 

 

m
:3}

‘ 'w
m- j.

v-‘. ;

dig:

@ﬁgfg i1
55%
‘5'. $3! A.” 1’
z .

a i ‘
"T‘s: .
gnu-3‘; vent-nay“
“"~ . 2" '

w“.

4w.“

v.
'Ll

“39:

513:

 

 

 

Qiﬂf
561 ”1/5? ’1 {a ‘0

This is to certify that the
dissertation entitled

The Effects of Non-Financial Performance Measures in a
Bonus Contract

presented by

Elizabeth Connors

has been accepted towards fulﬁllment
of the requirements for the

 

PhD. degree in Accounting

‘3 WW {Z #J.\

Major Professor’s Signature

[toll] 2,2, 1 2-00 ‘/

I

Date

MSU is an Afﬁnnative Action/Equal Opportunity Institution

 

LIBRARY '
Michigan State
University

 

 

 

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 c:ICIFiC/DateDue.p65-p.15

 

 

PERM

 

THE EFFECTS OF NON-FINANCIAL
PERFORMANCE MEASURES IN A BONUS CONTRACT

By

Elizabeth Connors

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Accounting

2004

 

 

PliRl

 

 

 

This stud} L'.\é
placedonqualil} and
and financial and non
sources and through I
pluties industries are
shown positixe LISMK
placed on internal qu
contracts. There is :1!
ﬁnancial performing

hints eontraets.
The ell

eels til

rte also examined. l"

ABSTRACT

THE EFFECTS OF NON-FINANCIAL
PERFORMANCE MEASURES IN A BONUS CONTRACT

By

Elizabeth Connors

This study examines the association between the association between the weight
placed on quality and environmental performance measures in managers’ bonus contracts
and ﬁnancial and non-ﬁnancial performance. Data collected from publicly available
sources and through mail surveys to plant managers in the chemical, paper, furniture and
plastics industries are used to examine several hypotheses. The results reported here
show a positive association between external quality performance levels and the weight
placed on internal quality performance and external quality performance in the bonus
contracts. There is also a positive association between external quality performance and
ﬁnancial performance when weight is placed on external quality performance in the
bonus contracts.

The effects of three additional factors on financial and non-ﬁnancial performance
are also examined. First, I examine whether the association between bonus weight and
quality and environmental performance will be more positive when managers report the
belief they are experiencing increasing marginal returns to investment. Contrary to

predictions, test results suggest a negative (no) association between external quality

 

(em-ironmentall Pt“

 

bonus contracts \\ h

 
 
   
  

Second. l ex
enrironmental pert}
budgeting decision l
relationship. FinulL
and quality and cm i‘
emironmental eost i
association betueen
emironmental eost '
between bonus \\ ei;

This study r
intemal and extern:

hints contracts dlik

(environmental) performance levels and high internal and external performance weight in
bonus contracts when managers perceive increasing marginal returns to investment.

Second, I examine whether the association between bonus weight and quality and
environmental performance is more positive when managers report having capital
budgeting decision rights for these measures. The tests reported here do not support this
relationship. Finally, I examine whether the association between ﬁnancial performance
and quality and environmental performance is more positive when reliable quality and
environmental cost information is available. The results suggest that a negative
association between environmental performance and ﬁnancial performance when
environmental cost information is available. Also, the results show a positive association
between bonus weight on environmental performance and ﬁnancial performance.

This study raises several interesting questions regarding the relationship between
internal and external quality performance and the relationship between the design of

bonus contracts and the design of annual review contracts for managers.

This dissertation is dedicated to my parents, Beverly and John Connors and to my sister,
Anne.

iv

 

  
  

l thank my t
lacohs and Dr. l: rat
acknott ledge the ill.

l am especiu

 

PhD. candidacy. S‘
Old Smith and Andr
lam also in i

W? long time.

ACKNOWLEDGEMENTS

I thank my dissertation committee, Dr. Susan Haka, Dr. Joan Luft, Dr. Fredrick
Jacobs and Dr. Frank Boster for their heroic patience and insightful guidance. I also
acknowledge the ﬁnancial assistance of Michigan State University.

I am especially thankful to my friends who have supported me throughout my
Ph.D. candidacy. Sally Ure, Jeff Ballard, Sandy Callaghan, Dan Heitger, Donna Booker,
Ola Smith and Andrea Drake have all been kind to me along the way.

I am also in debt to my children, Ian and Tristan, who have taken a back seat for a

very long time.

 

 

LlS’T 0F T:\lll-li.\

nrnonttt‘ritix

WK“
Wt
Multi-action l’riﬂ
The Value of Atlg
The Choice of l’e_
The Factors That
Ex .lanations for
QualitV and lim i:

Contracts ..

Contributions to ‘
\

 

   

will“ ~
JTllI'ﬁlzs
The Eiiects of \t

The Effects of l n

 

TABLE OF CONTENTS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

LIST OF TABLES ............................................................................................................ ix

CHAPTER 1

INTRODUCTION ............................................................................................................. 1

CHAPTER 2

LITERATURE REVIEW .................................................................................................... 4
Performance Measures in Incentive Contracts ............................................................... 4
Multi-action Principal-Agent Theory ........................................................................... 5
The Value of Additional Performance Measures in Incentive Contracts ........................ 6
The Choice of Performance Measures in Incentive Contracts ...................................... 11
The Factors That Affect Weights on Performance Measures in Incentive Contracts ...13
Explanations for Multi-action Agency Model Failure .................................................. 15
Quality and Environmental Performance Measures in Management Incentive

Contracts .............................................................................................................. 18
The Effects of Institutional Factors on Financial and Non-Financial Performance ..... 24
Contributions to the Literature ...................................................................................... 30

CHAPTER 3

HYPOTHESES ................................................................................................................. 33
The Effects of Non-Financial Measures in Incentive Contracts ................................... 33
The Effects of Institutional Factors on Financial and Non-Financial Performance ..... 35

CHAPTER 4 '

RESEARCH METHODS AND RESULTS ..................................................................... 37
Sample Selection and Research Methods ...................................................................... 37
Dependent Variable Construction .................................................................................. 40
Independent Variable Construction ............................................................................... 42
Descriptive Statistics ...................................................................................................... 46
Results: Hypotheses 1a, 3a, 3t; 5a and 5b ..................................................................... 48
Results: Hypotheses 2a and 2b ...................................................................................... 53
Results: Hypothesis 4 ..................................................................................................... 55

CHAPTER 5

CONCLUSION .................................................................................................................. 57
Discussion of Results ..................................................................................................... 57 ‘
Study Limitations ........................................................................................................... 59
Contributions to the Literature and Future Research Opportunities .............................. 61

 

vi

 

.ll’PENDlX B

%

 

 

 

.tPl’ENDlX C
SERVEY INS'I‘RI '3‘

BlBllOGRXl’l lY ..

TABLE OF CONTENTS (Continued)

 

 

 

 

 

 

APPENDIX A
SURVEY COVER LETTER EXAMPLE ......................................................................... 76
APPENDIX B
SAMPLE RESEARCH PROJECT SUPPORT LETTER ................................................. 77
APPENDIX C
SURVEY INSTRUMENT ................................................................................................. 78
BIBLIOGRAPHY .............................................................................................................. 88

 

vii

 

    
    

TABLE I: Pilllt‘lf.
Oblique Rotatior'

TABLE 3: Descri ‘

TABLE 3'. Ex Ante
\. .
hon-Financial l’e
\
RCVICW .....

TABLE 4: Ex Ante

Non-Financial l’c
TABLE 5: Descri ‘
\I;

TABLE 6: ()l.{\‘ R
\, .
etuhts on ()ua

(Zualitt‘ [)ecisiot

TABLE 7: 01.8 R

leieltts on (Jun

Ouaan Decisio

\l'eichts on Fm
lnr

CSIITTCDIS ant

TABLE 9: 01 S I
We I.

iuhts on No-

LIST OF TABLES

TABLE 1: Principal Components Analysis and Factor Loadings After
Oblique Rotation ............................................................................................................ 63

TABLE 2: Descriptive Statistics of Independent Variable Factors .................................. 64
TABLE 3: Ex Ante Expectations of the Weight Placed on Various Financial and
Non-Financial Performance Measures in the Managers’ Annual Performance

Review ........................................................................................................................... 65

TABLE 4: Ex Ante Expectations of the Weight Placed on Various Financial and

 

 

Non-Financial Performance Measures in the Managers’ Annual Bonus ....................... 66
TABLE 5: DescLiptive Statistics of Reggession Variables ............................................... 67

 

TABLE 6: OLS Regression of Internal mality Performance Levels on Bonus
Weights on Quality Performance MeasureLRetums to Quality Investments and
Quality Decision Rights ................................................................................................. 68

TABLE 7: OLS Regression of External Quality Performance Levels on Bonus
Weights on OualityPerfonnance Measures. Returns to Quality Investments and
Quality Decision Rights ................................................................................................. 69

TABLE 8: OLS Regssion of Environmental Performance Levels on Bonus
Weights on Environmental Performance Measures, Returns to Environmental
Investments and Environmental Decision Rights .......................................................... 70

TABLE 9: OLS Regression of Financial Performance on the use of Bonus
Weights on Non-Financial Performance and Non-Financial Performance ................... 71

TABLE 10: OLS Regression of Bonus Weight on Quality and Environmental
Performance Measures Levels and Availability of Quality and Environmental

 

 

 

Cost Information on Financial Performance .................................................................. 72
TABLE II: Variable Deﬁnitions ...................................................................................... 73
TABLE 12: Table of Results ............................................................................................ 74

 

viii

 

 

Multi-actiun
non-ﬁnancial perfor
an improrement of t
the association hem
measures. quality an
ﬁnancial and non-lit
between three instit

rertomtance. Them

enrrronmental in\ e-

CHAPTER 1

INTRODUCTION

Multi-action principal-agent theory suggests that, under certain circumstances,
non-ﬁnancial performance measures can be used in an agent’s bonus contract to induce
an improvement of both ﬁnancial and non-ﬁnancial performance. This study examines
the association between the weight placed on two speciﬁc non-ﬁnancial performance
measures, quality and environmental performance, in managers’ bonus contracts and
ﬁnancial and non-ﬁnancial performance. In addition, the study assesses the relation
between three institutional and process-related variables on ﬁnancial and non-ﬁnancial
performance. These variables are the level of decreasing returns to quality and
environmental investment experienced by the sample facilities, the amount of reliable
quality and environmental cost infomration available, and the level of control assigned to
the manager over quality and environmental capital investment decisions.

In general, previous studies on the value of the use of additional performance
measures in incentive contracts have been focused on determining whether certain non-
ﬁnancial performance measures are leading indicators of ﬁnancial performance. Another
related stream of literature investigates whether the use of incentive based or non-
traditional reward systems are associated with either current non-ﬁnancial or ﬁnancial
performance. This study extends the second stream of literature by testing the association
between the weights on non-ﬁnancial measures in a bonus plan and both ﬁnancial and

non-ﬁnancial performance. It also attempts to explicitly incorporate the effects of non-

 

   
  
 
  

linearity on the rcl-
studies hate doculi
specifically control
inlomiation and dc
performance.
Surt‘e} data
regarding the \xeiglt
annual rules and t
eight was placed .
than in their bonus
etTects of incentix‘c
relating to the insti
financial perfomia
atailahle informai
The result:
mcasures in the be
30mplex relations

"t”l' -‘ .
silent} lnkk’nttt'eg

linearity on the relation between ﬁnancial and non-ﬁnancial performance. Previous
studies have documented that these non-linear relations do appear to exist, but do not
speciﬁcally control for them. In addition, the study incorporates two contingency factors,
information and decision rights, that are hypothesized to have a positive association with
performance.

Survey data collected from plant managers in four industries provide information
regarding the weights assigned to several ﬁnancial and non-ﬁnancial measures in their
annual review and their bonus contracts for three years. Interestingly, on average more
weight was placed on non-ﬁnancial measures in the managers’ annual review contracts
than in their bonus contracts. This ﬁnding highlights the complexity of evaluating the
effects of incentive contracts on performance. The survey data also provided information
relating to the institutional and process-related variables of interest and on quality and
ﬁnancial performance. Environmental performance was measured using publicly
available information. '

The results indicate a positive association between the weight placed on quality
measures in the bonus contracts and external quality performance. The data also suggest a
complex relationship between internal and external quality performance and between
quality incentives and performance. Bonus weight on environmental performance is
positively associated with ﬁnancial performance levels. Also, external quality
performance is positively associated with ﬁnancial performance when weight is placed
on external quality performance in the bonus contract. Finally, the associations between
performance and experiencing decreasing returns to investment, information and decision

rights are mixed or non-existent, suggesting possible variable measurement issues or

 

complexities that .
questions.
The remain
literature ret'iett. (
I

research methods at

limitations of the st;

 

complexities that are not well understood and providing the basis for future research
questions.

The remainder of this paper is structured as follows: Chapter 2 presents the
literature review, Chapter 3 presents and discusses the hypotheses, Chapter 4 presents the
research methods and results, Chapter 5 concludes with a discussion of the results and

limitations of the study as well as directions for further research.

CHAPTER 2
LITERATURE REVIEW
Performance Measures in Incentive Contracts

Corporate managers are responsible for many aspects of ﬁrm performance. Some
of the decisions that they make and actions that they take affect current proﬁtability
directly. Examples of such actions often relate to cost reductions such as downsizing,
supplier changes, and product and process changes. Other decisions made and actions
taken by managers may only increase ﬁrm proﬁts in the future, sometimes at the expense
of current proﬁtability. Increased customer satisfaction, quality improvements, pollution
reduction, employee training, and investment in information technology are examples of
such actions. Management compensation contracts that reward performance measured
solely as short-term proﬁtability may discourage managers from making decisions and
taking actions that improve proﬁtability over a longer time horizon.

Firms are increasingly incorporating non-ﬁnancial performance measures in
management reward systems in order to focus attention towards long-term strategic
corporate goals and short-term customer satisfaction, particularly when achievement of
those goals is expected to be at the expense of near-tenn ﬁnancial performance (Ittner et
a1. 1994). Accordingly, there has been increasing interest in the performance
consequences of including non-ﬁnancial performance measures in executive
compensation plans as well as in the management of competing goals (Ittner and Larcker
1998).

The current study examines the financial and non-ﬁnancial performance effects of

incorporating non-ﬁnancial performance measures in bonus contracts of plant managers.

 

 

Economic theory _~
certain characterist
of performance on .
non-financial pcrti
leading indicator OT
indirect eliects of tl

ﬁnancial performan

 

 

 

marginal returns to c
information at'ailahl
int'estment decision:

There is an e
consequences of the
contracts. This char

studies relating to tl‘

determinants of the «

lllf‘ ~
ﬂoures. Explanat

lesslon 0f the ch
l

Economic theory suggests that inclusion of non-ﬁnancial performance measures that have
certain characteristics in the incentive contracts of managers should result in the increase
of performance on both the ﬁnancial and non-ﬁnancial measures. The improvement of
non-ﬁnancial performance is important because this performance is expected to be a
leading indicator of ﬁnancial performance. This study also examines the direct and
indirect effects of three institutional and process-related variables on ﬁnancial and non-
ﬁnancial performance. These variables are the perception of experiencing increasing
marginal returns to quality and environmental investment, the amount of reliable cost
information available and the control that a particular manager has over capital
investment decisions that affect his or her plant.

There is an emerging and diverse literature stream relating to the use and
consequences of the use of non—ﬁnancial performance measures in compensation
contracts. This chapter begins by reviewing the agency theory and supporting empirical
studies relating to the value of non-ﬁnancial measures in compensation contracts, the
determinants of the choice of such measures and the factors that affect the weights on the
measures. Explanations for multi-action agency model failure are presented. A
discussion of the choice of non-ﬁnancial measures to be included in this study follows. A
review of the literature relating to the institutional factors included in the model is
presented, followed by a discussion of the contributions of this study to the literature.
Multi-action Principal-Agent TheoLy

Agency models address the situation in which there is an incentive problem
because the agent’s (manager’s) actions are unobservable to the principal (owner). In

single action models the agent is typically risk-averse and effort-averse and the challenge

 

 

for the principal l>
principal desires v.
costly to the agent
such risk (Lambert
Multi-actiot‘
responsible fora se

measure of perform;

 

[he important action
action models suggc
may be added to an i
actions more congru
b} the contract on th

These multi-.
separated into three .
measures in incenti \-
incentive contracts. a

The Value OfA’klditii

for the principal is to create a contract that causes the agent to take the action that the
principal desires while limiting the risk imposed on the agent. Assumption of risk is
costly to the agent and he/she will require additional compensation in order to assume
such risk (Lambert 2001 ).

