It... 9:53;..5
. Eu
.2

 

 

 

 

 

 

i! x 24;. 1:.

.l1l
,2...,.r.nv..r: i.“ .
. . :2... 51.4mm“.

..!.z.n . :3.
he. imam“; .
€59§flfif

it

a

x:
.

flu...
“I... n
2 I 0‘

..
fl
s .3

.31115l5
3.5.x .. :0
I 1:: 1

A:
3.4

a.

HO
3 .
Ah 9

8.

I
nut-i=1

3:52.. i

. .1192 tia:.1l 1.9.10.1?

.5

xi! .ItmuoWNIXAainawlnl
xi 2...
a . . 4 1131.»

In}... '
7 651:1.

v}. v 3.1.1,. ANN an
o .s 5- v . .
'6‘ Ilv.‘ fi‘i’rn'V)v3
$331,591}.

I o. v .1.

 

TRISH'S

412’)?»

llBRARY
Michigan State
University

This is to certify that the

dissertation entitled

An Exploratory Study of the
Capabilities, Structures and Performance
of Environmental Management Systems

presented by

Frank Lloyd Montabon

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Operations Mmgamt

WW

Majot’profeskg

Date 14 December 2001

MSU i: an Affirmative Action/Equal Opportunity Institution 0-12771

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DateDue.065-p.15

AN EXPLORATORY STUDY OF THE CAPABILITIES, STRUCTURES AND
PERFORMANCE OF ENVIRONMENTAL MANAGEMENT SYSTEMS

By

Frank Lloyd Montabon

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Marketing and Supply Chain Management

2001

ABSTRACT

AN EXPLORATORY STUDY OF THE CAPABILITIES, STRUCTURES AND
PERFORMANCE OF ENVIRONMENTAL MANAGEMENT SYSTEMS

By

Frank Lloyd Montabon

Due to developments such as the propagation of the ISO 14000 series of
standards, increased consumer awareness and the continuing search for greater
efficiency, firms no longer have a choice regarding whether or not to become
environmentally responsible. Rather than a question of “why” (should my firm
become environmentally responsible), the question is now “how" (can my firm
become environmentally responsible). There are three primary approaches that
a firm can take in order to become more environmentally responsible. The first is
tools-based (e.g., Life Cycle Analysis, Design for Environment). The second is to
choose amongst various environmental options such as recycling and
remanufacturing. The third is a more systematic approach, through the use of an
environmental management system. This last option has recently become even
more important in the competitive arena given the introduction of the ISO 14000
certification series, which is aimed at the certification of environmental
management systems. Despite the importance of the competitive advantages
that environmental responsibility offers and the introduction of the ISO 14000
series of standards, the environmental management system has only recently

been the focus of the research literature. There are still many questions to be

answered; chief among them are the issue of how to measure an environmental
management system, the behavior of the environmental management system
and how the environmental management system effects operational
performance. Using the 1510 responses from a large-scale survey of American
manufacturers. this study will attempt to answer those questions, while
contributing to the literature a measurement model for environmental

management systems and an understanding of how these systems work.

Copyright by
Frank Lloyd Montabon
2001

I have always believed that to do anything well and to have a fulfilling life, you
must find something for which a fire burns inside you. This work, then, is
dedicated to those who have given me my fire for being a college professor:
To Gary Kern, who gave me the spark,

To my wife Sheryl, who provided the fuel
And to the Lord, who provides the air for us all.

ACKNOWLEDGMENTS

I would like to thank the members of my dissertation committee:
Dr. Steven A. Melnky, chair
Dr. Roger Calantone
Dr. Gary Ragatz
Dr. Ted Stank
The twin goals of science are the creation and dissemination of
knowledge. I would argue that nothing realizes these goals better than the
mentoring of new scientists. Thus, this process of mentoring is not to be taken
lightly. I would like to thank my committee for not merely for their time and efforts
regarding my dissertation, but for their commitment to furthering science by
taking this process very seriously in order to ensure that I would be prepared to

contribute to the growth of our field

On a much less serious note, I would like to thank eBay Inc., Diageo PLC,

and Sony Corporation for the reassurance, comfort and motivation, respectively,

that they provided to me during this process.

vi

TABLE OF CONTENTS

ACKNOWLEDGMENTS ...................................................................................... vi
TABLE OF CONTENTS ...................................................................................... vii
LIST OF TABLES .................................................................................................. x
LIST OF FIGURES .............................................................................................. xi
LIST OF ABBREVIATIONS ................................................................................ xii
Chapter 1 INTRODUCTION AND OVERVIEW OF RESEARCH ......................... 1
Motivations for This Dissertation ............................................................. 1
Research Objectives .............................................................................. 4
Research Methodology ........................................................................... 7
Contribution of Research ........................................................................ 8
Outline of the Work ............................................................................... 10
Summary .............................................................................................. 1 1
Chapter 2 LITERATURE REVIEW ..................................................................... 13
Introduction ........................................................................................... 13
Overview of Environmental Management ............................................. 14
Growing Interest in Environmental Issues ......................................... 22
Characterizing the Environmental Management Literature ............... 25
Defining and Measuring the Environmental Management System ....... 27
Relationship of EMS to EM ............................................................... 30
Measurement of Environmental Management Systems .................... 32
Is it a Management System? ......................................................... 33
Is it an Information System? .......................................................... 36
EMS Success Factors .......................................................................... 38
Concluding Remarks ............................................................................ 41
Unresolved Questions ....................................................................... 43
Chapter 3 MODEL DEVELOPMENT ................................................................. 45
Introduction ........................................................................................... 45

Link Between Environmental Management and Environmental
Management Systems .................................................................... 45
Model Development .............................................................................. 48
Measurement Model Hypotheses ..................................................... 52
Comparison Hypotheses ................................................................... 53
Summary .............................................................................................. 56
Chapter 4 RESEARCH METHODOLOGY ......................................................... 57
Introduction ........................................................................................... 57
Measurement Model Operaticnalizaticn ............................................... 57
Data Collection ..................................................................................... 61
The Survey ........................................................................................ 61
The Sample ....................................................................................... 63
Two Stage Analysis Methodology ......................................................... 64
Operaticnalizaticn of Comparisons ....................................................... 66

vii

Industrial Membership ................................................................... 66

Size of Firm ................................................................................... 69
Experience with Environmental Management System .................. 70
Presence of ISO 14000 Certification ............................................. 70
Statistical Methodology ..................................................................... 71
Stage 1 .......................................................................................... 71
Stage 2 .......................................................................................... 72
Primary Statistical Technique ........................................................ 72
Limitations of Proposed Methodology ............................................... 73
Summary .............................................................................................. 74
Chapter 5 ANALYSIS ......................................................................................... 75
Introduction ........................................................................................... 75
Pre-Processing of Data ........................................................................ 75
Stage One: Determination of Model ..................................................... 78
Management Functions Model .......................................................... 78
Systems Activities Model .................................................................. 81
Selection of Model ............................................................................ 84
Stage Two: Comparisons ..................................................................... 85
Measurement Models ........................................................................ 85
Baseline Structural Models ............................................................... 85
Industrial Setting ............................................................................... 89
Firm Size ........................................................................................... 91
Age of the Environmental Management System ............................... 94
The Presence (or Absence) of ISO 14000 Certification .................... 98
Conclusion ........................................................................................ 99
Chapter 6 DISCUSSION .................................................................................. 100
Introduction ......................................................................................... 100
Measurement Model ........................................................................... 100
Systems Activities Framework ........................................................ 101
Management Functions Framework ................................................ 104
Viability of Both Models ................................................................... 105
Single Group Model ........................................................................ 106
Multi-Group Comparisons ................................................................... 107
Industry Setting ............................................................................... 107
Firm Size ......................................................................................... 109
Age of Environmental Management System ................................... 110
Presence of ISO 14000 Certification ............................................... 112
Summary ......................................................................................... 1 14
Chapter 7 CONCLUSION ................................................................................ 117
Summary of Findings .......................................................................... 117
Limitations of the Study ...................................................................... 118
Future Research ............................................................................. 122
Implications for Management .......................................................... 129
Contribution of Research ................................................................ 130
Summary ............................................................................................ 131
APPENDIX A Questions for Management Functions Model ............................ 134

viii

APPENDIX B Questions for the Systems Activities Model ............................... 138
BIBLIOGRAPHY ............................................................................................... 142

LIST OF TABLES

TABLE 2.1: REPRESENTATIVE LIST OF OUTLETS FOR ENVIRONMENTAL MANAGEMENT

LITERATURE .................................................................................................. 26
TABLE 3.1: SUMMARY OF RESEARCH HYPOTHESES ................................................ 56
TABLE 5.1 CONFIRMATORY FACTOR ANALYSIS OF MANAGEMENT FUNCTIONS

MEASUREMENT MODEL .................................................................................. 80
TABLE 5.2 CONFIRMATORY FACTOR ANALYSIS OF SYSTEMS ACTIVITIES MEASUREMENT

MODEL ......................................................................................................... 82
TABLE 5.3 COMPARISON OF VARIOUS STRUCTURAL MODELS ................................... 87

LIST OF FIGURES

FIGURE 2.1: BAJWA, RAI, RAMAPRASAD 1998 RESEARCH MODEL ............................ 35
FIGURE 2.2 CHOE 1996 RESEARCH MODEL ............................................................ 40
FIGURE 3.1 GENERAL FRAMEWORK OF ENVIRONMENTAL MANAGEMENT RESEARCH ...46
FIGURE 3.1 THE FUNCTIONS OF MANAGEMENT AS TAKEN FROM MELNYK AND DENZLER

(1996) .......................................................................................................... 50
FIGURE 4.1 “CONCISE” RESEARCH MODEL ............................................................. 66
FIGURE 5.1 STATUS OF ISO 14000 AMONGST ALL RESPONDENTS ............................ 77
FIGURE 5.2 STRUCTURAL MODEL FOR MANAGEMENT FUNCTIONS MODEL ................. 86
FIGURE 5.3 STRUCTURAL MODEL FOR SYSTEMS ACTIVITIES WITH LOGSALES,

LOGFTE ERROR VARIANCES FIXED To .05 ..................................................... 88
FIGURE 5.4 INDUSTRY SETTING MULTI-GROUP ........................................................ 90
FIGURE 5.5 SIZE OF FIRM MULTI-GROUP ANALYSIS WTH DATASET SPLIT INTO THIRDS

..................................................................................................................... 92
FIGURE 5.6 SIZE OF FIRM MULTI-GROUP ANALYSIS WITH DATASET SPLIT INTO HALVES

..................................................................................................................... 93
FIGURE 5.7 AGE OF ENVIRONMENTAL MANAGEMENT SYSTEM MULTI-GROUP ANALYSIS

WITH DATASET SPLIT INTO THIRDS ................................................................... 95
FIGURE 5.8 AGE OF ENVIRONMENTAL MANAGEMENT SYSTEM MULTl-GROUP ANALYSIS

WITH DATASET SPLIT INTO HALVES .................................................................. 96
FIGURE 5.9 AGE OF ENVIRONMENTAL MANAGEMENT SYSTEM MULTI-GROUP ANALYSIS

WITH PERFORMANCE DISTURBANCE FIXED To 0 FOR “YOUNG” GROUP ............... 97
FIGURE 5.10 PRESENCE OF ISO 14000 MULTl-GROUP ANALYSIS ............................ 98

xi

LIST OF ABBREVIATIONS
This dissertation will not use any abbreviations in order to avoid confusion

over some acronyms that are the same for different terms. For instance, a
careful reader will note that the acronyms for environmental management and
environmental manufacturing are the same, as are the acronyms for
environmentally responsible management and environmentally responsible
manufacturing. For this reason, this dissertation will not use acronyms and

instead all terms will be spelled out.

xii

Chapter 1

INTRODUCTION AND OVERVIEW OF RESEARCH

Over the past decade, a confluence of events has focused both
practitioner and academic interest in environmental management systems. In
that time, what has been learned about this key component of environmentally
responsible management? Unfortunately, not a great deal. While there are
many descriptions of environmental management systems in the literature, there
have not been very many measurements of them. This leaves both practitioners
and academics with a lack of models and theory upon which to build best
practices. This dissertation will attempt to build theory regarding the
measurement Of environmental management systems and the business
environment factors that affect the efficacy of environmental management
systems. This will help the academic audience build its understanding of
environmental management systems, from which best practices for practitioners
can be hypothesized and tested.

This chapter will provide an overview to this dissertation, by describing the
motivations for this study, the particular research issues to be investigated and
the structure of the investigation. This study will also describe its contribution to

both the theory and practice of environmentally responsible management.

Motivations for This Dissertation
Recently, there has been an increase in interest in the environmental

practices and impacts of businesses. There are three primary reasons for this

increased interest. The first is consumer attitudes regarding the environment,
manifested interest in buying environmentally friendly products and consumer
concern over highly publicized industrial environmental accidents, such as the
Bhopal and the Exxon Valdez incidents.

Second, government environmental regulations have continued to
increase, with a recent emphasis on systematic management approaches.
Certainly, avoiding regulatory problems is one reason for a firm to become more
environmentally proactive. Somewhat paradoxically, Porter and van der Linda
(1995), have argued that environmental regulations can actually be helpful to
firms, as they can lead to innovation.

Perhaps most importantly, firms have recognized that becoming more
environmentally responsible may Offer a competitive advantage. This line of
argument states if the firm is better able to manage and control all forms of waste
that it produces, it should become more efficient and add more value to the
product This was essentially Porter's (1991) argument, albeit in the context of
how government regulation can spur innovation.

In fact, because of the strength Of these three phenomena, much of the
discussion about firms’ environmental practices and impacts has moved away
from “what” (is environmentally responsible management?) Over the past
twenty years, various programs and techniques have appeared that have Offered
reasonably good, though not always complete, answers to this question.

Perhaps more importantly, the discussion has also migrated away from “why”

(should firms become environmentally responsible?) The movement away from
the “why” argument is demonstrated in at least three prominent ways.

First, research is indicating that there is a definite correlation between
improved environmental performance and improved business performance, as
been shown by numerous studies (Klassen and McLaughlin 1996; Feldman,
Soyka, and Ameer 1997; Steger 2000). Second, large multinational firms have
now gone on record as being committed to the idea of environmental
responsibility (Anonymous 1999).

Third, and perhaps most important to this research, the idea that
environmental management is not only accepted but is now a permanent part of
the competitive landscape is attested to by the creation Of the ISO 14000 series
of standards, which aim to standardize environmental management just as ISO
9000 series standardized quality management. Specifically, the ISO 14000
standards are designed as certification guidelines for environmental
management systems. As described by multiple authors in a volume edited by
Hillary (1997), this standard went from non-existence to publication in just three
years. This is an indication of the importance of environmental management to
the intematicnal business community, given the difficulty on achieving consensus
on such a value-laden issue.

The propagation of the ISO 14000 series highlights the importance Of the
environmental management system to a firm’s environmental management
program and goals. As explained in an ISO 14000 handbook edited by Baker

and McKiel (1998), "An EMS is essential to an organization's ability to anticipate

and meet growing environmental performance expectations and to ensure
ongoing compliance with national and intematicnal requirements.”

To further explain the importance of an environmental management
system to accomplishing the goals of environmental management, consider the
three primary ways that a firm can become more environmentally proactive. One
way a firm can become more environmentally conscious is to incorporate the use
Of environmental tools in its processes, such as life cycle analysis or design for
environment (Conway-Schempf and Lave 1999). Other firms may choose to use
various recoverable manufacturing options (Guide, Jr. and Srivastava 1998) such
as remanufacturing or recycling.

It is argued in this dissertation that a critical element Of the environmental
management is the third primary approach, the environmental management
system. This is the system that allows a firm to monitor, analyze, manage and
control its environmental impacts. It is the finn’s formal system that enables
managers to achieve the objectives of environmental management and thus,

their business Objectives (Stapleton, Cooney, and Hix, Jr. 1996; Graft 1997).

Research Objectives
Until recently, the environmental management system has not been the

focus Of very many studies. Yet, with the recent introduction Of the ISO 14000

series, attention has now been focused on this system. Most of the research in
the emerging area of environmental management has focused on issues other

than the environmental management system. For instance, there has been

literature rationalizing the need for environmental management (Porter and van

der Linde 1995). There have also been some anecdotal stories (Epstein 1996)
and empirical research (Klassen and McLaughlin 1996) on some of the impacts
of environmental management. It should be noted that most of this research has
noted favorable impacts on the firm, both in terms of Operational and financial
performance.

There has been a concurrent increase in the amount of research on the
ISO 14000 series of certifications. However, this research has not really delved
into the makeup of the environmental management system and might be
characterized as treating the environmental management system as a "black
box“. That is, the focus was on the presence or absence Of an environmental
management system and its effects, but not on the capabilities and structure Of
the environmental management system (for instance, see Steger (2000)).

There appear to be at least three arguments that would indicate that there
should be more research on environmental management systems. First, as
noted above, it can be argued that having an environmental management system
is critical to the success of any environmental initiatives that the firm attempts to
undertake, as an environmental management system provides for an integrated,
holistic system for managing environmental impacts, rather than just a hodge-
podge Of tools or options. Second, there is now a series of standards, ISO
14000, for defining and certifying environmental management systems. This has
elevated the importance Of environmentally responsible management for
managers and has provided research with a measurement framework, both Of

which could be stimuli for increased research. Third, the information systems

literature has a wealth Of research frameworks (in addition to the ISO 14001
standard) that can be used to study environmental management systems. Thus,
there are very viable starting points for theory building.

The relative lack of research on environmental management systems has
two important research implications. First, there remain many questions in need
of investigation regarding environmental management systems. Second, a lack
of previous research allows researchers the opportunity to experiment with a
variety Of frameworks for building and testing theory.

SO, practitioners have clearly indicated that environmental management
systems are an important part of the competitive landscape. And academics
have begun to study this new phenomenon. Thus, much of the argument
surrounding environmental management systems has circled back from ‘why”
(do we need an environmental management system?) to “how'. In particular,
practitioners are anxious to know how to set Up their environmental management
system in order to achieve a competitive advantage and academics need to
know how to measure environmental management systems.

This dissertation is aimed at making a substantial contribution to the
emergent environmental management systems literature. The first Objective of
this study is to examine the extent to which an environmental management
system can be viewed as either an information system or a management
(problem solving) system. This discussion will attempt to describe the
capabilities of the environmental management system. Based on this discussion,

constructs will be developed for environmental management systems and two

models of how these constructs are related will be developed. These two models
will be compared to see which one does a better job, statistically and
theoretically, of measuring environmental management systems. Thus, the
second major objective of this study will be to determine a theoretically and
statistically valid measurement model for environmental management systems.
Once this model has been determined, a series of business context factors will
be examined in order to understand how they affect environmental management
systems. This will be done by splitting the sample based on these contextual
factors. These factors will come from conjectures that have been put forth in the
literature as having an effect on the environmental management system. This
analysis will comprise the third major Objective of the study.

Since the environmentally responsible management literature is still
growing and evolving, it would be useful to take the findings from these two
primary research Objectives and interpret them in order to help contribute to the
growing body of knowledge for this field. This interpretation and theory building
will comprise the fourth major Objective of this study. While this Objective may
seem unnecessarily vague, it is important if the field is to grow and develop to the
point where it is having practitioner impact, especially with regards to

environmental management system best practices.

Research Methodology
The stated Objectives will be accomplished via the use of a three stage

methodology. In the first stage, the two models will be compared to see which

fits best on statistical and theoretical grounds. The second stage will be a series

Of multi—group structural equation modeling analyses to see how different
business context factors affect environmental management systems. The large
number of responses will allow this research to split the response pool into
groups based on the conjectural arguments that are in the environmentally
responsible management literature. These types of multi-grcup analyses will be
used to examine the contexts of environmental management system success.
The third stage will be to synthesize the results for both practitioners and
academics, with the emphasis being on making a contribution to the growth of
environmental management systems theory. Structural equation modeling will
be the primary analysis tool.

The data set for the statistical analysis comes from a large project
regarding many aspects of environmental management. As part of this project, a
large survey was conducted of managers at American manufacturers, which
resulted in over 1500 responses. This large data set helps to increase

generalizability and statistical power.

Contribution of Research
This research will attempt to break new ground In terms of providing a

detailed, rigorous study Of a heretofore under-examined topic and will do so
using one of the first major surveys of environmental management systems.
Thus, this research should be considered as exploratory, or theory building,
rather than theory testing. This is due to the fact that research in the field Of
environmental management is in an early growth stage. Thus, there are not

many previously developed theories or models available for testing.

This lack of models, however, presents the primary opportunity for
contribution of this research. In particular, the primary contribution of this
research will be to help grow the relatively new field of literature concerning
environmental management systems by developing and testing constructs and
the relationship among those constructs for environmental management
systems. There has been limited previous research that has examined these
systems in any sort of empirical depth. This research will help the growth of this
relatively new research area and thus help future research develop theories
regarding why and how environmental management systems are adopted, what
causes certain capabilities to be favored over others, and which combination of
capabilities has the greatest efficacy. This is an important gap tO be filled given
the potential of environmental management systems to achieve the goals of
environmental management.

There is another related contribution for academics. This research is
based on the first large-scale survey of perceptions of, knowledge about and
reasons for the ISO 14000 series and environmental management systems. As
such, the questions themselves will be of use to future researchers as pre-tested
constructs. As will be discussed in more detail later, the survey used was quite
long. Much of the reason for this length was a lack of previously validated
constructs. It is anticipated that one Of the results from this research will be a
usable set of scales for measuring various attributes of environmental

management systems in future research.

The development and testing of the relationships in environmental
management systems leads to the next major contribution. By doing this, there
is the potential to cultivate empirically based guidelines for environmental
management system development and implementation for practitioners. Thus,
this research may help to indicate the contexts of effective environmental
management systems. Further, this research may help to provide empirical
evidence to support the adoption Of an environmental management system.

While the development of such guidelines will have Obvious bottom line
appeal for practitioners, the development of the constructs and relations among
the constructs will be the primary contribution of this dissertation. Thus,
academics are the primary audience for this research. This makes sense, as the
first step in developing practitioner guidelines is to take measure of what is
happening in the field. These measurements cannot happen without constructs

and models having been defined.

