether, and with whom, to benchmark.
EInathan. Lin and Young 41

•a i2
a tf) 6 «
•a
u Hi 5

q |.5r [0
.g 3 u V.a c
a a.
c (U O -M JJ
a u.
M« -t-^ 4-J t4_

« -s a g- ^ -a CO
t) 2 S °
•~. "J
cfi DU9 9 °-§ « £ £ feb

I ^s: cn m
- i oi cri o)
a: d X3 Q

e
i
a .4J
V
a
«tfl 3g
o d .2 o c
uo 5 ra
1^ pi) R
3 S ra I-" .s
t°§ o C u. c3 O
ffl o o a
o >> "> O ^ CJ S
OJ
o t: c S- B o
I) —'
bb
ro CO - • w o OJ
-^ « n Q
Is "S ^
^ 3 »j #
3u
Ui Ui
-« <N CO
42 Journal of Management Accounting Research. 1996

work provides researchers with a way to study and assess the relative suc-
cess of specific projects in which benchmarkors engage. The three sets of
variables are: (1) antecedent variables that set up the necessary precondi-
tions for success; (2} contextual variables that may modify the specific na-
ture of benchmarking, and (3) outcome variables by which the organiza-
tion gauges overall effectiveness of a benchmarking effort.^

Antecedent Variables
The following review of the literature identifies three general sets of
antecedent variables which organizations should consider before
benchmarking occurs. Those are: (1) results of a preliminary con^peiiiive
analysis, [2) degree of organizational commilment, and (3) prior benchmarking
experience. There are several factors under each of these three variables.

Results of a Preliminary Competitive Analysis

The impetus for benchmarking can arise from a preliminary competi-
tive analysis either internal to the organization or externcd to the organiza-
tion and may occur in varying degrees of formality.
Analyses internal to the organization can occur for at least three not
necessarily independent reasons. First, an organization could suspect that
a product docs not meet its design specifications. Thus, an assessment of
performance in relation to a target or a goal could occur. Second, the orga-
nization could be guided by a philosophy of continuous improvement that
leads them to assess how their products and processes are performing.
Third, the need for analysis could result for less systematic reasons such
as management intuition that their products, processes or strategies are
not competitive.
In many cases, analyses external to the organization are conducted.
For example, data from industry rankings of performance in trade journals
or industry comparisons via publications such as Consum.er Reports are
used. Another critical source of external information is customer and stake-
holder feedback regarding products, processes and strategies.

Degree of Organizational Comm.itm.ent

Like all new management innovations, obtaining senior management
support, is critical for the success of benchmarking. Shields and Young (1989)
note that high level support is necessary for legitimizing the innovation in
the eyes of skeptical managers and employees and for increasing a
subordinate's commitment to the innovation. Senior management support
has long been recognized as important for information systems implemen-
tation use (Lucas 1981; Ginzberg 1981; Jarvenpaa and Ives 1991) and
other management innovations (Shields and Young 1989).
Senior management support can manifest itself in numerous ways. First,
it can aid the benchmarking team by giving them the authority necessary to
motivate employees to take the benchmarking project seriously. Second, se-
nior management can authorize the funding which is required for training in
benchmarking, and for the costs of the benchmarking process itself. Such
investments will increase organizational experience and allow for better choices

^Variables in the framework are identified in the text via italics.

EInathan, Un and Young . 4 3

with respect to benchmarking venues, some of which may be quite expensive

and require a protracted process. Third, senior managers often have rela-
tionships with other firms and associations and can use these ties to solicit
others' participation in a benchmarking program. In many discussions of
benchmarking, the need for top management support, involvement and com-
mitment is imperative (SMAC 1993, 10).
Every management initiative requires developing a dear set of objec-
tives which serve as focal points of a project and allow the organization to
monitor its success. For example, in implementing new information sys-
tems, Ginzberg (1981) finds that clear pre-implementation expectations or
financial plans tend to have a greater chance of being successful. Shields
(1995) finds a similar result regarding activity based costing implementa-
tion. Further, as we suggested earlier, developing measures to assess the
effects of a benchmarking effort are critical for evaluation. Bean and Gros
(1992, 37) argue that it is critical for benchmarking teams to have a clear
idea of the organizational mission and customer needs they arc serving
before benchmarking commences. Without a clear set of objectives evalua-
tion is problematic.
The level of commitment to benchmarking has to be a long-, rather
than a short-term commitment. Long-term commitment is necessary, as
most significant organizational changes take up to three years, and some
even take longer (Shields and Young 1989), so the benefits of benchmarking
may take time to be realized. Organizations should begin to observe changes
in the way that people are thinking about issues and problems and mak-
ing decisions and processing information in the short-run. However, fi-
nancial gains (if they are obtained) often take longer to materialize since
the rippling effects of decisions and actions take time to be translated into
dollars.
Shields and Young (1989) point out that implementation of new cost
management systems is most easily accomplished when organizations have
a functional, well-defined culture that empowers employees. Such empow-
ering organizational cultures include allowing participation of employees in
decision making and more autonomy of action (see Birnberg et al. 1990) as
well as provide opportunities for organizational learning (Scnge 1990). Suc-
cessful benchmarking requires a work environment in which employees
can make suggestions and have the authority to make change.