Multi-action contracting models address the fact that managers are typically
responsible for a set of actions. An incentive contract that relies on a single, imperfect
measure of performance will not provide the principal with information regarding all of
the important actions that are taken, or not taken, by the agent. In general, the multi-
action models suggest that under certain circumstances, additional performance measures
may be added to an incentive contract to reduce welfare loss by making the agent’s
actions more congruent with the principal’s preferences and by reducing the risk imposed
by the contract on the agent.

These multi-action contracting models and corresponding empirical work can be
separated into three categories: those that address the value of additional performance
measures in incentive contracts, those that address the choice of performance measures in
incentive contracts, and those that address the choice of weights associated with
performance measures in incentive contracts.

The Value of Additional Performance Measures in Incentive Contracts

F eltham and Xie (1994) acknowledged the fact that performance measures can be
imperfect reﬂections of economic consequences of manager’s actions. A performance
measure may be incongruent where the degree of congruity is deﬁned as the relationship
between the action of the agent and the objectives of the principal. A performance

measure can also be noisy, where noise is the effect of random, uncontrollable events on

 

 

metal
intluc

thiﬁ

ShUUi
f. or
dgﬁﬁl

ha\ e

..“Lch

“tilted

the measure. The use of additional performance measures in an incentive contract can
inﬂuence the agent’s actions to be more congruent with the goals of the principal and/or
can reduce the noise in the overall contract, thereby reducing the risk to the agent.

F eltham and Xie (1994) argued that one setting in which additional measures
should provide incremental value is in a contract in which a myopic measure is used. A
myopic measure is one that does not fully reﬂect the economic consequences of an
agent’s actions. Financial measures of performance such as accounting earnings or ROA
have been criticized as being myopic because they are focused on short-run performance
and penalize current investments in long-term proﬁtability. Suppose, for example, a
principal rewards an agent solely on the basis of current earnings, but customer
satisfaction inﬂuences future earnings. If the agency relation ends before customer
satisfaction generates additional proﬁts, the agent will focus all of his or her efforts on
improving current earnings and none on customer satisfaction. The Feltham and Xie
(1994) model suggests that the addition of a customer satisfaction measure to this
contract could add value (reduce welfare loss) by encouraging the manager to focus his or
her actions on both ﬁnancial and customer service performance, thereby improving
current and future ﬁnancial performance. The addition of this non-ﬁnancial performance
measure may also reduce compensation risk to the agent given the appropriate weights on
the measures and their variance-covariance properties (for an example see Feltham and
Xie 1994, 438-440).

Results of empirical studies that attempt to identify the effect of the use of non-
ﬁnancial measures in reward systems on ﬁnancial and non-ﬁnancial performance are

mixed (Banker et al., 2000; Campbell and Soderstrom, 1998; Ittner and Larcker, 1995).

Banker et al. (2000) investigated whether some non-ﬁnancial performance measures were
leading indicators of future ﬁnancial performance and whether the introduction of an
incentive plan that included both ﬁnancial and non-ﬁnancial measures was followed by
an increase in performance on both measures. Their model acknowledged that incentive
plans that include non-ﬁnancial performance measures can improve ﬁnancial
performance directly or indirectly through improved non-ﬁnancial performance. This
study beneﬁted from access to a long time-series of performance data pre- and post-
introduction of the incentive plan as well as information regarding the weights attached to
the measures in the plan.

The authors found that customer satisfaction measures were signiﬁcantly
associated with future ﬁnancial performance and that both ﬁnancial and non-ﬁnancial
performance improved after the introduction of a management incentive plan that
included non-ﬁnancial performance measures. Interestingly, the improvements on both
ﬁnancial and non-ﬁnancial measures were achieved with very little increase in incentive
intensity. That is, the maximum bonus eligibility as a percentage of base salary remained
unchanged for most managers.

Campbell and Soderstrom (1998) found that there is a positive association
between the inclusion of an environmental performance component in CEO
compensation and ﬁnancial and environmental performance for a subset of S&P 500
ﬁrms. They also found that the incorporation of environmental performance in CEO
compensation was associated with a lower risk premium in compensation.

Unfortunately, the data source used for this study does not speciﬁcally determine whether

the level of management reported is indeed the CEO or if it is some lower level such as

the management of the environmental and legal department. The data also does not
reveal the weights applied to environmental performance in the compensation contracts
of interest.

Ittner and Larcker (1995) found that a greater use of nontraditional information
and reward systems by ﬁrms with less extensive formal TQM practices is associated with
better performance as measured by return on assets and product quality. Yet, they did not
ﬁnd the expected result that the interaction of extensive use of TQM and nontraditional
information and reward systems was associated with high ﬁnancial performance. The
authors measured the reported importance of non-ﬁnancial measure relative to ﬁnancial
measures in compensation contracts, but were not able to determine what measures were
used and to what extent.

Sim and Killough (1998) studied the relationship between the focus on non-
ﬁnancial performance measures and goals and non-ﬁnancial performance. They
investigated whether there were synergies between Total Quality Management (TQM)
and Just-in-Time (J IT) management practices, performance goals, and incentive-based
compensation plans that affect customer and quality performance. They found that there
were indeed synergies with the combination of high levels of TQM and J IT, incentive
compensation and extensive performance goals having the highest effect on the non-
ﬁnancial performance measures. This study does not directly relate to the model
presented by F eltham and Xie (1994) because Sim and Killough (1998) did not know
whether the incentive pay contracts were based on the performance measures of interest.
However, the complementary aspects of incentive compensation and non-ﬁnancial

performance highlighted in their study have implications for the current research.

CI,

p‘

 

 

Using survey data, Perera et al. (1997) tested whether there is an association
between the use of customer-focused manufacturing strategy and the use of non-ﬁnancial
measures in performance measurement systems. They found the hypothesized positive
association. They also tested whether there was an association between the use of non-
ﬁnancial measures in performance measurement systems and self-reported ﬁnancial
performance. They did not ﬁnd this association. Unfortunately, the authors were unable
to determine if the non-ﬁnancial measures used were actually incorporated into an
incentive compensation system. They also did not test for an association between the use
of such measures and non-ﬁnancial performance.

Empirical studies that test for an association between customer satisfaction and
future ﬁnancial performance also yield equivocal results. Foster and Gupta (1998)
attempted to determine if customer satisfaction performance was a leading indicator of
ﬁnancial performance in a wholesale beverage company. Their results varied depending
on the questions included in the customer satisfaction measure and depending on whether
they were testing a levels model or a percentage change model. Their results appear to
indicate diminishing returns to customer satisfaction performance at high levels of
performance.

Ittner and Larcker (1998) also studied the effect of customer satisfaction on future
ﬁnancial performance using three distinct samples of data; speciﬁcally, the customer
level, the business-unit level and the ﬁrm level. They found that customer satisfaction
measures were associated with future customer purchase behavior, growth in the number
of customers and ﬁnancial performance. However, they found that these associations

move through stages and appear to diminish at high satisfaction levels.

The empirical studies discussed above investigated the relationship between the
use of non-ﬁnancial measures in reward systems and non-ﬁnancial and ﬁnancial
performance. The current study is an extension of this stream of literature. It attempts to
examine the same relationships using the weights placed on non-ﬁnancial performance
measures in bonus contracts and incorporating other institutional and process related
variables that are expected to have an impact on those relationships.

The Choice of Performance Measures in Incentive Contracts

Feltham and Xie (1994) suggested that the addition of a performance measure to
an incentive contract adds value if the measure is sufﬁciently congruent with the
principal’s goals, sensitive on the agent’s actions, and is not too noisy. A noisy measure
imposes additional risk to the agent for which he or she will require compensation. This
additional cost may outweigh the beneﬁt of the additional congruent measure. Therefore,
the precision of the measure (the inverse of the variance) is important to the choice of
performance measure and the weight (incentive intensity) assigned to it in a linear
contract.

Dikolli and Kulp (2002) addressed the fact that performance measures may be
interrelated and an agent’s-efforts may be interrelated. They showed that these substitute
or complementary relationships between both performance measures and agent efforts
affect the choice of, weight and use of measures in compensation contracts. This result is
independent of the sensitivities (the mean change in the measure due to an agent action),
precision, or congruity of the measures.

Bushman et al. (1995) presented a model in which interdependencies within a

ﬁrm affected the type of performance measures chosen for compensation contracts of

individual managers. Speciﬁcally, they showed that the use of aggregate measure rather
than more localized measures of performance increases with intraﬁrm interdependencies.
This is a single action model, however. In a multi-action model, the weights applied to
aggregate and localized measures depend on the type of interdependencies (Lambert
2001).

Bushman et al. (1995) examined their theoretical model by investigating the
association between ﬁrm interdependencies and the use of aggregate performance
measures in the bonus contracts and long-term compensation contracts of division and
group CEO’s of a sample of 246 public ﬁrms. They found that the use of aggregate
performance measures in these contracts is positively associated with interindustry
segment sales and intergeographic segment sales (proxies for interdependencies) and
negatively associated with product-line diversiﬁcation and geographic diversiﬁcation
(proxies for divisional independence).

Ittner and Larcker (2002) extended Banker et al. (1995) by providing an empirical
test of the determinants of the choice of speciﬁc performance measures in worker
compensation plans. They found that the factors that inﬂuence the choice of speciﬁc
measures in the contracts vary and are complex. For example, they found that plans that
supported a continuous improvement strategy made more use of both ﬁnancial and non-
ﬁnancial performance measures such as volume, quality and attendance, but not safety.
Plans that supported innovation-oriented strategies were negatively associated with the
use of volume, safety and attendance measures.

Ittner and Larcker (2002) also found that regulation affected the choice of

performance measures in worker contracts. In addition, they found that the reasons for

adoption of a certain compensation plan, such as organizational change, affected
performance measure choices as well as did unionization and management participation
in plan design. The authors concluded that studies that aggregate performance measures
into classiﬁcations such as ﬁnancial and non-ﬁnancial for the purpose of studying
performance measure choice should acknowledge these complexities when making
inferences from results.
The Factors That Affect the Weights on Performance Measures in Incentive Contracts

The ﬁnal set of studies relating to contracting in a multi-action setting investigate
the determinants of the appropriate relative weights to be applied to performance
measures in a linear contract. Banker and Datar (1989) constructed a model in which the
weights assigned to each measure were a proportional to sensitivity/precision.
Empirically, Core et al. (2002) found that the relative weight on stock price and non-
stock price measures in total CEO compensation was an increasing function of the
relative variances, while the relative weights in cash compensation were a decreasing
function of the relative variances. They suggest that their ﬁndings may differ from
economic theory because the incentives from cash compensation, as opposed to salary
and stock option grants for example, may have a stronger effect on motivation than the
incentives from non-cash compensation. However, the authors admit, this explanation
seems counterintuitive because the equity portion of the compensation would impose
costly risk to the CEO without observed incentive beneﬁts.

Datar et al. (2001) extended the work of Feltham and Xie (1994) by investigating
the determinants of the optimal weights on performance measures in a multi-action

framework. Their model indicated that the principal chooses the weights in order to

maximize the congruity between the performance measures and the principal’s payoff.
The weights are also affected by the interactions between the measures in the contract.
Datar et al. (2001) also showed that there is a cost/beneﬁt tradeoff between the congruity
of the contract and the (costly) risk imposed on the agent. Increased sensitivity of the
measure to the actions of the agent does not necessarily lead to increased weight on the
measure in a contract due to congruity and variance effects as well as performance
measure interactions. The current study does not speciﬁcally address these effects or
interactions. However, the results of the study may be affected by these variables which
are, unfortunately, difﬁcult to identify and measure.

Ittner et al. (1997b) identiﬁed factors that affect the weight placed on ﬁnancial
and non-ﬁnancial measures in fonnula-based CEO annual bonus contracts. They found
that the weight of non-ﬁnancial measures increased with the level of regulation, the
extent to which the ﬁrm followed an innovation-oriented strategy, adoption of strategic
quality initiatives and the noise in the ﬁnancial performance measures. The authors
identiﬁed these particular institutional factors to examine because they represent
situations in which non-ﬁnancial performance measures are expected to be relatively
more informative about managerial actions. Due to the constraints that are inherent in the
use of public data to generate information relating to the compensation contracts of
CEOs, the (Ittner et al. (I997b) study was unable to address the question of the
determinants of the weights in the contracts on speciﬁc non-ﬁnancial performance
measures.

The model in this study tests for positive associations between the weight placed

on two speciﬁc non-ﬁnancial performance measures in the bonus contracts of plant

managers and concurrent ﬁnancial and non-ﬁnancial performance. As such, this model
relates most closely to the multi-action agency theory literature stream regarding the
value of certain measures in performance contracts. This study does not attempt to
identify the factors that affected the choice of these measures for the sample of plants, nor
the choice of weights on those measures in the bonus contracts. However, a discussion of
those two particular multi-agent literature streams is instructive because it illustrates the
complexity of the factors involved in creating an incentive contract and may provide
explanations for weak or nonexistent observed associations between the use of
performance measures and performance in practice.

Explanations for Multi-action Agency Model Failure

Recall that the Feltham and Xie (1994) model suggests that the appropriate use of
additional performance measures in an incentive contract will cause the actions of a
manager to be more congruent with the goals of the ﬁrm. Speciﬁcally, the use of the
measures will improve concurrent performance on ﬁnancial and non-ﬁnancial measures
where there is an assumption that the non-ﬁnancial performance is a leading indicator of
ﬁnancial performance. There are several reasons why ﬁnding empirical evidence of
these relationships using survey data and archival data may be difﬁcult.

In order to encourage improvement on both ﬁnancial and non-ﬁnancial
performance the addition of non-ﬁnancial measures in an incentive contract must reduce
the overall risk to the manager, alternatively, the incentive intensity of the contract must
increase (F eltham and Xie, 1994). Otherwise, the manager may simply shift effort away
from actions that affect ﬁnancial performance towards affecting non-ﬁnancial

performance, thereby reducing current ﬁnancial performance assuming no

 

 

 

U:
Ax.

complementarities. In practice, ﬁrms may underestimate the risk imposed on the
manager by multi-measure performance contracts or may underestimate the future
beneﬁts of current non-ﬁnancial performance and accordingly under-provide ﬁnancial
incentives to the manager. In either event, a negative effect on ﬁnancial and/or non-
ﬁnancial performance may occur.

Dikolli and Kulp (2002) suggested that interactions between the agent’s efforts
and the interactions between performance measures should be taken into account in the
design of incentive contracts. In practice, these interactions could be difﬁcult to identify
and measure and these difﬁculties could result in suboptimal contract design. For
example, process changes adopted in order to reduce costs may simultaneously affect
product quality, pollution emissions and safety measures. Further, it is not clear that
ﬁrms choose the weights on performance measures in accordance with agency theory
because compensation contracts are complex and include both cash and non-cash
incentive components (Core et al., 2002).

As previously discussed, tests of the value of non-ﬁnancial performance measures
as leading indicators of ﬁnancial performance exhibit mixed results. Despite the claims
of quality and customer satisfaction improvement proponents, the appropriate functional
form of this suggested relationship has not been determined (Ittner and Larcker, 1998).
This lack of a predictable relationship could cause ﬁrms to under(over)-estimate the
future ﬁnancial beneﬁts of improved non-ﬁnancial performance and therefore
under(over)-provide incentives to managers.

It is also important to understand that multi-action agency models rely on the

assumption of linear contracts. Even in cases when this assumption holds, there may be

judgement and discretion used in compensation decisions that essentially make the
contracts non-linear (Lambert, 2001). These adjustments to compensation may not be
observable in available data but may have an impact on the relationship between
performance measure use and performance. Indeed, it is not clear that formula-based
contracts are the best contract choice for all ﬁrms. Gibbs et al. (2002) pointed out that
ﬁrms may avoid the use of formula based contracts because of their incompleteness,
inﬂexibility and because of the risk associated with target achievability. Using data
collected from department managers in automobile dealerships, they showed some
evidence that subjective bonuses are used to complement perceived weaknesses in bonus
contracts based on quantitative performance measures.

There is also an argument that the use of ﬁnancial rewards in certain
circumstances can be counterproductive. For example, Crosby and Deming warned
against the use of ﬁnancial rewards for quality performance because such a system could
adversely affect the teamwork required to successfully implement a quality program

(Crosby, 1996). Instead, public recognition was suggested as a sufﬁcient reward for
outstanding performance.

Finally, even though ﬁnancial rewards may be used in a linear compensation
contract to elicit desired behavior from managers, other aspects of rewards evident in
organizations may be more important to the manager. Raises, promotional opportunities
and rePutation could affect the choices that managers make and could be an inﬂuential

factor in the relationship between the use of certain performance measures in contracts

and the resulting performance effects.