Outline Of the Work
This chapter summarized the problem to be discussed in the rest of the

dissertation. The second chapter will review the relevant literature in order to
highlight previous efforts in environmentally responsible management research
and to note as yet unanswered questions. Chapter three will develop the
theoretical structures of the environmental management system. This chapter
will explicate the relationship between environmental management and the
environmental management system and develop the constructs and relationships

for the environmental management system. This development will be grounded

10

in existing theory. This chapter will also describe the research hypotheses.
Chapter four will discuss the research methods. This chapter will discuss the
research methodology to be used for both collecting and analyzing the data.
Chapter five will present the results of the analysis. The primary analysis tool will
be structural equation modeling. Finally, the sixth and seventh chapters will
discuss the analysis with a goal of contributing to theory development in the

environmentally responsible management field.

Summary
This chapter has introduced the problem to be investigated and what

contribution this investigation will make to both academics and practitioners.
This research plan aims to make an important contribution to the body of
environmental research literature. It is intuitive that an environmental
management system is essential for a firm if it intends its environmental
management effort to be complete, integrated and effective. Despite this
criticality, little research on the environmental management system has been
done, so there is very little in terms of previous constructs, models or theories to
build upon and test. Thus, this research will be a starting point for theory
development about environmental management systems.

Vlfith a large data set already collected, this research will have a strong
empirical base from which to draw. Statistically speaking, the size of the data set
gives the research a lot of power. Further, the survey itself is a valuable
contribution to theory development, as it investigates latent constructs that have

previously gone uninvestigated. Overall, then, this dissertation will mark an

11

important starting point in the development Of theory regarding environmental
management, environmental management systems, measurement constructs,
and prescriptions for environmental management system success.

In the next chapter, the relevant literature is reviewed in order to gain an
understanding of where the field currently is and what issues are in need of
investigation. As the chapter will demonstrate, environmental management is a
relatively new field of inquiry, with a literature base that is fractured into many
disciplines. This review will illustrate how we have arrived at the issue to be

studied in this dissertation.

12

Chapter 2

LITERATURE REVIEW

Introduction
This literature review has four main goals. The first is to underline the fact

that environmental management is a relatively new field. The primary implication
of this is that there are precious few previous models that have been tested. As
interest in environmental management and related topics has grown, the amount
of research literature has increased. However, the growth of the literature in
environmental management has occurred in a variety of disciplines, such as
ecology, natural resource management, public policy, economics, and the
business fields of marketing, corporate strategy and operations management.
Thus, it would be fair to say that while there is a growing amount of literature
regarding environmental management, this literature has been spread out among
a number of fields. This means that the amount of literature in any one particular
field may appear small relative to more established topics in that field. Because
of this phenomenon, this literature review and this research study will have more
references to journals and research outside of the operations management field
than might be expected.

The second goal of this literature review is to go through the definitions
and models that have been put forth. The definitions will help to outline the field
and help to draw boundaries around the multi-disciplinary concept of

environmental management. Accomplishing this second goal will help to setup

13

chapter three, where the specific definitions for this dissertation will be explicated
and the theoretical structures to be tested will be developed.

The third goal is to examine what the literature has had to say about the
relationship between environmental management and the environmental
management system. Essentially, this examination seeks to come to an
understanding if one is a subset of the other or if one is necessary and/or
sufficient condition for the other. Based on this review, chapter three will develop
a construction of this relationship that will be one of the major premises of this
dissertation.

The final goal of this literature review is to position the research study
within the stream of environmentally related literature. By first reviewing the field
and its gaps, this literature review will establish the need for this Specific research

study and spotlight how this dissertation will contribute to the field.

Overview of Environmental Management
One of the issues that is still under discussion regarding environmental

management is exactly how to define it. As Walley and Whitehead (1994)
expressed the problem, “The common rallying cry of many environmental
thinkers is that the environment must be integrated into everyday business
decisions, yet few specify what that means.” Though pessimistic, to a certain
extent they identify a goal, that of working environmental concerns into all
business decisions, and the reality that how to do this is still a matter of some
argument. This argument is reflected in the fact that the field has seen a variety

of different definitions divided up among many disciplines. Though it can be

14

argued that each of these definitions is related in some way, the various
definitions typically differ in how much falls under their individual umbrellas.
Klassen and McLaughlin (1996) defined environmental management as
“all efforts to minimize the negative environmental impact of the firm’s products
throughout their life cycle.” This definition is useful in that it is relatively easy to
understand, but it should be noted that this definition does not explicitly state that
all business decisions must incorporate a concern over environmental matters.
In a similar way, Wu and Dunn (1995) defined integrative environmental
management as “every element in the corporate value chain involved in
minimization of the firm’s total environmental impact from start to finish of the
supply chain and also from beginning to end of the product life cycle.” Again, this
refers more to the idea of a collection of practices rather than an integration into
the decision making process. Brown and Karagozoglu (1998) discuss four
stages of firm responses to the increased interest in environmental issues. In the
fourth stage, firms achieve “a focus throughout all activities on managerial
practices for a sustainable environment.” Their discussion again shows an
emphasis on environmental management being integrated into all firm activities.
fimmins (1998) set up three goals for environmentally responsible
managements. The first is to minimize the impact of the operation on the
environment. The second is to use resources in the most efficient way possible.
The third is to set up systems and motivate employees to make the necessary

changes. This definition is primarily goals or output based.

15

Shrivastava and Hart (1994) propose the concept of total environmental
management, which is how the firm should deal with the environmental problems
that arise in their organizational activities. In describing this concept, they set up
goals for the “'gmen' organization” for inputs, throughput (processes) and
outputs. With regard to inputs, they state that the “green“ organization should try
to minimize the use of virgin materials and non-renewable forms of energy. The
goal for throughputs is to eliminate emissions, effluents and accidents.
Minimizing the life cycle cost of the organization’s products and services is the
goal for outputs. These authors also place a special emphasis on what they term
“systems thinking“ for firms who desire to have total environmental management.
By this, they are referring to the idea that a firm should strive to achieve all three
goals (input, throughput and output) and that the entire system must adopt this
perspective. This highlights the importance of a fully integrated approach. As
they mention later in their article, “Only when environmental concerns are
integrated into day-to-day operations can an organization be ‘green”' (p. 624).

As can be seen in all these definitions, a common theme is the idea that
environmental management is a finnwide construct. That is, it is concept that
cuts across all areas of a firm and thus requires attention and resources from all
of a finn's functional areas.

However, in attempting to elucidate the components of environmental
management, the definitions reviewed above are somewhat vague. When it
comes to defining this concept, terms like “all efforts" or “all elements“ are

popular, but this does not provide a more concrete listing of what these efforts or

16

elements might be. Despite this lack of an explicit listing of the components of
environmental management, it is reasonable to impute such a list from the
environmental management literature.

The first component of environmental management would be that a firm
incorporates a concern for its environmental impacts in its business practices.
This concern may be finnwide or may be limited to particular functions (the
functional applications of environmental management will be discussed below).
From a research perspective, the concept of “concern” may be a little vague or
fuzzy. To state this idea in more concrete terms, this first component of
environmental management is the adoption (either formally or informally) of
policies and the procedures which flow from those polices that are aimed at
integrating an understanding of the firm’s environmental impacts of its business
practices. Thus, in terms of something that might be more easily measured, the
key would be the policies and the procedures that flow from those policies. Note
that these policies do not necessarily have to state a firm's commitment to
environmental impact goals (such as X% less effluent release in year Y).

When considering procedural components of environmental management,
one could categorize some of these procedures as options or tools. Tools would
include such concepts as life cycle analysis (see Miettinen and Hamalainen
(1997) and Conway-Schempf and Lave (1999) for good discussions of life cycle
analysis) and design for environment (F iskel 1996). Options would include

concepts like recycling or remanufacturing (Horrnozi 1999).

17

One of the characteristics of environmental management is that it does not
necessarily require that a firm avail itself of all possible tools and options. There
are a number of potential approaches that can be followed. Though it might be
argued that this flexibility renders environmental management too vague to be
taken seriously as a construct, it is easy to see why this is the case if you
consider a simple example. For instance, a restaurant is unlikely to avail itself of
remanufacturing, for obvious reasons.

Another important component of environmental management is some type
of information system. In this instance, “information system“ does not
necessarily mean “computer system”, but refers simply to the procedures and
tools used to collect, hold and disseminate the relevant data. In the field of
environmental management, multiple such systems have appeared, such as
environmental accounting (Gray 1992), environmentally responsible logistics
systems (Wu and Dunn 1995), and environmental management systems (which
will be described in great detail shortly). The importance of these systems to
environmental management is that they allow the firm to track its environmental
performance, understand how the costs related to environmental impacts affect
the firm and setup evaluation mechanisms.

Though environmental management is usually considered a finn-wide
construct, there have been many functional applications of its principles.
Essentially, some firms have concentrated on adapting environmental
management concepts to specific functions within their organizations. The

research literature has also tracked this phenomenon as well, with a series of

18

new definitions and related research. Reviewing these definitions will yield
further insight into the larger rubric of environmental management.

One of the most commonly discussed functional applications is for
operations/manufacturing. Honnozi (1999) quotes the President’s Council on
Sustainable Development for a definition of environmentally conscious
manufacturing. This definition requires reduction in material use, water use,
energy use and waste generation while manufacturing continues to develop
technologies to further accommodate these reductions. Obviously, this is a very
output oriented definition. Sarkis and Rasheed (1995) define this very same term
as primarily consisting of what they call the “3 Rs” - reduce, remanufacture and
reuse/recycle. This particular definition is thus focused on some of the
environmental options that firms have for managing their environmental impacts.
Owen (1993) defined environmental manufacturing as the determination of the
effort involved in disassembly, the logistics of reusing the material and the
difficulty of reprocessing it to a quality level suitable for manufacturing. Again,
this definition focuses on environmental options, similar to Sarkis and Rasheed.
Gupta and Shanna (1996) defined environmental operations management as
“the integration of environmental management principles with the decision-
making process for the conversion of resources into useable products.” This
particular definition is very basic in that it directs a function to adopt

environmental management for its function’s ends.

19

Handfield and Melnyk (1995) offer one of the most complete functional
application definitions. They defined environmentally responsible manufacturing
as:

A corporate system that integrates product and design
issues with issues of production planning and control and supply
Chain management in such a manner as to identify, quantify,
assess, and manage the flow of environmental waste with the goal
of reducing and ultimately minimizing its impact on the
environment, while also trying to maximize resource efficiency.

Underlying this definition is the implicit assumption of a positive
relationship between environmental and corporate performance. Associated with
this definition are several important assumptions and premises (Handfield and

Melnyk 1995):

o Environmentally responsible manufacturing decisions are
always present and are integrated with (and part of) the
business process.

0 The ultimate goal of environmentally responsible manufacturing
is waste elimination.

0 To be ultimately successful, environmentally responsible
manufacturing must be viewed as a strategically-driven decision
that is evaluated by comparing its relative costs and benefits
(and making appropriate decisions based on this analysis).

0 Effective environmentally responsible manufacturing systems
must be viewed as corporate systems, not as manufacturing,
engineering or purchasing decisions.

. Effective environmentally responsible manufacturing systems
must focus on the three PS - Product, Process, and Packaging.

0 To be ultimately SUOOBSSfUI, environmentally responsible
manufacturing must be integrated into the product and process
design and development system.

0 To be ultimately successful, environmentally responsible
manufacturing must be considered within the context of both the
internal factory (the manufacturing system found within the four
walls of the enterprise) and the external factory (the
transformation process as it involves the supply chain).

20

. Environmentally responsible manufacturing is ultimately a cross-
functional undertaking.

. Environmentally responsible manufacturing is dynamic.

Amongst the discussion of functional applications of environmental
management, the manufacturing or operations function predominates. However,
other functional applications have also been described. One example would be
Wu and Dunn’s (1995) description of environmentally responsible logistics as
“(expanding) the manager's horizon by adding another objective to the system:
minimizing total environmental impact.“ Again, this could be described as an
output-oriented view. Carter, Ellram and Ready (1998) devised a definition for
environmental purchasing, which is “purchasing’s involvement in supply chain
management activities in order to facilitate recycling, reuse and resource
reduction.“

In reviewing these descriptions of the functional applications of
environmental management, the theme of integration is prevalent. This theme
has two facets. The first is the integration of environmental management
principles into all phases of a particular function's operations. The second facet
is the integration across functions that is required for environmental management
to be successful in any one function. This facet comes through most clearly in
Handfield and Melnyk's (1995) definition of environmentally responsible
manufacturing. Looking at the other definitions, it is also easy to see that
integration is required across functions for environmental management to be
successful at any one function. For instance, Carter, Ellram and Ready’s (1998)

definition of environmental purchasing discusses how purchasing activities within

21

supply chain management affect the finn’s use of various environmental options.
Given this, it is quite easy to envision that the decision to implement
environmental purchasing will affect the manufacturing (through the potential use
of different parts and materials), logistics (via the logistics systems required for
recycling internal materials and customer returns) and finance (due to the
potential of different payment streams caused by recycyling) functions. Thus,
even in these functional applications of environmental management, a key
implication appears to be some level of finn-wide integration, which has already
been identified as key aspect of environmental management.

Growing Interest in Environmental Issues

The recent increase in research interest in environmental management
can be traced to the work of Porter and van der Linde (Porter 1991; Porter and
van der Linde 1995). In these articles, they discussed how environmental
regulations, rather than hurting competitiveness, can actually help
competitiveness by inspiring innovation. Though they were primarily referring to
environmental regulations, which is only one piece of the environmental
management research stream, these pieces helped to spur other researchers to
investigate this emerging competitive issue.

There are market forces that have increased the interest in environmental
management as well. Dechant, Altman, Dowining and Keeney (1994)
summarized three of these forces. The first is the need to stay ahead of
environmental regulations. As the authors noted, governments continue to put

out new environmental regulations. Further, a firm that is more involved in

22

environmental planning may avoid having to adopt a “fire-fighting” mentality for
dealing with the regulations.

The second force Dechant et al summarize is stakeholder activism. This
refers to environmental activist groups putting pressure on firms to clean up their
production processes, community groups demanding less pollution, stockholder
groups or mutual funds set up specifically to make environmentally sound
investments, and even employees demanding cleaner workplaces. There are
also any number of surveys that indicate consumers are willing to buy “cleaner”
products (e.g., Rosewicz (1990)), but it has been noted that sometimes
consumers seem to be more willing to say they will buy cleaner products than to
actually buy the products (Melnyk and Denzler 1996).

The third force is competitive pressures. Simply put, if a firm sees its
competitor gaining an edge via environmental management, it is likely to attempt
to replicate its competitor’s actions. As Melnyk and Denzler (1996) noted, “waste
is waste”. Thus, if one firm is using environmental management to eliminate
waste (and thereby perhaps decrease cost and increase quality), Dechant et al
have simply noted that this finn’s competitors are likely to adopt a similar course
of action. In essence, this third force is at the heart of the primary problem in
operations management - trying to make your processes more efficient while still
delivering what the customer wants. F inns are now beginning to see
environmental management as not only a way to lessen pollution and the
attendant risks of regulatory violations, but to make their firm more efficient in its

use of resources.

23

As evidenced by the increasing amount of literature on the topic of
environmental management, this concept is becoming more prevalent. In
particular, the survey that this dissertation uses found significant interest and use
of environmental management programs (Melnyk, Calatone, Handfield,
Tummala, Vastag, Hinds, Sroufe, Montabon, and Curkovic 1999). Steger (2000),
in a meta-analysis of 24 different surveys, also noted the spread of
environmental management, in particular the increase in certified environmental
management systems. The tone that many researchers have adapted in their
writing is one that perceives the acceptance of environmental concerns as a
standard part of doing business. The only question according to most authors, is
the degree to which firms have embraced proactive programs in order to take
more control over their environmental impacts. (See Hart (1997), Ben'y and
Rondinelli (1998), and Brown and Karagozoglu (1998) for just a few examples of
this tone of writing.)

This new literature stream has started to produce some empirical
evidence that environmental management is beneficial for firms. This type of
evidence will most likely stimulate future research and practitioner interest.
There are many anecdotal accounts of how firms have used environmental
management to achieve various benefits. Gupta (1995) recounts many of these
stories. WIth regard to more systematic research, Klassen and McLaughlin
(1996), analyzing environmental award announcements, and Feldman, Soyka
and Ameer (1997), using a proprietary environmental rating system, both found a

positive relationship between environmental management and stock price.

24

Karagozoglu and Lindell (2000) found support for their hypothesis that a finn’s
achievement of environmental competitive advantage would have a positive
impact on the company’s financial performance. Even if “today’s managers lack
a framework that will allow them to turn their good intentions into reality“ as
Walley and Whitehead (1994) claimed, these studies are the type that get the
attention of top management, as they demonstrate positive bottom-line effects of
environmental management.
Characterizing the Environmental Management Literature

In reviewing the environmental management literature, two important
characteristics of the literature are worth noting, as they have a very large effect
on the field. The first is that environmental management can be described as a
fractured field. That is, the literature covers a great variety of disciplines. To a
certain extent, this is a reflection of the fact that environmental management is by
its very nature a cross-functional undertaking. This fragmentation can be seen in
the discussion of the functional applications of environmental management

above. This phenomenon can also be illustrated by referring to Table 2.1.

25

 

Academy of Management Executive
Accounting Horizons

 

 

American Business Law Journal

 

American Industrial Hygiene Association Journal
BioCycle
Engineering Management Journal

 

 

 

Environmental Quality Journal

 

Harvard Business Review

 

Journal of Environmental Planning and Management

 

Journal of Investing

 

Journal of Operations Management

 

Management Accounting

 

Management Science

 

Professional Safety

 

Public Administration Review

 

Research Technology Management

 

Review of African Political Economy

 

Risk Management

 

Scientific American

 

The British Journal of Administrative Management
The International Journal of Quality and Reliability

 

 

 

 

TABLE 2.12 REPRESENTATIVE LIST OF OUTLETS FOR ENVIRONMENTAL MANAGEMENT

LITERATURE

Table 2.1 represents just a small sample of the journals where
environmental management articles have appeared. As can be seen, there are a
variety of journals across many business disciplines, as will as journals from the
hard sciences and public policy. To distill this feature of the environmental

management literature down to one word, it could be described as “wide”.

26

The second characteristic about the environmental management literature
that is important to note is that it is a relatively new field in terms of academic
research. As indicated by Steger (2000), US. firms themselves did not get very
serious about environmental issues until the mid-eighties. As he notes, the early
nineties saw the “Earth Summit’ (held in Rio de Janeiro) and the creation of the
ISO committee which created the ISO 14000 series. Thus, the vast majority of
the research literature in environmental management did not start appearing until
the nineties and in fact most of the empirical pieces did not appear until the mid-
nineties. This gives the environmental management literature a “shallow”
appearance, in that many issues have not been investigated very deeply.

There are two important implications of these two characteristics. The first
is that the amount of environmental literature is relatively small compared to
other fields. This appearance is exacerbated by the fragmentation of the
literature. To speak of “gaps” in this literature is almost superfluous, as there are
many areas that have not been subject to rigorous examination. The second
implication is that there is a relative paucity of previous research models, so
researchers in this field have many avenues for developing new models. In sum,

due to these factors, this dissertation will be an important contribution to the field.

Defining and Measuring the Environmental Management
System
In order to define the environmental management system, one can turn to

ISO 14000, which is series of intematicnal standards for environmental
management systems. Cascio (1996) offers a very detailed description of how

the ISO 14000 series evolved. This description makes two things very clear.

27

First, the coordinated effort to provide firms a structure for managing
environmental impacts resonated with business and government leaders from
around the world. Second, by releasing the standards for the environmental
management system first, the lntemational Organization for Standardization
highlighted the importance of the environmental management system to firms’
environmental efforts.

Later in his book, Cascio provides a very detailed description of an
environmental management system, taken from the official definition. He says
that

...(environmental management systems) should provide

organizations with the framework to do the following:

. Establish an appropriate environmental policy, including
a commitment to prevention of pollution;

. Determine the legislative requirement and environmental
aspect associated with the organization's activities,
products and service;

0 Develop management and employee commitment to the
protection of the environment, with clear assignment of
accountability and responsibility;

0 Encourage environmental planning throughout the full
range of the organization’s activities, from raw material
acquisition through production distribution;

0 Establish a disciplined management process for
achieving target performance levels;

0 Provide appropriate and sufficient resources, including
training, to achieve targeted performance level on an
ongoing basis;

. Establish and maintain an emergency preparedness and
response program;

0 Establish a system of operation control and maintenance
of the program to ensure continuing high levels of system
performance;

28

0 Evaluate environmental performance against the policy,
objectives and targets, and seek improvement where
appropriate;

. Establish a management process to review and audit the
environmental management system and to identify
opportunities for improvement of the system and resulting
environmental performance;

. Establish and maintain appropriate communications with
internal and external interested parties; and

0 Encourage contractors and suppliers to establish an
environmental management system

AS can be seen, this definition is basically a checklist of what an
environmental management system should do. This type of definition should not
be surprising, as the idea of ISO 14001 is that firms will be audited against it in
order to determine if they have met the certification guidelines.

The very detailed Checklist offered by Coscia has been boiled down by
some authors to its five key elements. As noted by Fenante and Cotter (1999),

these elements are:

A

. Establishing an environmental policy;

2. Planning a course of action;

3. lmplenting the environmental management system;
4. Checking and conective action;
5. Conducting management review.

Note how these items closely minor the primary functions of management
as one might find them in an introductory management textbook. In a similar
way, Stapleton, Cooney and Hix (1996), in their guide for small and medium —
sized organizations, defined an environmental management system as “a

continual cycle of planning, implementing, reviewing and improving the actions

29

that an organization takes to meet its environmental obligations.“ Here is another
definition that relies upon the standard activities of management.

In their research piece, Feldman, Soyka and Ameer (1997), listed “several
key elements“ of an environmental management system. These were given as
policy, planning process, resources and implementation, progress measurement,
performance results and periodic review and reporting. While this appears to be
a reasonable list, it seems likely that these authors were somewhat unwilling to
give their complete definition due to a need to not disclose the proprietary rating
process they used in their study. Again, however, a certain similarity to the basic
functions of management can be seen.