Prior Benchmarking Experience

As noted above, strong organizational commitment will result in in-
creased efforts to improve the extent of benchmarking experience of an or-
ganization through greater commitment of resources for training and re-
tention of benchmarking experts. Greater organizational experience would
improve the ability within the organization to identify appropriate areas to
be benchmarked, and to employ the most effective information gathering
and sharing methods. Furthermore, experienced organizations also have
the ability to choose the best set of benchmarking partners, given the area
to be benchmarked and benchmarking method. In addition, the experi-
enced organization will be considered an attractive partner by others, thus
providing it another opportunity to further improve its own results. For
example, Xerox's reputation as an experienced and sophisticated
44 Journal of Management Accounting Research. 1996

benchmarking organization makes it an attractive benchmarking partner,

and Xerox can use this leverage when it chooses firms with whom to
benchmark.
Accumulated organizational experience in any area is an attribute which
is expected to contribute to the success of new implementations in that
area. Argyris (1977) and Raymond (1985) find that previous experience
with information systems led to more successful projects. In our frame-
work, we argue that the preparation for benchmarking projects and the
ability to effectively interpret the benefits from learning improves over time.
As the organization develops benchmarking experience, the time required
for these activities will be shorter (thus lowering cost), and the actual pro-
cess of benchmarking will be more efficient (thus increasing likelihood of
success). The costs of all activities depend on the level (within the organi-
zation) of the personnel involved and their benchmarking skills.
Commitment to long-term benchmarking success is evidenced by the
permanent employment of benchmarking professionals. Generally,
benchmarking is an iterative process with participants defining and ac-
counting for the practice, while quantifying operating results for the pur-
pose of comparison. It is not unusual to have unique, firm-specific ac-
counting for internal purposes, and thus, common definitions need to be
derived by benchmarking organizations for both the practices and their
accounting measures. In practice, there is a need to resolve the conflict
between participants who emphasize quantitative and financial measures
of comparison, and participants who emphasize nonfinancial and/or quali-
tative operating measures. In these situations, a strong and experienced
coordinator can adjudicate such situations by developing a balanced set of
measures.
Organizational success in program implementation includes the avail-
ability and extent of training to employees who manage and run the pro-
gram. In implementing cost management systems. Shields and Young (1989)
and Shields (1995} note that a lack of training often will lead to the failure
of the implementation. Thus, we argue that the actual level of training in
benchmarking that employees receive is critical for successful projects, as
they improve employees' understand of the entire process, its benefits and
limitations. ;

Contextual Variables
Three general sets of contextual variables should be considered when
engaging in benchmarking. These are: (1} scope and areas selected: (2) irifor-
mation gathering and sharing methods; and (3) partner{s) selected. Choices
on many of these variables will be made based on the results of the prelimi-
nary competitive analysis.

Scope and Areas Selected

The scope of analysis is critical since many firms enter into
benchmarking without a clear idea of what they want to accomplish. Initial
efforts at benchmarking should clearly identify a manageable proeess or a
product that needs improvement. Attempting to engage in a benchmarking
effort that is unreasonable in size can reduce motivation and performance.
As is true of many types of management innovations, the first attempt to
EInathan. Un and Young 45
I

implement an innovation will significantly influence future implementa-

tions. If the scope of the first benchmarking process is too large and it
fails, then the tone may be set unfavorably for future efforts.
Recently, it was pointed out that, "In selecting benchmarking func-
tions, it is more beneficial to make major improvements to the firm's weak-
est functions rather than making smaller improvements to the areas in
which the firm is already strong" {Industry Week 1990, 39). This suggests
that first a value chain analysis needs to be conducted together with a
functional analysis to determine the key area(s) for study and the signifi-
cance of the study for each area(s). These analyses will identify the weakest
area by function, as well as the function's impact from value chain and
performance perspectives. Typically, the greater the benchmarking gap—
the difference between the organization's performance and the benchmarked
observations in a critical area—the greater the benefit from a benchmarking
study.