17

Quality and Environmental Performance Measures in Management Incentive Contracts

The non-ﬁnancial performance measures of interest in the model proposed in this
study are quality performance and voluntary pollution reduction performance. These two
measures are considered to be appropriate for this study for two reasons. First,
researchers disagree regarding whether the use of ﬁnancial incentives and rewards will
encourage quality improvement and pollution reduction. Crosby and Deming argue that
ﬁnancial rewards are counterproductive in a Total Quality Management (TQM) setting
(Crosby, 1996). On the other hand, Juran (1989) argues that recognition and rewards
through salary increases, bonuses and promotion are essential to engender motivation for
quality performance. Wruck and Jensen (1994) criticize the arguments of Deming and
Crosby as being too simplistic. They suggest that properly designed and implemented
ﬁnancial reward systems are essential to TQM success. .

Similar differences are expressed in the environmental performance literature.
Makower (1994) is ambiguous in his discussion of the value of ﬁnancial incentives and
rewards for Environmental Management Systems (EMS) and towards pollution reduction
goals. He provides examples of ﬁnancial reward use and the use of more intangible
rewards such as public recognition in various companies, but does not suggest that either
type of rewards are preferable. However, Epstein (1996) emphatically argues that stated
corporate environmental goals must be reﬂected in the performance measurement used
for salary, bonus and promotion purposes. He speciﬁcally argues that the use of solely
ﬁnancial performance goals in the evaluation of performance for compensation purposes

undermines stated goals of pollution reduction.

18

 

 

These disagreements raise interesting issues regarding the performance effects of
including these performance variables in bonus contracts. TQM and EMS performance
measures have also been chosen for this study because of striking similarities between
TQM and EMS (Klassen and McLaughlin, 1996). The fundamental goal of each is to
reduce waste. Both TQM and EMS pursue long-term goals that integrate either product
quality or pollution reduction in all phases of decision making. Such phases include
product and process design. supplier evaluation, marketing, product delivery and use,
customer service and post-consumer product disposition (Hunt and Auster, 1990). The
theories of EMS are heavily drawn from the theories of TQM and the programs may
interact (Epstein, 1996). Both TQM systems and EMS can be internationally certiﬁed
under the ISO 9000 and ISO 14000 standards, respectively.

Second, quality performance is arguably a leading indicator of ﬁnancial
performance (Banker et al., 2000; Foster and Gupta, 1998; Ittner and Larcker, 1998;
Ittner and Larcker, 1995). The TQM literature proposes that the beneﬁts from
investments in product quality result from internal cost reductions and increased revenues
from improved customer satisfaction. Increased product quality should lead to lower
warranty costs and greater production efﬁciencies. Direct productivity beneﬁts result
from lower costs of scrap, rework, inspection, and repair. Indirect productivity beneﬁts
result from lower downtime due to fewer defects in purchases from suppliers and
production, lower buffer inventories, higher machine utilization, and lower schedule
changes and congestion (Ittner, 1994).

Increased customer satisfaction should reduce marketing costs and increase

revenues through higher customer retention and loyalty, higher proﬁt margins, and lower

19

vulnerability to competitive threats (Ittner and Larcker, 1996). The logic of this model is
quite compelling. So much so that, since the mid-1980’s, most American companies
have initiated some type of quality program (Ittner and Larcker, 1996). It is expected that
the beneﬁts of quality investments will be recognized over time (Banker et al., 2000) and
may not be captured entirely by current ﬁnancial performance measures. Therefore,
quality performance appears to exhibit characteristics that should make it appropriate for
inclusion in a model that tests the effects of non-ﬁnancial performance measures in bonus
contracts on ﬁnancial and non-ﬁnancial performance.

In contrast to the TQM model, standard economic theory suggests that, at the ﬁrm
level, voluntary pollution reduction cannot be ﬁnancially justiﬁed because pollution is an
extemality. Absent regulation, the costs of pollution are borne by the public and not by
individual polluters (Milgrom and Roberts 1992). Of course, in the U. S. and many other
countries environmental regulations are fairly strict and the penalties for violation can be
harsh. Thus, regulation imposes additional production costs on many manufacturing and
some service ﬁrms. There is general agreement that this cost is quite high (Jaffe et al.,
1995).

There has been a rather lively debate among economists about the theoretical
possibility of a positive relationship between environmental performance and economic
performance (Jaffe et al., 1995; Palmer et al., 1995; Porter and van der Linde, 19953,
1995b). Most of the differences between among the researchers hinge on the reference to
either static or dynamic economic environments. Porter has been the most outspoken
supporter of a dynamic view of environmental regulation that provides economic beneﬁts

to ﬁrms through increased learning and innovation over time (Porter, 1991; Porter and

20

 

 

 

 

van der Linde, 1995a, 1995b). There is some anecdotal evidence to support this claim
(Fisher and Schot, 1994; Lanen, 1999). Empirically, however, no systematic relationship
between environmental compliance costs and successful patent applications has been
found (Jaffe and Palmer, 1997) and the effects of environmental performance on
productivity are inconclusive (Jaffe et al., 1995).

Russo and F outs (1997) found a positive relationship between environmental
performance and Return on Assets (ROA) with evidence of higher returns to
environmental performance in high growth industries. Overall, however, empirical
studies have found that the relationship between environmental performance and
concurrent ﬁnancial performance is negative, or at best zero, as many economists would
predict (Freedman and Jaggi, 1992; Jaggi and F reedman,l992).

Despite these ﬁndings, there is some empirical evidence supporting the argument
that increases in environmental performance reduce future regulatory and liability risks
relating to accidents (Klassen and McLaughlin, 1996) and are, therefore, priced by the
market (Johnson et al., 1996). For example, Konar and Cohen (1997) found that
environmental performance measured as levels of indexed pollution emissions and
environmentally related lawsuits pending is negatively related to intangible asset value in
a subset of S&P 500 ﬁrms.

Static extemality models (eg. Palmer et al., 1995) predict that ﬁrms should reduce
pollution emissions only in response to regulations, but, voluntary over-compliance with
environmental regulations does exist. The EPA hails its voluntary 33/50 program as an
enormous success and sponsors several other voluntary emissions and energy use

reduction programs. This voluntary program was initiated in 1991 and encouraged ﬁrms

to reduce their emissions of seventeen toxic chemicals by thirty-three percent by 1992
and by ﬁfty percent by 1995. In addition, the 1996 Toxics Release Inventory (TRI)
public data release shows a voluntary reduction of nearly 50% in the releases of toxic
chemicals over the previous decade.l

Some voluntary emissions reductions may be simply a by-product of investment
decisions, such as equipment upgrades or customer-required product changes that also
happen to have a positive environmental effect. Reductions could also be a function of
the fact that compliance investments can be lumpy in their effects resulting in
unintentional over-compliance (Arora and Cason, 1995). First-mover ﬁrms may be able
to reap the beneﬁts of selling pollution technology and may intentionally over-comply in
order to guide regulatory authorities to set tighter standards for the industry, thereby
raising the cost of compliance for other ﬁrms and restricting competition (Caimcross,
1990; Klassen and McLaughlin, 1996). Compliance in anticipation of future regulations
gives managers ﬂexibility to make cost-effective emission reductions without the threat
of non-compliance ﬁnes and penalties (Boyd, 1998a; Caimcross, 1990).

Other strategic beneﬁts may result from voluntary reductions. Pressure from
environmental groups, customers, and community members, at least in certain
geographical regions and industry sectors, can be extremely high. The reputations of poor

performers may make it harder for them to win permission for expansion, to recruit and

 

‘ The Emergency Planning and Right-to-Know Act (EPCRA) was enacted in response to the Union Carbide
chemical release accident in Bhopal, India. Section 313 of EPCRA established the Toxics Release
Inventory (TRI) program. TRI is a national database that identiﬁes facilities and chemicals manufactured
and used, released and otherwise managed at each facility. The TRI Program has collected and made this
information public since 1987. As of 3/24/03, the 2000 Public Data Release can be accessed at
http://wwwepagov/tri/

22

motivate staff, and to attract customers (Caimcross, 1990).

A study by Cram (1997) supports this notion of public inﬂuence. He founds that

a large, unidentiﬁed, U. S. chemical company implicitly prices public commitments to
pollution reductions in capital budgeting decisions. These positive prices existed when
controlling for the ﬁnancially quantiﬁable Net Present Value (N PV) of the individual
investments. In other words, the company placed a signiﬁcant positive value on public
promises to reduce emissions which, in turn, inﬂuenced their capital budgeting decisions.

Also, representatives of the EPA have publicly stated that, given the agency’s
scarce resources, scrutiny is focused on the worst polluters. Facilities with good track
records are less likely to be audited by the agency. In addition, positive environmental
performance on the whole may reduce ﬁnes and penalties when accidents occur and may
be an advantage in court (Epstein, 1996). The sentences imposed on ofﬁcers of
corporations convicted of federal environmental offenses may be reduced if it can be
shown that an environmental compliance program has been instituted. No reduction in
sentences is allowed if there are no explicit incentives for environmental compliance
within the ﬁrm (Hamner and Stinson, 1995).

The existence of voluntary pollution reduction goals and achievements within
ﬁrms suggest that there is a positive ﬁnancial performance effect of improved
environmental performance beyond simple compliance with regulations. Admittedly,
this ﬁnancial effect may occur in the distant future. As with quality performance,
voluntary environmental performance also appears to exhibit characteristics that should

make it an appropriate non-ﬁnancial performance measure for inclusion in the model.

The Effects of Institutional Factors on Financial and Non-Financial Performance

The model in this study includes three institutional factors that may affect
ﬁnancial and non-ﬁnancial performance. These factors are the stage of the process
toward quality and environmental performance goals, the amount of reliable cost
information available and the control that a particular manager has over capital
investment decisions that affect his or her own plant. These three factors were chosen for
this study because they were highlighted in personal and telephone interviews with plant
managers in the chemical, automobile and automobile parts industries and because of the
importance of the factors in the studies discussed below.
The Stage of process toward quality and environmental performance goals

Executives often make reference to the phenomenon of the “low-hanging fruit”
when discussing the inevitable tradeoffs that result from improved environmental
performance. It seems to be an accepted, but untested, phenomenon that the "farther
companies travel along the road to zero emissions, the smaller the returns will become"
(Caimcross, 1990, p. 9). Comments such as the following are typical:

"We picked the low-hanging fruit ﬁrst, in the 70's and 80's," emphasized Edward

E. Quick, the environmental manager for Celanese. "As we move into the 1990's

we're ﬁnding the available technology much more costly." (Meyerson, 1998, l)

“Enlightened companies have exhausted many of the relatively easy energy,

waste, and resource-efﬁciency options. They are now into the harder, longer term

investment commitments in which conventional economic and environmental

criteria are not necessarily in harmony.” (Fischer and Schot, 1994, 46-47)

As described, this phenomenon implies that investments made in the early stages

of an environmental management program are generally not costly and often yield high

ﬁnancial and environmental payoffs. In later stages, investments become expensive and

24

 

 

 

 

complex and the payoffs become highly uncertain at best. Thus, it appears that a non-
linear relationship may exist between investment and payoff that could affect the
outcome of performance evaluation choice. Lanen (1999) studied waste reduction at 3M
Corporation. He found that performance improvements were negatively associated with
baseline performance and plan age, but was not able to measure returns to investment for
his sample.

The “low-hanging fruit” phenomenon has also been documented in the quality
literature (Ittner and Larcker, 1996; Atkinson et al., 1994). Ittner and Larcker (1996)
found decreasing returns to quality improvement programs over time, with returns
measured as self-reported costs-of-quality. The results are similar when customer
satisfaction is used as a proxy for product quality. Ittner and Larcker (1998) found that
customer satisfaction was a leading indicator of future accounting earnings, growth in
customers and customer purchase behavior. This relationship appears to be non-linear
with diminishing returns to customer satisfaction at high levels of satisfaction.

In contrast, multi-period dynamic quality—based learning models suggest that
ﬁrms may actually achieve increasing quality returns to investment over time (Fine 1986;
Marcellus and Dada 1991). Empirically, Ittner (1994) examined the reported quality
costs of 49 plants over a mean period of nearly ﬁve years. He found that 29 of the 39
plants that reported reduced nonconformance costs (a return measure) also reported
reduced expenditures on conformance activities (an investment measure). Individual
time series analysis of the quarterly cost data of 24 of the plants showed that reductions in
failure costs were associated with simultaneous reductions in prevention and appraisal

costs. Ittner et al. (1998b) examined 13 plants of a single company and found that the

prevention and appraisal costs of the plant appeared to be contingent on the levels of
quality experienced during the period. Their ﬁndings are consistent with the quality ,
learning models that suggest that quality expenditures can be reduced as quality
improves.

The empirical evidence of increasing returns to quality expenditures provided by
Ittner (1994) and Ittner et al. (1998b) is not inconsistent with the “low-hanging fruit”
story. Neither study examines the behavior of quality costs at very high levels of quality.
As Ittner et al. (1998b) point out, quality cost tradeoffs and performance implications
may be different at higher quality levels than those achieved by their research site. Given
empirical and anecdotal evidence regarding the “low-hanging fruit”, or stage effect, it is
expected that non-ﬁnancial performance will decrease as a ﬁrm approaches the later
stages of quality or environmental programs and experiences decreasing returns to
investment.
Information and Decision Rights

Jensen and Meckling (1992) argued that decision rights should be co-located with
agents that have speciﬁc information. Simultaneously, reward systems that are designed
to align the interests of the principal and the agent should be implemented in order to
make sure that the agents do not misuse the information and decision rights.
Accordingly, Wruck and Jensen (1994) argued that incentives must complement the co-
location of decision rights and information in order to implement a successful TQM
program. By the same argument, it would not be optimal to provide incentives to agents

without ensuring that the agents have appropriate information and decision rights.

26

 

Even though a compensation system may align the interests of the agent with
those of the principal, the agent may not be able to take appropriate actions due to poor
information or a lack of decision rights. It is possible that the rights assignment and
information problem may have adverse effects on quality and environmental decision
making and outcome performance (Epstein 1996; Joshi, et al. 2001). This study will
focus on the effects of two contingency factors that have been highlighted in recent
academic, practitioner and EPA literature as obstacles to increased ﬁnancial, quality and
environmental performance: traditional cost accounting and capital budgeting decision-
making systems.

The Amount of Reliable Cost Information Available

Traditional cost accounting systems may not provide sufﬁcient or appropriate
information to support effective TQM or EMS decision-making. Previous studies have
suggested that new manufacturing practices require new sources or types of information
(Ittner and Larcker 1995; Wruck and Jensen 1994). This is particularly true in the case of
environmental management. Recall that the effects of pollution are not fully realized by
polluting ﬁrms. Without market prices or explicit bargaining with affected parties,
managers may not have the information on the costs and beneﬁts of their actions that is
necessary help them to accomplish environmental goals (Milgrom and Roberts, 1992).
At the very least, managers will require accounting information regarding the costs of
pollution emissions within their own facilities in order to make informed decisions
regarding pollution reduction.

Traditional accounting systems tend to hide environmental costs in overhead

accounts and fail to consider all of the potential environmental costs associated with the

27

ﬁrm's activities (Epstein 1996; McLaughlin and Elwood 1996; Walley and Whitehead
1994). A survey of management accountants found that 70-80% of the respondents
allocated environmental costs to overhead. These overhead costs were usually reallocated
to products or processes based on labor hours, production volume and other volume
related cost drivers (White et al. 1995). This leads to the standard problem of under/over-
costed products that may result in inefﬁcient resource allocation.

In addition, ﬁrms tend to assign a value of zero to these “hidden costs” in capital
budgeting, product pricing, product and process design, and purchasing decisions
(Epstein, 1996). This practice results in a systematic bias against decisions that will
produce environmental beneﬁts (Porter and van der Linde 1995a, 1995b).

Joshi, et al. (2001) analyzed plant level data from 55 steel mills and found that
“hidden” environmental costs are 9 — 10 times higher than those identiﬁed by the plants
costing systems. The authors conclude that the underestimation of costs may lead to the
aforementioned distortions in decision making.

The under-reporting of the costs and beneﬁts of quality performance is also a
problem within ﬁrms (Ittner et al. l998b; Ittner and Larcker 1996; Nandakumar 1993).
Cost-of-poor-quality systems tend to track only direct and easily measured costs. In a
survey of manufacturing plants, Ittner and Larcker (1996) found that internal costs such
as scrap, rework, inspection and testing, planning and analysis, warranties and returns and
training were most often tracked by accounting systems. External costs were often not
measured, reﬂecting the difﬁculty of estimating customer-related costs and beneﬁts

associated with quality improvement. Ittner et al. (1998b) determined that the reported

 

quality costs of thirteen plants signiﬁcantly understate the ﬁnancial beneﬁts of quality
improvements.