Steger (2000) broadly defined this as “a transparent, systematic process
known corporate-wide, with the purpose of prescribing and implementing
environmental goals, policies and responsibilities, as well as regular auditing of
its elements.” As he stated in his article, he wanted to offer a broad definition
and he has done that, while still indicating what some of the specific elements of
an environmental management system are (goals, policies, etc).

Relationship of EMS to EM

It is important to clarify the relationship between environmental
management and the environmental management system in order to understand
which is or is not a subset of the other and which is or is not necessary and/or
sufficient for the other. Unfortunately, accomplishing this delineation may not be
as straightforward as one might surmise. For instance, in Environmental

Manager, an interview appeared with an expert on environmental management

30

(Anonymous 1995). In an article entitled “The Elements of Environmental
Management,” under a sub-heading of “Key Components of an Environmental
Management System”, the expert listed what the unknown author referred to as
“...elements that every environmental program should have.“ As can be seen,
this unknown author appears to be using the terms environmental management
and environmental management system (and for that matter, environmental
management program) somewhat interchangeably. (By the way, the elements
were a data management system, a facility audit system, internal planning
program, capital expenditure review system, willingness to address problems that
are discovered, regulatory and legislative action system and management
accountability).

Thankfully, other authors have been able to offer some clarification about
the relationship. As noted by a joint publication of CEEM, Inc. and the American
Society for Quality (1998), “An EMS is essential to an organization’s ability to
anticipate and meet growing environmental performance expectations and to
ensure ongoing compliance with national and lntemational requirements.“ This
would seem to indicate that an environmental management system is a
necessary, but not necessarily sufficient condition for environmental
management.

Shrivastava and Hart (1994) describe the relationship in a similar way.
They discuss the need for a firm to have “systems thinking“ in order for the firm to
be considered a “green“ organization. They go on to specifically describe how

formal systems are a key element of a “green“ organization. From their

31

argument, one can again View the environmental management system as a
necessary but not necessarily sufficient condition for environmental
management.

The key to understanding the importance of the environmental
management system within an environmentally responsible management is the
word system, as this connotes a coordinated, integrated effort. This is different
than the idea of choosing among various environmental tools and options. The
place of the environmental management system in environmental management
is analogous to the place of a materials requirements planning system or
enterprise resource planning system in the finn’s planning function.

This research will focus on the idea of environmental management as the
way in which a firm manages and controls its environmental impact. The
environmental management system is the coordinated, finn-wide effort that gives
the firm the ability to manage and control its environmental impact.
Measurement of Environmental Management Systems

In turning to the literature to develop a framework for studying
environmental management systems, one will find that to date there has not
been much work on this subject. This is very puzzling, given its critical
importance to a finn’s overall environmental management efforts. Further, the
recent release of the ISO 14000 series, which describes the standard for an
environmental management system, should emphasize the significance of an
environmental management system to a finn’s environmental management

efforts. As discussed above, an environmental management system is to

32

environmental management what a materials requirements planning system is to
requirements planning. Thus, reviewing frameworks from the materials
requirements planning literature (or similar streams, such as accounting
information systems, decision support systems, etc.) is expedient as a starting
point for developing an environmental management system research framework.
Reviewing frameworks from other fields in this manner is consistent with the idea
of “speculation” as discussed by Hunt (1991), whereby assumptions (in this case
that other information systems models are relevant to the study of environmental
management systems) and hypothetical models (the models that will be derived
from the information systems literature) are used to deduce generalizations.

Is it a Management System?

An environmental management system can be viewed as a bundle of
capabilities, much like other management systems (materials requirements
planning, enterprise resource planning, etc.). By capabilities, this dissertation
means the ability to accomplish management functions, rather than specific
technological features. Potential implementers of an environmental management
system must determine which capabilities to implement. Unfortunately for firms
considering implementing an environmental management system, there is not as
much literature describing success factors as there is for other kinds of systems.
This is in direct contrast to other fields, where such research is readily available.
(For instance, see Sum, Ang and Yeo (1997) for an example of a prescriptive

materials requirements planning implementation piece.)

33

Bajwa, Rai and Ramaprasad (1998) used the model in Figure 2.1 as their
research framework. This model is especially useful as a building block for this
research due to two things. First, they were interested in the adoption (or lack
thereof) of executive information system capabilities. Thus, rather than focusing
on specific tools or hardware, there were more interested in the management
capabilities of a system. Second, in focusing on the capabilities of the executive
information system, they categorized these capabilities into the different functions
of management (communication, coordination, control and planning, in their
terminology). Thus, rather than grouping capabilities by more traditional
functional groupings (accounts receivable, inventory, shipping/receiving, etc.) or
by technical specifications (such as file Size limits, number of concurrent users,
etc.) the authors used groupings that reflected the tasks that management as a
whole had to accomplish. This type of measurement model typifies what is
meant by what will be refened to as the “management functions“ framework in

this dissertation.

 

Slze Related Factors

 

 

 

 

 

 

-Finn Size
olS Department Size
EIS Adoption for
Indus Sect r
.335,” 0 Communication 32;??th
oManufacturing —> ’COOfdlnatIOI'l H ~Dis't‘rirgulfed
oGovemment ’COWO' .
-Planning rDecentralIzed

 

 

 

 

 

 

 

 

 

I

Environmental

Characteristlcs
~Dynamice
0Heterogeneity
-Hostility

 

 

 

 

FIGURE 2.1: BAJWA, RAI, RAMAPRASAD 1998 RESEARCH MODEL

In their empirical investigation of the relationship between environmental
management system, environmental performance and firm performance,
Feldman, Soyka and Ameer (1997) used a model of environmental management
systems that was similar, in that it viewed the system as a tool for accomplishing
various management functions. Though they did not explicate their model
completely, it is apparent that their choices of management functions were
guided by the ISO 14001 specification. Their model is also noteworthy as it
appears to be one of the few operationalizations of an environmental
management system in the research literature.

The idea of modeling environmental management systems in terms of the
functions of management has the advantage that these functions and thus this

type of model are familiar to both academics and practitioners. In fact, this type

35

of management functions framework has been around at least since Fayol
(1930), who labeled the functions planning, organization, command, co-
ordination and control. This sort of framework is now commonly found in
introductory management texts.

The primary disadvantage of this type of model would be lack of previous
applications of this type of framework to environmental management systems.
This problem is endemic to the field and was anticipated when the research was
launched. This dissertation intends to overcome this potential drawback simply
by offering a theory-grounded new model.

This type of model for an environmental management system would
simply be operationalized in terms of the functions of management. Some
difficulty might be encountered when determining which items to include for each
function, but if the items (for a survey) are written clearly enough, this difficulty
can be mitigated.

Is it an lnfonnation System?

One theme that is quite common in the environmental management
systems literature is to view it as an information system, as compared to a
management system. This theme comes through quite clearly from a number of
authors. For instance, Shrivastava and Hart (1994) discuss the idea of “’green'
information systems“. Fitzgerald (1993; 1994; 1997; with Orlin and Swalwell
1993) appears to rather consistently view environmental management systems
as information systems. In this view, the environmental management system is

perceived as having (or not having) specific technical capabilities (e.g., ability to

36

compile certain reports) rather than accomplishing management functions (e.g.,
planning).

The primary advantage of this framework is that there are many examples
of operationalizing different systems (executive information systems, accounting
lnfonnation systems, decision support systems, etc.) in the information systems
literature. Thus, researchers would most likely be comfortable with the idea of
operationalizing environmental management systems in this manner. Also, it is
argued that the word “system“ for many managers is often connoted as
“computer system“. Thus, for this group as well, such an Operaticnalizaticn
would be relatively easy to comprehend.

The primary disadvantage of this framework is that it may be selling the
environmental management system short. That is, the environmental
management system can definitely be viewed as more than just an information
system. A quick read of the ISO 14000 documentation (well, perhaps a quick
read of a summary of the ISO 14000 documentation) would lead many to
conclude that an environmental management system is a finn-wide program, not
just the information system for a program.

For this type of framework, one would expect to see the environmental
management system operationalized in terms of specific capabilities. Examples
of these might include the ability to produce reports, the ability to track best in
class environmental lnfonnation and the types of lnfonnation stored in the

database.

37

EMS Success Factors
It is apparent from the lack of model development that to date the

research literature has not investigated environmental management systems
very thoroughly. Despite this, the research literature has made some attempts to
explicate what factors might affect the presence, structure and success of
environmental management systems.

Given the different operating characteristics and material flows different
industries have, industry membership could have a great beefing on a firm's
environmental risks. Rondinelli and Vastag (1996) discuss one method of
evaluating a finn’s environmental risks and prescribe environmental policy
according to where firms fit on a two by two grid that they developed. Though
they do not explicitly say so, it could be argued that the industry a firm is in has a
great bearing on what Rondinelli and Vastag refer to as endogenous and
exogenous environmental risks. Based on the evaluation, the authors suggest
policies with labels of crisis preventive, reactive, strategic and proactive. As the
names imply, different levels of effort and resource expenditures are associated
with each type of policy. Obviously, these different levels of resource allocation
would affect the finn’s decision to implement an environmental management
system and the capabilities included in the implementation.

When evaluating the strength and efficacy of any system, there is a long-
held belief that larger firms have an advantage relative to smaller firms due to the
larger firms’ ability to commit a greater absolute amount of resources to the
particular system under study. Choe (1996) discussed this concept when he

noted the relationship between firm size and lnfonnation system performance.

38

The basic argument is that larger firm size should be conelated with higher
information system performance. In the area of environmental management,
Spicer (1977) analyzed the relationship between market risk and pollution control
performance and found that his results seemed to indicate that larger firms have
more resources to purchase and install pollution control equipment. Choe did
note, however, that some studies have contradicted this relationship between
firm size and system performance. In fact, Rappaport and Flaherty (1992)
indicated no particular relationship between firm size and development of
environmental management systems, which would contradict the conventional
wisdom on the relationship between firm size and system performance. Perhaps
it is the case that smaller firms, by virtue of the greater flexibility they have due to
a small bureaucracy, can be more aggressive in implementing environmental
management systems as a way to gain a competitive advantage over their larger,
less nimble competitors.

Another intuitive argument would be the relationship between the age of a
system and its efficacy. Basically, this is the learning curve argument that the
more experience a firm has with a system, the more effective it is. For instance,
Choe (1996) presented the research model found in Figure 2.2 for studying the
implementation and performance of accounting information systems. This model
looks at the influences of accounting information system success. Note the
inclusion of “evolution level of IS” in the model. This is basically a measure of
where on the learning curve the firms information systems are. The evolution

level was found to have a great influence in his study.

39

Influence Factors

 

 

 

 

Top management support rAIS P rform ance
Technical capability 0f 13 personnel a User AIS satisfaction I
User involvement A User AIS use

User training & education A

Steering committees

Location of IS department

Formalization of system development

Organization size

 

 

 

 

 

 

Evolution level of IS ‘-—-~

 

 

 

FIGURE 2.2 CHOE 1996 RESEARCH MODEL

In reading the environmental management literature, one might assume
that all environmental management systems must be certified, as it often seems
that authors rarely talk about environmental management systems outside of the
context of the ISO 14000 series. However, it is definitely not the case that firms
are required to certify their environmental management system, any more so
than they are required to certify their other systems. In fact, the argument that a
corporate-wide certification may be inappropriate for some firms fueled an article
by Rondinelli and Vastag (1996). In a later article (2000), they go on to note that
many firms have chosen not to have their environmental management system
certified. lntuitively, it might seem that attempting to achieve one of the available
certifications (BS7750 - United Kingdom; EMAS — European Union; ISO 14000 -
world) might lead a firm to take its implementation of an environmental

management system more seriously and thus do a better job of implementing

4O

and using it. However, this proposition has not been tested in the literature.
Feldman, Soyka and Ameer (1997) get close to testing it, but since they do not
divulge how they measured environmental management systems, it appears safe
to say that this question is yet to be answered. Rivera-Camino (2001) tested this
type of hypothesis, but it is unclear how he measured the concept of
environmental management system implementation.

In general, how the above factors affect the implementation, structure and
performance of environmental management systems has not been explored
much. As there have not been many attempts at even measuring or modeling
environmental management systems, this is not surprising. This dissertation

aims to fill this important gap in the study of environmental management.

Concluding Remarks
This literature review has demonstrated the increased interest in the

research area of environmental management and the fact that it is a relatively
new field. The “newness” of the field means that there are still many questions
waiting for investigation. Despite this, there have been some efforts to begin
investigations. For instance, a number of authors have attempted to define what
is meant by “environmental management“. As was seen, the consensus
definition is a finnwide effort to minimize environmental impact of the finn’s
processes throughout the supply chain and the life cycle of the finn’s products
and services. This literature review also noted the attempts that have been
made at defining “environmental management system“. While a very detailed

definition is available in the ISO 14000 documents, a more concise consensus

41

definition would be that this is the system that allows a firm to plan, review,
control and change its environmental impacts.

This literature review also discussed the relationship between
environmental management and environmental management system. Despite
some blurring of definitional lines that occurs in the literature, after some
consideration this relationship is relatively easy to understand. As is discussed,
either explicitly or implicitly in most of the environmental management literature,
the environmental management system is a necessary but not necessarily
sufficient condition for environmental management.

This chapter also discussed two general frameworks for measuring an
environmental management system. The first views the environmental
management system as a management system that desires to accomplish the
general functions of management (as defined by Fayol (1930)). This model
offers the nice feature that there is much familiarity with these functions of
management. The second framework was based on the idea that an
environmental management system is an lnfonnation system. This idea has
appeal due to the reams of previous literature on information systems that can
help guide its formation. Both frameworks appear to be theoretically legitimate.
That is, it is perhaps difficult to say that one is better than the other. Rather, it is
probably best to characterize both frameworks as being valid views of the same
phenomenon.

While this is a relatively new field, some authors have attempted to put

forth some arguments as to what factors might affect the implementation,

42

structure and performance of environmental management systems. This
literature highlighted some of the key arguments here. In particular, industry
membership, firm size, the age of the environmental management system and
the presence (or absence) of ISO 14000 certification have been discussed as
being large influences on environmental management systems.

Unresolved Questions

As noted earlier, environmental management is still a relatively new
research field. Thus, when speaking of “gaps“ or “contradictions“ in this field, this
issue of the newness of the field dominates the discussion. That is, since the
field is so new, there are many gaps in the field, as many important research
questions have yet to be fully explored, if they have even been explored at all.
Further, since there are so many gaps, there are really no contradictions in the
literature. This is not saying that there is consensus in the field, rather, there is a
lack of baseline theory with which to argue. However, it is still appropriate and
reasonable to point out which are some of the most important areas in
environmental management in need of further study.

For instance, this literature review discussed some studies that linked
environmental management to stock price. While stock price is certainly an
important measure of a firm, this construct does not specifically test the finn’s
competitive position, especially the four dimensions of operations strategy - cost,
quality, speed and flexibility. While it is fair to say that a number of conceptual
arguments have appeared regarding environmental management’s effect on the

dimensions of operations strategy, there has been very limited empirical testing

43

of these arguments. As was demonstrated by a series of articles in the late
1980s, culminating in Ferdows and DeMeyer (1990) these dimensions are critical
to the operations management field. Yet, the effect of environmental
management on these dimensions has not been given the attention it deserves.
Though this dissertation is not Specifically aimed at investigating this question, it
is highlighted here due to its importance to the field.

Another issue critical to this study also appears to be relatively
unanswered in the current literature. Specifically, there have been very few
attempts to measure environmental management systems. In research that
includes environmental management systems as part of its research model, the
environmental management system is typically treated as a binary phenomenon -
either the firm has one or it does not. A very exhausting, if not exhaustive,
literature search found that attempts to measure the capabilities and structures of
environmental management systems have been very limited.

Also, despite the increasing importance of environmental concerns in the
business world, there have been very few efforts in the research literature to test
what affects the structure and performance of environmental management
systems. This hole in the literature is particularly glaring due to its importance to
practitioners. The next chapter will start the process of answering some of these

questions by developing some models of environmental management systems.

Chapter 3

MODEL DEVELOPMENT

Introduction
This chapter will present the theoretical frameworks that will drive the

dissertation research. As will be seen, these frameworks will be based on
concepts already discussed in the literature review. After a further discussion of
the relationship between environmental management and the environmental
management system, the two models that will be used in the analysis will be

developed.

Link Between Environmental Management and
Environmental Management Systems
The previous chapter elucidated many of the components of

environmental management. These included tools like life cycle analysis and
options like recycling. In order to round out a full picture of what the framework
of instituting environmental management might look like, one could add many of
the usual prescriptions for program implementation, such as top management
support and an information system. (See Rivera-Camino (2001) for a discussion
of this.) As was noted in the literature review, firms can, to a certain extent, pick
and choose which components of environmental management they are going to
use. That is, they do not have to have every possible component in order to say
that they use environmental management. The literature review also noted that

one of the motivations for firms to consider environmental management is

45

competition, so there is definitely a potential return in unit performance that firms
need to weigh when considering implementing an environmental management
program.

Combining these ideas into one framework shows how environmental
management fits into the general competitive environment. Figure 3.1 has been
developed to do just this. It is a modified version of a previously developed
framework (Montabon, Calatone, Melnyk, and Sroufe 2000).

 

External Factors
- Competitors/Industry
- Government Regulations

 

 

 

 

 

 

 

 

 

 

 

- Stakeholders
Environmental Management Unit Performance
-EnVIronmental Management System . .
-F1nanCIal
-Tools -D .
. ~Operational
°OptIons .
. °Env1ronmental
Internal Factors dnformatron
- Resources available/size of
firm

0 Culture (commitment to
environmental responsibility)

0 Management style
(willingness to take risks)

0 Top management support

a Past experience with other
initiatives/programs

 

 

 

FIGURE 3.1 GENERAL FRAMEWORK OF ENVIRONMENTAL MANAGEMENT RESEARCH

AS can be seen, this framework shows some of the various factors, both
internal and external, that influence a finn’s decision regarding implementation of
environmental management. Depending on these factors, a firm may choose to
fully implement environmental management, implement it partially or gradually, or

may choose to ignore it completely. Whatever form of environmental

46

management is chosen, it will have an effect on various aspects of unit
performance.

As might be surmised from the literature review, the testing of which
components are most important to the success of a finn’s environmental
management efforts has yet to be fully explored, though this will make for some
interesting future research. Despite this, as the previous Chapter highlighted, one
of the themes in the literature is the need for an environmental management
system for any firm attempting environmental management.

AS was also mentioned in the literature review, many firms have chosen to
implement environmental management along more functional lines. One could
argue that this is merely an example of a firm picking and choosing which
components of environmental management to implement. (This argument will be
left for future research.) One of the themes in the literature regarding functional
applications of environmental management is the need for an environmental
management system. For instance, in Honnozi’s (1999) article on
remanufacturing and environmentally conscious manufacturing, there is a great
deal of discussion about ISO 14000. From this article, it is hard to escape the
implicit conclusion that ISO 14000 certification (of a finn’s environmental
management system) will at the very least be necessary to fully take advantage
of remanufacturing and environmental conscious manufacturing.

Though the environmental management system is best conceptualized as
a firm wide construct, it can certainly serve as the basis of a functional

application of environmental management. However, because the environmental

47

management system is a firm wide construct, it is unlikely to be contained within
just one function’s application of environmental management.

When one considers the firm wide nature of an environmental
management system and its importance to any finn’s environmental
management effort, a key premise of this dissertation becomes clear. To wit:

Premise 1: Every complete implementation of environmental management
must have an environmental management system, but having an
environmental management system is not sufficient to achieve
environmental management.

In other words, the environmental management system is a necessary,
but not sufficient condition for environmental management. It is recognized that
while this premise is reasonable and in fact based on arguments from the
research literature, this premise itself could be tested in future research. It is
argued here that this premise would be borne out by such future research. Given
this premise, a usable framework for measuring environmental management

systems must be derived.

Model Development
This research intends to examine the capabilities of environmental

management systems. By “capability”, this research is refening to the system’s
ability to assist the firm in accomplishing certain tasks or functions necessary for
the management and control of environmental impacts. The capabilities used in
this research are latent variables that are defined by various manifest variables.

Thus, the capabilities in this research are perceptual and do not have objective

48

measures, such as one might use to describe a computer system (e.g., capable
of storing 3O gigabytes of data).

As originally described by Fayol (1930), management must accomplish
certain functions. The first measurement framework to be developed in this
research defines the capabilities in terms of the functions of management, similar
to the Bajwa et al (1998) model (see Figure 2.1). Bajwa et al defined
management in terms of four management functions. The revised measurement
framework, which will be refened to as the “management functions” model,
defines management in terms of five functions, taken from Melynk and Denzler
(1996) (please see Figure 3.1). These five functions (plan, analyze, organize,
implement and control) will be the latent constructs in the model. Appendix A
displays how the questions on the survey were used to measure these latent
constructs. (The survey will be described in great detail in the next chapter).
Thus, this model views the environmental management system as a tool for
accomplishing these functions of management relative to the finn's

environmental impacts.

49

    
   
   

Planning

Controlling

  
  

Directing!
Implmenting

   
 

Organizing

FIGURE 3.1 THE FUNCTIONS OF MANAGEMENT AS TAKEN FROM MELNYK AND DENZLER

(1996)

The second approach uses the following definition of an environmental
management system from the ISO 14001 specification to guide the development
of the latent variables (Lamprecht 1997).

That part of the overall management system which includes
organizational structure, planning activities, responsibilities,
practices, procedures, processes and resources for developing,
implementing, achieving, reviewing and maintaining the
environmental policy.

Based on this definition, eight latent constructs have been formed

(database, measurement, documentation, feedback, problem solving,

50

procedures, reporting, supply chain). This will be refened to as the “systems
activities“ model. Appendix B displays how the survey questions were assigned
to each of the latent constructs. Essentially, this model views the environmental
management system as batch of capabilities. When a firm chooses to adopt a
particular system, it can pick and choose amongst the available capabilities. (For
instance, consider a materials requirements planning systems, where it has been
shown that the choice of capabilities varies among firms (Cerveny and Scott
1989).) This model of the environmental management system views
management as picking and choosing amongst the relevant capabilities.