Information Gathering and Sharing Methods

Two dimensions relating to information gathering and sharing emerge
from the literature. These are: (1) type of information which benchmarking
organizations collect, and (2) method of information collection.
First, there are three major types of information on which firms inter-
ested in benchmarking can focus. The information gathered includes physi-
cal and cost attributes of the product or process under consideration. Prod-
uct Benchmarking is the long standing practice of carefully examining other
organizations' products. Not only is the cost of competitors' products
examined, but often methods such as reverse engineering are used. For
example, each year Toyota purchases 160 automobiles from around the
world and analyzes each one, component by component.-' Functional (Pro-
cess) Benchmarking is the study of other organizations' practices and costs
with respect to functions or processes, such as assembly or distribution.
Strategic Benchmarking is the study of other organizations' strategies and
strategic decisions (Miller et al. 1992) such as why a low-cost versus dif-
ferentiation strategy was chosen. Most pertinent to this paper is the idea
that costing systems, such as activity based cost management, are now an
integral part of organizational strategy for many organizations and can pro-
vide a competitive advantage. Thus, many organizations are now involved
in benchmarking each other's costing systems such as through the IMA's
CIC. This issue will be discussed in depth later in the paper.
Second, there are two major methods of information collection for
benchmarking. The most common can be described as Unilateral (Covert)
Benchmarking, in which companies independently gather information about
one or several other companies that excel in the area of interest. Data can
be obtained from industry trade associations, clearinghouses of informa-
tion sueh as the American Productivity and Quality Center's International
Benchmarking Clearinghouse or through other means.** One popular ap-

^ This information comes from an interview with the General Manager of Accounting at Toyota,
Japan (Toyota City. Japan. November 1994).
•^ We acknowledge that industrial espionage is another method of information gathering, but
since the activity is illegal, we do not consider it in our discussion.
46 Journal of Management Accounting ResearcK 1996

proach is to study information pertaining to the Malcolm Baldrige Award

winners as they arc required to share their information with other firms
(Miller 1992).
A seeond method is called Cooperative Benchmarking. Cooperative
benchmarking involves the voluntary sharing of information through mu-
tual agreements. From the process, participating companies hope to gain
a better understanding of their own operations and information on how
others gain leadership positions in performing similar functions or pro-
ducing simiiar products. The major advantage of cooperative benchmarking
is that information sharing occurs both within and across industries. Within
cooperative benchmarking there are three sub-categories which we call
Database, Indirect/Third Party, and Group.
Companies that use the Database approach (and for whom it is avail-
able) typically pay a fee and, in return, gain access to information from a
database operator who collects and edits the information prior to making
it available to users. In most cases, no direct contact with other firms is
established, and the identity of the source of the data often is not revealed.
The database at the CIC has developed along these lines but has since
added a "best practices" sharing mechanism. The Indirect/Third Party ap-
proach involves the hiring of an outside consultant to act as a liaison among
two or more firms which arc engaged in benchmarking. The consultant
supplies information from one party to the others, and handles all com-
munications. Often, the consultant participates in the selection of part-
ners. For example, Andersen Consulting, Deloitte & Touche, A.T. Kearny,
Towers Perrin, and Price Waterhouse have set up groups of companies
that benchmark each other regularly. Since the members may be competi-
tors, these groups pass information through the consultant so that
members can learn about best practices but the actual sources remain
confidential.
Participants following the Group approach meet openly to discuss their
methods. They coordinate their efforts, define common terminology, visit
each other's sites, and generally have a long-run association. For example,
Xerox has organized several such groups and acted as a group coordina-
tor. We assume that firms will engage in any benchmarking efforts only if
the benefits outweigh the costs. It follows that if firms do not believe that
they can trust each other, they will perceive the benefits to be low or the
costs to be high, and, in any case, will not engage in benchmarking. Typi-
cally, firms that engage in cooperative benchmarking abide by a code of
conduct which they agree upon prior to the beginning of the study.
Of the three methods listed above, the database and the indirect/third
party approaches involve little or no direct contact with the other organiza-
tions being benchmarked. Cooperative benchmarking that uses the group
approach, though, allows each participant to assess the level of commit-
ment of the other participants. As in most interactions, direct contact of-
fers the opportunity for better understanding of the other parties involved
and usually is the most effective benchmarking method. However, this
method also is the most costly to implement and firms need to evaluate the
cost-benefit tradeoffs.
The database method has the advantage of being able to assemble a
large amount of information in one place. However, insights regarding what
EInathan. Lin and Young 47