While the use of non-ﬁnancial performance measures in bonus contracts may
increase the willingness on part of an agent to invest in quality enhancing or pollution
reducing investments, the agent may not know the most cost efﬁcient means to do so
because he does not have the appropriate information. Therefore, it is expected that the
availability of such information will be associated with higher ﬁnancial performance.
Control Over Capital Investment Decisions

As previously discussed, the beneﬁts "of pollution reduction and quality
investments are largely underestimated by accounting systems and the costs of pollution
and poor quality typically are not adequately tracked within organizations. Capital
budgeting decision-making processes that do not correct for these underestimates tend to
be biased against the acceptance of quality and environmental investments. It is often the
case that quality and environmental projects compete for scarce funds with all other types
of investment opportunities. This practice combined with the use of all unadjusted
traditional methods of screening and approval of capital investments, such as Net Present
Value (NPV), Internal Rate of Return (IRR), Return on Investment (ROI) and Payback
may bias against approval of quality and environmental projects given scarce investment

I‘CSOLII‘CCS.2

 

2 Four chemicals and automotive parts managers were interviewed in person or over the telephone during
the preliminary stages of this study. Each manager stated that the capital budgeting practices of their
respective companies were too rigid to adjust for the difﬁculties inherent in estimating the beneﬁts of
environmental investments. Discussions with two EPA consultants conﬁrmed that this is a common
problem with the companies that they work with.

29

Flexibility in capital budgeting decision making systems should provide managers
additional decision rights necessary to make investment choices that adjust for this bias.
Atkinson et al. (1994) proposed a system for evaluating investment opportunities which
incorporates both ﬁnancial and non-ﬁnancial quality information. The assignment of
non-zero weight to non-ﬁnancial beneﬁts in the investment approval process is also
proposed by Boyd (1998b) as an improvement upon the standard ﬁnancial analysis of
environmental investments.

Decision rights over investment choices should give managers the ﬂexibility to
take actions that are suggested by their incentive systems, thereby increasing quality and
environmental performance. Such management decision inﬂuence may take the form of
weighted non-ﬁnancial characteristics, separate evaluation of quality and environmental
investments, or other adjustments to the capital budgeting decision making process.
Contributions to the Literature

This study examines the association between the weight placed on quality and
environmental performance measures in the bonus contracts of plant managers and
ﬁnancial and non-ﬁnancial performance. It also attempts to measure the direct and
indirect effects of three institutional and process related variables on ﬁnancial and non-
ﬁnancial performance. This study will contribute to several literature streams in various
ways.

The study extends previous work on the associations between non-ﬁnancial
performance measures, non-ﬁnancial performance and ﬁnancial performance. Ittner and
Larcker (2001) outline several limitations in the work on this issue to date. They state

that non-linearity in various relations has not been adequately incorporated into

30

performance measurement models. This study addresses this issue by attempting to
measure the process stage (of environmental and quality performance for each plant and
including that measure explicitly in the model.

They state that results in this area seem to vary by industry and suggest that tests
should be conducted over a broad range of industries. They also point out that, due to
data limitations, little is known about the-incentive systems of lower-level (below CEO)
employees even though their impact on performance can be substantial. This study
addresses both of these issues by creating a sample from four manufacturing industries
and by focusing on the bonus plans of plant managers.

Ittner and Larcker (2001) also state that past studies of the value relevance of non-
ﬁnancial performance measures tend to look at only one value driver and ignore potential
interactions with other value drivers. They suggest that it is possible that non-ﬁnancial
performance measures could be highly correlated as substitutes or complements. This
study includes two non-ﬁnancial performance measures which are, in theory, positively
correlated.

The also point out that previous empirical work in this area has ignored
contingent factors that may moderate the relation between the value driver and
performance. They add that previous studies have tended to overlook the quality of
information used for decision making and control even though this aspect of information
may affect the quality of decisions made and the effects of incentive and control systems.
This study addresses both of these issues. The amount of reliable environmental and
quality cost information available and the capital budgeting decision making control that

a manager has are both included as variables in the study’s model.

31

This study will also contribute to the quality and enviromnental program design
literature by investigating whether the inclusion of quality and environmental
performance measures explicitly into the bonus contracts of managers has a positive or
negative effect on quality and environmental performance. This has been an area of
much disagreement.

Finally, this study should contribute to the quality and environmental literature on
the association between non-ﬁnancial and ﬁnancial performance. Recall that the results
of the diverse studies in this area are mixed. Few of the previous studies have examined
these relations at the plant level, particularly in the environmental literature. This study

should serve to enrich this small, but growing, stream of literature.

32

CHAPTER 3
H YPOTHESES
The Effects of Non-Financial Measures in Incentive Contracts

This study proposes an empirical model that is based on the multi-action agency
model of Feltham and Xie (1994). In particular, it provides evidence on the theory that,
under certain conditions, non-ﬁnancial measures can add value to an incentive contract in
which a myopic measure is used. This study assumes that short-term measures of
ﬁnancial performance are myopic and tests whether the non-ﬁnancial performance
measures in a bonus contract will make the agent’s actions more congruent with the goals
of the principal.

One key assumption of this and other studies that attempt to measure the
performance effects of performance measurement and other accounting choices is that
individual ﬁrms are exhibiting sub-optimal behavior at any particular point in time. The
cross-sectional variation across ﬁrms inform us about the relation between accounting or
measurement choice and performance (Ittner and Larcker, 2001). This study not only
relies on this assumption, it also assumes that goals of the sample ﬁrms include increased
quality and environmental performance. This assumption does not appear to be
unreasonable given the previously discussed theoretical beneﬁts of improved
perfomiance on these two measures such as reduced costs, increased revenue due to
improved customer satisfaction, reduced penalties due to environmental accidents,
beneﬁts of improved reputation and product and process improvements.

The ﬁrst model of this study tests whether the inclusion of quality and

environmental bonus plans of plant managers will improve the plant’s performance on

33

those measures. Speciﬁcally, it is expected more heavily weighted non-ﬁnancial
measures in bonus contracts will be associated with higher non-ﬁnancial performance.

Hla: There is a positive association between the weight placed on quality
performance in the annual bonus contract and performance.

Hlb: There is a positive association between the weight placed on
environmental performance in the annual bonus contract and
environmental performance.

The model constructed to test H1 assumes that quality and environmental
performance are leading indicators of future ﬁnancial performance. Unfortunately, this
expected association between current non-ﬁnancial performance and future ﬁnancial
performance cannot be tested due to the data limitations of the sample. Nonetheless, it is
possible for current ﬁnancial performance measures to reflect the beneﬁts of current
quality and environmental performance. Indeed, Feltham and Xie (1994) model implies
that this result is possible under certain circumstances; namely risk reduction to the agent
or increased incentive intensity.

Firms may use additional non-ﬁnancial performance measures in compensation
contracts in order to increase the motivation for, or reduce the risk of, actions to improve
non-ﬁnancial performance. If this is the case, it is expected that ﬁnancial performance
will be positively associated with quality and environmental performance when explicit
quality and environmental performance measures are included in bonus contracts.

H2a: There is a positive association between ﬁnancial performance and
quality performance when weight is placed on quality performance
measures in annual bonus contracts.

HZb: There is a positive association between ﬁnancial performance and

environmental performance when weight is placed on these
environmental performance measures in annual bonus contracts.

34

The Effects of Institutional Factors on Financial and Non-Financial Performance
The Stage of Process Toward Quality and Environmental Performance Goals

There is anecdotal evidence of decreasing ﬁnancial returns to environmental
investments. In addition, some empirical evidence suggests diminishing returns to
customer satisfaction performance and mixed evidence of a non-linear relationship
between quality performance and ﬁnancial performance. This study does not attempt to
specify a non-linear relation between ﬁnancial performance and quality and
environmental performance. However, the study does address whether experiencing
decreasing returns to quality and environmental investments makes improvement of
quality and environmental performance more difﬁcult. Speciﬁcally, it is expected that
there will be a positive association between non-ﬁnancial performance and the interaction
between the weight on quality and environmental performance in the bonus contract and
perceived increasing marginal returns to quality and environmental investments.

H3a: The weight placed on quality performance measures in annual bonus
contracts will have a more positive effect on quality performance
when increasing marginal returns to investment are experienced.

H3b: The weight placed on environmental performance measures in annual
bonus contracts will have a more positive effect on environmental
performance when increasing marginal returns to investment are
experienced.

Incentives, Information and Decision Rights

Jensen and Meckling (1992) suggested that there were three institutional factors
that are complimentary within a ﬁrm. Speciﬁcally these factors are the location of
decision rights and speciﬁc knowledge or information and the incentive system. Demers

et al. (2002) found that these factors are generallyjointly determined in a subset of

intemet companies. The current study does not measure the interdependencies of these

three factors. However, complimentary effects of these factors are predicted to be

associated with improved ﬁnancial and non-ﬁnancial performance. Speciﬁcally, it is

expected that there is a positive association between ﬁnancial performance and the

interaction between availability of reliable quality and environmental cost information

and the weight on quality and environmental performance in the annual bonus contract.

H4a:

H4b:

The weight on quality performance measures in annual bonus
contracts will have a more positive effect on quality performance
when reliable quality cost information is available.

The weight on environmental performance measures in annual bonus
contracts will have a more positive effect on environmental
performance when reliable environmental cost information is
available.

It is also expected that there is a positive association between quality and

environmental performance and the interaction between speciﬁc decision rights, namely

capital investment decision rights, and the weight on quality and environmental

performance in the annual bonus contract.

HSa:

HSb:

The weight placed on quality performance measures in annual bonus
contracts will have a more positive effect on quality performance
when capital investment decision rights can be exercised.

The weight placed on environmental performance measures in annual
bonus contracts will have a more positive effect on environmental
performance when capital investment decision rights can be exercised.

CHAPTER 4
RESEARCH METHODS AND RESULTS

Sample Selection and Research Methods

In the early stages of this study, executives of several companies in various
industries were personally contacted and asked to participate in a ﬁeld study. Each of the
executives expressed an interest in the topic of this study and the results. They refused to
participate, however, due to the sensitive nature of the data and institutional decision
making information requested. They were particularly concerned about the potential of
attracting the attention of the EPA towards their companies. One senior manager stated
that he was speciﬁcally concerned that the EPA would subpoena the collected data and
use that as a steppingstone to examining the internal documents of the company.
Contacts at the EPA conﬁrmed that the concerns of the executives were indeed
appropriate.

Consequently, data for this study was collected from mail surveys of
plant/operations managers in four industries. The industries were chosen based on
Toxics Release Inventory (TRI) performance because this public data is a key objective

measure of environmental performance in this and previous research.3 There is no

 

3 A facility 'must report releases of listed chemicals if it belongs to certain SIC codes industry sectors, has
l0 or more full-time equivalent employees and manufactures or processes more than 25,000 pounds or uses
more than 10,000 pounds of any listed chemical per year. The EPA’s listed chemicals are considered to be
the most toxic chemicals in use, not all chemicals. Listed chemicals change periodically, but not during the
time period covered by this study. Additional information regarding TRI data can be found at
http://www.epagov/tri/tridata/triOOlpress/overview as of March, 2003.

known source of public information on quality performance at the ﬁrm or plant level.

The chemical (SIC 28) and paper (SIC 26) industries were chosen because they are
relatively high polluting industries. Over the time period of 1988 - 1998 the chemical
industry reduced its TRI releases by 57%, while the paper industry reduced its releases by
only 14%. The furniture (SIC 25) and plastics (SIC 30) industries were chosen because
they are medium-level polluters. Over the same time period the furniture industry
reduced its releases by 73% and the plastics industry reduced its releases by 39% (US.
EPA, 1998). This cross section of industries provides variation in environmental
performance.

A list of names and addresses of plant/operations managers in the four industries
was collected using the TRI database. This list was manually ﬁltered to target plants with
the highest levels of pollution and to eliminate plants that were no longer in production.4
Each plant was then contacted by telephone in order to conﬁrm the name of the
plant/operations manager and the mailing addresses. This exercise also served as a
second way of identifying plants that had closed. The ﬁnal sample consisted of 1,257
plants of which 278 were in the furniture industry, 319 were in the paper industry, 549
were in the chemical industry and 111 were in the plastics industry. Of the total sample,

682 were plants in where a speciﬁc name of a plant/operations manager was obtained.

 

4 The TRI database provides data on the pounds of chemicals released. This data is not scaled by
production levels or full-time equivalent employees. Therefore, in its raw form it does not provide
information regarding facility performance in general or relative to other facilities in its industry. A high
level of releases does not necessarily imply poor environmental performance. The decision to target plants
with the highest levels of pounds of emissions was made in order to survey plants where environmental
performance might matter enough to include incentives in a bonus plan. It was not intended to limit the
variability of the environmental performance variable and should not cause bias because the dependent
variable, pounds of emissions scaled by full-time equivalent employees, was not known until the completed
survey was received.

In order to maximize response rates to the survey instrument, a modiﬁed version
of the Dillman (1991) method was used. Two mailings of the survey were sent to the full
sample of plants. The second mailing occurred approximately two weeks after the ﬁrst
mailing. The cover letter of introduction emphasized that all responses would be kept
conﬁdential and would be analyzed in the aggregate (See Appendix A). A letter written
by APICS-The Education Society for Resource Management was included with each
survey (See Appendix B). This society is widely recognized and highly regarded in the
manufacturing community. The letter expressed support for the project and encouraged
participation.

A total of l 19 survey responses were received of which 117 were usable. Twenty
surveys were returned as undeliverable. Accordingly, the response rate to the survey was
approximately 9% of the full sample. While this response rate is low, it is consistent with
the rates of other recent independent mail surveys regarding business practices. Of the
total usable responses 22 were from managers of furniture plants, 31 were from managers
of paper mills, 55 were from managers of chemical plants and 9 were from managers of
plastics plants. Sixty-nine responses were from managers that had been contacted

directly by name and seventy-ﬁve were responses from the ﬁrst mailing.5

 

5 Tests for non-response bias were performed on respondent/non-respondent emission levels for the years
1996 and 1997 by industry. Two sample unpaired t-tests were performed under the assumption of unequal
variances. 1n the furniture industry the respondents mean emissions were smaller than non-respondents in
1997 and there were no differences at the .05 signiﬁcance level in 1996. In the paper, chemicals and plastic
industries for both years the mean emissions of the respondents were larger than the non-respondents
(p<.05). It cannot be determined if the larger emissions are a result any particular plant being heavier
polluter or simply a larger facility. There is no measure of size available for the non-respondents.

39

 

Dependent Variable Construction

In the survey design phase of this study a pilot survey was sent to 329 plant
managers in the automotive parts industry. The pilot survey asked the managers to
provide either levels or change from prior period information on speciﬁc quality and
ﬁnancial performance measures for their plants. Of the 41 responses received, most
refused to provide this information citing data sensitivity issues. Therefore, the survey
constructed for this study asked the respondents to provide the year-to-year levels of
quality and ﬁnancial performance for their plant but they were not asked to disclose the
type of measure that they were using (see Appendix C for the survey instrument). So, the
survey data provides levels of performance on the dependent variables, but what those
measures represent (except in a few cases) was not collected. This method of response
provides a way of addressing conﬁdentiality issues and thereby increasing response rates,
but results in noisy measures of the dependent variables relating to quality and ﬁnancial
performance. As discussed below, outlier performance measure responses were trimmed
resulting in a loss if data. Most importantly, interpretation of the results is difﬁcult
because it must be assumed for the statistical tests that the responses reﬂect comparable
measures. Though this measure of data collection introduces considerable noise into the
data, it is not an obvious source of bias.

The respondents were asked to provide the levels of their most important external
quality measure on an annual basis. Examples of external quality performance measures
are number of on-time deliveries, customer satisfaction and defective parts per million.
They were also asked to indicate whether their performance on this variable had

improved from year to year. Nearly all respondents indicated that an increase in this

40

measure was an improvement. Thirty-three responses (9%) were eliminated from this
analysis because they indicated that an increase in this measure indicated a decrease in
performance. This indicated that these respondents’ external quality performance
measures were different from the external quality performance measures of the majority
of the respondents and, therefore, were not comparable. The range of the omitted
responses was .92 — 4,500 with nineteen in the range of .92 — 54. The average of the
retained responses was 89.5 with a range of 70 — 99. Again, these responses represent the
annual level of the most important external quality measure. The measure itself may not
be the same across responses.

The respondents were also asked to provide the levels of their most important
internal quality measure on an annual basis. Examples of internal quality performance
measures are scrap or rework rates. They were also asked to indicate whether their
performance on this variable had improved from year to year. Nearly all respondents
indicated that a decrease in this measure was an improvement. Of the total responses,

1 13 (32%) were eliminated from this analysis because they indicated that an increase in
this measure indicated an increase in performance. Again, this implied that these
respondents’ measure of internal quality performance was different from the majority of
the respondents and, therefore, not comparable. The range of the omitted responses was
20 - 350 with 102 omitted responses in the range of 60 — 99. The average of the retained
responses was 4.8 with a range of .005 - 20. This represents the annual level of the most
important internal quality measure. The measure itself may not be the same across

responses.