Premise 2: For each model, all elements exist.

This premise is put forth as it is argued by this research that as the models
have been developed from existing literature, there is little reason to believe the
proposed latent constructs will not be found. Certainly, this premise could be
tested, but this is not the focus of this research, so it will be left to future
research. F urther, future research could also test the strength of the different
constructs. That is, future research could endeavor to test which constructs are
used more and which have greater efficacy.

Premise 3: Each model may be viable.

One of the primary contributions of this dissertation is to develop and test
a measurement model for environmental management systems. To this end, two
potential models have been developed and will be tested. The reason for

producing more than one model is to help ensure that the best possible model is

51

found. It is quite possible that the analysis will show that that both models are

viable, though one may be shown to have superior statistical properties.

Measurement Model Hypotheses
Formally stated, the hypotheses regarding the measurement models

developed above are as follows.

H1 nu": Neither of the proposed frameworks adequately measures
the environmental management system.

H1....m.....: The management functions framework best measures
the environmental management system.

H1...m.,...,: The systems activities framework best measures the
environmental management system.

Hum: Both frameworks are equally good at measuring the
environmental management system.

The specifics of how these models will be tested and how they will be

compared to see which one is superior both statistically and theoretically will be
explained in detail in the next chapter. It is expected that Hm... will be rejected,
however, it is unclear as to which alternate will be accepted because the models
are not orthogonal. That is, they are not using mutually exclusive sets of
manifest variables.

This hypothesis is important for both practitioners and researchers. For
researchers, these hypotheses are of crucial importance, as the ability to
measure an environmental management system will be an important starting
point for theory development. For practitioners, defining and understanding the

structure of environmental management systems is important as this will help to

52

further the knowledge of environmental management systems, and will help to
improve firms’ implementations of environmental management systems.

Again, it is expected that mm... will be rejected and thus provide this
crucial starting point to both practitioners and researchers. However, even if
Ht..." cannot be rejected, this is still an interesting result. This type of result
would lend evidence to the idea that environmental management systems cannot
be measured, whereas the measurement of other types of systems (ERP, MRP,
etc.) abounds in the research literature.

Comparison Hypotheses

Once the measurement model frameworks have been analyzed, one of
them will be used to test a series of comparisons. These comparison tests will
be based on arguments that were presented in the literature review. These tests
are aimed at understanding how certain factors affect the implementation and
efficacy of environmental management systems.

H2m...: The industrial setting of a firm has no influence on the
structures and capabilities of a finn’s environmental
management system.

One of the more popular arguments in the environmental management

literature stream has been that firms that are in industries that have a greater
potential environmental impact (whether perceived or actual) will be more
concerned about environmental issues and will thus devote more resources to
environmental issues. Similarly, this research hypothesis aims to test how
industry membership effects the capabilities of a finn’s environmental

management system. The expectations of the literature are that this hypothesis

53

would be rejected, because the industrial setting of a firm does in fact influence
its environmental management system.

H3...,..: F inn size has no influence on the structure or efficacy of
environmental management systems.
As discussed in the literature review, there is a long-held belief that there

is a relationship between firm size and information system performance.
Specifically, it is theorized that larger firm Size should be conelated with higher
lnfonnation system performance. The usual explanation for this is that a larger
firm has greater resources to allocate to any particular issue than a smaller firm
does. Thus, this hypothesis has been formulated to test this issue for
environmental management systems. This question is important for
practitioners, as it will help them understand the relationship better and possibly
help them to understand how firm size may effect the implementation of and
results from an environmental management system. For researchers, this
proposition, no matter which way the results turn out, will help to clarify the
debate over firm size and system development. As indicated in the literature
review, most researchers would expect this hypothesis to be rejected due to the
fact that larger firms Should have more resources with which to build their
environmental management system.

H4Mn: Greater experience with an environmental management
system will have no influence on operational performance.
The genesis of this hypothesis is the well-documented idea of the learning

curve. This hypothesis looks to measure its effects on environmental

management systems. The expectation of the learning curve would be that firms

that have had their environmental management system in place for a longer
period of time would demonstrate better operational performance. That is, it
would be expected that this hypothesis would be rejected. However, the entire
area of environmental management has not been a top priority for firms until
recently. Thus, the analysis may in fact show results that are not in line with
learning curve theory. A result of this nature would be a good motivator for future
research to find out an explanation. Note that this hypothesis will be of great
interest for practitioners, as it will give them some indication of how long the pay-
off period for an environmental management system is.

H5nuuz The absence of ISO 14000 certification is associated with no
differences in the relationships among environmental
management system capabilities and performance.

The number of ISO 14001 certificates that have been issued in the US. is

relatively low when compared to other industrialized nations. Yet even a cursory
preliminary analysis of the database reveals that a number of firms have an
environmental management system in place. Thus, one interesting comparison
will be to determine the differences between those firms that have certified their
environmental management system and those that have not. Either way, the
result of this particular analysis is interesting, as it will invigorate the argument of
whether or not attempting ISO 14000 certification is worthwhile. Logically, one
would expect that this hypothesis would be rejected, as the ISO 14000
documentation provides a blueprint for a finn’s environmental management
system and thus this documentation should help Speed the implementation

process.

55

Summary
This chapter has presented a development of two measurement models

and four comparison tests, all of which are based on existing literature. Now that
the models and hypotheses have been set up (see Table 3.1 for a summary of
the hypotheses), the next chapter will describe in detail the methodology used to
analyze these hypotheses, as well as a complete description of the data

gathering process.

 

H1,,,,..: Neither of the proposed frameworks adequately measures the
environmental management system.

 

H2”: The industrial setting of a firm has no influence on the structures and
capabilities of a finn’s environmental management system.

Ham“: Firm Size has no influence on the structure or efficacy of environmental
management systems.

H4”: Greater experience with an environmental management system will
have no influence on operational performance.

 

 

 

H5”: The absence of ISO 14000 certification is associated with no differences
in the relationships among environmental management system
capabilities and performance.

 

 

 

TABLE 3.12 SUMMARY OF RESEARCH HYPOTHESES

56

Chapter 4

RESEARCH METHODOLOGY

Introduction
In this chapter, the operationalizations, data gathering methods and

analysis methods will be described. First, the operationalization of the two
measurement models will be described in more detail. The data source and how
the data was gathered will be described in detail. The latter sections of this
chapter will deal with the analysis methods to be used. As was briefly mentioned
in the previous chapter, the analysis will have two distinct stages, which will be

described in more detail in this chapter.

Measurement Model Operationalization
In order to operationalize the management functions framework, each

question (refened to by its variable name) was placed in the appropriate latent
construct. This was a relatively straightforward process, as many of the survey
questions were designed to look at how various aspects of environmental
management interact with the general functions of management. Thus, it was
relatively easy to determine which questions related to planning, analyzing, etc.
Figure 4.1 Shows the results of this operationalization. Each of the five latent
constructs appears in bold. The manifest variables associated with each latent
construct appear below it. A cross-reference of the variable name to survey

items is in Appendix A.

57

 

 

Plan
EMSeinf
EMSbict
EMSprod
EMSproc
EMSequip
EMSpdprc
EMSlay
EMSloc

Analyze
EMSccost
EMSqualC
EMSsched
EMSefect
EMScause
EMSreasn
EMSsupp

Organize
EMSforml
EMSdoc
EMScirC
EMSdept
EMSrpt
EMSpos
EMSachv
EMStrain
EMSdist
EMSaware

Implement
EMSdata
EMSbicc
EMSvis

Control

EMSeperf
EMS goals
EMSrecyc
EMSedept
EMSepers
EMSesupp

 

Figure 4.1 Proposed Constructs for Measuring Environmental Management

System - Management Functions Model

The hypotheses associated with this model are below. These specific

hypotheses were inspired by the idea of viewing a system as a way of

accomplishing the specific functions of management. This idea is exemplified by

the Bajwa et al (1998) model discussed in the literature review. In this study, the

five functions of management specified by Melnyk and Denzler (1996) form the

five hypotheses for this measurement model.

Hmf1: There exists a plan capability of an environmental

management system.

Hmf2: There exists an analyze capability of an environmental

management system.

Hmf3: There exists an organize capability of an environmental

management system.

Hmf4: There exists an implement capability of an environmental

management system.

Hmf5: There exists a control capability of an environmental

management system.

58

 

For the systems activities framework, a similar process was followed.
Some of the survey questions were designed to track the different aspects of the
ISO 14001 specification. Thus, determining which variables to use in this
framework and where they should be placed was relatively straightforward.
Figure 4.2 shows how the systems activities model was operationalized. Again,
the latent constructs appear in bold, with the manifest variables for each latent
construct listed below it. Appendix B contains an excerpt of the survey with each

question mapped to a variable.

 

Database Measurement Documentation Feedback

 

EMSsumm EMSecost EMSforml EMSprod
EMScacct EMSqualc EMSdoc EMSproc
EMSeperf EMS goals EMSequip
EMSdata EMSrecyc EMSpdprc
EMSeinf EMSlay
EMSbict EMSloc
Problem Solving Procedures Reporting Supply Chain
EMSefect EMScirC EMSdept EMSsdec
EMScause EMSbicc EMSrpt EMSsupp
EMSreasn

 

 

Figure 4.2 Proposed Constructs for Measuring Environmental Management
System — Systems Activities Model
The hypotheses associated with this model are below. These hypotheses
were derived from the ISO 14000 definition of an environmental management
system.

Hsa1: There exists a database capability of an environmental
management system.

59

A database capability allows the firm to process and review performance
relating to environmental impact.

Hsa2: There exists a measurement capability of an environmental
management system.
As has been stated by any number of scholars, it is difficult to manage

something that is not measured. This is also indicated in the idea of “reviewing
and maintaining” from the ISO 14000 definition. This measurement capability
construct follows directly from the definition.

Hsa3: There exists a documentation capability of an environmental
management system.
The documentation capability relates to the procedures and processes of

the system which allow for the development, review and maintenance of the
environmental management system.

Hsa4: There exists a feedback capability of an environmental
management system.
The feedback mechanism of an environmental management system

allows management to review environmental performance.

Hsa5: There exists a problem solving capability of an environmental
management system.
Problem solving is a critical procedure for an environmental management

system. Environmental issues are often connoted as “problems”, thus the
environmental management system needs to be able to help solve the problems.

Hsa6: There exists a procedures capability of an environmental
management system.
This hypotheses comes directly from the ISO definition.

60

Hsa7: There exists a reporting capability of an environmental

management system.

The reporting capability refers to the structure in place to allow lnfonnation
to flow to the proper people within the firm so that they can review and maintain
the system.

Hsa8: There exists a supply Chain capability of an environmental

management system. '

Certainly, environmental concerns for any particular firm do not end at the

factory walls. To a certain extent, this is part of the thinking behind life cycle
analysis. Thus, firms need in place an environmental management system

capability that accounts for this.

Data Collection

The Survey
A sixteen page, 250 item survey was developed in 1997 by a team of

researchers. The survey was designed to gather lnfonnation about:

a Knowledge of ISO 14000 at firm

0 Degree of firm commitment to ISO 14000

. Use and capabilities of environmental management system

. Environmental options and programs in use

0 Benefits and costs associated with various environmental initiatives

. Barriers to implementation of various environmental management

programs

This survey was pre-tested with a group of fifteen managers using a three-

round process over a period of two months in 1997. The pre-test was used to

give the research team a “reality check" of the questionnaire to make sure that

61

the questions were clearly worded and that the survey would have interest for
those in industry. The primary concern noted by the pre-test group was the
length of the questionnaire. Based on this feedback, the research team worked
to pare down the questionnaire to what was believed to be a minimum number of
questions. A lengthy survey was necessary, as most of the concepts covered in
the questionnaire were not well defined in the research literature.

It should be noted that the research team was aware that the issue of the
environment might cause socially desirable responding by potential informants.
Social desirability bias, as it is also known, is the tendency in respondents to
create a favorable impression, either consciously or unconsciously (Zikmund
1991). To help mitigate socially desirable responding, the research team used
two particular techniques suggested in the literature. First, the questionnaire and
related material quite clearly stated that responses were anonymous. As noted
by Zerbe and Paulhus (1987), anonymity reduces socially desirable responding,
so the respondents were given the opportunity to remain completely anonymous
if they so desired. Also, each question was formulated with the goal of ensuring
that each item was neutral with respect to social desirability, as Zerbe and
Paulhus also recommended. In particular, questions were formulated in a way
that emphasized business issues, rather than social or public policy issues.
Given these measures and that fact that some researchers have found social
desirability contamination to be not very widespread (Ganster, Hennessey, and
Luthans 1983), it is argued here that it is unlikely that this bias affected the

results of the analysis.

62

The survey consisted of five major sections. The first section asked for
information about the respondent, their position, professional affiliations (if any),
and extent of involvement in various corporate initiatives. The second section
focused on the business unit and detail about it such as products manufactured,
extent of uncertainty facing the business unit and its personnel, and the status of
various initiatives. The third section dealt with the perceived impact of the
lSO/QS 9000 certification process on the business unit and its competitive
position in the market place. In section four, the respondent was asked to
evaluate a series of questions pertaining to ISO 14000. These questions
assessed the level of knowledge of the respondent on the ISO 14000 certification
process, as well as the factors affecting its implementation and use. The fifth
and final section gathered lnfonnation about the business unit’s environmental
management system, the effectiveness and efficiency of this system and the
types of options used to improve environmental performance. This section will
be of primary importance to this research. At the very end of the questionnaire,
respondents were given some free-fonn space to describe any obstacles,
potential or realized, to their firm implementing ISO 14000.

The Sample

A mailing list of 5000 names each were obtained from three professional
associations (National Association of Purchasing Management, American
Production and Inventory Control Society and one group who wishes to remain
anonymous) for a total of 15,000 names. The lists were checked for duplicate

names and the few that were identified were eliminated. Where possible, the

63

professional associations were asked to provide names of managers who worked
for manufacturers, i.e., those in the two-digit Standard Industry Classification
code range of 20 to 39.

A major American manufacturer provided an additional list of 104
managers at six of their facilities. Three waves of mailings were sent out, in what
is often called the modified Dillman (1978) method. The survey was sent out in
the fourth quarter of 1997 and responses were received well into 1998. 1510
usable responses were obtained, for a response rate of 10.35%. Mile this is
lower that the 20% that researchers strive to achieve, it is possible that the length

of the survey discouraged some potential respondents.

Two Stage Analysis Methodology
The analysis will be done in two stages. Stage one will determine which

of the two potential models of measuring environmental management systems
has a better fit. The primary determinant will be the results of the statistical
analysis from the measurement model analysis. In particular, the fit indices from
the structural equations modeling analysis will be used to determine statistical
superiority.

As indicated in the discussion of the measurement model hypotheses, it is
a possibility that both models will prove to be viable based on this first criterion.
In this case, the fit indices will be examined to see if there is noticeable gap
between the two models. However, it could be the case that the fit indices

indicate that both models do an equally good job in measuring environmental

management systems. For example, if both models have fit indices above .90,
this would be a clear indication that both models are useful.

If this were to become the case, some type of tiebreaker will be needed.
The first tiebreaker that will be used is the to determine which model is more
theoretically tenable. At this point, however, that is very difficult to determine, as
both models were developed based on existing literature and a preference for
one over the other would be based mostly on personal taste, rather than
theoretical arguments. Perhaps once all model fitting and modification is done,
one model will be able to be considered more theoretically tenable (for purposes
of tie breaking), but this is impossible to predict at this point.

Using both the statistical results and the issue of theoretical tenability
points to the issue of examining the baseline structural models. Certainly, the fit
indices from this analysis will help to show which model is more useful. Further,
if any computational problems appear in this analysis, this would be an indication
that the model with the problems should fade from further consideration for use in
the multigroup analyses.

A second tiebreaker would be to examine the structural models to see
which offers more explanation for the variance in firm performance. If one model
does a better job of explaining variance in performance, this model would be
considered superior, in accordance with the usual rules for evaluating the

usefulness of theory (Bacharach 1989).

65

Operaticnalizaticn of Comparisons
Once this is done, the general research model shown in Figure 4. 1 will be

used in stage two. The figure is referred to as “concise” as the manifest
variables are not shown for sake of saving space, and the environmental
management system capabilities have been collapsed into one “meta-latent”

referred to as “EMS”.

 
   
 
 
     

Firm Factors

  

Operational
Performance

FIGURE 4.1 “Concrse” RESEARCH MODEL

This model is the very familiar “systems-performance” model, as seen in
numerous studies involving any type of system, be it information systems or
quality systems. Basically, this type of model looks to study the relationship
between a particular system and some measure of that system’s performance or
the effect of that system on some measure of firm performance. For other
examples of this type of model, see Arthur (1994) on human resource systems
and manufacturing performance, Grover, Teng, Segars and Fielder (1998) on
information systems and productivity, Choe (1996) on accounting information
systems and accounting information systems performance, Forza (1995a; 1995b)

on quality information systems and quality performance and Flynn and Flynn

66

(1999) on manufacturing information systems and manufacturing performance.
Similar to many of these studies, the current research model incorporates the
firm size construct in order to reflect the fact that the environmental management
system is unlikely to account for all the variance in operational performance.

The specific hypotheses to be tested are listed below.

H2nu..: The relationships of environmental management system
capabilities to operational performance will not be
significantly different across industry membership groups.

H3nuuz The relationships of environmental management system
capabilities to operational performance will not be
significantly different across different firm sizes.

H4...» The relationships of environmental management system
capabilities to operational performance will not be
significantly different based on how much experience a firm
has with its environmental management system.

H50“: The relationships of environmental management system
capabilities to operational performance will not be
significantly different based on the presence or absence of
ISO 14000 certification.

For these hypotheses, rejecting the null would mean that all of the

relationships from the environmental management system capabilities to
operational performance are variant. Partial support would refer to the situation
where at least one, but not all of the relationships from the environmental
management system capabilities to operational performance are variant.

It should be noted that many other comparisons, based on conjectures in
the literature, could also be analyzed. For instance, one interesting comparison

might be to analyze the presence or absence of ISO 9000 certification on a firm’s

67

environmental management system. This comparison would argue that a firm
that is already experienced in implementing a formalized process system such as
ISO 9000 may be able leverage its experience in creating its environmental
management system. This and many other possible comparisons will be left for

future research.

Industrial Membership
As was noted earlier, a popular argument in the environmental

management literature has been the idea that industry setting has an effect on a
firm’s environmental policies and performance. Unfortunately, few articles have
articulated which specific industries are considered more high risk. Further, the
current literature has not offered much guidance as to how researchers could
parse industries into various risk categories. However, it should be noted that
the issue of placing a rating on an individual finn's environmental performance
has certainly been covered in the literature. For instance, the Toxics Release
Inventory database, compiled by the US. government has been used by a
number of studies as a measure of environmental performance (Clelland, Dean,
and Douglas 2000; Dooley and Fryxell 1999; Sigman 1996; Hamilton 1995).
Kinder, Domini and Lyndenberg (1996) use a more subjective rating technique
for a firm’s environmental performance in their compilation of corporate social
performance (for an interesting discussion of their technique, see (Sharfman
1996)). Their database has also been used in a number of studies, though most
of them were concerned with corporate social performance (Simerly and Bass

1998; Griffin and Mahon 1997). Unfortunately, these types of individual firm

68

ratings could not be used with this research’s database, as the respondents were
anonymous. This being said, it was necessary to devise a way to break up the
data set into industry groups that could be viewed as having different
environmental issues.

Fortunately, the standard industrial classification (usually known as “SIC')
coding system provides a reasonable argument for dividing up the data set.
Manufacturing firms are classified from 20XX to 39XX. As the numbers increase
from 20XX to 39XX, the types of manufacturing included in each category move
from the more basic to the more complicated. A reasonable breaking point is
3400. Firms from 2000 to 3399 tend to be involved in more basic industries,
such as lumber, food and clothing. Firms from 3400 to 3999 tend to be involved
in relatively more complex products, such as vehicles, industrial machinery and
electronic equipment.

Two data files were created based on this breakpoint. In the original data
file, 326 responses were from standard industrial classifications 2000 to 3399
and 970 responses were from standard industrial classifications 3400 to 3999.
(Note that the remaining 163 responses did not give an indication of their industry
membership.) After list-wise deletion, 266 cases remained in the “low SIC" data
and 736 cases remained in the “high SIC” data.

Size of Firm

Of the 1459 firms in the database, 1413 reported their number of full time

equivalent employees, while only 1272 reported their annual sales. In order to

maximize the size of the data sets for the firm size multi-group analysis, full time

69

equivalent employees was used to determine the breakpoint for categorizing the
data set into two groups, referred to as “small firms“ and “large firms". The
dataset was split into thirds. Those firms with 219 or less employees were put
into the small firmgroup and those firms with 800 employees or more were put
into the large firm group. These cut—offs meant that 369 firms were in the “small

firms" group and 366 firms were in the “large firms" group.

Experience with Environmental Management System
To create the data files for the comparison based on age of environmental

management system, the entire dataset was divided into thirds based on the self-
reported ages of the firms environmental management systems. 817 firms
reported this value. The three groups created had an environmental
management system from zero to three, four to nine, or ten or more years. The
group with environmental management systems that were zero to three years old
are referred to as the “young“ group, while the group whose environmental
management system was 10 or more years old is referred to as “old“. After
listwise deletion, there were 222 firms in the “young” group and 216 firms in the
“old” group.
Presence of ISO 14000 Certification

The survey asked respondents to indicate the status of various initiatives
at their firm. They were asked to use the following scale for each initiative:

1 = Not applicable

2 = Not being considered
3 = Future consideration

4 = Assessing suitability

70

5 = Planning to implement

6 = Currently implementing

7 = Successfully implemented

Two of the initiatives respondents were asked about were an

environmental management system and ISO 14000. Firms that responded with
a 5, 6 or 7 regarding environmental management system and a 2, 3 or 4
regarding ISO 14000 were placed into one group (“EMS-Yes ISO 14000-No“).
After list-wise deletion, 406 firms were in this group. Firms that responded with a
5, 6, or 7 for both environmental management system and ISO 14000 were
placed in another group (“EMS-Yes ISO 14000-Yes”). After list-wise deletion, 88

firms were in this group.