the data mean for the firm and how the information can be used often are
not available. Without detailed discussion, the database approach can only
provide rough guidance for an organization. The information provided is
relatively nonspecific (mean measures across the sample) and the timeli-
ness of the data is in question. Of the cooperative forms of benchmarking,
the database method usually is the least effective, but also the least costly.

Partneris) Selected
Currently, the knowledge accumulated in practice does not provide
definitive answers to questions about: (1) size of partners; (2) number of
partners; [3) relative position of the partners within and across industries;
and (4) degree of trust among partners.
Some practitioners argue that benchmarking partners should be of com-
parable size (Tonkin 1991), while others claim that this is not a critical
requirement. All else being equal, we would expect, within an industry,
that firms of similar size are more likely to benchmark than those of differ-
ent sizes, although this may not be the case across industries.
The Industry from which partners are selected appears to be important
as well, but optimal choices are unclear. When firms benchmark func-
tional areas, they minimize the costs of proprietary information disclosure
by benchmarking with companies that are not their direct competitors in
the product market. However, on some occasions, companies actually re-
quest that competitors be brought into the group, arguing that in special-
ized areas (e.g., "clean room" maintenance) the only meaningful bench-
marks will be direct competitors. In this situation, the tradeoff between
costs (e.g. disclosing proprietary information to direct competitors} and
benefits (e.g. acquiring relevant information from direct competitors) has
to be considered favorable.
With regard to the number of partners in a project, the law of diminish-
ing returns applies. Benchmarking partners benefit from initial increases
in the number of participants, but as these numbers increase further, is-
sues of coordination, timeliness, and concern over proprietary information
disclosure will dominate. For example, definitions of operating and cost
items vary across companies, so higher costs are associated with the
need to integrate a larger number of alternative definitions and
operationalizations.
EInathan and Kim (1995) model how cooperative benchmarking groups
are formed among a set of firms that possess different amounts of techno-
logical information contained in their operations. They show that there is
a unique equilibrium group structure characterized by grouping among
firms with similar amounts of technological information. In a compara-
tive, static analysis, EInathan and Kim (1995) show that group size and
the number of firms participating in cooperative benchmarking tend to
increase as learning becomes more efficient, as technology becomes more
complementary, as firms' technological information uniformly increases,
and as the cost of benchmarking is reduced. They argue that today's chang-
ing business environment is likely to encourage cooperative benchmarking
and increase group size because increased competition and technological
progress in information processing increase benchmarking benefits rela-
tive to costs.
48 Journal of Management Accounting ResearcK 1996

Another factor is the relative position of the organization within an in-

dustry. We contend that, within an industry, industry newcomers and those
whose performance on leading indicators has declined, are more likely to
seek a wider variety of benchmarking partners than those who are estab-
lished industry leaders. Those who are industry leaders, though, have dif-
ferent incentives regarding benchmarking. Because they are on top, in-
dustry leaders often receive praise from the business press if they appear
to be helping less capable, or less advanced, organizations. Thus, it is in
the industry leader's best interest to be a supplier of information for other
firms desiring to benchmark. However, when industry leaders are looking
for other organizations with which to benchmark, they can afford to be
highly selective. A leading organization in one industry may be quite will-
ing to benchmark with leaders from other industries, believing that each
will gain from the other and allow them to maintain their lead within their
own industries.
Another critical variable is the degree of trust among parties. From the
benchmarker's point of view, obtaining truthful and timely information
seems to be absolutely essential if benchmarking is to be successful. For
instance, when it comes to benchmarking management accounting sys-
tems, firms will be very hesitant to discuss their cost structures with "just
any" firms without the proper incentives. Most organizations, including
industry leaders, operate on a quid pro quo basis. Ultimately, this means
that the benchmarkor would give the bcnchmarkee some information that
it can use. However, there is never a guarantee for the benchmarkee that it
will be able to realize any value from the information provided by the
benchmarker. In considering this dilemma, we expect that benchmarkees
often provide dated information to benchmarkors whom they do not yet
trust. While the "lagged" information may be useful to the benchmarker
and lead it in the right direction, the benchmarkee does not believe that it
realized a loss if the benchmarker does not reciprocate later with useful
information. In this win/win situation, the benchmarkee gains a potential
benchmarking partner without divulging the information from which it may
derive its competitive advantage.