41

Quality performance was dichotomized into extemal quality measures and
internal quality measures because production facilities tend to measure them separately.
Ponemon et al. (1994) found that companies in the early stages of quality improvement
tend to focus on external failure rates at the expense of internal failure rates.

Data on environmental performance has been collected from the EPA’s TRI
database. This database provides levels of chemical emissions measured in pounds on a
facility level. While types of chemical emissions are categorized in the TRI reports, they
have been aggregated for the purposes of this study because it is assumed that the plants
are attempting to reduce overall emissions and not speciﬁc categories such as air
emissions. The reported emissions have been scaled by the self-reported full-time
equivalent employees for each facility as a proxy for production levels.

Independent Variable Construction

To measure the weight placed on quality and environmental performance two
different measures were combined. First, the survey asked the respondents to estimate
their ex ante expectations of the percentage of their bonus that was based on the quality
management (internal quality) and customer satisfaction (external quality) performance
of their plants for each of the years 1997 — 1999. The respondents were also asked to
estimate the percentage of their bonus that was based on environmental performance for
the same time period as well as other ﬁnancial and non-ﬁnancial performance measures.
The bonus percentages were forced to sum to one-hundred percent. Respondents
allocated percentages to other ﬁnancial and other non-ﬁnancial categories in order to
reach that 100%. All bonus responses correctly summed to a total of one-hundred

percent.

Second, in three separate questions the respondents were asked to rate the
signiﬁcance of their bonus to them on a six — point Likert scale. This is intended to
represent a crude measure of incentive intensity because the actual dollar amount of the
respondents’ bonuses is not known. The conﬁrmatory factor analysis factor loadings on
the three bonus signiﬁcance questions ranged from .64 to .77 (Table 1, A) and the
Cronbach alpha of the three bonus signiﬁcance items was .74 (Table 2). The three items
that measure intensity were averaged and multiplied by the bonus percentages for internal
quality, external quality and environmental performance in order to create the
independent variables INTQWTSIG (internal quality), EXTQWTSIG (external quality)
AND ENVWTSIG (environmental quality). These variables proxy for the weight placed
on internal quality, external quality and environmental performance in the respondents’
bonus contracts. Descriptive statistics for these variables are reported in Table 5. Recall
that Hla and Hlb propose a positive association between the weight placed on these non-
ﬁnancial performance measures and the respective non-ﬁnancial performance.

The remaining independent variables are reported decreasing returns to quality
(H3a) and environmental (H3b) investments, availability of reliable quality cost (H4a)
and environmental cost (H4b) information, and quality (H5a) and environmental (H5b)
capital investment decision rights. These variables were measured by answers on a six-
point Likert scale anchored by completely disagree and completely agree. Table 1
reports the factor loadings from Principal Components Analyses performed on responses
relating to these variables. A Principal Components Analysis was performed separately
for each of these variables. Table 2 reports the Cronbach alphas, means, standard

deviations, theoretical and actual ranges of the variables.

The stage items attempted to measure the stage at which the plant was operating
in regard to the quality and environmental improvement process. These stage measures
relate to H3a and H3b. The process stage for each of the quality and environmental
processes was measured by four questions. Table 1, Panel A shows the results of
conﬁrmatory factor analyses in which these four questions loaded on two separate
constructs for both quality and environmental performance stage. The ﬁrst two questions
relating to quality stage loaded on a construct labeled QPOSl. These questions related to
the existence of inexpensive and cost effective opportunities for product quality
improvement. The second two questions loaded on a construct labeled QPOSZ. These
questions related to the existence of decreasing returns to quality investments. Table 2
reports Cronbach alphas for these measures of .82 and .48, respectively. The ﬁrst two
questions relating to environmental stage loaded on a construct labeled EPOS 1. These
questions related to the existence of inexpensive and cost effective opportunities for
pollution reduction. The second two questions loaded on a construct labeled EPOSZ.
These question related to the existence of decreasing returns to pollution reduction
investments. The Cronbach alphas for these measures are .90 and .57, respectively.

The information systems questions provided a measure of the availability of
extensive and reliable quality and environmental cost information within the plant which
is used for examining H4a and H4b. Four items measured quality information systems
and their underlying construct is labeled QUALIS. Table 1, Panel B shows that the factor
loadings on these four questions range from .54 to .68. Table 2 reports a Cronbach alpha
for this measure of .89. Five items measured environmental information systems and

their underlying construct is labeled ENVIS. The factor loadings on the third and ﬁfth

44

 

questions are quite low (.37 and .49, respectively). Nevertheless, the questions have been
retained in this measure because omitting them does not change the outcome of the OLS
regressions reported in this section. Also, the Cronbach alpha for the ENVIS measure
including all ﬁve questions is .79 and is acceptable.

The capital budgeting practice questions provided a measure of the degree of

inﬂuence or control that the manager had over the plant-level quality and environmental

 

investment choices used to examine H5a and H5b. Four questions (see Appendix C,
Section D, the ﬁrst four questions) measured control over quality investment decisions
and the underlying construct is labeled QUALC B. The factor loading on the fourth
question is .45. The question has been retained in the QUALCB measure for three
reasons. First, it is the question that most directly addresses the issue of control over
quality capital budgeting decisions. Second, omitting this question from the measure or
including it as a separate measure in the subsequent OLS regressions does not
signiﬁcantly alter the results. Finally, the Cronbach alpha on the QUALCB measure
including all four questions is .71 and is acceptable.

Five questions (see Appendix D, Section D, questions 8 — 12) measured control
over environmental investment decisions. The factor loadings on the ﬁrst two questions
are low at .38 and .34, respectively. Including these two questions as a separate measure
in the OLS regression test of H5b does not alter the overall results, so the questions have
been omitted from the analysis. The last three questions have been retained and the

underlying construct is labeled ENVCB. The Cronbach alpha of this measure is .64.

45

Descriptive Statistics

A maximum of three years of levels data was collected from each respondent
when possible. This is because respondents were asked to provide estimates of their ex
ante expectations of the weight placed on various ﬁnancial and non-ﬁnancial measures in
their annual performance review and in their annual bonus for the years 1997-1999 to
collect key independent variable data. They were also asked to provide quality and
ﬁnancial performance levels information for the years 1996-1999 for dependent variable
data. The amount of information provided by the respondents on these variables varied
based on years in the position and memory constraints. Therefore, sample size varies
between the models included in this study. As discussed above, the respondents also
provided current measures of the independent variables relating to bonus signiﬁcance,
process stage, cost information availability and capital budgeting processes. It is
assumed that the responses to these questions would not change for the years 1996-1999.
To the extent that this is not true. additional noise is introduced into the measurement of
these variables.

Table 3 provides descriptive statistics relating to the structure of the managers’
annual performance reviews. The data in this table is provided in order to compare the
structure of annual reviews with the structure of bonuses in Table 4. Annual review
weights differed among industries. Managers in the paper and plastic industries reported
the highest weight placed on quality management in their annual performance review
(11% and 10%, respectively). Managers in the furniture and chemicals reported the
highest weight placed on customer satisfaction in annual reviews (13% and 11%,

respectively). Managers in the paper and chemicals industries reported the highest

46

weight placed on environmental performance (6% and l 1%, respectively). Overall, the
managers in the chemical industry reported a total of 56% of weight in the annual
performance reviews was placed on non-ﬁnancial performance measures.

Table 4 provides the respondents ex ante expectations of the weight placed on

 

various ﬁnancial and non-ﬁnancial measures in their annual bonus. Consistent with the
annual review weights, the managers in the chemical and plastic industries reported the
most weight placed on quality management (6% and 17%, respectively) for their annual
bonus. Consistent with annual performance reviews, managers in the furniture and
chemicals industries reported the most weight placed on customer satisfaction (6% for
each). Managers in all four industries reported that very little weight was placed on
environmental performance with the largest weight of only 7% reported by managers in
the chemicals industry. Overall, the same managers reported that 35% of the weight in
their bonuses was placed on non-ﬁnancial measures.

In order to test whether the weight on each of the performance measures in Table
3 was signiﬁcantly different from the weight on the measures in Table 4 two sample,
paired t-tests were perfomted under the assumption of unequal variances. The mean
weights in the annual performance review were signiﬁcantly different from the weights in
the annual bonus for all categories except for plant ﬁnancial performance and other non-
ﬁnancial performance measures (p<.05).

Table 5 reports the means and standard deviations of the dependent and
independent variables used to examine Hl — H5. Interpretation of the means of the
dependent variables is difﬁcult due to the previously discussed inherent measurement

difﬁculties. The mean of the independent variable ENVWTSIG, representing the weight

47

placed on environmental performance in the bonus contract multiplied by bonus
signiﬁcance, is small and consistent with the low weights reported in Table 4. The means
of the variables EXTQWTSIG and INTWTSIG, representing the weight placed on
external and internal quality performance, respectively, multiplied by bonus signiﬁcance,
are larger but still quite small.

The descriptive statistics for the balance of the independent variables reported in
Table 5 represent the mean values of constructs that are measured by items that can take
on values ranging from 1 - 6. On average, the managers reported fewer cost effective and
inexpensive opportunities exist for environmental improvement (EPOS 1) than for quality
improvement (QPOSl). Consistent with this result, the managers reported that they were
closer to facing diminishing returns to investment for environmental investments
(EPOS2) than for quality investments (QPOS2). The managers also reported that their
ﬁnancial quality information (QUALIS) was slightly more reliable and useful that their
ﬁnancial environmental information (ENVIS). Finally, the managers reported that they
had slightly more control over their quality investment choices (QUALCB) than their
environmental investment choices (ENVC B).
Results: Hypotheses la, 3a, 3b, 5a and 5b

The ﬁrst set of tests examine models of the non-ﬁnancial performance effects of
the weight placed on quality and environmental performance in bonus contracts, stage
effects and the interaction of capital investment decision rights and quality and
environmental bonus weights. H la predicts that there will be a positive association
between the weight placed on quality performance measures in a bonus plan and quality

performance. H3a predicts that there is a positive association between quality

48

performance and the interaction between the weight on quality performance measures in
the bonus contract and reported increasing marginal returns to quality investments. H5a
predicts that the interaction between capital investment decision rights relating to quality
investments and the weight on quality performance measures in the bonus contract will
be positively associated with quality performance. Model 1(a) represents these
hypotheses.

Model 1(a):

QPERF, = a0 + (1| QPERFH + a2 QWTSIG, + a3 QPOS, + a4 QWTSIGR‘QPOS, +
asQUALCBl + abQUALCBt* QWTSIG, + ayAGEt

Where:

QPERF = Quality performance levels.

QWTSIG = The weight attached to quality performance*bonus
signiﬁcance.

QPOS = Two measures of decreasing returns to quality investments.

QUALCB = Measure of quality capital investment decision rights.

AGE = Age of facility.

Quality performance for the previous year (QPERF 1-.) is included in this levels

model in order to control for time-series trends (Banker et al 2000; Ittner and Larcker

1998; Sim and Killough l998).6‘ 7

 

6 Prior studies have also tested models with performance as a dependent variable as change models (Banker
et al 2000; Ittner and Larcker 1998; Sim and Killough 1998). That is, both the dependent variable and
independent variables are measured as changes. This model provides econometric beneﬁt of reducing bias
relating to omitted variables and spurious correlations (Banker et al. 2000, Wooldridge 2000). Some
independent variable changes have not been measured for this study in order to reduce the length of the
survey and thereby increase the response rate. Therefore, OLS regressions with performance changes as
the dependent variable measured as changes and the independent variables measured as levels were
performed. The results were not signiﬁcantly different from those of the levels models and are not reported
here.

7The Cook-Weisberg test for heteroscedasticity and visual inspection of residual plots revealed
heteroscedasticity for all of the levels models in this study. The results of these regressions are reported
using Huber/White/Sandwich adjusted standard errors. Examination of Variance Inﬂuence Factors (VIP)
for all models revealed the expected multicollinearity of the interaction terms with their interacted
variables. In order to reduce the effects of variance inﬂation, interaction terms are dropped in regressions
where their coefﬁcients are not signiﬁcant.

49

A control variable which measures the age of the plant is included in all models
with non-ﬁnancial performance as the dependent variable. Speciﬁcally, age is expected
to be negatively associated with non-ﬁnancial performance levels because of the
difﬁculty of improving performance in older facilities.8

Tables 6 and 7 provide the results of the tests of Model 1(a). The internal quality
performance results are presented in Table 6. Recall that the respondents indicated that a
decrease in the reported internal quality performance measure was an improvement. In
order to clarify the interpretation of regression results, the internal quality performance
data has been coded to be negative. So an improvement in internal quality performance
is an increase in that measure and the predicted signs on the INTQWTSIG and
EXTQWTSIG variables are positive. Recall that Hla predicts that there will be a
positive association between the weight placed on quality perfomtance measures in a
bonus plan and quality performance. Table 6, a model with internal quality performance
and the dependent variable, shows that the coefﬁcients on these terms are not signiﬁcant
at conventional probability levels.

The external quality performance results are presented in Table 7. The regression

results show a positive association between external quality performance levels and the

 

8 Industry control variables were included in all of the models presented but were not signiﬁcant.
Therefore, they have been excluded from the analyses currently presented.

50

weight placed on internal quality performance measures (p<.05) and external quality
performance measures (p<.10) after controlling for the effect of the interaction between
the bonus weights and experienced decreasing returns to quality investments.

Tests of H3a are presented in Tables 6 and 7. H3a predicts that there is a positive
association between quality performance and the interaction between the weight on
quality performance measures in the bonus plan and reported increasing marginal returns
to quality investments. The QPOS2 variable has been reverse coded in order to present a
clearer interpretation of the results on the variable and its associated interaction term.
Accordingly, a high response (say 6) on this variable represents reported low perception
of decreasing returns to investment. Thus, a high response on this variable combined
with high bonus weight is expected to be positively associated with performance levels.
The interaction term is not signiﬁcant in the internal quality performance model
presented in Table 6 which does not support H3a.

The results of the external quality performance model presented in Table 7 do not
support H3a either. The coefﬁcients on the interactions between QPOSI and the weight
placed on internal and external quality performance in the bonus contract are not
signiﬁcant. The coefﬁcients on the interaction terms of QPOSZ and the weight in internal
and external quality measures in the bonus contract are negative and signiﬁcant (p<.01
for both). This suggests that managers in the early stages of the external quality
improvement process that have relatively high weight placed on internal and external
quality in their bonus contracts are experiencing low external quality performance levels.

It also suggests that managers that perceive high decreasing returns to quality investment

51

 

and have relatively lower weight placed on quality performance in their bonus contracts
are experiencing high external quality performance levels.

H5a predicts that the interaction between capital investment decision rights
relating to quality investments and the weight on quality performance measures in the

bonus contract will be positively associated with quality performance. The results

 

provide no support for H5a. The interaction of quality capital investment decision rights
(QUALCB) and the weight on internal quality (INTQWTSIG) or external quality
(EXTQWTSIG) in the bonus contract is not signiﬁcantly associated with internal (Table
6) or external (Table 7) quality performance levels.

Model 1(b) provides a test of Hlb, H3b and H5b. Speciﬁcally, H1 predicts that
there will be a positive association between the weight placed on environmental
performance measures in the bonus contract and environmental performance levels
measured as pounds of TRI emissions weighed by the reported number of full-time
equivalent employees. Note that a lower level of ENVPERF means lower levels of
emissions, an indicator of good performance. This levels data has been coded negative to
aid in the interpretation of the results. Therefore, all associations with this ENVLEVEL
are predicted to be positive. H3b predicts that there is a positive association between
environmental performance levels and the interaction of high environmental bonus
weight and reported low experience of decreasing returns to environmental investment.
H5b predicts that the interaction between capital investment decision rights relating to
environmental investments and the weight placed on environmental performance in the

bonus plan will be positively associated with environmental performance.

Model 1(b):

ENVPERF. = (10 + on ENVPERFH + a2 ENVWTSIG, + a3 EPOS, + a4 ENVWTSIG, *
EPOS, + (15 ENVCB, + a6 ENVCB,*ENVWTSIG, + 0L7AGE,

Where:
ENVPERF = TRI emissions levels scaled by number of full-time equivalent
employees.
ENVWTSIG = The weight attached to environmental performance*bonus
signiﬁcance.

EPOS = Measures of decreasing returns to environmental investments.
ENVCB = Measure of enviromnental capital investment decision rights.

AGE = age of facility.