Statistical Methodology

Stage 1
Confirrnatory factor analysis will be used via structural equation modeling.

Note that there are two potential factor models to be analyzed. One will be
based on a definition taken from the ISO 14000 documentation (the systems
activities model), the other will be based on the five activities of management (the
management functions model). Both confirmatory factor analyses will be
examined to determine which has the better fit, which will be determined by the
usual and customary structural equation modeling standards of fit indices. As
discussed earlier, fit indices which are not noticeably different will require that the

theoretical tenability and structural model computational suitability be assessed.

71

Stage 2

This stage of the research will use structural equation modeling multi-
group analysis. The method for creating the groups was discussed above.
When looking at the relationships across the groups, the betas will be examined.
Multi-group structural equation modeling is an appropriate technique for this type
of analysis due to the use of latent constructs and the desire to compare the

relationships among the latent constructs across groups.

Primary Statistical Technique
The primary statistical analysis method for this study will be structural

equation modeling. As described by Hoyle (1995), “(structural equation
modeling) is a comprehensive statistical approach to testing hypotheses about
relations among observed and latent variables.” Structural equation modeling is
a very comprehensive approach that allows the user to specify a measurement
model and the relationships between latent constructs. According to Hoyle, one
of the advantages of structural equation modeling is “that it requires researchers
to think carefully about their data and to venture hypotheses regarding each
variable.” Thus, this particular analysis methodology requires that the researcher
have a model design a priori, rather than attempting a fishing expedition based
on exploratory type of analysis.

As noted by West, Finch and Curran (1995) structural equation modeling
has become very popular over the last 20 years for researchers in social and
behavioral sciences. It can be argued that this popularity tracks the rise in
personal computers closely, as structural equation modeling analysis does call

for computing power. West et al go on to note that this enthusiasm has led some

72

to jump to using structural equation modeling without applying it correctly. Given
this, it is important to note that structural equation modeling is well suited to this
particular research due to the need to formulate and test a measurement model
and then take that measurement model to test the relationship among latent

constructs.

Limitations of Proposed Methodology
However, this is not to say that the use of structural equation modeling will

be entirely problem-free. Structural equation modeling is a very complex
technique. This complexity does offer the advantage of providing a large amount
of analysis and related statistics, however, there are still a number of
disagreements about technical matters regarding various estimating techniques
and interpretation. Some of these issues will be reflected in the discussion of the
analysis.

The primary limitations of this proposed methodology have to do with the
survey itself. As was noted, the survey was created without the benefit of very
many pre-tested environmental management constructs. Thus, it is fair to say
that many of the questions were brand new inventions. A danger of this
approach is that once the results are analyzed, it may become obvious that there
are some problems with the questions. This was the reason for the extensive
pre-test that the research team did on the questionnaire. Despite this, however,
one concern that the research team could not mitigate was the perception that
the general level of knowledge about environmental management generally and

ISO 14000 specifically was low among American managers.

73

Summary
This chapter has laid out the plan for collecting and analyzing the data.

Two distinct measurement models were developed for environmental
management systems, along with hypotheses to test them. The multi-group
comparisons were described along with their origins in the literature. Further, a
description of the survey and mailing protocol was given. From the data

collected in this survey, the hypotheses will be tested in the next chapter.

74

Chapter 5

ANALYSIS

Introduction
This chapter will simply present the results of the various analyses and will

discuss the procedures used to obtain the results. Discussion and interpretation

of the results will be left for Chapter Six and Seven.

Pre-Processing of Data
As was mentioned in the discussion of the survey, a major American

manufacturer supplied the research project with a list of potential respondents at
six of their locations in addition to the lists obtained from the three professional
associations. Of the 104 sent to this manufacturer, 57 were returned. Figure 5.1
shows the status of ISO 14000 at the all the respondents’ firms. The variable
TBLISO14 refers to a survey question that inquired about the status of ISO
14000 at the respondent’s firm. This variable has values from one to seven, with
the following meanings:

1 = not applicable

2 = not being considered

3 = future consideration

4 = assessing suitability

5 = planning to implement

‘ 6 = currently implementing

7= successfully implemented

75

As shown in Figure 5.1, a total of 204 firms were either planning to
implement, currently implementing or had successfully implemented ISO 14000.
Of these 204 responses, 56 (27.5%) came from the one major American
manufacturer. Thus, while 13.5% of the firms as a whole fell into one of these
three categories, 98.2% (56 of 57) of the responses from this particular firm were
in these categories. This was due to the commitment that this manufacturer had
made to ISO 14000. It can be argued that the presence of these 56 responses in
the data set gives this particular firm undue influence on the analysis of the entire
data set, especially any analyses that focus on those firms which have made a
commitment to achieving ISO 14000 certification. Thus, the responses from this
major American manufacturer were averaged by the six locations from which
they came. This resulted in six data points from this firm appearing in the final
data set, rather than 57. Thus, the sample size for the data set was 1459 rather

than 151 O.

76

 

500'

3100'l

2004

100i

 

 

 

 

Frequency

 

TBLISO14

FIGURE 5.1 STATUS or= ISO 14000 AMONGST ALL RESPONDENTS

In the preliminary data analyses, the data for the variables representing
full time equivalent employees (FT E) and annual sales (SALES) were
demonstrated to be highly nonnorrnal. SALES had a skewness of 7.788 and a
kurtosis of 71.576. FTE had a skewness of 25.621 and a kurtosis of 808.240.
These descriptive statistics are highly indicative of a nonnonnal distribution
according to West, Finch and Curran (1995). This caused all types of problems
in the preliminary analyses using structural equation modeling. As was noted by
West et al, one of the generally accepted methods for this situation is to use a
power function of the variable to produce a new transformed variable that more
closely approximates normality. In this case, a simple log transformation was
used for both SALES and FTE, resulting in variables that very closely

approximated normality.

77

Stage One: Determination of Model
Structural equation modeling was used to determine which of the two

potential models fit better. In the preliminary analysis of the data, it became quite
apparent that the data set was multivariate nonnormal. Depending on the
particular analysis being done, it was common to obtain Mardia’s Coefficients of
400 to 500. Normalized Mardia coefficients were typically 150 to 250. As Bentler
(1999) has indicated, a normalized Mardia cofficient above four indicates
multivariate nonnormality. Given this, elliptical re-weighted least squares
estimation was used for all structural equation modeling analysis, as indicated by

Sharma, Durvasula and Dillon (1989).

Management Functions Model
In the confirmatory factor analysis for this model, boundary constraints

were found to exist between the latent constructs for Organize and Implement.
The Lagrange Multiplier test also indicated problems with the relationship
between these two constructs. Thus, these latent constructs were combined into
one construct. This is deemed to be theoretically tenable, as the constructs are
very much related to one another, as they are both aimed at getting ready to
implement and actually implementing. The name of the combined construct is
Organize/Implement, in order to lessen confusion.

Table 5.1 shows the factor loadings and associated tvalues for this
measurement model. As can be seen, all of the factor loadings are quite high
and the (values are also quite high. The Lagrange Multiplier test did not indicate

any significant cross-loadings. The comparative fit index for this model was .945,

78

measurement model works well.

which is above the usual cutoff of .900. All of these statistics indicate that this

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Factor Factor

Item Loadan tVaIue Item Loadlng tVaIue
EMSEINF, .796 26.390 EMSFORML, .842 28.873
Plan Organize!

Implement
EMSBICT, .756 24.598 EMSDOC, .846 29.084
Plan Organize!

lmplernent
EMSPROD, .700 22.215 EMSCIRC, .826 28.072
Plan Organize!

Implement
EMSPROC, .802 26.668 EMSDEPT, .762 25.093
Plan Organize!

Implement
EMSEQUIP, .788 26.015 EMSRPT, .781 25.973
Plan Organize!

Implement
EMSPDPRC, .861 29.585 EMSPOS, .863 29.944
Plan Organize!

Implement
EMSLAY, .838 28.444 EMSACHV, .846 29.106
Plan Organize!

Implement
EMSLOC, .734 23.626 EMSTRAIN, .840 28.780
Plan Organize!

Implement

EMSDIST, .878 30.726

Organize!

lmglement
EMSECOST, .872 30.194 EMSAWARE, .848 29.186
Analyze Organize!

Implement
EMSQUALC, .877 30.490 EMSDATA, .860 29.783
Analyze Organize!

Implement
EMSSCHED, .746 24.158 EMSBICC, .835 28.511
Analyze Organize!

Implement
EMSEFECT, .518 15.486 EMSVIS, .866 30.087
Analyze Organize!

Implement
EMSCAUSE, .729 23.456
Analyze
EMSREASN, .722 23.142 EMSEPERF, .898 31.797
Analyze Control
EMSSUPP, .560 16.942 EMSGOALS, .923 33.166
Analyze Control

EMSRECYC, .547 16.622

Control

 

 

 

79

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ACTCOST, .861 29.360
Performance
ACTLT, .904 31 .606
Performance
ACTQUAL, .910 31.965
Performance
ACTBENE, .686 21.491
Performance

 

 

 

TABLE 5.1 CONFIRMATORY FACTOR ANALYSIS OF MANAGEMENT FUNCTIONS

 

 

 

EMSEDEPT, .812 27.337
Control
EMSEPERS, .762 25.028
Control
EMSESUPP, .708 22.175
Control
LOGFTE, .915 22.522
Firm Size
LOGSALES, .821 20.756
Firm Size

 

 

 

MEASUREMENT MODEL

The hypotheses for the management functions framework are:

Hmf1: There exists a planning capability of an environmental

management system.

Hmf2: There exists an analyze capability of an environmental

management system.

Hmf3: There exists an organize capability of an environmental

Hmf4: There exists an implement capability of an environmental

management system.

management system.

Hmf5: There exists a control capability of an environmental

These hypotheses were all supported, with the caveat that the organize
and implement capabilities were combined into one construct. This support for
the hypotheses was demonstrated in the results by high factor loadings and a

lack of cross-loadings, indicating a measurement model that was both strong and

valid.

management system.

80

 

Systems Activities Model
The confirmatory factor analysis for this model indicated that there were

problems with the Procedures latent construct. In particular, this construct was
shown to have a bound constraint in its relation to the Database latent construct.
The Lagrange Multiplier test suggested that there were some cross-loadings
between the Database latent construct and the manifest variables for the
Procedures construct. However, these cross-loadings are theoretically
untenable. After trying a number of adjustments to the measurement model to
rectify this situation, it was determined that dropping the Procedures construct
altogether eliminated the problems.

After this adjustment, the Lagrange Multiplier test indicated no significant
cross-loadings. The comparative fit index for the measurement model was .980,

which is indicative of an excellent fit.

81

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Factor Factor
Item Loadlnl tvalue Item Loading tvalue
EMSSUMM, .926 33.417 EMSEFECT, .608 18.796
Database Problem
Solvirg
EMSCACCT, .768 25.345 EMSCAUSE, .979 36.164
Database Problem
Solving
EMSEPERF, .913 32.678 EMSREASN, .939 33.803
Database Problem
Solving
EMSDATA, .857 29.640
Database
EMSEINF, .852 29.354 EMSDEPT, .902 31.209
Database Repoang
EMSBICT, .821 27.812 EMSRPT, .897 30.977
Database , Remrting
EMSECOST, .875 30.447 EMSESUPP, .905 30.391
Measurement Supply Chain
EMSQUALC, .876 30.514 EMSSDEC, .882 29.346
Measurement Supply Chain
EMSGOALS, .929 33.449
Measurement
EMSRECYC, .528 15.910 LOGFTE, .914 23.630
Measurement Firm Size
LOGSALES, .822 21.639
Firm Size
EMSFORML, .953 34.627
Documentation
EMSDOC, .952 34.565 ACTCOST, .856 29.147
Documentation Perforrnanoe
ACTLT, .905 31.718
Performance
EMSPROD, .704 22.340 ACTOUAL, .916 32.288
Feedback Performance
EMSPROC, .820 27.474 ACTBENE, .683 21.398
Feedback Performance
EMSEQUIP, .825 27.726
Feedback
EMSPDPRC, .890 31.058
Feedback
EMSLAY, .884 30.748
Feedback
EMSLOC, .732 23.473
Feedback

 

 

 

 

 

TABLE 5.2 CONFIRMATORY FACTOR ANALYSIS OF SYSTEMS ACTIVITIES MEASUREMENT

MODEL

The hypotheses for this model are:

82

Hsa1: There exists a database capability of an
environmental management system.

Hsa2: There exists a measurement capability of an
environmental management system.

Hsa3: There exists a documentation capability of an
environmental management system.

Hsa4: There exists a feedback capability of an
environmental management system.

Hsa5: There exists a problem solving capability of an
environmental management system.

Hsa6: There exists a procedures capability of an
environmental management system.

Hsa7: There exists a reporting capability of an environmental
management system.

Hsa8: There exists a rewards capability of an environmental
management system.

Hsa9: There exists a supply chain capability of an
environmental management system.

The results from the measurement model would indicate that there is
support for these hypotheses, with the exception of Hsa6, the procedures
capability. However, this support is tempered by this model’s failure in the
structural analysis. This may be due to the analysis technique used. Thus, it
would seem that while this operationalization of an environmental management
system does have some merit, it might be the case that this particular
operationalization needs further modification in the form of additional manifest

variables if it is to be used in a structural equation modeling analysis.

83

Further modification of this model was tried in an attempt to overcome its
computational difficulties. These modifications were designed to rearrange the
manifest variables into a set of latent variables that was in line with the definition
used for the systems activities framework and would help to decrease the
potential risk of Heywood cases. Unfortunately, these attempts were not
successful, primarily due to lack of theoretical tenability. These issues will be
discussed further in the discussion in chapter six.

Selection of Model

Both measurement models had excellent fit statistics. Picking between
the two models based on fit statistics alone is very difficult, as there is really little
difference between a comparative fit Indices of .945 (management functions) and
.980 (systems activities). To a great extent, this situation requires a tiebreaker,
as was discussed in chapter 4.

Both models undenlvent some revision during the measurement model
analysis. The management functions model had two latent constructs combined.
The systems activities model had a latent construct dropped. This would argue
for using the management functions framework, as it is not only measuring more
concepts, but it has stayed closer to its original formulation.

Further, it can be anticipated that the systems activities model will have
computational difficulties when the structural models are examined. This is
because this model has more latent constructs that are measured by just two

manifest variables. As Anderson and Gerbing (1984) noted, this type of model

specification runs the risk of having an improper solution. This would argue
against going fonnard with the systems activities model.

Based on the superior theoretical tenability and computational suitability,
the management functions model will be used in the multigroup comparisons.
However, before delving into those analyses, the structural models of both
measurement frameworks will be examined, in order to provide a baseline for the

multgroup comparisons.

Stage Two: Comparisons

Measurement Models
For all of the multi-group comparisons, measurement model results are

not reported here. This is because there is no reason to believe that the
measurement models will behave differently for each sub-group than they did for
the entire sample. This was borne out in the muIti-group measurement model
analyses that were run. Based on this, for all multi-group analyses presented
here, the betas were constrained, while the lambdas were not.
Baseline Structural Models

Though the management functions framework will be used to investigate
the multigroup comparisons, both measurement frameworks’ structural models
were analyzed in order to further ensure that the correct model had been chosen
and to get a baseline for comparison to the multigroup analyses. Figure 5.2
shows the structural model using the management functions framework. The
factor loadings were very strong and the model has an excellent comparative fit

index of 0.943.

85

 

EMS . I I EMSbict EMSeinf

.802 .752
.700 .793

864
' .33I m =
lama -735 @
369 m
EMSecost /

m .885 -5.067
EMSefect .903
h.
0

.708

EMScui EMS '
.788

ii

All factor loadings significant I p < .OI
‘° = significant at p < .05
‘ = significant at p < .I0
CFI = .943

 

FIGURE 5.2 STRUCTURAL MODEL FOR MANAGEMENT FUNCTIONS MODEL

The analysis of the systems activities structural model was somewhat
problematic. As discussed previously some of the latent constructs are

measured by only two manifest variables, thus running the risk of improper

86

solutions. That is what happened, with the result being some negative variances,
commonly referred to as “Heywood cases“. Further analysis and investigation of
this outcome indicated that the problem was most likely in the Firm Size
construct. (Though as Dillon, Kumar and Mulani (1987) note, “it may be
extremely difficult to detect the exact cause of the problem”) Various attempts
were made to eliminate the problem with the improper solution. These attempts
are summarized in Table 5.3. In this table, the term “condition codes“ refers to
the warning messages provided by the E08 software, thus indicating that the

solution cannot be trusted.

codes

inequalities to

error variance for
and to .
error variance
logsales and logfte to .05,
them

error variance
logsales and logfle to .05
for

SA = Systems Activities model
MF = Management Functions model

 

TABLE 5.3 COMPARISON OF VARIOUS STRUCTURAL MODELS

One method of overcoming this problem was to fix the error variances for
the LOGFTE and LOGSALES variables to .05. The theoretical argument for this
is that while the other manifest variables used are mostly perceptual measures
and thus should have their error variances freely estimated, it is reasonable to
suspect that most respondents were reasonably accurate in reporting their firms

full time equivalent employees and annual sales, as these are objective numbers

87

readily available in any number of firm reports. When these error variances were

fixed, the results in Figure 5.3 were obtained.

 

 

    
 

.857 9'3 .932 53
7" '
.858 463”
EM .321 o
.958
m 952
m ~
.4“.
m .950 393,,
normal
m 955 Documentation
EMSproc IEMSprod I
.82 .
m ’°’ 36m III,
.829
m 8,, MM 896m
- .357"
.638
m 480 .724
.273" m
m
m -°°°
.997
.923 .mee
All factorloadingssignificantatp<m
. "=3i3nificaflatp<.05
m 9'9 ...Wmmlo
m err-.804
m:
.877 .273”
- a:
.909 SuppIyChain
m

FIGURE 5.3 STRUCTURAL MODEL FOR SYSTEMS ACTIVITIES WITH LOGSALES,

LOGFT E ERROR VARIANCES FIXED TO .05

A review of Table 5.3 Clearly points to the superiority of the management

functions structural model. Though it is true that some of the systems activities

88

models that were analyzed do have superior fit statistics, these models cannot be
used due to the existence of improper solutions. Further, even when a “fix” was
used to eliminate the Heywood case, the model was still inferior to the unaltered
management functions model. It is interesting to note that when this same “fix“
was applied to the management functions model, the management functions
model was still superior. This being said, it should be noted that the
Confirrnatory Factor Analyses for both models indicated that the measurement
models have good potential for the proposed latent constructs.

AS this analysis of the baseline structural model indicates, the
management functions model will be used going forward. One result of the
single group management functions model structural analysis that should be
noted at this point is the large and statistically significant beta values from the
Firm Size construct to the four environmental management systems constructs.
This would be a clear indication that firm size plays an important role in
determining the structure and strength of the environmental management
system.

Industrial Setting

In the multi-group analysis, a lower bound constraint was found for the
disturbance term of the Control latent construct for the “low SIC” group only. This
was unusual because no such constraint appeared in the vast majority of all the
analyses, including a single group analysis using just the “low SIC” group.
Further analysis revealed that the relative size of the problem was quite small. It

was concluded that this lower bound constraint was a function of the starting

89

values used in the equations. Unfortunately, despite many attempts, a
combination of starting values that yielded a clean solution could not be found.
Thus, the value of this particular disturbance term was fixed to .073, which was
taken from the single group analysis of the “low SIC” group. This disturbance
term was fixed for the “low SIC“ group only in the multi-group analysis. The

results of this analysis are shown in Figure 5.4.

.946“

Plan

 

2. 686
2. 752

 

856” Perf
Orgmizellmplemen -, 88 6"
4.436
4.48!
.997” Top number = SIC 2000 - 3399
.990” Bottom number = SIC 3399 - 3999

” significant at p < .05

‘ significant at p < .IO
Italics = invuiant across groups
Bold = significantly different across groups
CFI - .938
Disturbmoe term for Control set to .073 for “low SIC” group

FIGURE 5.4 INDUSTRY SETTING MULTI-GROUP

This mUIti-group analysis differed in its results from the analysis of the
entire dataset. The relationships from Firm Size to Performance and from
Control and Performance in the industry setting multi-group analysis have

different directions and a smaller size than their counterparts in the analysis of

90

the entire database. The relationship from Plan to Performance is a different
direction in this multi-group analysis, but it is still small. The relationship from
Analyze to Performance is smaller in this mUIti-group analysis than in the
analysis of the entire dataset. The beta between Organize/Implement and
Performance is larger, but still negative in the multi-group analysis.

With regard to the hypothesis being tested here, the results could not be
more Clear. There appears to be no difference in the relationships between the
environmental management system and performance based on industry
membership. From the point of view of the lntemational Organization for
Standardization, this may be a very useful result, as it indicates uniformity in how
firms use environmental management systems. This type of uniformin is the
idea behind the ISO 14000 series of standards. This result may also be an
indication that no matter what the perceived environmental risks that a firm may
face due to its line of business, the environmental management system from firm
to firm will perform in a similar manner. This does not necessarily mean that it
will be implemented in the same manner, just that the environmental
management system’s relationship to performance will be similar from firm to firm

across industries.

Firm Size
Though firm size is being used as the firm contextual factor in all the

models, it is of interest to explicitly do a multi-group analysis based on firm Size.
The small firm group exhibited a boundary condition for the Performance

construct. In comparison to the totality of all the analyses presented in this

91

research, this was an anomalous result. In keeping with Dillon, Kumar and
Mulani’s (1987) suggestion, this disturbance term was fixed to 0. The results of
the analysis are presented in Figure 5.5.

.905“
Plan .905”

x: v
o.“

.956”
.959“

Perfomumce
. -l. 103 ”
Organize/I mplement _. 994..