Outcome Variables
The decision to benchmark is not entirely different from other efficiency-
oriented decisions made by organizations and individuals within them. For
benchmarking to be suecessful, it is crucial to understand the incentives
for benchmarking organizations, as decision makers will only select courses
of action where the expected benefits exceed the expected costs. Defining
the most appropriate performance measures (benefits) is critical for firms
to affect change (see figure 1). For example, Kaplan and Norton (1992. 74),
writing about the balanced scorecard as a comprehensive performance
measurement system, comment that, "Benchmarking procedures arc yet
another technique companies use to compare their performance against
competitors' best practices."

Benefits
We suggest that the overarching measure is whether benchmarking
objectives were met. Attaining benchmarking objectives requires setting
EInathan. Un and Young 49

up performance measures. Performance measures can be divided into fi-

nancial measures or nonfinancial measures q/" outcomes or benefits. Non-
financial benefits can be assessed with cither (or both) nonfinancial quanti-
tative measures or nonfinancial qualitative measures. In most instances,
nonfinancial variables will be affected first by benchmarking efforts.
Nonfinancial quantitative measures such as improved quality (regarding
engineering specifications, for instance), greater yield, reduced defectives,
increased speed to market, faster on-time delivery, increased functionality
often will result before financial gains can be measured. Equally as impor-
tant, however, are nonfinancial qualitative measures of change and im-
provement. For example, benchmarking the ways that others do things
often can lead to changes in employee decisions concerning ways to work
or how to solve problems. Other qualitative measures of benefits can arise
such as increased n^otivation and satisfaction, improved cooperation and
coordination among workgroups and employees, better understanding of
operations and support systems, and an expanded opportunity set due to
the additional information derived through benchmarking. While these ef-
fects are more difficult to assess compared to obtaining quantitative mea-
sures, a plethora of psychometric methods arc available (Nunnally 1983).
Financial measures such as reduced cost, increased sales or increased
income usually will result after nonfinancial measures have been affected.
Since most financial gains may take a significant amount of time, it is
critical that organizations monitor the nonfinancial variables in the short
term. Simply judging the effects of a benchmarking effort in the short term
based on financial indicators may often lead to premature abandonment.
One key difficulty with assessing whether benchmarking, or any man-
agement innovation for that matter, will be successful is being able to control
for other contemporaneous changes in the benchmarking unit. For instance,
an organization may be employing a total quality management program as it
begins to alter processes as a result of benchmarking. Isolating the effects of
benchmarking in such a situation is difficult and will require multiple mea-
sures and longitudinal analysis of the organizational unit.

Costs
On the cost side of benchmarking, we distinguish between traceable
costs and non-traceable costs. Traceable costs include out-of-pocket ex-
penditures such as the staff time devoted to benchmarking from the prepa-
ration for the study to the actual time used to implement it, and the time
required for analysis and interpretation. Included are the costs of addi-
tional time and effort needed to coordinate among participants in order to
obtain a comparable set of data. More problematic to evaluate are costs
which may be significant, yet not easily traceable or measurable such as
those associated with the cultural change in the organization and the po-
tential resistance to change.

BENCHMARKING AND MANAGEMENT ACCOUNTING

In this section, we discuss the interrelationship between benchmarking
and management accounting. Then, we illustrate how researchers can use
the framework developed in this paper when conducting a study to deter-
mine whether benchmarking an activity based cost management system
has been successful.
50 Journal of Management Accounting ResearcK 1996