Table 8 presents the results of tests of Model 1(b). Hlb, H3b and H5b are not
supported in this model. Speciﬁcally, the coefﬁcient on ENVWTSIG is not signiﬁcant
(Hlb). The coefﬁcients on the interactions between EPOSl and bonus weight on
environmental performance as well as EPOS2 and bonus weight on environmental
performance are not signiﬁcant (H3b). Finally, the coefﬁcient on the interaction between
ENVCB and bonus weight on environmental performance is also not signiﬁcant (H5).
Results: Hypotheses 2a and 2b

H2a predicts that ﬁnancial performance will be positively associated with quality
performance when quality performance measures are included in the bonus plan of the
manager. H2b predicts that ﬁnancial performance will be positively associated with
environmental performance when environmental performance measures are included in
the bonus plan of the manager. Model 2b is constructed to assess the effects of the
interaction between quality and environmental performance and weight placed on those

performance measures in a bonus contract on ﬁnancial performance.

 

Model 2:

FPERF. = the + ¢IFPERFH + in ENVPERF. + 4), QPERF, + q). ENVWTD. + q), QWTDt +
(p, ENVPERF*ENVWTD, + «p, QPERF*QWTD,

Where:

FPERF = Financial performance levels.
ENVPERF = Environmental performance levels.
QPERF = Internal and external quality levels.
ENVWTD = Dummy variable where:

1 = Positive weight placed on environmental performance in the bonus
contract.

0 = Zero weight placed on environmental performance in the bonus
contract.
QWTD = Dummy variable where:
1 = Positive weight placed on quality performance in the bonus contract.
0 = Zero weight placed on quality performance in the bonus contract.
The results presented in the ﬁrst column of Table 9 show that the coefﬁcient on
the interaction of internal quality performance and a dummy variable representing
positive weight on internal quality performance in the bonus contract is not signiﬁcant.
This result does not support H2a. The second column of Table 9 presents OLS regression
results without the internal quality variables. Removal of these variables increases the
sample size signiﬁcantly. In this reduced model, 112a is supported for external quality
performance. The interaction term EXTQPERF*EXTQWTD is positive and signiﬁcant
(p<.15). These results also show that H2b is not supported. The interaction term
ENVPERF*ENVWTD is not signiﬁcant. However, there is a signiﬁcantly negative

relationship between ﬁnancial performance and the dummy variable representing positive

weight on environmental performance in the bonus contract (p<.15).

54

 

Results: Hypothesis 4

 

Model 3 provides a test of the association between the use of quality and
environmental performance measures in bonus contracts and ﬁnancial performance.
Model 3 also provides a test of H4, which predicts that there is a positive association
between ﬁnancial performance and the interaction between the weights on quality and
environmental performance and the availability of extensive and reliable cost
information.

Model 3 .'

FPERF. = [30+[31FPERFH+1321£NVWTSIGt + [howrsrot + B4ENVIS, +
BSQUALISI + B6ENVWTSIG,*ENVIS, + B7QWTSIGI*QUALIS,

Where:
FPERF = Financial performance levels.
ENVWTSIG = The weight attached to environmental performance*bonus
signiﬁcance.
QWTSIG = The weight attached to quality performance*bonus
signiﬁcance.
ENVIS = Degree to which reliable environmental cost information is available.
QUALIS = Degree to which reliable quality cost information is available.
Table 10 shows the results of Model 3. These results suggest that there is a
positive association between the weight on environmental performance in the bonus
contract and the levels of ﬁnancial performance (p<.15). That is, higher weight on
environmental performance is associated with higher ﬁnancial performance after
controlling for the effects of the interaction of weight on enviromnental performance in
the bonus contract and the availability of reliable environmental cost information. H4,
however, is not supported. The coefﬁcients on the interaction between bonus weight on

quality performance and the availability of reliable quality cost information are not

signiﬁcant. The interaction between the weight placed on environmental performance

55

 

and the reported availability of reliable environmental cost information is negatively
associated with ﬁnancial performance (p<.10). This result implies that managers that
have relatively high bonus weights on environmental performance and report high
availability of reliable environmental cost information are experiencing relatively low

ﬁnancial performance.

56

CHAPTER 5
CONCLUSION

Discussion of Results 9

This study proposes that the weight placed on quality and environmental
performance in managers’ bonus contracts will be positively associated with quality and
environmental performance and with ﬁnancial performance. The data suggest a
relationship between the bonus weight placed on quality management and external
quality performance. Speciﬁcally, there is a positive association between the weight
placed on internal and external quality measures in the bonus contract and external
quality performance levels after controlling for interactions between the weights and
experiencing decreasing returns to quality investments. The corresponding associations
were not found relating to either internal quality bonus weight and performance or
environmental bonus weight and performance. It is interesting that internal quality bonus
weight is positivelyassociated with external quality performance levels but not internal
quality performance levels. Measurement error on the internal quality performance
measure could explain this result or it is possible that internal and external quality
performance are related in a manner that is not captured by these models.

Tests of the data also show that there is a positive relationship between external
quality performance and ﬁnancial performance when weight is placed on external quality

performance in the bonus contract. This association may be due to the lagged effect of

 

9 A summary of the results of tests of the hypotheses presented in this study can be found in Table 12.

1.10
\l

 

past quality performance on current ﬁnancial performance or due to current quality
performance on current ﬁnancial performance. Unfortunately, this distinction cannot be
evaluated due to data limitations.

The data also reveal a positive association between bonus weight on
environmental performance and ﬁnancial performance levels. There is no similar
relationship between the bonus weight and environmental performance, though. One

possible explanation for these results is that higher ﬁnancial performers can afford to

shift bonus weight from ﬁnancial performance to environmental performance, and do so.

The effects of this shift in incentives may occur over a long time horizon and therefore
are not captured in a cross-sectional study such as this.

The effects of certain institutional factors on performance were also investigated
in this study. The results of the tests of these effects do not generally support the
proposed hypotheses. For example, high bonus weight on environmental performance
combined with high environmental cost information availability is associated with low
ﬁnancial performance levels. Perhaps, the hypothesized positive association cannot be
captured in a short time horizon study. This result also suggests that low bonus weight
on environmental performance combined with low environmental cost information
availability is associated with high ﬁnancial performance levels. This is counterintuitive
and reveals questions for further study. 1

The effects of decreasing returns to investment on quality and environmental
performance are not clear. On average, managers that report the existence of high levels
of cost-effective and inexpensive opportunities for quality improvement combined with

higher weights on quality measures in their bonus contracts manage plants with low

 

external quality performance levels. One interpretation of this result is that more bonus
weight is placed on quality measures in bonus contracts of managers in plants that are
early on in their quality improvement process. The converse would also apply. Low
bonus weight is placed on quality measures in bonus contracts of managers in plants that
are experiencing decreasing returns to quality investment. The non—linearity in
performance measures certainly merits further attention. It is likely that the measures of
non-linearity used in this study were too crude to actually capture the effects on
performance and interactive effects with bonus weights particularly relating to internal
quality and environmental performance. Further studies with more reﬁned measures may
ﬁnd the hypothesized relationships.

Finally, the respondents in this study report that there is more weight placed on
non-ﬁnancial performance measures in their annual review contracts than in their bonus
contracts. This further complicates the relationship between incentives and performance
because annual reviews may affect overall compensation and most likely affect other
career aspects that are important to managers such as promotion and reputation.

Study Limitations

There are several limitations to this study that are worthy of discussion. First,
there is considerable noise in the measurement of the dependent variables. Because the
study focused on tests of hypotheses at the plant level, public quality and ﬁnancial
performance was not available and data sensitivity caused managers to respond to very
general questions regarding this performance. Assumptions have been made regarding
the cohesiveness of the responses that may not hold true. The use of a survey

methodology to collect this data will probably always result in this problem. One

59

ll

possible solution to this problem is to convince a single or a few companies to provide
internal plant-level performance data.

The environmental performance data is public and is considered to be credible.
However, this data has been normalized using a proxy for production levels that is
imperfect. Also, it is not known if the TRI data used in this study is the performance
measure used for the managers’ bonus contracts. It is possible that the plants have
internal environmental performance goals that are only marginally reﬂected in TRI
performance.

There may also be considerable noise in the in the independent variables. The
respondents were asked questions regarding cost information reliability, investment
decision rights and program stage for 1999, not for the years 1997-1999. It is assumed
that these institution factors are stable over that period, but that may not be the case. In
retrospect, the respondents should have been asked if there was a change on these
variables over the period, when, and in what direction. Incorporating this information
into the model may have reduced some of the noise.

Second, the theory supporting this study has an incentive intensity component that
could not be measured due to data sensitivity issues. It is highly unlikely that the
respondents would have revealed the size of their bonuses or the size relative to salary.
This piece of information could surely provide a clearer understanding of the effects of
incentives on performance. Obtaining internal data for several plants within the same
company would reduce the noise in the incentive intensity measure and might yield
considerable insights into the general incentive aspects of bonuses at the plant manager

level.

60

 

Further, a cross-sectional study based on survey data over a short time horizon
may not be the appropriate research design to capture complex interactions. Plant level
performance data over an extended period of time could provide interesting information
regarding the interactions between several non-ﬁnancial performance measures and
between those measures and ﬁnancial performance.

Finally, this study did not test for cause and effect relationships, only associations.
Contributions to the Literature and Future Research Opportunities

This study ﬁnds a positive association between the bonus weight on quality
measures and external quality performance and accordingly extends the literature on the
performance effects of incentives. Similar associations between bonus weight on quality
measures and internal quality performance and between bonus weight on environmental
measures and environmental performance were not found in this study. The study has
contributed to this literature stream in other ways, though. It investigates the issue at the
plant manager level rather than the CEO level as in most previous studies. This provides
the ability to identify the weights used in the bonus contracts and to speciﬁcally test the
relationships between those weights and performance. The study has also provided
insight into the design of plant managers’ bonus and performance review contracts. It
has raised also raised interesting issues regarding the effect of the composition of
performance review contracts and performance that are worthy of further investigation.
It may also prove interesting to try to determine why these two sets of contracts are
designed to place different weights on the same measures.

The study has also expanded the literature on the performance effects of

incentives by examining the relationship between bonus weights and environmental

6|

 

IJ

performance as well as quality performance. Little is known about the factors that affect
voluntary pollution reduction and this study provides a basis for discussing measurement
issues as well as arguments and theory regarding what factors might affect environmental
performance. Further investigations into how ﬁrms manage environmental risks,
performance goals and costs should enhance our understanding of the relationship
between environmental incentives and performance.

This study has attempted to incorporate the non-linear relations between
incentives and non-ﬁnancial performance into a performance model with limited success.
The results indicate that, however difﬁcult it may be, further research is needed on the
non-linear relations between incentives and both non-ﬁnancial and ﬁnancial performance.
This study also included two contingency factors that are expected to affect performance.
The lack of results regarding the association between performance and the amount of
reliable cost information available or capital investment decision rights suggest that
reﬁnements to these measures will likely be required. It is very difﬁcult to determine
what constitutes good cost information. Perhaps, extensive discussions with managers
across several industries would provide insights into alternative measures of this variable.

It may not be true that greater control over capital investment decision making
will lead to higher performance. For example, investment decision making at the
corporate level may be (and probably should be) perfectly in line with the incentives
offered to plant managers. Further research could address how to determine whether this
is the case within organizations and how to measure the performance effects of this

alignment.

l}\Bll{l(PaneL\)

Principal Components .»\nalysis

l’actor loadings After Oblique Rotation

 

Factor

SIG ons 1

()P()S2

1?. P08 1

li PO S 2

 

Question 1
Question 2
Question 3

Question 1
Question 2
Question 3
Question 4

Question 1
Question 2
Question 3
Question 4

.04534
.76532
.68728

.77153
.77213
.26460
.13978

.10234
.17265
.39752
.39503

TABLE 1 (Panel B)

.84618
.85542
.16921
.07070

.23722
.12727
.51792
.48656

 

Factor

Question 1
Question 2
Question 3
Question 4

Question 1
Question 2
Question 3
Question 4
Question 5

Question 1
Question 2
Question 3
Question 4

Question 1
Question 2
Question 3
Question 4
Question 5

QUALIS
.68373
.68037
.53749
.61604

ENVIS

.64334
.57546
.30896
.6485]
.49081

\*'ariables are deﬁned in 'l'able l 1.

Q U A L C B

.56782
.05442
.06369
.44513

ENVCB

.38607
.33832
.6497:
.66069

45123

TABLE 2

Descriptive Statistics of Independent Variable Factors

 

 

(n=1 [7)

Cronbach Actual Theoretical
Factor Alpha Mean SD. Range Range_
SIG .74 4.172 1.269 1 — 6 1 - 6
QPOSl .82 4.333 1.235 1.5 — 6 l - 6
QPOS2 .48 3.564 1.106 1 — 6 1 - 6
EPOS] .90 3.197 1.417 1—6 1-6
EPOS2 .57 4.385 1.090 1 —— 6 1 - 6
QUALIS .89 3.694 1.058 1.5 — 6 1 - 6
ENVIS .79 3.506 0.954 1 - 6 l - 6
QUALCB .71 3.985 0.914 1.25 - 6 l - 6
ENVCB .64 3.875 1.057 1 — 6 1 — 6

 

Variables are deﬁned in Table 1 1.

64

Ex Ante Expectations of the Weight Placed on Various Financial and Non-Financial

TABLE 3

Performance Measures in the Managers’ Annual Performance Review

Mean Weight

(Standard Deviation)

 

 

 

 

 

 

 

 

 

 

 

 

 

SIC 25 SIC 26 SIC 28 SIC 30
Furniture Paper Chemicals Plastic Total
Plant Financial .327 .404 .166 .404 .270
Performance (.213) (.440) (.140) (.210) (.282)
Firm Financial .124 .058 .086 .054 .084
Performance (.183) (.087) (.165) (. 107) (.151)
Cost .133 .134 .173 .130 .153
Control (.1 13) (.090) (. 106) (.122) (. 106)
Other .027 .020 .019 .017 .021
Financial (.060) (.084) (.034) (.028) (.055)
Performance
Cycle Time .037 .003 .024 .011 .021
Management (.060) (.014) (.044) (.018) (.043)
Safety .064 .156 .193 .121 .151
(.110) (.105) (.115) (.069) (.118)
Quality .067 .107 .090 .104 .090
Management (.080) (.062) (.067) (.095) (.072)
Customer .137 .076 .106 .097 .104
Satisfaction (.144) (.065) (.090) (.072) (.099)
Environmental .041 .064 . 107 .043 .079
Performance (.056) (.063) (.080) (.046) (.075)
Other Non- .043 .033 .037 .013 .036
Financial Perf. (.067) (.053) (.060) (.021) (.058)
N 62 74 147 23 306

 

 

 

 

 

 

 

65

TABLE 4

Ex Ante Expectations of the Weight Placed on Various Financial and Non-Financial

Performance Measures in the Managers’ Annual Bonus

 

 

 

 

 

 

 

 

 

 

 

 

 

Mean Weight
(Standard Deviation)
SIC 25 SIC 26 SIC 28 SIC 30

Furniture Paper Chemicals Plastic Total
Plant Financial .306 .403 .180 .559 .287
Performance (.282) (.331) (.018) (.337) (.285)
Firm Financial .293 .257 .332 .115 .290
Performance (.331) (.288) (.275) (.156) (.289)
Cost .099 .076 .122 .055 .102
Control (.137) (.123) (.283) (.104) (.217)
Other Financial .033 .019 .022 .090 .029
Performance (.101) (.084) (.058) (.150) (.085)
Cycle Time .000 .003 .013 .004 .007
Management (.000) (.030) (.034) (.010) (.028)
Safety .035 .010 .139 .040 .101

(.107) (.118) (.134) (.062) (.128)
Quality .037 .048 .057 .166 .059
Management (.085) (.073) (.070) (.334) (.119)
Customer .064 .033 .059 .013 .050
Satisfaction (.123) (.050) (. 104) (.021) (.095)
Environmental .003 .021 .071 .007 .041
Performance (.012) (.041) (.082) (.016) (.068)
Other Non- .032 .041 .027 .010 .031
Financial Perf. (.130) (.111) (.084) (.023) (.099)
N 62 74 149 23 308

 

 

 

 

 

 

 

TABLE 5

Descriptive Statistics of Regression Variables

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

j Uni—f Mean Standard Plant-Year
I l\leasure _ Deviation N
ENVWV'I‘SIG "/0 ueight’“bonus .172 .31 l 309
signiﬁcance
I EXTQWTSIG weight*bonus .217 .486 309
sigtiﬁcance
INTQWTSIG % weight *bonus 230 .470 309
F signiﬁcance
QPOSI Range 1 - 6 4.337 1.242 312
QPOS2 Range 1 - 6 3 551 1.091 312
EPOS] Range 1 —6 3.187 1.424 313
EPOS2 Range 1 — 6 4.359 1.076 312
QUALIS Range 1 —- 6 3 735 1.046 312
ENVIS Range 1 — 6 3.502 .937 311
QUALCB Range 1 - 6 4.028 .893 307
ENVCB Range 1 — 6 3.900 1.003 309
ENVLEVEL Lbs. of 3292.628 8118.296 395
emissions per
employee
EXTQLEVEI. Unknown 89.517 14.185 351
INTQLEVEL Unknown 4.586 3.448 240
FPLEVEI. Unknown 1 1.080 8.593 309

 

 

 

 

 

 

 

Variables are deli ncd in ~I’able 1 l.