-5.0[l “
4.285”

6.810”
5.849”

 
    
 

 

. 993 ”
.997” Top number : S 2l9 employees (n = 369)
Bottom number : 2 800 employees (n = 366)

‘° significant at p < .05
’ significant at p < .10
Italic: = immiant across groups
Bold = significmtly different across groups

CF 1 a .939
Disturbance term for Performmoe for small firms fixed to 0

FIGURE 5.5 SIZE OF FIRM MULTI-GROUP ANALYSIS WTH DATASET SPLIT INTO THIRDS

Due to the slight computational problems found when the dataset was cut
into thirds, another analysis was run using the dataset cut in half. The median
value for full time equivalent employees was 400. Thus the “small firm” group
was made up of those firms with less than 400 employees and the “large firm”
group was made up of those firms with 400 or more employees for this analysis.
After list-wise deletion there were 547 in the small firm group and 553 in the large

firm group.

92

In this analysis, the disturbance term for the Control latent construct
exhibited a lower bound constraint for the small firm group. Similar to the
Industry Setting analysis, this result confounded the result obtained in the single
group analysis and it is concluded again that this is an artifact of the starting
values used. Despite many attempts, a sequence of starting values could not be
found which yielded a Clean solution. Thus, this disturbance term was fixed to
.075, which was its value in the single group analysis. The results of the multi-

group analysls are shown in Figure 5.6.
.90! “

.959”

ov/ a

1

3.854”

 

.974” 3.382
.962“
709
Organize/Implement __ 630
- 2.963
- 2.593
'996" Top number: < 400 employees

.999" Bottom number : 2 400 employees

” significant at p < .05
‘ significant at p < .10
Italic: = invariant across groups
Bold == significantly different across groups

CF I = .943
Disturbance term for Control fixed to .075

FIGURE 5.6 SIZE OF FIRM MULTI-GROUP ANALYSIS IMTH DATASET SPLIT INTO HALVES

93

In comparison to the analysis of the entire dataset, the betas in this multi-
group analysis tend to be smaller. Also, the relationship between
Organize/Implement is larger, thought still negative in this multi-group analysis.

The hypothesis being tested here has no support, as none of the
relationships were shown to be variant across the two groups. Similar to the
previous result, this may be encouraging to the lntemational Organization for
Standardization, because this results again points toward a uniformity of
approach that is embodied in the ISO 14000 series of standards. Combining
these results to the results of the analysis of the baseline management functions
model, it appears that though size does have an effect on the environmental
management system, the multigroup analysis does not give an indication of
effect Size.

Age of the Environmental Management System

In completing the multi-group analyses for invariance of betas, an unusual
result occurred. On the last iteration of testing for invariance, the disturbance
term for Performance was shown to have a lower bound constraint for the
“young“ group. This particular result was especially anomalous given that the
previous iterations indicated that this particular estimate showed no signs of
having any boundary constraint problems. In order to obtain a Clean solution,
this disturbance term was fixed to 1.126 for the “young” group in what turned out
to be the last iteration of invariance analyses. This value was the average for

this particular disturbance term from the previous invariance analyses. This

value itself is an indication that the boundary constraint is anomalous. The
results of the multi-group analysis are in Figure 5.7.

.870“
Plan .847”

.957“
337“ v
4m-
m... .949“ 4375
I905“ J“

 

Organize/lmpl . ..029
3.884“
2.331
.995" Top number= EMS 3orless yerl's old(n= 222)
.992“ Bottom manber-EMS IOOrmore yous old (n=2|6)

“ significant at p < .05

‘significant a p < .10
Italics - invariant across groups
Bold - significantly different across groups
CFI - .9I4

Disturburoe term for Performance fixedto I126

FIGURE 5.7 AGE OF ENVIRONMENTAL MANAGEMENT SYSTEM MULTI-GROUP ANALYSIS

WITH DATASET SPLIT INTO THIRDS

In order to take another look at this analysis, the dataset was divided in
half based on the age of the environmental management system. For this
stratification, those firms with an environmental management system from one to
Six years old are considered “young”, while those from seven years old and
above are considered “old”. This breakpoint was chosen as it divided the entire

data set roughly in half, thus maximizing the number of firms in each group. After

95

listwise deletion, there were 276 cases in the “old” data set and 337 cases in the

“young” data set. The results of the multi—group analysis are in Figure 5.8.

.815“

Pin

 

 

Orgmiullmpl « -.734"
- 1.888
- 1.648
.986“
.984" Topnunber- EMS I-6yeas old

Bottom nunber -= EMS 7 or more yen old

” significant II p < .05
‘ significant I p < .I0
Italic: - immimt across groups

Bold - significantly different across groups
CFI - .900

FIGURE 5.8 AGE OF ENVIRONMENTAL MANAGEMENT SYSTEM MULTI-GROUP ANALYSIS

WITH DATASET SPLIT INTO HALVES

Another method of dealing with the computational difficulties encountered
in the analysis displayed in Figure 5.7 is to fix the offending estimate to zero, a
technique recommended by Dillon, Kumar and Mulani (1987). When this
technique is applied, the results displayed in Figure 5.9 are obtained.

.870”
Mm .847“

232:: ( ®
onQ 3.21?

 

.968“
.905“
Orgmrzell mpleme _ 135
6.230‘
3. 727‘
.995 ” Top number = EMS 3 or less years old (n - 222)
.99!” Bottom number=EMS IOormoreyerus old(n-216)

“ significant a p < .05

‘significant at p < .10
Italics = invariant across groups
Bold = significartly different across groups

CFI - .914
Disturbmoe term for Performance fixed to 0 for young group

FIGURE 5.9 AGE OF ENVIRONMENTAL MANAGEMENT SYSTEM MULTI-GROUP ANALYSIS

WITH PERFORMANCE DISTURBANCE FIXED TO 0 FOR “YOUNG” GROUP

With respect to the hypothesis that was being tested across the three
different analyses, only partial support has been found. In this case, the Analyze
capability appears to have a different relationship to operational performance
depending Upon how much experience the firm has with its environmental
management system. This may indicate a perception among firms that this
particular capability becomes less Important the more experience the firm has
with their environmental management system. This does not necessarily mean

that a finn’s ability to analyze issues related to environmental impacts becomes

97

less important over time, just perhaps that firms rely less on their environmental
management system for this particular capability.
The Presence (or Absence) of ISO 14000 Certification

As described previously, all firms in this analysis indicated that they had
an environmental management system. This subset of firms was further divided
into those who did not have nor were not attempting to achieve ISO 14000

certification and those who had or were attempting to get the certification. The

results of this analysis are displayed in Figure 5.10.

o .891”

 

 

TopnranbeerolSO I4000¢ertific¢ion

Bottom number: Has or mempting to obtain ISO 14000 ocrtific‘ion
” significant at p < .05
‘ significmt at p < .10
Italics 8 inverimt across groups

Bold - significmtly different across groups
CFI = .92I

FIGURE 5.10 PRESENCE OF ISO 14000 MULTI-GROUP ANALYSIS

This multi-group analysis exhibited a few differences in comparison to the
analysis of the entire dataset. The betas for both the relationship from Firm Size

to Performance and from Control to Performance are smaller in the multi-group

98

analysis than they are in the analysis of the entire dataset. The relationship
between Organize/Implement and Performance is both positive and large in this
mUIti-group analysis compared to the Single group analysis.

With regard to the hypothesis being tested, there is only partial support.
The Analyze capability does have a different relationship to Performance
depending upon whether or not the firm has or is attempting to obtain ISO 14000
certification. This could be an indication that firms that obtain the certification do
not place as much importance on this particular capability as those firms who do

not have the certification.

Conclusion
As can be seen, a few bumps in the road were encountered during the

analysis of the data. These “bumps” along with the implications of the results of
the analysis will be discussed in the next Chapter. Overall, one theme that has
emerged at this point is the consistency of the environmental management

systems across the different settings that were tested.

99

Chapter 6

DISCUSSION

Introduction
This Chapter will discuss the results of the analysis and highlight those

results that are interesting or unusual. For instance, perhaps the most interesting
result is the uniformity found among firms’ environmental management systems,
despite the fact the multigroup comparisons placed the firms in groups that had
some significant differences. The implications of the results for both
environmentally responsible manufacturing theory and for actual manufacturing
managers will be discussed. Throughout the chapter, special attention will be
paid the issue of possible alternative explanations for the results, especially with
regard to whether or not the results were influenced by certain biases in the data

or by the analysis method used.

Measurement Model
One of the primary objectives of this research was to develop and test a

measurement framework for environmental management systems. This
objective was key to the entire research as there have been very limited previous
efforts to measure environmental management systems. In an effort to ensure
that the best possible measurement model was being used, two environmental

management system measurement frameworks were derived.

100

Systems Activities Framework
This framework was derived by reviewing the official definition of an

environmental management system taken from the ISO 14000 standards and
attempting to operationalize that definition. The operationalization of the
definition was very straightforward and was theoretically appealing due to its
grounding in the ISO 14001 standard. The results of the measurement model
analysis were quite good. However, this model was deemed to be inferior to the
management functions model despite comparable measurement model analysis
statistics. This was due to its lessened theoretical tenability, which came about
due to the need to drop a latent construct. Further, this model raised concerns
about the possibility of improper solutions during the structural model analysis.

The actual testing of this framework in the structural analysis revealed
these concerns to be legitimate. As was noted in the analysis Chapter, this
measurement framework had great difficulty converging to a Clean and proper
solution when the structural analysis was done.

In such a situation, there are a number of possible reasons that a model
might encounter such difficulties. The first is the data used to test the model.
While the data set did Clearly exhibit multivariate nonnorrnality, it is unlikely that
this caused any problems for this model, especially as elliptically reweighted
least squares estimation was used.

With respect to the analysis method, another issue to consider when
reviewing the results of the analysis for the systems activities measurement
framework is the use of structural equation modeling itself. When just the

measurement model of the systems activities framework was analyzed, it ran

101

cleanly and produced a very good solution with strong loadings and fit indices.
This would indicate that further development of the systems activities model
could be very fruitful, as it seems to have reasonable measurement properties.
However, when the structural model was analyzed, this framework yielded
improper solutions. Given this, it is reasonable to consider if other types of
analysis methods would work with this measurement framework. As just one
example, perhaps a form of step-wise regression using this type of model would
work and yield useful insights. As mentioned above though, the use of elliptically
reweighted least squares estimation should mitigate concerns about structural
equation modeling being the wrong choice for the analysis of this measurement
framework.

Given this, the other potential problem to examine is the adequacy of the
variables. The results from the confirmatory factor analysis give a strong
indication that the variables were adequate in the sense that they were
appropriate measures. What is most likely not adequate is the number of
variables per latent construct. This left the model open to the possibility of
“Heywood cases“, which Is in fact what happened. Among the many alterations
to this framework that were tried in an effort to find a Clean structural solution was
to look at only the latent constructs that made Up the environmental management
system. That is, the constructs for Firm Size and Operational Performance were
dropped and the remaining model analyzed. Unfortunately, even this variation of

the model resulted in improper solutions.

102

In hindsight, it could certainly be argued that such a framework should not
even be tested in the first place, due to the risk of improper solutions. Before
siding with that argument, however, some issues should be carefully considered.
The first issue to consider here is the relative paucity of previous efforts to
measure environmental management systems. Given this, it is important to
derive alternative models in an effort to find the one that fits best. This was the
reasoning behind presenting alternative models in this research. Further, when
deriving these alternative models, a natural and logical place to start is with the
official definition of an environmental management system as taken from the ISO
14000 documents. This was done to create the general framework of the model
and to make sure the grounding of the model was solid. The next step was to
create survey questions that could be used in such a model. The survey and
related research product was and is a group effort looking into a relatively new
and growing field in Operations Management. As such, this meant that there
were many questions the researchers had about all aspects of environmental
management and very little in the way of tested and validated questionnaires on
which to build. Ideally, each of the latent constructs in the systems activities
framework would have at least three observables, or put another way, three
questions in the survey. Due to the various needs of all the researchers involved
with the survey, unfortunately that did not happen. Perhaps a future survey
focused solely on the environmental management system will be able to include
enough questions or observables related to this type of model to overcome the

problems encountered in this analysis.

103

Though it was known that this particular model might exhibit some
problems, it was hoped that the sheer size of the sample could overcome the
potential for improper solutions. Obviously, this did not happen and is an
indication that the Heywood case problem is more a function of model
specification than it is of empirical underidentification (see (Bollen 1989) for a
discussion of empirical underidentification). Again, this would point to the idea
that the structural model is the issue here, not the data or method of analysis. To
the extent that the research team was unable to include more queStions that may
have been useful for the systems activities framework, one could view the
problems with this framework as being caused by missing variables. “Missing“ in
this case refers to the fact that certain questions were not asked, not the sense

that questions on the survey went unanswered.

Management Functions Framework
Given the previous discussion, it is logical that the management functions

framework yielded cleaner results, especially given the fact that the analysis
method was structural equation modeling. Structural equation modeling tends to
reward models that have multiple observables for each latent construct, which
was definitely the case for the management functions framework.

Given the problems described above, it should not be surprising that in
Table 5.3, which displays the results of the baseline structural models, the
management functions structural model is clearly preferred. As can be seen, the
comparative fit index achieved by the management functions model “MF1” is a

very strong 0.943. As was noted above, the systems activities models had

104

difficulty coming to a clean solution. When the variances for both the log of sales
and the log of full time equivalent employees were fixed to 0.05 for the systems
activities model, this model only achieved a comparative fit index of 0.804 (see
the listing for model “SA1k”). As an interesting and insightful means of
comparison, this same “fix“ was applied to the management functions model
(“MF1a'). The comparative fit index for this model was 0.861, another indication
of the superiority of the management functions model over the systems activities

model.

Viability of Both Models
It can be argued that both models have excellent potential for use in

further studies. This argument is based on the results that both frameworks
showed when only the measurement models were analyzed, which Clearly
indicated strong fit indices and high factor loadings for both frameworks.
Certainly, if structural equation modeling is to be used as the analysis method, it
would be appropriate to add more manifest variables to the systems activities
model in order to lessen the risk of Heywood cases. However, the fact that both
measurement models appear to exhibit good measurement qualities is an
important finding for this research as it is an advancement of measurement
technique for the environmentally responsible manufacturing literature.

A possible alternative explanation for why the two frameworks did perform
differently in this analysis is that the set of questions (i.e., manifest variables) that
were used for the systems activities model were understood differently from the

questions used for the management functions models. That is, the managers

105

answering the questionnaires interpreted the questions differently and responded
in kind. However, it is unlikely that this is the case, as many of the manifest
variables (i.e., questions) that were used for the systems activities framework
were also used for the management functions framework. Thus, the issue of
different interpretations by respondents between the two frameworks is not
deemed to be an issue.

Single Group Model

The single group (i.e., entire sample) analysis in Figure 5.2 is worthy of
discussion in order to set up a baseline for comparison to the multi-group
analyses. The relationships between Firm Size and the environmental
management system capabilities are all large and significant. This is an
indication that Firm Size plays an important role in determining the capabilities of
a firm’s environmental management system. These strong relationships are
probably also due somewhat to the fact that in this model, Firm Size is the only
exogenous construct.

Of the relationships between environmental management system
capabilities and Operational Performance, only Analyze has a relationship that is
large, positive and significant. This would be an indication that this particular
capability is of singular importance to the environmental management system.
Though Control exhibits a large positive relationship with Operational
Performance, this relationship was not statistically significant. Similarly, Firm
Size had large, but negative, relationship to Operational Performance, but again,

this relationship was not statistically significant. In the other analyses, it is

106

interesting to note how these two large but statistically not significant

relationships will change depending on the group comparison being made.

Multi-Group Comparisons

Industry Setting
The results clearly indicated that there were no significant differences in

the relationships of the finn’s environmental management system to its
operational performance across the industry groupings. This result is interesting
because it can be argued that this demonstrates that there is no significant
difference on how firms setup and use their environmental management system
based on their industry membership. This analysis and argument is specific to
the environmental management system, so it may still be the case that the
environmental policies that the firms set up may differ, but their environmental
management systems behave in a similar manner.

It is interesting to note that the results showed that at the .10 level of
significance, five of the nine betas were variant across the two groups. These
variant relationships (at the .10 level) are:

0 Plan - Firm Size

0 Analyze — Firm Size

. Control - Performance

0 Firm Size - Performance

0 Control — Firm Size
A reflexive reaction to this result would be to release these constraints and

continue the analysis. However, the structural equation modeling literature is

107

very, very consistent in its use of .05 as the cut-ofi‘ for variance of constraints
(Bollen 1989; Byme 1994). In general, this cut-off appears to be almost a
standard in statistical analysis. For instance, Cohen 8: Cohen (1983) describe
the .05 standard in statistical analysis as “...so widely used as a standard in the
behavioral sciences that it has come to be understood to govern when a result is
said to be statistically significant in the absence of a specified or value.“ As they
go on to explain later in their book, this is primarily due to the need to balance the
Type I and Type II error. This result is still worth noting, however, especially as
four of the five variant relationships (again, at the .10 level) involve firm size.
Perhaps it is the case that there is some type of interaction among firm size and
industry membership that is associated with differences in how firms use their
environmental management systems.

As discussed at the beginning of this section, there have been very limited
efforts in the literature to categorize whole industries along a risky to safe
continuum. It could be argued that the results reported here are merely a
reflection of that fact. That is, the method used to break down the database into
groups was not truly able to distinguish industries based on environmental
riskiness. Though this argument may have some merit, it is argued here that the
method Chosen definitely splits the data base along the lines of how far on the
“earth to final product“ continuum a firm is. That is, one group is involved in more
basic types of industry, producing simpler products, while the other group was
involved in more complex types of manufacturing. Each of these groups has a

distinctly different set of manufacturing techniques and related environmental

108

issues with which to deal. This split in the database along these lines produced

two distinctly different groups.

Firm Size
As discussed previously, the concept of firm size is often used in systems

research, usually due to the argument about the relationship between firm size
and systems development. That was the reason for its inclusion in this research,
though as the analysis showed, the use of this construct proved somewhat
problematic. This was due to the extreme nonnorrnality exhibited by the firm size
data in the database. As with many statistical analysis techniques, structural
equation modeling strongly prefers to have multivariate normal data. For this
analysis, the firm size data were transformed using a log function. This helped to
put the data set closer to multivariate normality, though the Mardia’s coefficients
of the analyses continued to indicate that the overall data set was quite clearly
multivariate nonnormal. It should be noted again that this log transformation is in
keeping with standard structural equation modeling practice (West, Finch, and
Curran 1995).

Firm size was found to have very strong correlations to environmental
management system capabilities (as demonstrated in the single group analysis),
but the relationships were invariant across the two groups (as seen in the
multigroup analysis). This would be an indication that firm size does have an
effect, but the multigroup analysis was unable to shed any light on the effect size.
Thus, trying to determine what this effect size is would be a good topic for future

research. These results appear to be in line with the resource base view of the

109

firm (Wemerfelt 1984; Barney 1991), but this analysis was unclear as to the
effect Size.

The questionnaire used for this survey did not collect lnfonnation
regarding how many resources were expended on overall environmental efforts
and environmental management systems more specifically, which might have
allowed some interpretation as to the relative strengths of the firms'
environmental management system. Perhaps the use of other statistical
techniques on this database might be able to suggest how strong the firms
consider their environmental management systems to be.

It should be noted that using firm size as a basis for splitting up groups to
use in a structural equation model that includes firm size as a construct may
seem unusual, as it could be argued that the inclusion of the firm Size construct
already controls for firm size effect. Despite this, it is argued here that this
particular division of the database is useful in helping to uncover further insights
into how firm size affects firms’ environmental management systems.

Age of Environmental Management System

This analysis showed that there were some differences among the
relationships tested based on the age of the finn’s environmental management
system. In particular, the Analyze capability was Shown to have a different
relationship to operational performance across the two groups. Firms with a
relatively new environmental management system had a strong positive
statistically significant relationship between Analyze and Performance, while

firms with older environmental management systems had a small positive, not

110

statistically significant relationship between these two constructs. This could be
indicative of the fact that firms that have just started using an environmental
management system place a high importance on this particular capability, while
firms who have had their environmental management system in place for a while
do not see this capability as important and rely on it less.

Perhaps these results were due to the age cut-offs used. That is, perhaps
the use of different age cut-offs would have yielded different results. This is a
reasonable expectation, but the arbitrary age cut-offs used here (three or less
years and ten or more years) are quite reasonable, as they yielded two distinctly
different groups in terms of environmental management systems experience.
Perhaps one might object to the use of arbitrary age cut-offs, rather than cut-offs
grounded in some particular theory. Unfortunately, the literature has very little to
say on the matter of what would be appropriate age cut-offs to use.

Perhaps one could argue that a grounded cut-off date would be 1996, as
this was when the ISO 14000 standards were first released. The use of this date
as a cut-off was rejected for this research as this would have not maximized the
systematic variance in the two groups (the MAXMINCON principle described by
Kerlinger (1986)). If this cutoff had been used, there would have been only 40
firms in the group from post fall 1996. This small sample size would have most
likely had negative effects on the structural equation modeling analysis. The
issue of cut-off dates will be considered further in the section regarding future

research in chapter seven.

111

Presence of ISO 14000 Certification
It could be argued that this particular analysis had the potential to provide

the most interesting results. This is because the previous analyses show a
certain uniformity in the relationships between the environmental management
system and performance, no matter what the stratification used. Alter reviewing
those analyses, it would appear that environmental management systems do not
vary much from firm to firm. At this point, a natural question might be whether or
not environmental management systems will continue to display this uniformity
whether or not ISO 14000 is in place at the firm.

Perhaps the most interesting result of this multi-group analysis is the
statistically significant difference in the relationship between Analyze and
Performance between the two groups. In the group that did not have ISO 14000
certification, this relationship was reasonably strong and statistically significant.
In the group that had or was attempting to achieve ISO 14000 certification, this
relationship was very small and statistically insignificant. This points to the idea
that for the group with the certification, the Analyze capability of their
environmental management system does not appear to be a major point of
emphasis. One possible explanation for this discrepancy between the two
groups is that in firms that have not achieved ISO 14000, this particular capability
is given emphasis because these firms are attempting to understand what their
environmental risks and costs are. Once the certification is in place, firms may
de-emphasize this capability as they have become more familiar with the range

and scope of their environmental issues over the years.