Roles for Benchmarking in Management Accounting

The role of management accounting across a wide variety of organiza-
tional contexts has been expanded greatly since the mid-1980s.
Benchmarking has emerged as a central process through which manage-
ment accounting systems and specific types of information used and gen-
erated by the system can be compared and improved.
In many organizations, management accounting falls under the aus-
pices of the finance function. Within this domain, benchmarking is being
used in two ways. Most recently, benchmarking has been directed towards
planning and budgeting processes, billing, accounts receivable, account-
ing systems development, pajToll, credit and collections, financial analysis
and internal auditing. Benchmarking these functions has been occurring
over the past three years (Malcolm 1995).
The other, and more established, use of benchmarking is related to the
operations level of both manufacturing and service organizations. For in-
stance, activity-based cost management (ABCM) is being made more use-
ful to organizations by information derived through the benchmarking of
functions, products or processes as described in Coburn et al. (1995) Ex-
ternal benchmarks and best practices are identified within and across in-
dustries and incorporated into the design and functioning of ABCM sys-
tems (Atkinson et al. 1995; Horngrcn et al. 1993). Below, we expand the
discussion of the association between benchmarking and ABCM. The im-
portance of benchmarking as part of the current management accounting
environment provides opportunities for systematic studies of benchmarking
given the number of strategic information issues involved. ,|

Benchmarking An ABCM System

In this section, we illustrate how our research framework could be ap-
plied to a study of benchmarking an ABCM system. Developing an ABCM
system is done internally, often based on historical cost information, and
most of the discussion of ABCM systems in the literature seldom refers to
obtaining external benchmarks for cost pools, activity centers and drivers,
value-added versus non-value-added costs or burden rates, or to the pro-
cess by which firms can benchmark their costs with others (see Coburn et
al. 1995 for an exception}. We argue, though, that the information pro-
vided by an internal ABCM analysis alone is not sufficient, as the bench-
marks derived from such a process may be suboptimal and noncompeti-
tive. Thus, the organization needs to seek appropriate external benchmarks
(depending on the function, product or process under analysis) in order to
improve performance and to use ABCM information more effectively.
Below, we will illustrate how our research framework can be used as-
suming that both the benchmarker and the benchmarkee are involved with
the cooperative, group form of information gathering.^ In studying the
other forms of information gathering and sharing methods may require a different mode of
inquiry to determine whether the benchmarker is using a particular form effectively. For
example, Coburn. et al. (1995) describe benchmarking an ABCM system using the database
approach to gather information. In this case, we expect partner(s) selection aspects to be of
secondary importance as the benchmarker cannot choose the participants to the database.
Also, costs and benefits will be different with this approach as discussed earlier in the
paper.
EInathan. Lin and Young 51

benchmarking process with our framework, initially, researchers should

determine whether the benchmarker has managed the key antecedent vari-
ables of the degree of organizational commitment and the extent of prior
benchmarking experience. This can be accomplished through structured
interviews and well-designed surveys. Since ABCM often will alter the per-
formance evaluation and measurement systems of an organization, deter-
mining whether these conditions have been met is critical. Regarding orga-
nizational commitment, researchers should determine the highest level of
senior management support and whether this support has filtered down to
those involved with the specific organizational function, process or prod-
uct being benchmarked. Also, researchers should judge whether the orga-
nizational culture is one that will empower employees to encourage change
and improvement, and take and reward initiative. A determination of
whether a clear set of objectives for the project exists, and if there is a long-
term commitment on the part of the benchmarking organization to the project
also is necessary.
The extent of benchmarking experience, in general, and ABCM in par-
ticular, and the existence of an experienced coordinator in the organization
are essential to ensure that both insights from previous studies and ad-
ministrative strength are evident. Researchers should evaluate the extent
of training of those involved with the benchmarking effort through the
amount of previous technical and practical experience.
In studying the scope and areas selected for benchmarking, ascertain-
ing the key areas for study (such as key activity centers, activity param-
eters or specific cost pools) and the significance of the study for each area
(such as whether all activity centers need to be redesigned) must be done.
Measuring the benchmarking gap between current operations, product or
process characteristics may be accomplished using the benchmarker's
methods or an independent set of methods that the researcher develops.
For accounting researchers, benchmarking research on ABCM focus-
ing on the methods of information gathering and sharing probably is the
most insightful aspect of the research study. As mentioned earlier, since
an organization's cost structure is a key source of competitive advantage,
we would not expect the benchmarkee to casually reveal their current cost
structures readily to the benchmarker. Rather than reveal actual cost data
to a group, organizations often are more willing to discuss the overall struc-
ture of activity centers, cost pools, or the cost drivers that they have cho-
sen for specific activities. While specific data may be provided in the form
of databases, confidentiality often is maintained.
We would expect cooperative information sharing through the group
mechanism to yield very useful data since each organization believes that
there are mutual benefits to sharing. Initially, researchers should focus on
understanding the criteria for partner selection such as size, number of
partners, and assessing the degree of trust between parties. Next, investi-
gating the incentives to reveal information truthfully and the organization's
strategy and relative position in its industry may determine how it strategi-
cally reveals or withholds costing information. Clearly, research designs
involving multiple organizations could be used to determine which method
of information collection was deemed the most useful. >
52 Journal of Management Accounting ResearcK 1996