TABLE 6

OLS regression of internal quality performance levels on bonus weights on quality
performance measures, returns to quality investments, and quality decision rights

Test olea, H3a andH5a

Coefﬁcient
(t-statistic)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

INTQPERF.-.(+) .79 0.80
(8.84)*** (9.30)***
INTQWTSIG (+) -.21 .08
(-0.07) (0.45)
EXTQWTSIG (+) 4.81 .17
(2.24)** (0.56)
QPOS] (+) .04 .05
(0.32) (0.35)
QPOS] *INTQWTSIG (+) .03
(0.08)
QPOS] *EXTQWTSIG (+) -. 18
(-053)
QPosz (+) .43 .30
(0.99) (1.14)
QPOS2*INTQWTSIG (+) -.26
(-0.48)
QPOSZ*EXTQWTSIG (+) -.71
(-140)
QUALCB (+) .06 .08
(0.24) (0.55)
QUALCB*INTQWTSIG (+) .28
(.25)
QUALCB*EXTQWTSIG (+) -.26
(-099)
AGE (+) -.01 -.00
(-041) (-0.22)
CONSTANT -2.38 -202
(-1.60)* (-1.54)*
R2 .67 .67
N 159 159

 

 

 

68

***, **, * and # = signiﬁcant at less that 0.01, 0.05, 0.10 and 0.15 levels (one-tailed),
respectively.

The second column presents OLS regression results omitting insigniﬁcant interactions.
Variables are deﬁned in Table 11.

 

 

11

TABLE 7

OLS regression of external quality performance levels on bonus weights on quality
performance measures, returns to quality investments, and quality decision rights
Test olea, H3a andH5a

Coefﬁcient
(t-statistic)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

EXTQPERF1-1(+) .85 .84
(l8.00)*** (18.12)***
INTQWTSIG (+) -.55 e 4.52
(-0.08) (2.15)**
EXTQWTSIG (+) 1.49 2.12
(0.43) (2.49)*
QPOSI (+) -.61 -.45
(-1.97)** (-1.97)**
QPOSI *INTQWTSIG (+) .49
(0.85)
QPOS1*EXTQWTSIG (+) .29
(0.49)
QPOSZ (+) 1.08 1.05
(3.14)*** (3.16)***
QPosz*1NTQWTSIG (+) 3 -1 .38 -127
(-2.82)** (—2.54)***
QPOS2*EXTQWTSIG (+) -.77 -.94
(1.75)* (-2.59)***
QUALCB (+) -.04 -.04
(-0.09) (-0.08)
QUALCB*INTQWTSIG (+) .71
(0.67)
QUALCB*EXTQWTSIG (+) -.27
(-0.62)
AGE () —. 14 -.13
(-3.29)*** (-3.36)***
CONSTANT 18.04 17.29
(3.92)*** (3.85)***
1? .89 .89
N 234 234

 

 

 

 

***, **, * and # = signiﬁcant at less that 0.01, 0.05, 0.10 and 0.15 levels (one-tailed),
respectively.

The second column presents OLS regression results omitting insigniﬁcant interactions.
Variables are deﬁned in Table 11.

69

   

OLS regression of environmental performance levels on bonus weight on environmental
performance measures, returns to environmental investments, and environmental decision

TABLE 8

rights

Test oleb, H3b and H5b

Coefﬁcient
(t-statistic)

 

 

 

 

 

 

 

 

 

 

 

 

ENVPERF.-. (+) .95 .95
(17.66)*** (19.80)***
ENVWTSIG (+) -6535.87 -1770.76
(-1.11) (-127)
EPOSl (+) -78.61 124.52
(-0.80) (0.99)
EPOSl *ENVWTSIG (+) 1080.92
(1.22)
EPOS2 (+) 132.54 -81.59
(0.83) (-.48)
EPOS2*ENVWTSIG (+) -1287.53
(-104)
ENVCB (+) -230.96 -9297
(1.90)* (0.69)
ENVCB*ENVWTSIG (+) 1386.14
(1.04)
AGE (+) 4.20 12.21
(0.29) (0.68)
CONSTANT 638.48 -43.62
(0.81) (-051)
R2 .84 .83
N 270 270

 

 

 

 

 

***, **, * and # = signiﬁcant at less that 0.01 , 0.05, 0.10 and 0.15 levels (one-tailed),
respectively.

The second column presents OLS regression results omitting insigniﬁcant interactions.
Variables are deﬁned in Table 11.

70

 

 

TABLE 9

OLS regression of ﬁnancial performance on the use of bonus weights on non-ﬁnancial

performance and non-ﬁnancial performance
Tests of H2a and H2b

Coefﬁcient
(t-statistic)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

FPERF.-. (+) .88 .85
(16.42)*** (22.59)***
ENVPERF (+) -0003 .00
(-2.11)** (0.09)
ENVWTD (+) .29 -1.12
(0.31) (-1.44)#
ENVPERF*ENVWTD (+) .0003 -.00
(2.10)** (-030)
INTQPERF (+) .05
(0.60)
INTQWTD (+) -.54
(-057)
INTQPERF*INTQWTD (+) .18
(0.77)
EXTQPERF (+) .03 .01
(1.04) (0.78)
EXTQWTD (+) 17.9 -9.06
(1 .88)** (-1.26)
EXTQPERF*EXTQWTD (+) -. 19 .11
(-1.85)** (1.39)#
CONSTANT -.31 .93
(-024) (1.87)
R2 .83 .79
N 1 16 190

 

 

***, **, * and # = signiﬁcant at less that 0.01, 0.05, 0.10 and 0.15 levels (one-tailed),

respectively.

The second column presents OLS regression results omitting insigniﬁcant internal quality

measures.

Variables are deﬁned in Table 1 1.

71

TABLE 10

OLS regression of bonus weight on quality and environmental performance measures
levels and availability of reliable quality and environmental cost information on ﬁnancial

performance
Test of H4

Coefﬁcient
(t-statistic)

 

 

 

 

 

 

 

 

 

 

 

 

FPERF“ (+) .90 .90
(27.35)*** (27.75)***
ENVWTSIG (+) 11.04 10.72
(1.51)# (1.58)#
INTQWTSIG (+) -222 .21
(-105) (0.56)
EXTQWTSIG (+) -217 .72
(-054) (0.95)
ENVIS (+) .18 .19
(0.66) (0.70)
QUALIS (+) .28 .33
(0.70) (0.90)
ENVWTSIG*ENVIS (+) -2.81 -2.78
(-l.46)# (-1.62)*
INTQWTSIG*QUALIS (+) .60
(1.19)
EXTQWTSIG*QUALIS (+) -.46
(-0.67)
CONSTANT -.67 -.88
(-042) (-059)
R2 .81 .81
N 216 216

 

 

 

 

 

***, **, * and # = signiﬁcant at less that 0.01, 0.05, 0.10 and 0.15 levels (one-tailed),
respectively.

The second column presents OLS regression results omitting insigniﬁcant interactions.
Variables are deﬁned in Table 11.

TABLE 1 1
Variable Deﬁnitions

SIG -— the self-reported signiﬁcance of the bonus to each manager.
ENVWTSIG — the weight attached to environmental performance in the bonus contract *
bonus signiﬁcance.
EXTQWTSIG — the weight attached to external quality performance in the bonus
contract * bonus signiﬁcance.
INTQWTSIG — the weight attached to internal quality performance in the bonus contract
* bonus signiﬁcance.
QPOSI — Inexpensive and cost-effective opportunities for product quality improvements
exist. A stage measure.
QPOSZ — Experiencing or expecting to experience decreasing returns to quality
investment. A stage measure.
EPOSI — Inexpensive and cost-effective opportunities for environmental improvements
exist. A stage measure.
EPOSZ — Experiencing or expecting to experience decreasing returns to environmental
Investment. A stage measure.
QUALIS — The availability of reliable and useful quality cost information.
ENVIS — The availability of reliable and useful environmental cost information.
QUALCB - The ability to control quality capital budgeting decisions.
ENVCB —— The ability to control environmental capital budgeting decisions.
ENVPERF — The annual levels of TRI emissions scaled by number of full-time
equivalent employees.
EXTQPERF — The self-reported performance level on each plant’s most important
measure of external quality performance.
INTQPERF - The self-reported performance level on each plant’s most important
measure of internal quality performance.
F PERF— The self-reported performance level on each plant’s most important
measure of ﬁnancial performance.
ENVWTD = Dummy variable where:
l = Positive weight placed on environmental performance in the bonus
contract.
0 = Zero weight placed on environmental performance in the bonus
contract.
INTQWTD = Dummy variable where:
l = Positive weight placed on intemal quality performance in the bonus
contract.
0 = Zero weight placed on internal quality performance in the bonus
contract.
EXTQWTD = Dummy variable where:
l = Positive weight placed on external quality performance in the bonus
contract.
0 = Zero weight placed on external quality performance in the bonus
contract.
AGE — The self-reported age of the plant.

73

Table 12
Table of Results

Hypotheses
Results

 

Hla: There is a positive association between the weight placed on quality
performance in the annual bonus contract and quality performance.

Hlb: There is a positive association between the weight placed on environmental
performance in the annual bonus contract and environmental performance.

 

Results: Hla-is partially supported. There is a positive association between external
quality performance levels and the weight placed on internal quality performance
measures (p<.05) and external quality performance measures (p<.10). Hlb is not
supported.

 

H2a: There is a positive association between ﬁnancial performance and quality
performance when weight is placed on quality performance measures in annual
bonus contracts.

H2b: There is a positive association between ﬁnancial performance and
environmental performance when weight is placed on environmental
performance measures in annual bonus contracts.

 

Results: H 1a is partially supported. There is a positive association between external
quality performance and ﬁnancial performance when weight is placed on external
uality performance in the bonus contract. H2b is not supported.

 

H3a: The weight placed on quality performance measures in annual bonus
contracts will have a more positive effect on quality performance when marginal
returns to investment are experienced.

H3b: The weight placed on environmental performance measures in annual
bonus contracts will have a more positive effect on environmental performance
when marginal returns to investment are experienced.

 

Results: H3a is not supported. There is a negative association between external
quality performance and the interactions between internal and external quality
measures in the bonus contract and perceived decreasing returns to investment
(p<.01) for both. H3b is not supported

 

 

H4a: The weight on quality performance measures in annual bonus contracts
will have a more positive effect on quality performance when reliable quality
cost information is available.

H4b: The weight on environmental performance measures in annual bonus
contracts will have a more positive effect on environmental performance when
reliable environmental cost information is available.

 

74

 

Table 12
(Continued)

Table of Results

Hypotheses
Results

 

Results: H4a and H4b are not supported. Tests of H4b show that there is a negative
association between the interaction of t he weight placed on environmental
performance in the bonus contract and the availability of reliable environmental cost
information and ﬁnancial performance.

 

H5a: The weight placed on quality performance measures in annual bonus
contracts will have a more positive effect on quality performance when capital
investment decision rights can be exercised.

H5b: The weight placed on environmental performance measures in annual
bonus contracts will have a more positive effect on environmental performance
when capital investment decision rights can be exercised.

 

 

Results: H5a and H5b are not supported.

 

 

75

APPENDIX A

Survey Cover Letter Example

Date

Name/Production Manager
Title Associated with Name
Company

Address

Dear Name or Sir/Madam:

APICS-The Educational Society for Resource Management and Michigan State
University are supporting a survey regarding the effects of compensation design and
information and decision making systems on operational performance. Your
participation in the project will ensure that your experiences as a manager are included in
a study that should provide important insight into the areas of compensation design and
non-ﬁnancial performance measurement. You will have access to a summary benchmark
report when all responses have been received and analyzed.

The survey has been enclosed with this letter. Managers that have tested the
questionnaire reported that it required about twenty minutes of their time to complete the
survey. The also reported that the survey was easy to follow and that the questions were
not too difﬁcult to answer.

Individual responses to this survey will be kept strictly conﬁdential. The responses will
be aggregated and analyzed in order to create a report of current practices. This
benchmarking report will allow you to compare your facility by industry and against best
practice. If you would like a copy of the report to be sent to you, please attach your
business card to the last page of the survey. ’

If you have any questions about the survey, please contact me at (517) 353-8754 or at
connorse@pilot.msu.edu. Questions regarding the commitment of Michigan State
University toward protection of the rights and privacy of research participants may be
directed to David Wright at the Ofﬁce of Research and Graduate Studies, (517) 355-
2180. Thank you in advance for your participation in this project.

Sincerely,
Elizabeth Connors
Research Principal

Enclosure

76

 

 

APPENDIX B
Sample Research Project Support Letter

APICS Educational & Research Foundation, Inc.
5301 Shawnee Road
Alexandria, VA 22312-2317

Date
Dear Valued Colleague:

On behalf of the APICS Educational and Research Foundation, I am asking for your
support and assistance in helping to complete an important business research project.
Michigan State University is conducting a survey of the chemical, furniture and paper
industries in order to evaluate the effects of compensation design on ﬁnancial and non-
ﬁnancial performance at the plant level.

Effective compensation design is a very important issue facing management at all
organizational levels. The issue is complex and there are many theories proposed by
consultants and academics that have yet to be tested. This research project proposes to
determine the types of compensation design used at the plant management level and to
establish a relationship between the designs and plant-level ﬁnancial and non-ﬁnancial
performance. We believe that this is very important research because it will provide a
test of current executive compensation theory and will allow ﬁrms to see where they
stand on this issue relative to their compensation.

We strongly encourage you to participate in this important research. We have reviewed
the survey and the overall methodology, and we ﬁnd them to be of sound business
research practice. All individual responses will be kept conﬁdential and all results will
be reported in the aggregate. As noted at the end of the survey, you may request a copy
of the results, which could be useful for benchmarking the effects of your ﬁnn’s
compensation plans on ﬁnancial and non-ﬁnancial performance relative to your
competitors.

If you have any questions about APICS involvement in this research, or how APICS can
serve you, please call me at (800) 444-2742. For questions about the survey itself, please
contact Elizabeth Connors at (517) 353-8754 or connorse@pilot.msu.edu.

Thank you very much for your time and attention. We hope that you are looking forward
to the results of the survey as much as we are at APICS.

Sincerely,

Mike Lithgoe
Director

77

APPENDIX C
Survey Instrument

The Effects of Managers’ Compensation Design
Michigan State University
The Eli Broad Graduate School of Management
East Lansing, Michigan

This survey is part of a study that examines how the design of managerial compensation
relates to ﬁrm characteristics. The differences between the structures of compensation
plans of individual plant managers and the corresponding ﬁnancial and non-ﬁnancial
outcomes are of primary importance in the study. Other plants in your industry
(chemical, furniture, paper, plastic, and rubber) are also participating and the importance
of your response cannot be overemphasized. Aggregate results of this study will be
provided in a benchmark report to APICS-The Educational Society for Resource
Management and to you, if you so wish.

Your careful consideration of all of the questions in this survey will provide a
comprehensive understanding of the practices used and outcomes experienced by your
plant. Please answer the questions as accurately and completely as possible.

The survey covers the period 1996 — 1999. Please complete the survey for the yearsthat
you have held your position as manager of this plant or have the appropriate information
about the previous manager in your position.

Two non-ﬁnancial performance measures are stressed in this survey. The ﬁrst is product

quality and the second is voluntary pollution reduction. For the purposes of this study,
voluntary pollution reduction efforts are those that are not motivated by regulations that
are currently in effect. Thus, efforts to comply in advance of future regulations are
considered to be voluntary.

ALL INDIVIDUAL RESPONSES WILL BE KEPT STRICTLY CONFIDENTIAL
AND ONLY SUMMARY RESULTS WILL BE REPORTED.

If you have any questions or concerns, please feel free to contact:

Elizabeth Connors
Michigan State University
N250 North Business Complex
East Lansing, MI 48824
Tel.: (517)353-8754
Fax.: (517)432-1101

E-Mail: connorse@pilot.msu.edu
Thank you for your time and attention.

78

Section A: Compensation Structure

Before your review, what did you think would be the weight placed on the following

measures in your annual performance review?