112

Another potentially intriguing result in this analysis is the difference
between the two groups in their relationships between Control and Performance.
Though this relationship was found statistically to be invariant across the groups,
the size difference in the betas is quite noticeable none the less. This result may
be hinting at the idea that firms without the ISO certification are more concerned
about being able to control their environmental impacts because they are still in a
Ieaming mode trying to understand the size and scope of these impacts, whereas
the firms with or about to have the certification are more comfortable about the
levels of control that their environmental management system gives them.

Perhaps it is the case that the results obtained in this particular analysis
were caused by the fact that the groups were of very different size. The group
without ISO 14000 certification had 406 firms, while the group with or about to
have ISO 14000 certification had only 88 members. Various discussions on the
SEMNET listserv (archive available at http://bama.ua.edularchiveslsemnet.html
as of September 2000) indicate that this type of situation may influence the
results obtained. In particular, it has been suggested on SEMNET that the larger
group will have a disproportionate effect on fit indices and that the LM tests on
the constraints may yield some results that may appear unusual at first glance,
such as suggesting that similar betas are released and dissimilar beta constraints
should remain. A review of the peer-reviewed literature has thus far not turned
up much on the topic of unequal group sizes. Perhaps this issue is something

yet to be resolved in the structural equation modeling field.

113

Summary
Looking at the five analyses as a whole, some interesting results present

themselves. The size of the firm seems to not have an effect on the relationships
between environmental management system capabilities and performance. This
would indicate that no matter what size the firm, the same emphases are being
placed on the various capabilities, and by extension, perhaps the same relative
level of resource allocation. This result would seem to contradict the idea from
the systems literature (Choe (1996) discusses this) that bigger firms have greater
success with their systems due to superior resources. The results presented
here would seem to indicate perhaps it is not the amount of resources dedicated
to an environmental management system that ensure its performance, but
perhaps the skill with which it is managed.

With regard to industry setting, the analysis indicated that whether a firm is
in a relatively basic manufacturing industry or a more technically complex
industry, there appears to be little difference in how either of these groups use
their environmental management system. This would appear to be a partial
contradiction of the idea that based on some measurement of environmental risk,
firms will behave differently with regard to their environmental policies, tools and
systems (such as Rondinelli and Vastag (1996) discussed).

When the firms were divided by age of their environmental management
system, some differences appeared in the relationships between environmental
management system capabilities and operational performance. The strongest
relationship in this particular analysis was the one between firm size and

operational performance, perhaps indicating that firm size has more to do with

114

how an environmental management system affects performance than does the
age of the environmental management system.

F inns that have or are trying to achieve ISO 14000 certification appear not
to concentrate on the Analyze and Control capabilities quite as much as the firms
without the certification. This may be a reflection of the comfort level the firms
with the certification have in their environmental management systems.

Overall, it would appear that there are not as many differences in firms’
environmental management systems as some of the current conjectures in the
literature would argue. This in and of itself is an interesting result. As the various
analyses are reviewed, there is a consistency to the betas obtained. As these
results are interpreted, there does not appear to be much difference in the
relationships between environmental management systems and operational
performance no matter how the dataset is split. To the extent that the ISO 14000
standards attempt to promote a uniformity in how firms implement and use their
environmental management systems, this evidence would seem to argue that
this particular goal is being achieved. However, as was noted, this data set
contains a relatively low amount of firms that have or are attempting to achieve
ISO 14000 certification. This being the case, it would appear that this uniformity
of approach would exist whether or not the ISO 14000 standards did.

The uniformity found in the results could well be a function of the analysis
used. This uniformity points to the idea that the structural equation modeling
analysis attempted to fit the same set of parameters to each of the analyses, no

matter which way the dataset was cut. This would argue for other potential

115

models to be examined. In particular, Firm Size was the only exogenous
construct. This most likely affected the results obtained. The next Chapter will
discuss how the research might be adjusted in order to overcome the potential

weaknesses of the current approach.

116

Chapter 7

CONCLUSION

Summary of Findings
This study had two major aims. The first was to attempt to successfully

find a method of measuring an environmental management system. The second
was to use this measurement framework in order to test some of the early
arguments that have appeared in the environmental management literature.
Both were accomplished with a fair deal of success, yielding insights for both
academics and practitioners.

For the first Challenge, two measurement models were derived based on
existing literature and then both were used in a structural equation modeling
analysis to see which model worked better. The model named “management
functions” worked better, based on the fit statistics. The other model, named
“systems activities”, did show some promise, but was hampered by a lack of
manifest variables available to it. There was indication that this model would also
be viable for future research. The fact that this research was able to Show the
viability of a measurement model and the potential for another model is a major
step forward for the field. Future research will be able to build off of these
models as environmentally responsible manufacturing research continues to
grow.

The four multi—group comparisons were an attempt to empirically test

some of the arguments that have appeared in the new and growing field of

117

environmentally responsible management. The age of the environmental
management system was shown to have little difference in its relationship to
operational performance. This may be a signal that the environmental
management system has benefits immediately. The size of the firm also
appeared to have no effect on the relationship between the environmental
management system and operational performance, thought the single group
analysis did indicate an effect on the relationship from firm size to the
environmental management system. The industry setting multi-group
comparison was enlightening for the research community as it highlighted the
difficulty in categorizing whole industries based on the amount of environmental
risk. Further, the analysis showed that industry setting did not affect the
relationship between the environmental management system and operational
performance. The presence of ISO 14000 certification did appear to have an
effect the environmental management system-perfomrance relationship. In
particular, it could be argued that firms with the certification are more comfortable
with their system and thus have confidence that the system is identifying and

measuring environmental impacts properly.

Limitations of the Study
Much of the discussion of the limitations of this study should focus on the

questionnaire itself. The questionnaire was developed by a team of researchers
as part of a large research effort investigating environmental management. At
the time that the questionnaire was devised the field was still growing and had a

very splintered literature. This introduced some limitations to this research. First,

118

due to the lack of previous research, the research team had to derive questions
without having the benefit of previously tested questions. In order to ensure
validity of measurement, the team used multiple questions to ask about the same
items. Though this is standard practice, this meant that the survey length was
increased, which most likely discouraged some managers from responding.

Further, as the questionnaire was part of a larger research project in a
new and emerging field, this meant that the research team was looking to gather
lnfonnation on a number of different topics. This had two effects. First, this was
another reason to increase the length of the questionnaire, which again is a
reason the response rate was relatively low. Second, this meant that some
issues did not get as much coverage as individual researchers on the team
would have preferred.

A further limitation regarding the questionnaire was the timing of its mail-
out. The survey was conducted relatively soon after the ISO 14000 standard
was officially released. The term “ISO 14000“ was used prominently on the
survey and related mailing materials. It is most likely the case that many of the
managers who received a questionnaire did not have a complete knowledge of
ISO 14000. Given this, many may have chosen not to answer the survey, which
would have decreased the response rate. Further, the fact that ISO 14000 was a
relatively new phenomenon in the business world at the time of the survey meant
that not very many firms had achieved the certification. This was shown in the

responses, as typified by Figure 5.1. This led to a data set in which the vast

119

majority of the firms did not have the certification, even though this was a fair
representation of reality.

Despite these issues, it is argued that the questionnaire did not cause any
systematic problems with the research. For instance, though the response rate
was lower than most surveys in the field of operations management, the raw
number of returned surveys was quite large. In fact, it can be argued that the
size of the database is much larger than those found in similar research. Further,
given the circumstances, it is argued that the research team constructed a valid
questionnaire. The results from the confirmatory factor analysis bear this out.

In any case, these limitations present a guide for future research. In
particular, the fact that ISO 14000 was relatively new at the time of the mail out
would argue strongly for a replication of this research. A replication of this survey
today or in the future would most likely find a higher percentage of firms that
have achieved ISO 14000 certification or are intending to achieve this
certification. This would result in a more “balanced” dataset, in the sense that the
firms without the certification would not dominate the dataset as they did in this
research.

In this replication, it would be prudent to use a more focused
questionnaire. This should be possible given that there is now a data set on
hand to indicate the effectiveness of the set of questions used the first time. This
should result in a shorter questionnaire, which may help garner a higher

response rate.

120

The research has all of the usual limitations associated with the use of a
single respondent (per firm) and the use of perceptual measures. It should be
noted that as part of the larger research project, the research team attempted to
do what was referred to as a “cohort survey“ where multiple copies of the
questionnaire were to be mailed to a firm. The idea was that key contacts from
multiple different functional areas would be surveyed in order investigate whether
or not responses would be consistent within firms and among respondents in the
same functional area. Unfortunately, this idea did not generate any enthusiasm
from the firms contacted to be a part of this cohort survey, so the project was not
completed.

With regard to the use of the perceptual measures, this was something
that could not be avoided for this research, as in many cases there is no
accepted objective way to measure some of the items that the questionnaire
asked about. Whereas you can objectively measure whether or not a
computerized accounting system allows application of cash payments across
multiple invoices, it Is more subjective or perceptual as to whether or not a firm’s
environmental management system allows them to accomplish the managerial
task of planning.

One of the perceptual measures, the Firm Performance construct, bears
more discussion here. This construct was included as it is typically found in
“systems-performance“ models. However, it may be the case that there is a lag
in the relationship between the environmental management system and firm

performance. If this lag does occur, it would probably be because any effects on

121

firm performance caused by the environmental management system may not
have yet been noticeable to the respondent. This could happen either because
the respondent was unable to parcel out effects that had already occurred, or
because these effects had not actually occurred yet. It is argued here that there
is no particular reason to believe this lag occurred, but it is important to note the
possibility.

In considering both the single respondent and perceptual measures
issues, it is argued that neither of these caused systematic problems with the
research. For instance, the dataset indicates the respondents come from a
variety of functional areas and hold a variety of titles. Thus, there does not seem
to be a reason to be concerned that any one particular functional area or any one
particular level of management dominated the dataset. Also, the use of
perceptual measures is so common as to be non-controversial, as long as the
questions are constructed properly. The confirmatory factor analyses would
indicate that they were.

Future Research

As this research was conducted, many Ideas for future research were
revealed. Some ideas concern the general field of environmental management,
while others are more specific to the questions being investigated in this
dissertation. The discussion will move from the more general environmental
management issues to specific dissertation issues, and will highlight the

importance of each idea.

122

In general, many of the relationships described in Figure 3.1 (the general
framework of environmental management research) have not been fi.Illy
investigated. Not only the relationships among the primary constructs (external
factors, internal factors, environmental management, and unit performance) in
the model need further investigation, but the relationships that exist among the
lower-level constructs bear further investigation as well. Studies that investigated
these relationships would be very important for practitioners as they would show
the linkages among various environmental management issues and unit
performance. This is critical to investigate the linkages to unit performance in
order to really capture practitioners’ attention.

One phenomenon in the field of environmental management that needs to
be explored is the difference in growth rates between ISO 9000 certifications and
ISO 14000 certifications. As was shown by the dataset used for this research,
relatively few American firms have achieved or are attempting to achieve ISO
14000 certification. It appears that ISO 9000 was much more readily accepted
by the United States business community than ISO 14000. The research
literature has not Shed much light on this issue of the different growth curves.
Perhaps it is the case that the expense associated with achieving ISO 9000 has
made firms leery of attempting the ISO 14000 certification. However, there is a
counter-argument to the problem of expense, as there is much similarity in the
ISO 9000 and 14000 specifications. Thus if a firm has achieved ISO 9000
certification, the experience and familiarity it gained in the 9000 process should

allow it to achieve what might be termed “economies of scale“ when attempting

123

the ISO 14000 certification. This is an argument that has been quite popular in
the literature (Renzi and Cappelli 2000; Minner1997; Cichowicz 1996). Perhaps
it is the case that firms simply do not view ISO 14000 as being worth the effort.
Though the literature does contain some empirical work on the potential
advantages of environmentally responsible manufacturing in general (see for
instance Klassen and McLaughlin (1996)), there has been very little work on the
financial impact of ISO 14000 specifically. Thus, these arguments point to the
idea that research which attempts to investigate the reasons for the difference in
the growth of ISO 9000 and ISO 14000 in the United States could shed some
light on how valuable United States managers perceive ISO 14000 to be.
Further, empirical research as to the financial impact of ISO 14000 would also be
useful. Both of these investigations are quite feasible and would be of great
interest to US. managers, as they would increase the understanding of how ISO
14000 affects competitiveness. As an extension, a comparison of the growth
rates of ISO 14000 in the US. and other parts of the world could help to highlight
the differences in the nature of the competitive environment in various parts of
the world, something in which managers would most likely have a great interest.
This study used the presence or absence of ISO 14000 as a stratification
variable in the analysis. Related to the above discussion, it might be useful to
examine how the presence or absence of ISO 9000 certification affects the
implementation and performance of environmental management systems. Both
ISO 9000 and environmental management systems are designed to a allow a

firm to manage and control specific aspects of its performance (quality and

124

environmental impacts, respectively). As such, they both take a systematic
approach to the performance issues under their purview. As there might be
similarities in the way that these programs are approved and implemented, it
could be interesting to see whether or not firms that have previous experience
with ISO 9000 have a distinctly different experience with environmental
management systems than those firms without the ISO 9000 certification. This
type of study would yield insights into the economies of scale idea discussed
above.

As this analysis was being done, one of the measurement issues to spring
forth was the difficulty in deciding what constitutes an industry that is
environmentally risky versus an industry that is environmentally safe. The
current state of the literature was not found to be of much help in this regard. So,
though arguments about industry membership effect on a firm’s environmental
operations and performance are prevalent, there was not a Clearly defined path
for testing these arguments. It is argued by this research that the Standard
Industrial Classification codes provide a reasonable way to divide the database.
However, further attempts to measure the environmental riskiness of industries
must be made. The lack of measures for this is hampering researchers’ ability to
test the theories that are based on the idea of environmental riskiness. Thus, in
order to help advance the field, more work needs to be done in developing this
type of measurement.

Men the cut-offs for the age of environmental management system

multigroup analysis were considered, another interesting measurement issue

125

came to the fore. Consider that the ISO 14000 regulations were issued by the
lntemational Organization for Standardization in the fall of 1996. It could be
argued that using that particular date as a cutoff might be more theoretically
appropriate due to the fact that firms implementing their environmental
management system after that date would have the advantage of the ISO
“blueprint“, whether or not they intended to attempt to achieve ISO 14000
certification. This cut-off was not used In this research due to the fact that there
would have been a small number of firms in the post 1996 group. Future
research will most likely have an easier time collecting acceptable sample sizes
based on the 1996 cut-off. Research of this type would be useful in order to
determine what differences exist in environmental management systems that
were started before and after the publication of the ISO 14000 standards. This
research would be important as it would yield insights into what the effects of the
publication of the ISO 14000 standards have been. For instance, perhaps it is
the case that pre-ISO 14000 environmental management systems exhibit a
better ability to limit certain types of pollution. By studying these types of
differences, many insights into the general environmental management field
could be gained.

The results of the age of environmental management system multigroup
analysis pointed at some other ideas for future research. In particular, a Ieaming
curve type of effect was noticed for only one of the four capabilities and only with
the particular age cut-off used. Perhaps it is the case that different age cut-offs

would have produced distinctly different results. Further, it may be the case that

126

testing only one capability at a time might yield some insights useful for the
systems literature.

This leads to another line of future research. This would be to use a
technique that measures the strength of each of the capabilities. Using this type
of technique would be more in line with Ieaming curve theory. Combining this
type of technique with the multiple possible age cut-offs available in this data set
could yield some interesting results that might help to form a “map“ of how
environmental management systems develop over time.

In general, the issue of measuring the strength of environmental
management systems capabilities should be investigated further. This would
require another method of analysis, such as using a structured means approach.
Unfortunately, even as Bentler himself has noted, multi—group structured means
analyses are “difficult to apply in practice” (Bentler 1993). They are difficult to do
as the number of parameters to be estimated is greatly increased as compared
to a standard structural equation modeling analysis. In fact, this type of analysis
was tried as part of this research on an exploratory basis. Bentler’s concerns
turned out to be justified, as the structural equation modeling software returned
many condition codes. Perhaps with a respecification of the model used in this
research, a structured means approach might work. Other methods that might
work well for this type of analysis include the various forms of stepwise
regression.

These types of analysis would be interesting as it would give some idea as

to which of the environmental management system capabilities tends to be

127

strongest. Given a longitudinal database, this type of analysis might be able to
identify how the mpabilities grow stronger or weaker over time. This would be of
use to practitioners in helping them to decide where to concentrate resources as
they develop their environmental management system over time. For instance, a
firm that was concerned about making sure one particular capability was strong
could use the results of this type of longitudinal analysis to allocate resources to
that capability as suggested by the study. For academics, this type of study
would be a strong contribution to the systems literature regarding systems
development.

Perhaps one of the most important questions that should be investigated
in future research is the markers of successful versus unsuccessful
implementations of environmental management systems. This research is based
on a dataset that looks at the systems already installed and was not focused on
the implementation process. While this research has shown what managers
might expect in the way of environmental management system capabilities
relationships to operational performance, this research did not investigate (nor
was it designed to investigate) the implementation process. This research has
shown that an environmental management system can be measured, so a
project could build on the measurement ideas presented here in order to
investigate the environmental management system implementation process.

This type of investigation would be of great interest to practitioners who
are considering implementing an environmental management system. To a

certain extent, this type of research would be searching for “best practices“ for

128

the implementation. For academics, this type of study would be a solid
contribution to the environmental management systems literature. This type of
study would be quite feasible, especially if a case study approach were to be
used.

Implications for Management
One of the interesting features of the idea of an environmental

management system Is that though it now has a formal definition (due to ISO
14000), it is perhaps a more amorphous system than the other management
systems a firm may use. That is, typically one can define and draw boundaries
around exactly where the production scheduling systems (for instance) begins
and ends. An environmental management system, as even the formal definition
implies, ranges across the entirety of a firm’s operations. Because of this, there
are many different ways in which to construct an environmental management
system, though the ISO 14000 definition is quite clear in what the system must
do. Despite this amorphousness, this research has shown that an environmental
management system can be operationalized for research purposes. In fact, this
research has indicated that the two measurement frameworks presented both
appear to be reasonable ways for a firm to define and measure its environmental
management system.

One of the more striking implications of this research is that for the most
part, environmental management systems behave in a very similar manner from
firm to firm. This occurs in spite of some arguments in the literature to the

contrary. This is an important finding for managers as it indicates that the

129

behavior of the system will not differ that greatly from situation to situation. The
managerial actions suggested by the environmental management system and
the operational results obtained from the environmental management system
may vary, but the behavior of the system appears relatively stable In most
settings. This should be reassuring to firms considering implementing an
environmental management system as it removes a possible source of variance

in the firm.

Contribution of Research
One of the primary contributions of this research was to demonstrate that

an environmental management system can be measured. In fact, this research
pointed the way to two reasonable measurement models. This is a major
contribution to the literature as there have been only very limited attempts to
measure environmental management systems before. While one of the two
proposed measurement models clearly worked better in the research presented
here, it is argued that the other measurement model has some very good
potential if it is further developed in future research.

The other primary contribution of this research is to give managers an
understanding of what they might expect from the behavior of an environmental
management system. In general, it appears that the behavior of environmental
management systems is fairly consistent across different settings. This lack of
variance in the behavior of the systems could potentially eliminate an objection to
implementing an environmental management system that some firms might

have. The similarity of behavior exhibited by the environmental management

130

systems in this study may encourage more firms to consider implementing an
environmental management system. This could be significant given the
increasing importance of environmental management to firms and the increasing
visibility of ISO 14000.

This apparent uniformity in the behavior of environmental management
systems should also strike a chord with those who favor a formalized
programmatic approach to managing a certain aspect of performance. One of
the precepts of the creation of the various ISO standards is that they are in fact
standards designed to produce a uniformity of approach and allow comparability
of systems across industries and national boundaries. This research would

appear to indicate that this goal has been achieved.

Summary
This research endeavored to create and test a model of environmental

management systems. The goal was not only to derive and validate a useful
measurement model, but also to use this model in a series of multi-group
examinations in order to get a sense of the contexts of environmental
management system’s relationship to operation performance. It was argued that
these analyses would be a major contribution to the literature on environmental
management systems, which is still in an early stage of growth.

In many respects, this research was a success with relation to its goals.
Two measurement models were derived and tested. Though one of the models
Clearly performed better in the structural equation modeling analysis, both show

promise as latent constructs for measuring environmental management systems.

131

The multi-group comparisons helped to clarify some of the conjectures now
found in the literature. Further, me research has pointed to some useful future
directions for research into environmental management systems. It is hoped that
this research will be a launching pad for further investigations into environmental

management systems.

132

APPENDICES

133

 

APPENDIX A

QUESTIONS FOR MANAGEMENT FUNCTIONS MODEL

The questions used to measure the environmental management system
appear as worded in the original questionnaire. Item number indicates the order
of the question as it appeared in the original questionnaire. In the following table,
the questions are grouped by the constructs used in the management functions

model.