Determining and classifying the types of outcome variables used by the

benchmarker is critical for the researcher because there are many ways to
determine benchmarking success. Strictly focusing on financial performance
measures often is too short-sighted as any significant change resulting
from benchmarking will take time to become evident. Both nonfinancial
quantitative and norifinancial qualitative measures most likely will show
results before financial measures. For instance, measures such as increas-
ing a product's functionality, speed to market, or improving manufactur-
ing and service processes with regard to quality, yield and defects can be
determined earlier than in changes in costs. Assessing changes in the ways
that employees make decisions or whether motivation, cooperation and
coordination have increased or an expanded opportunity set has become
clear are intermediate steps that researchers can assess using reliable psy-
chometric scales.
Traceable costs such as those directly out-of-pocket should be avail-
able. However, there will be a number of non-traceable costs such as the
cost of lost productivity, coordination problems, and other time-involved
delays that may make the costs grow. Overall, researchers can conduct
rigorous cost/benefit analyses to conclude whether the effort was worth-
while, which again may require interviews and structured surveys.
A variety of research approaches and methods can be used to test spe-
cific relations within the framework. For example, depending on the
nature of the research question, long-term field studies, laboratory experi-
ments, carefully designed surveys, analytical modeling, experimental mar-
kets research, or combinations of these methods can be undertaken (see
Birnberg et al. 1990 for a discussion). Statistical methods, such as LISREL
or time-series analysis, could be employed when exploring causal or tem-
poral relations. On a larger seale, management accounting researchers may
be able to test hypotheses about benchmarking, derived from our frame-
work, with large samples when databases become available. Currently, there
is no cost management equivalent of the CRSP or COMPUSTAT tapes, but
benchmarking clearinghouses are beginning to gather data on cost infor-
mation and eventually these databases may allow even more research on
benchmarking of management accounting systems.

SUMMARY AND DISCUSSION

In this paper, we have integrated the literature on benchmarking and
developed a research framework to provide guidance for more systematic
research. Benchmarking is playing an increasingly important role in ad-
vancing organizational competitiveness. Variations in practice have led to
a number of unanswered questions which have created a number of op-
portunities for research. Currently, very little empirical evidence exists to
address these questions.
To date, management accounting research on benchmarking has not
been widespread; however, from our review of the literature there appears
to be a number of researchable ideas that would benefit from scientific
analysis. We presented a research application derived from our framework
to study how researchers might assess the implementation of a
benchmarking study by a benchmarker. Within this plan, there are a num-
Elnathan. Lin and Young 53

ber of variations that could be researched. For instance, there is a strong

need for empirical investigation of the relative weights of variables, or the
best combination of variables in our framework, that determine
benchmarking success. Research can identify and model which variables
are critical for a successful benchmarking effort. Further, research is needed
to study various types of benchmarking relating to products, functions
and strategies as they relate to management accounting systems. Another
central issue pertains to the most effective method of information sharing
and gathering, and partner selection. Nowhere in the literature is there a
comprehensive, comparative study of which information sharing method
is most effective under a specific set of conditions. For example, is the
cooperative, group form of benchmarking cost effective, versus the data-
base form. Only a cost-benefit analysis can reveal the answer to questions
such as these.
Future research also could address the modeling of functional, cross-
industry cooperative benchmarking, and research designs could be devel-
oped to test predictions from these models. Finally, studies which capture
combinations of nonfinancial (quantitative and qualitative) benchmarking
measures, financial measures and those which assess both costs and ben-
efits will advance both our theoretical and practical understanding of
benchmarking effectiveness and success. With the increased internation-
alization of business, and the development of the information super high-
way, we predict that benchmarking opportunities across organizations,
industries and nations will explode as more and more firms seek compara-
tive information in order to compete and survive.

REFERENCES
Argyris. C. 1977. Organizational learning and management information systems.
Accounting, Organizations and Society: 113-124.
Atkinson, A. A., R. Banker, R. S. Kaplan, and S. M. Young. 1995. Management
Accounting. Englewood Cliffs. NJ: Prentice Hall.
Bean, T. J., and J. G. Gros. 1992. R&D benchmarking at AT&T. Research Technol-
ogy Management (July-August): 32-37.
Berliner. C , and J. Brimson. 1988. Cost Management Boston. MA: Harvard Busi-
ness School Press.
Birnberg. J. G.. M. D. Shields, and S. M. Young. 1990. The case for mulUple meth-
ods in empirical management accounting research (with an illustration from
budget setting). Journal oJ Management Accounting Research: 33-66.
Camp, R. C. 1989. Benchmarking: The Search for Industry Best Practices That Lead
to Superior Performance. Milwaukee. WI: American Society for Quality Control
Quality Press.
Coburn. S.. H. Grove, and C. Fukami. 1995. Benchmarking with ABCM. Manage-
ment Accounting (January): 56-50.
Elnathan. D.. and O. Kim. 1995. Partner selection and group formation in coopera-
tive benchmarking. Journal of Accounting and Economics: 345-364.
Ginzberg. M. A. 1981. Key recurrent issues in the implementation process. MIS
Quarterly 5 (2): 47-59.
Horngren. C . G. Foster, and S. Datar. 1993. Cost Accounting—A Managerial Em-
phasis. 8th edition. Englewood Cliffs, NJ: Prentice Hall.
54 Journal of Management Accounting Research. 1996

Industry Week. 1990. The art of benchmarking (November 5).

Institute of Management Accountants (IMA). 1993. IMA continuous improvement
center. IMA FOCUS (July): 1, 7.
Jarvenpaa. S. L., and B. Ives. 1991. Executive involvement and participation in the
management of information technology. MIS Quarterly (June): 205-227.
Kaplan, R. S.. and D. P. Norton. 1992. The balanced scorecard—measures that
drive performance. Harvard Business Review (January-February): 71-79.
Kearns, D. T.. and D. A. Nadler. 1992. Prophets in the Dark. New York: Harper
Business.
Kharbanda, M. 1993. Benchmarking: Making it work. CMA Magazine (March): 3 0 -
33.
Kirby. A. 1988. Trade associations as information exchange mechanisms. RAND
Journal of Economics (Spring): 138-146.
Lucas, H. C , Jr. 1981. Implementation: TheKey to Successjul Information Systems.
New York, NY: Columbia University Press.
Main, J. 1992. How to steal the best ideas around. FORTUNE (October 19); 102-
106.
Malcolm, L 1995. Benchmarking finance function activities. Management Account-
ing (U.K.) (February): 22.
McCamus, D. R. 1991. Performance measurement and the quality voyage. CMA
Ma^fozine (January): 8-12.
Miller, G., A. De Meyer, and J. Nakane. 1992. Benchmarking Global Manufacturing.
Illinois: Business One Irwin.
Miller, J. A. 1992. Benchmarking performance. CMA Magazine (June): 23.
Nunnally, J. C. 1983. Psychometric Methods. New York: McGraw-Hill.
Raymond, L. 1985. Organization characteristics and MIS success in the context of
small business. MIS Quarterly (March): 37-52.
Rivest. G. 1991. Make your business more competitive. CMA Magazine (May): 16-
19.
Senge, P. M. 1990. The Fifth Discipline. New York, NY: Doubleday.
Shields, M. D. 1995. An empirical analysis of firms" implementation experiences
with activity based costing. Journal of Management Accounting Research: 148-
166.
, and S. M. Young. 1989. A behavioral model for implementing cost manage-
ment systems. Journal of Cost Management for the Manufacturing Industry 2;
17-27.
Society of Management Accountants of Canada (SMAC). 1993, Implementing
Benchmarking. Hamilton, Ontario: SMAC.
Spendolini, M. J. 1992. The Benchmarking Book. New York, NY: AMACOM (Ameri-
can Management Association).
Tonkin, L. 1991. Tbe power of benchmarking at Seitz Corporation. Target (Winter):
6-12.
Young, S. M., and F. H. Selto. 1991. New manufacturing practices and cost man-
agement: A review of the literature and directions for research. Journal of Ac-
counting Literature 10: 265-298.