134

Instructions on survey:

For the following statements, you are asked to indicate your response by
Circling a number between the anchor points of the zero-tO-ten scales, using the
following scale:
Strongly Disagree (- 0 1 2 3 4 5 6 7 8 9 10 -) Strongly Agree

Your company’s environmental management system:

Item
number

17

18

38

19

20

21

33

Item

Performance measures quantify costs
associated with environmental performance

The performance measures identify qualitative
costs associated with environmental
performance and problems

The effectiveness of different production
schedules is evaluated considering possible
environmental problems created by those
schedules

Wren an environmental prome is identified,
attention is primarily focused on correcting the
effects of that problem

When an environmental problem is identified,
attention is focused on the underlying cause
and corrective measures

When an environmental problem Is identified,
the major effort within your company is on
determining the underlying reasons for that
problem and then attacking it

Environmentally responsible suppliers tend to
be more cost effective suppliers

Your company has a “fonnal' environmental
management system in place

The environmental management system's
procedures are formally documented

The environmental management system’s
procedures are widely circulated and available
to anyone who needs access

There is a specific department within your
company which is formally responsible for
environmental affairs

There is a formal reporting position between
those responsible for environmental affairs
within your company and one of the
cornpany’s senior executives (director level or
above)

The environmental position of your company
is given prominent visibility within annual
reports and other publications distributed by
your company

135

EMSecost

EMSqualc

EMSsched

EMSefect

EMScause

EMSreasn

EMSsupp

EMSfoanI

EMSdoc

EMScirC

EMSdept

EMSrpt

EMSpos

Variable Name Management

function
Analyze
Analyze

Analyze

Analyze

Analyze

Analyze

Analyze

Organize

Organize

Organize

Organize

Organize

Organize

11

12

16

35

13

14

28

30

31

32

37

10

22

23

The environmental achievements of your
company are given prominent visibility within
annual reports or other company publications
Environmental issues, policies and
procedures are included in formal company
training programs

Environmental performance results are widely
distributed within your company

Your company has been very successful in
creating company wide awareness of
environmental problems, costs and
opportunities

Environmental performance is formally
tracked and monitored within the company

Various performance goals have been
developed and implemented which report
environmental performance on different
dimensions

Environmental performance has been
captured primarily in terms of recyclability i.e.,
the amount of recycled material used in the
productorthedegreetowhichtheproductis
recyclable

Environmental performance is considered
during the evaluation of individual
departments or areas

Performance on environmental dimensions is
considered during personnel evaluations

Performance on environmental dimensions is
considered during supplier evaluations

Your company has a well developed
environmental data base for tracking and
monitoring environmental issues

Environmental practices, procedures and
systems within your company are compared
with the best-in-class on a regular basis

Top management has taken a highly visible
public position in support of improved
environmental performance
Environmentally-related information (e.g.,
changes in regulations, tracking of which
companies have been fined and how much) is
hacked and monitored on an on-going basis
lnfonnation about best-in-class environmental
performance in your industry is tracked and
recorded within your company

Product designs have been modified in
response to environmental problems
attributable to a specific design flaw

Process designs and plant layouts have been
modified in response to external
environmental problems attributable to a
specific design flaw

136

EMSachv

EMStrain

EMSdist

EMSaware

EMSeperf

EMSgoals

EMSrecyc

EMSedept

EMSepers
EMSesupp

EMSdata

EMSbicc

EMSvis

EMSeinf

EMSbict

EMSprod

EMSproc

Organize

Organize

Organize

Organize

Control

Control

Control

Control

Control

Implement

Implement

Implement

Plan

Plan

Plan

Plan

24

25

26

27

New equipment has been acquired or existing
equipment modified in response to external
environmental problems attributable to that
equipment

Environmental issues are included in the
process design process

Environmental issues are considered when
laying out the plant or specifying equipment
for use on the plant floor

Environmental issues influence plant location
decisions (i.e., where we locate a new plant)

137

EMSequip

EMSpdprc

EMSlay

EMSloc

Plan

Plan

Plan

Plan

APPENDIX B

QUESTIONS FOR THE SYSTEMS ACTIVITIES MODEL

The questions used to measure the environmental management system
appear as worded in the original questionnaire. Item number indicates the order
of the question as it appeared in the original questionnaire. In the following table,

the questions are grouped by the constructs used in the systems activities model.

138

Instructions on survey:

For the following statements, you are asked to indicate your response by
Circling a number between the anchor points of the zero-to-ten scales, using the
following scale:

Strongly Disagree (- 0 1 2 3 4 5 6 7 8 9 10 -) Strongly Agree

Your company’s environmental management system:

Item Item Variable Name System

number Activity

15 Environmental performance, as captured by EMSsumm Database
various measures, is periodically summarized

29 Cost accounting has been extensively EMScacct Database

involved in your company in capturing and
reporting environmental costs

13 Environmental performance is formally EMSeperf Database
tracked and monitored within the company
37 Your company has a well developed EMSdata Database

environmental data base for tracking and
monitoring environmental issues

8 Environmentally-related information (e.g., EMSeinf Database
changes in regulations, tracking of which
companies have been fined and how much) is
hacked and monitored on an on-going basis

9 Information about best-in-class environmental EMSbict Database
performance in your industry is hacked and
recorded within your company

17 Performance measures quantify costs EMSecost Measurement
associated with environmental performance
18 The performance measures identify qualitative EMSqualc Measurement

costs associated with environmental
performance and problems

14 Various performance goals have been EMSgoals Measurement
developed and implemented which report
environmental performance on different
dimensions

28 Environmental performance has been EMSrecyc Measurement
captured primarily in terms of recyclability i.e.,
the amount of recycled material used In the
product or the degree to which the product is
recyclable

1 Your company has a “Iorrnal' environmental EMSfonnl Documentation
management system in place

2 The environmental management system's EMSdoc Documentation
procedures are formally documented

22 Product designs have been modified in EMSprod Feedback
response to environmental problems
attributable to a specific design flaw

139

23

24

25

26

27

19

20

21

10

33

Process designs and plant layouts have been EMSproc
modified in response to external
environmental problems attributable to a

specific design flaw

New equipment has been acquired or existing EMSequip
equipment modified in response to external
environmental problems attributable to that

equipment

Environmental issues are included in the EMSpdprc
process design process

Environmental issues are considered when EMSlay

laying out the plant or specifying equipment

for use on the plant floor

Environmental issues influence plant location EMSIoc
decisions (i.e., where we locate a new plant)

When an environmental problem is identified, EMSefect
attention is primarily focused on correcting the

effects Of that problem

When an environmental problem is identified, EMScause
attention is focused on the underIying cause

and corrective measures

When an environmental problem is identified, EMSreasn
the major effort within your company is on
determining the underlying reasons for that

problem and then attacking it

The environmental management system’s EMSCirC
procedures are widely circulated and available

to anyone who needs access

Environmental practices, procedures and EMSbicc
systems within your company are compared

with the best-in-class on a regular basis

There is a specific department within your EMSdept
company which is formally responsible for

environmental affairs

There is a formal reporting position between EMSrpt
those responsible for environmental affairs

within your company and one of the

company’s senior executives (director level or

above)

Environmental concerns play a significant role EMSsdec
in the supplier selection and/or retention

decisions

Environmentally responsible suppliers tend to EMSsupp
be more cost effective suppliers

140

Feedback

Feedback

Feedback

Feedback

Feedback

Problem
solving

Problem
solving

Problem
solving

Procedures
Procedures

Reporting

Reporting

Supply chain

Supply chain

BIBLIOGRAPHY

141

BIBLIOGRAPHY

Anderson, James C. and Gerbing, David W. (1984). ”The Effect of Sampling
Error on Convergence, Improper Solutions, and Goodness-of-fit Indices
for Maximum Likelihood Confirrnatory Factor Analysis,” Psychometrika,
49(2), 155-173.

Anonymous (1995). "The Elements of Environmental Management,”
Environmental Manager, 6(12), 1-3.

Anonymous (1999). ”Ford and GM Demand ISO 14000 Compliance,” Quality,
38(13), 12-16.

Arthur, Jeffrey B. (1994). ”Effects of human resource systems on manufacturing
performance and turnover,” Academy of Management Journal, 37(3), 670-
687.

Bacharach, Samuel B. (1989). ”Organizational Theories: Some Criteria for
Evaluation," Academy of Management Review, 14(4), 496-515.

Bajwa, Deepinder S., Rai, Arun, and Ramaprasad, Arkalgud (1998). ”The
Structural Context of Executive lnfonnation Systems Adoption,”
lntemational Journal of Operations and Production Management, 11(3),
28-38.

Baker, Mark B. and MCKiel, Mary (eds.) (1998). ISO 14000 Questions and
Answers. Reston, Virginia: CEEM, Inc. (with ASQC).

Barney, Jay B. (1991). ”Firm Resources and Sustained Competitive Advantage,”
Jouma! of Management, 17(1), 99—120.

Bentler, Peter M. (1993). 508 Structural Equations Program Manual. Los
Angeles: BDMP Statistical Software.

Bentler, Peter M. (1999). "ML Robust," SEMNET@BAMA.UA.EDU from
bentler@UCLA.EDU, 8 June 1999.

Berry, Michael A. and Rondinelli, Dennis A. (1998). ”Proactive corporate
environment management: A new industrial revolution," Academy of
Management Executive, 12(2), 38-50.

Bollen, Kenneth A. (1989). Structural Equations With Latent Variables. New York:
John IMley 8 Sons.

Brown, Warren B and Karagozoglu, Necmi (1998). ”Current Practices in
Environmental Management,” Business Horizons, 41 (4), 12-18—

142

Byme, Barbara M. (1994). Stnrctura! Equation Modeling With E08 and
EQS/Wrndows: Basic Concepts, Applications, and Programming.
Thousand Oaks, CA: SAGE Publications.

Carter, Craig R., Ellram, Lisa M., and Ready, Kathryn J. (1998). "Environmental
Purchasing: Benchmarking Our German Counterparts,” lntemational
Joumal of Purchasing and Materials Management, 34(4), 28-38.

Cascio, Joseph (1996). The ISO 14000 Handbook. Fairfax, Vrrginia: CEEM
Information Services.

Cerveny, Robert P. and Scott, Lawrence W. (1989). ”A Survey of MRP
Implementation,” Production and Inventory Management Joumal, 30(3),
31-34.

Choe, Jong-Min (1996). "The Relationship Among Performance of Accounting
Information Systems, Influence Factors, and Evolution Level of
lnfonnation Systems,” Journal of Management lnfonnation Systems,
12(4), 215-239.

Cichowicz, Judith A. (1996). ”Should ISO 14000 be Linked with ISO 9000?,”
Environmental Quality Management, 6(1), 77-80.

Clelland, lain J., Dean, Thomas J., and Douglas, Thomas J. (2000). ”Stepping
Towards Sustainable Business: An Evaluation of Waste Minimization
Practices in US Manufacturing,” Interfaces, 30(3), 107-124.

Cohen, Jacob and Cohen, Patricia (1983). Applied Multiple
Regression/Correlation Analysis for the Behavioral Sciences. Hillsdale,
New Jersey: Lawrence Erlbaum Associates, Publishers.

Conway-Schempf, Noellette and Lave, Lester (1999). ”Green Design Tools for
Environmental Management,” Environmental Quality Management, 8(4),
35-46.

Dechant, Kathleen, Altman, Barbara, Dowining, Robert M., and Keeney, Timothy
(1994). "Environmental Leadership: From Compliance to Competitive

Advantage (Executive Commentary),” Academy of Management
Executive, 8(3), 7—20.

Dillman, Don A. (1978). Mail and Telephone Surveys: The Total Design Method.
New York: Wiley.

Dillon, William R., Kumar, Ajith, and Mulani, Narendra (1987). "Offending
Estimates in Covariance Structure Analysis: Comments on the Causes of
and Solutions to Heywood Cases," Psychological Bulletin, 101(1), 126-
135.

143

Dooley, Robert S. and Fryxell, Gerald E. (1999). ”Are Conglomerates Less
Environmentally Responsible? An Empirical Examination of Diversification
Strategy and Subsidiary Pollution in the US. Chemical Industry," Journal
of Business Ethics, 21(1), 1-14.

Epstein, Marc J. (1996). "You’ve Got a Great Environmental Strategy - Now
What?,“ Business Horizons, 39(5), 53-59.

Fayol, Henri (1930). Industrial and General Administration. Croyton, Great
Britain: H. R. Grubb, Ltd.

Feldman, Stanley J., Soyka, Peter A., and Ameer, Paul G. (1997). ”Does
Improving a Firm’s Environmental Management System and
Environmental Performance Result in a Higher Stock Price,“ Journal of
Investing, 6(4), 87-97.

Ferdows, Kasra and De Meyer, Amoud (1990). ”Lasting Improvements in
Manufacturing Performance: In Search of a New Theory,” Journal of
Operations Management, 9(2), 168-184.

Ferrante, John G. and Cotter, Jamie (1999). "The ISO 14001 Environmental
Management System: A Prescription for Enviommental and Financial
Health," Environmental Quality Management, 9(1), 63-74.

Fiskel, J. (1996). Design for Environment: Creating Eco-Efficient Products and
Processes. New York: MCGraw-Hill.

FitzGerald, Chris (1993). "Forging New Links Between Environmental
Management, lnfonnation Systems, and Business Processes,"
Environmental Quality Management, 3(1), 97-107.

FitzGerald, Chris (1994). "Environmental Management lnfonnation Systems:
New Tools for Measuring Performance," Total Quality Environmental
Management, 4(2), 21-34.

FitzGerald, Chris (1997). "Systems vs. Metasystems: EMIS at the Crossroads,"
Environmental Quality Management, 7(2), 71-80.

Flynn, Barbara 8. and Flynn, E. James (1999). "lnfonnation-Processing
Alternatives for Coping with Manufacturing Environment Complexity,"
Decision Sciences, 30(4), 1 021 -1 052.

Forza, Cipriano (1995a). ”Quality Information Systems and Quality Management:

A Reference Model and Associated Measures for Empirical Research,"
Industrial Management + Data Systems, 95(2), 6-9.

144

Forza, Cipriano (1995b). ”The Impact of Information Systems on Quality
Performance: An Empirical Study,“ lntemational Journal of Operations and
Production Management, 15(6), 69-73.

Ganster, Daniel C., Hennessey, Harry W., and Luthans, Fred (1983). ”Social
Desirability Response Effects: Three Alternative Models,” Academy of
Management Joumal, 26(2), 321 -331.

Graff, Susan (1997). ”ISO 14000: Should Your Company Develop an
Environmental Management System?," Industrial Management, 39(6), 19-
22.

Gray, Rob H. (1992). "Accounting and Environmentalism: An Exploration of the
Challenge of Gently Accounting for Accountability, Transparency and
Sustainability,” Accounting, Organizations and Society, 17(5), 399-425.

Griffin, Jennifer J. and Mahon, John F. (1997). ”The Corporate Social
Performance and Corporate Financial Performance Debate: Twenty-Five
Years of Incomparable Research,“ Business and Society, 36(1), 5-31.

Grover, Varun, Teng, James T., Segars, Albert H., and Fiedler, Kirk D. (1998).
"The Influence of Information Technology Diffusion and Business Process
Change on Perceived Productivity: The IS Executive’s Perspective,”
Information & Management, 34(3), 141-159.

Guide, V. D. R., Jr. and Srivastava, Rajesh (1998). "Inventory Buffers in
Recoverable Manufacturing,” Joumal of Operations Management, 16(5),
551-568.

Gupta, Mahesh C. (1995). "Environmental Management and Its Impact on the
Operations F unction," lntemational Journal of Operations and Production
Management, 15(8), 34-51.

Gupta, Mahesh C. and Sharma, Keshav (1996). "Environmental Operations
Management: An Opportunity for Improvement," Production and Inventory
Management Journal, 37(3), 4046.

Hamilton, James T. (1995). "Pollution as News: Media and Stock Market
Reactions to the Toxics Release Inventory Data," Journal of
Environmental Economics and Management, 28(1), 98-103.

Handfield, Robert and Melnyk, Steven A. (1995). "Environmentally Responsible
Manufacturing: New Challenges, Hidden Opportunities," American
Production and Inventory Control Society.

Hart, Stuart L. (1997). ”Beyond Greening: Strategies for a Sustainable World,”
Harvard Business Review, 75(1), 66-

145

Hillary, Ruth (ed.) (1997). Environmental Management Systems and Cleaner
Production. Chichester, West Sussex, England: John Wiley & Sons.

Horrnozi, Amir M. (1999). ”Make It Again," IIE Solutions, 31(4), 39-44.

Hoyle, Rick H. (1995). ”The Structural Equation Modeling Approach: Basic
Concepts and Fundamental Issues," in Hoyle, Rick H., Structural Equation
Modeling: Concepts, Issues and Applications. Thousand Oaks, CA: SAGE
Publications, InC., 1-15.

Hunt, Shelby D. (1991 ). Modem Marketing Theory: Critical Issues in the
Philosophy of Marketing Science. Cincinnati, Ohio: South-Westem
Publishing.

Karagozoglu, Necmi and Lindell, Martin (2000). "Environmental management:
Testing the win-win model,” Joumal of Environmental Planning and
Management, 43(6), 817-829.

Kerlinger, Fred N. (1986). Foundations of Behavioral Research. Chicago: Holt,
Rinehart and Winston, Inc.

Kinder, Lydenberg Domini & Co. (1996). Socrates: The Corporate Social Ratings
Monitor, version 2.03.

Klassen, Robert D. and McLaughlin, Curtis P. (1996). "The Impact of
Environmental Management on Firm Performance,“ Management
Science, 42(8), 1199-1214.

Lamprecht, James L. (1997). ISO 14000: Issues and Implementation Guidelines
for Responsible Environmental Management. New York: AMACOM.

Melnyk, Steven A., Calatone, Roger, Handfield, Robert, Tummala, R. L., Vastag,
Gyula, Hinds, Timothy, Sroufe, Robert, Montabon, Frank, and Curkovic,
Sims (1999). ISO 14000: Assessing Its Impact on Corporate Effectiveness
and Efficiency. Tempe, Arizona: Center for Advanced Purchasing Studies.

Melnyk, Steven A. and Denzler, David R. (1996). Operations Management: A
Value-Driven Approach. Chicago: Irwin.

Miettinen, Pauli and Hamalainen, Raimo P. (1997). "How to Benefit from
Decision Analysis in Environmental Life Cycle Assessment (LCA),”
European Journal of Operational Research, 102(2), 279-294.

Minner, Jill F. (1997). "The Quality Pro is Key in Pursuit of ISO 14000," Quality,
36(1), 40-41.

146

Montabon, Frank, Calatone, Roger, Melnyk, Steven A., and Sroufe, Robert
(2000). "A Model of the Decision to Achieve ISO 14000 Certification,”
Decision Sciences Institute.

Orlin, Judy, Swalwell, Peg, and FitzGerald, Chris (1993). "How to Integrate
lnfonnation Strategy Planning with Environmental Management
lnfonnation Systems (Part I),” Environmental Quality Management, 3(2),
193-202.

Owen, Jean V. (1993). "Environmentally Conscious Manufacturing,"
Manufacturing Engineering, 111(4), 44-48.

Porter, Michael E. (1991). ”America’s Green Strategy," Scientific American,
264(4), 168-168.

Porter, Michael E. and van der Linde, Class (1995). ”Green and Competitive:
Ending the Stalemate,” Harvard Business Review, 73(5), 120-134.

Rappaport, Ann and F laherty, Margaret F. (1992). Corporate Responses to
Environmental Challenges. New York: Quorum Books.

Renzi, M. F. and Cappelli, L. (2000). ”Integration Between ISO 9000 and ISO
14000: Opportunities and Limits,” Total Quality Management, 11(4-6),
$849-$858.

Rivera-Camino, Jaime (2001). "The implementation of environmental
management systems standards in Europe: An organizational approach,”
lntemational Journal of Management, 18(1), 3-10.

Rondinelli, Dennis A. and Vastag, Gyula (1996). "lntemational Environmental
Standards and Corporate Policies: An Integrative Framework," California
Management Review, 39(1), 106-122.

Rondinelli, Dennis A. and Vastag, Gyula (2000). ”Panacea, common sense, or
just a label? The value of ISO 14001 environmental management
systems,” European Management Journal, 18(5), 499-510.

Rosewicz, Barara. Americans Are WIlling to Sacrifice to Reduce Pollution, They
Say. Wall Street Journal , A1. 4-20-1990. New York.
Ref Type: Newspaper

Sarkis, Joseph and Rasheed, Abdul (1995). ”Greening the Manufacturing
F unction," Business Horizons, 38(5), 17-27.

Sharfman, Mark (1996). "The Construct Validity of the Kinder, Lydenberg 8
Domini Social Performance Ratings Data," Jouma! of Business Ethics,
1 5(3), 287-296.

147

Shanna, Subhash, Durvasula, Srinivas, and Dillon, William R. (1989). ”Some
Results On The Behavior Of Alternate Covariance Structure Estimation
Procedures in the Presence of Nonnonnal Data," Joumal of Marketing
Research, 26(2), 214-221.

Shrivastava, Paul and Hart, Stuart L. (1994). "Greening organizations - 2000,”
lntemational Journal of Public Administration, 17(3, 4), 607-635.

Sigman, Hilary (1996). "The Effects of Hazardous Waste Taxes on Waste
Generation and Disposal,” Journal of Environmental Economics and
Management, 30(2), 199-

Simerly, Roy L. and Bass, Kenneth E. (1998). "The Impact of Equity Position on
Corporate Social Performance,” lntemational Journal of Management,
15(1), 130-135.

Spicer, Barry H. (1977). "Market Risk, Accounting Data and Companies’ Pollution
Control Records,” Joumal of Business Finance and Accounting, 5(1), 67-
83.

Stapleton, Philip J., Cooney, Anita M., and Hix, William M., Jr. Environmental
Management Systems: An Implementation Guide for Small and Medium-
Sized Organizations. 1996. Ann Arbor, Michigan, NSF lntemational.
Ref Type: Report

Steger, Ulrich (2000). ”Environmental management systems: Empirical evidence
and further perspectives,” European Management Journal, 18(1), 23-27.

Sum, Chee-Chuong, Ang, James S. K., and Yeo, Lei-Noy (1997). ”Contextual
Elements of Critical Success Factors in MRP Implementation," Production
and Inventory Management Journal, 38(3), 77-83.

Timmins, Mike (1998). ”SO You Really Think Industry Alone is to Blame for
Pollution?,” The British Journal of Administrative Management, 20-21.

Walley, Noah and Whitehead, Bradley (1994). "It's Not Easy Being Green,”
Harvard Business Review, 72(3), 46-52.

Wemerfelt, Birger (1984). ”A Resource-based VIew of the F inn,” Strategic
Management Journal, 5(2), 171-180.

West, Stephen G., Finch, John F., and Curran, Patrick J. (1995). ”Structural
Equation Models With Nonnonnal Variables: Problems and Remedies,” in
Hoyle, Rick H., Stnrctural Equation Modeling: Concepts, Issues and
Applications. Thousand Oaks, CA: SAGE Publications, 56-75.

148

Wu, Haw-Jan and Dunn, Steven C. (1995). ”Environmentally Responsible
Logistics Systems,” lntemational Journal of Physical Distribution and
Logistics Management, 25(2), 20-38.

Zerbe, Wilfred J. and Paulhus, Delroy L. (1987). ”Socially Desirable Responding
in Organizational Behavior: A Reconception,” Academy of Management
Review, 12(2), 250-264.

Zikmund, William G. (1991). Business Research Methods. Chicago: The Dryden
Press.

149