 

1%

Plant-level ﬁnancial performance
Finn-level ﬁnancial performance
Cost control

Other ﬁnancial performance
Cycle time management

Safety

Quality management

Customer satisfaction
Environmental performance
Other non-ﬁnancial performance

Total 100 319

79

l 98 1997

100 % 100%

Before your review, what did you think would be the weight placed on the following

measures in determining your annual bonus?

 

ﬂ

Plant-level ﬁnancial performance
Firm-level ﬁnancial performance
Cost control

Other ﬁnancial performance
Cycle time management

Safety

Quality management

Customer satisfaction
Environmental performance
Other non-ﬁnancial performance

Total 100 %

80

199

100 %

1997

ICE/9

Scale points are completely disagree (1), disagree (2), tend to disagree (3), tend to agree
(4), agree (5), and
completely agree (6).

completely completely
disagree agree
My bonus is a signiﬁcant portion of my
compensation package. 1 2 3 4 5 6
Achieving my bonus is important to me 1 2 3 4 5 6
The structure of my bonus affects my decisions 1 2 3 4 5 6

Please circle your estimate of how much of the volatility of your bonus is due to plant-
level factors that are out of your control:

0-20% 21-40% 41-60% 61 -80% 81-100%

Section B: Position in Processes

Scale points are completely disagree (1), disagree (2), tend to disagree (3), tend to agree
(4), agree (5), and completely agree (6).

 

At our plant:
completely completely
disagree agree
Inexpensive opportunities for product
quality improvement exist. 1 2 3 4 5 6
Cost-effective opportunities for product
quality improvement exist. 1 2 3 4 5 6

In the future it will be more costly to achieve a given level of
product quality improvement than it has in the past. 1 2 3 4 5 6

Our quality efforts yield diminishing returns. 1 2 3 4 5 6

Inexpensive opportunities for pollution reduction exist.
1 2 3 4 5 6

81

 

 

Section B: Position in Processes

Scale points are completely disagree (I), disagree (2), tend to disagree (3), tend to agree
(4), agree (5), and completely agree (6).

 

At our plant:
completely completely
disagree agree
Cost-effective opportunities for pollution
reduction exist. 1 2 3 4 5 6

In the future it will be more costly to achieve a given level of
pollution reduction than it has in the past. 1 2 3 4 5 6

Our pollution reduction efforts yield diminishing returns.
1 2 3 4 5 6

Please indicate the approximate levels of your most important measure of external
quality performance (e.g. delivery times, customer satisfaction, defective parts per
million at ﬁnal inspection):

For example: A score of 95% on customer satisfaction measures would be a level of 95.
A score of 95% on delivery time would also be a level of 95.

Note: We do not need to know the type of performance measure that you use. We are
only asking for the levels of your most important measure.

 

 

 

 

Improvement over Did the

Measurement method

previous year? change?
Levels for 1999 Yes No Yes No
Levels for 1998 _ Yes No Yes No
Levels for 1997 Yes No Yes No

 

—F——-— .—

Levels for 1996

 

82

 

Please indicate the approximate levels of your most important measure of internal
quality performance (eg. scrap rates, prime rates, rework rates) :

For example: A scrap rate of 4% would be a level of 4. A rework rate of 4% would also
be a level of 4.

Note: We do not need to know the type of performance measure that you use. We are
only asking for the levels of your most important measure,

 

 

 

 

 

Improvement over Dld the

measurement method

previous year? change?
Levels for 1999 Yes No Yes No
Levels for 1998 Yes No Yes No
Levels for 1997 Yes No Yes No

 

_— ——

Levels for 1996

 

What is the status of the following programs within your plant?
Please place an “X” in one cell per program.

Not Planning
Being Assessing to Success-
Considered Suitability Implement Currently fully
Implementing(ed)

 

ISO/QS 9000

 

Total Quality
Management

 

Company-speciﬁc
Environmental
Management
System

 

ISO 14000

 

 

 

 

 

 

 

 

83

" J

Section C: Informatiop Systems

Scale points are completely disagree (1), disagree (2), tend to disagree (3), tend to agree
(4), agree (5), and completely agree (6).
At our plant:
completely completely
disagree agree

It is easy to relate quality efforts directly to production
cost reductions. ‘ 1 2 3 4 5 6

I can easily determine the ﬁnancial cost of quality
related process and product changes. 1 2 3 4 5 6

I know substantially all of the ﬁnancial effects of
quality-related decisions. 1 2 3 4 5 6

Our cost-of-quality information is reliable and useful for
decision making. 1 2 3 4 5 6

It is easy to relate pollution reduction efforts directly to
production cost reductions. 1 2 3 4 5 6

It is easy to relate pollution reduction efforts directly to
reduced future compliance costs. 1 2

b.)
A
U1
ON

I can easily determine the ﬁnancial cost of pollution reduction
related process and product changes. 1 2 3 4 5 6

I can easily determine the ﬁnancial beneﬁts of pollution reduction
related process and product changes. 1 2 3 4 5 6

Our environmental cost information is reliable and useful for
decision making. ' 1 2 3 4 5 6

84

 

Section D: Capital Budgeting Practices

Scale points are completely disagree (1), disagree (2), tend to disagree (3), tend to agree
(4), agree (5), and completely agree (6).

 

At our plant:
completely completely
disagree agree
Product quality investments compete on a
plant-wide basis for funds based on ﬁnancial
factors only. 1 2 3 4 5 6

Both ﬁnancial and non-ﬁnancial factors are given weight
in product quality investment decisions. 1 2 3 4 5 6

Non-ﬁnancial factors are heavily weighted in product
quality investment decisions. 1 2 3 4 5 6

I have a great deal of control over product
quality investment choices. 1 2 3 4 5 6

We are more likely to invest in quality improvement
projects during times of good
ﬁnancial performance. 1 2 3 4 5 6

Quality performance depends on
ﬁnancial performance. 1 2 3 4 5 6

Financial performance depends on
quality performance. 1 2 3 4 5 6

Voluntary environmental investments compete on a
company-wide basis for funds based on
ﬁnancial factors only. 1 2 3 4 5 6

 

Voluntary environmental investments compete on a
plant—wide basis for funds based on

ﬁnancial factors only. 1 2 3 4 5 6
Both ﬁnancial and non-ﬁnancial factors are given weight

in voluntary environmental investment decisions. 1 2 3 4 5 6
Non-ﬁnancial factors are heavily weighted in voluntary

environmental investment decisions. 1 2 3 4 5 6

85

Scale points are completely disagree (1), disagree (2), tend to disagree (3), tend to agree
(4), agree (5), and completely agree (6).

At our plant:

completely completely
disagree agree
I have a great deal of control over voluntary
environmental investment choices. 1 2 3 4 5 6

We are more likely to invest in voluntary pollution reduction
projects during times of good
ﬁnancial performance. 1 2 3 4 5 6

Voluntary environmental performance depends
on ﬁnancial performance. 1 2 3 4 5 6

Financial performance depends on voluntary
environmental performance. 1 2 3 4 5 6
Section E: Plant Characteristics and Management Information
What is the average age of your plant’s production equipment? years.
Is your plant considered to be (please check one):
A proﬁt center

A cost center
Other (please describe)

 

86

“"1

Please indicate the approximate levels of your r_nost important measure of ﬁnancial
performance (eg. Return on Assets, Return on Assets Employed, Total Costs):

For example: A 7% Return on Assets Employed would be a level of 7. A cost reduction
of 7% would also be a level of 7.

Note: We do not need to know the type of performance measure that you use. We are
only asking for the levels of your most important measure.

 

Improvement over Did the
previous year? measurement
method chapge?

 

 

 

Levels for 1999 Yes No Yes No
Levels for 1998 Yes No Yes No
Levels for 1997 Yes No Yes No

 

Levels for 1996

 

Please indicate the approximate number of full-time equivalent employees:

during 1999
during 1998
during 1997
during 1996

 

 

 

 

 

87

I3 I B I... I ()0 RA PH Y

.~\memi_va. T. 1985. Advanced Eeononzetrics. Cambridge. MA: Harvard University Press.

Arora. S and T. N Cason. 1995. An Experiment in Voluntary Environmental Regulation:
Participation in EPA's 33/50 Program. Journal ofEnvironmental Economics and
11 Ianagemenl. 28: 271-286.

Atkinson. H., Hamburg. I. and C. Ittner. 1994. Lin/ring Quality to Pro/its. Montvale, NJ:
Institute of Management Accountants.

Banker. R. D. and S. M. Datar. 1989. Sensitivity. Precision, and Linear Aggregation of
Signals for Performance Evaluation. Journal of Accounting Research. 27(1):21-
9.

b)

Banker. R. D., Potter, G. and D. Srinivasan. 2002. An Empirical Investigation of an
Incentive Plan Based on Nonﬁnancial Performance Measures. The Accounting
Review. 75(1):65-92.

Boyd. .1. l998a. Searching for the Proﬁt in Pollution Prevention: Case Studies in the
Corporate Evaluation of Environmental Opportunities. Working Paper.

1998b. The Beneﬁts of Improved Environmental Accounting: An Economic
Framework to Identify Priorities. Resources for the Future Discussion Paper 98-
49. September.

Bushman, R. M., Indjejikian. R. .1. and A. Smith. 1995. Aggregate Performance Measures
in Business Unit Manager Compensation: The Role of Intraﬁrm
Interdependencies. Journal of.~1ecounting Research. 33 (Supplement)1101-128.

Cairncross. F. 1990. Survey: The Environment. The Ecmzomist. (Sept. 8): l- 26.

Campbell. K. and N. S. Soderstrom. 1998. Executive Compensation and Environmental
Fxpr‘rsurc. \V'orking Paper.

Core. .1. 1;}. Quay. V1". and R. E \t’errecchia. 2002. Price vs. Non-Price Performance
Measures in Optimal CEO Compensation Contracts. Working Paper. University
of Pennsylvania.

Cram. 1). P. 1997. Implicit Price Estimation: l\'leasuring the Impact ofNonﬁnancial

factors in [Environmental l\1an:’1gement Decision-making. Dissertation. Stanford
University.

88

 

Crosby, P. B. 1996. Quality is Still Free. New York: McGraw-Hill.

Datar, S., Kulp, S. C. and R. A Lambert. 2001. Balancing Performance Measures.
Journal of Accounting Research. 39 (1): 75-92.

Delavigne, K. T. and J. D Robertson. 1994. Deming’s Profound Changes. Englewood
Cliffs, N. 1.: Prentice-Hall, Inc.

Demers, E., Shackell, M. and S. K. Widener. 2002. Complementarities in Organizational
Design: Empirical Evidence from the New Economy. Working Paper.

Dikolli, S. A. and S. C. Kulp. 2002. Interrelated Performance Measures, Interactive
Effort, and Optimal Incentives. Working Paper.

Dillman, D. A. 1991. The Design and Administration of Mail Surveys. Annual Review
of Sociology. 17: 225-249.

Eccles, R. G. 1991. The Performance Measurement Manifesto. Harvard Business
Review. Jan.-Feb.: 131-137.

Epstein, M. J. 1996. Measuring Corporate Environmental Performance. Montvale, NJ:
Institute of Management Accountants. '

Feltham, G. A. and J. Xie. 1994. Performance Measure Congruity and Diversity in
Multi-Task Principal/Agent Relations. The Accounting Review. 69 (3): 429-453.

Fine, C. H. 1986. Quality Improvement and Learning in Productive Systems.
Management Science 32 (10): 1301-1315.

Fischer, K. and .1. Schot. 1994. The Challenge of Going Green. Harvard Business Review.
July/August: 37-50.

Foster, G. and M. Gupta. 1998. The Customer Proﬁtability Implications of Customer
Satisfaction. Working Paper.

Freedman, M. and B. Jaggi. 1992. An Investigation of the Long-Run Relationship
Between Pollution Performance and Economic Performance: The Case of Pulp
and Paper Firms. Critical Perspectives on Accounting 3: 315-336.

Gibbs, M., Merchant, K. A., Van der Stede, W. A. and M. E. Vargus. 2002. Causes and

Effects of Subjectivity in Incentives. Working Paper. University of Southern
California.

Hunt, C. B. and E. R. Auster. 1990. Proactive Environmental Management: Avoiding the
Toxic Trap. Sloan Management Review. 31 (2): 7-18.

89

 

Institute of Management Accountants and Society of Management Accountants of

Canada. 1996. Tools and Techniques of Environmental A ccounting for Business
Decisions. Statement on Management Accounting, Statement 4Z. Montvale, NJ:
Institute of Management Accountants.

Ittner, C. D. 1994. Exploratory Evidence on the Behavior of Quality Costs. Operations

Research 44(1): 114-130.

and D. F. Larcker. 1995. Total Quality Management and the Choice of

Information and Reward Systems. Journal of A ccounting Research.
33(Supplement): 1-34.

, and D. F. Larcker. 1996. Measuring the Impact of Quality Initiatives on Firm

Financial Performance. Advances in the Management of Organizational Quality.
1: 1-37.

, and D. F. Larcker.1998. Are Non-Financial Measures Leading Indicators of

Financial Performance? An Analysis of Customer Satisfaction. Journal of
Accounting Research. 26 (Supplement): 1-35.

, and D. F. Larcker. 1999. Innovations in Performance Measurement: Trends

and Research Implications. Journal of Management Accounting Research. 1 1:
206-238.

, and D. F. Larcker. 2001. Assessing Empirical Research in Managerial

Accounting: A Value-Based Management Perspective. Journal of Accounting
and Economics. 32: 349-410.

S

, and D. F. Larcker. 2002 Determinants of Performance Measure Choices in

Worker Incentive Plans. Journal of Labor Economics. 20 (2):858-S90.

, Larcker, D. F. and M. W. Meyer. 1997a. The Anatomy of a Balanced Scorecard

Bonus Plan. Working Paper. University of Pennsylvania.

, Larcker, D. F. and M. V. Rajan. 1997b. The Choice of Performance Measures in

Annual Bonus Contracts. The Accounting Review. April:23 1-255.

, Nagar, V. and M. V. Rajan. 1998. An Empirical Analysis of Reported

Quality Costs. Working Paper.

Jaffe, A. B. and K. Palmer. 1997. Environmental Regulation and Innovation: A Panel

Data Study. The Review of Economics and Statistics. 79 (4): 610-619.

Jaffe, A. B, S. R. Petersen, P. R. Portney and R. N. Stavins. 1995. Environmental

Regulation and the Competitiveness of US. Manufacturing: What the Evidence
Tells Us. Journal of Economic Literature 33 (March): 132-163.

90

 

 

 

Jaggi, B. and M. Freedman. 1992. An Examination of the Impact of Pollution
Performance on Economic and Market Performance: Pulp and Paper Firms.
Journal of Business Finance & Accounting 19(5): 697-713.

Jensen, M. C. and W. H. Meckling. 1992. Speciﬁc and General Knowledge, and
Organizational Structure. In Contract Economics, edited by L. Werin and H.
Wijlander. Oxford: Basil Blackwell.

Johnson, M. F., Magnan, M. and C. Stinson. 1996. Nonﬁnancial Measures of
Environmental Performance as Proxies Environmental Risks and Uncertainties.

Working Paper.

Joshi, S., Krishnan, R. and L. Lave. 2001. Estimating the Hidden Costs of Environmental
Regulation. The Accounting Review. 76(2):]71-199.

Juran, J. M. 1989. Juran on Leadership for Quality. New York: MacMillan, Inc.

Kaplan, R. S. and D. P. Norton. 1996. The Balanced Scorecard. Boston, MA: Harvard
Business School Press.

Klassen, R. D. and C. P. McLaughlin. 1996. The Impact-of Environmental Management
on Firm Performance. Management Science. 42 (8): 1199-1214.

Konar, S. and M. A. Cohen. 1997. Does the Market Value Environmental Performance?
Working Paper. Vanderbilt University.

Lambert, R. A. 2001. Contracting Theory and Accounting. Journal of A ccounting and
Economics. 32 (1-3): 3-87.

Lanen, W. N. 1999. Waste Minimization at 3M Corporation: A Field Study in Non-
Financial Performance Measurement. Working paper.

Makower, J. 1994. Beyond the Bottom Line. New York: Simon & Schuster.

Marcellus, R. L. and M. Dada. 1991. Interactive Process Quality Improvement.
Management Science 37 (11): 1365-1376.

McLaughlin, S. and H. Elwood. 1996. Environmental Accounting and EMSS. Pollution
Prevention Review (Spring): 13-20.

Milgrom, P. and J. Roberts. 1992. Economics, Organization and Management.
Englewood Cliffs, NJ: Prentice-Hall.

Myerson, A. R. 1998. U. S. Splurging on Energy After Falling Off It’s Diet. The New
York Times. October 22, Section A, p. 1.

91

 

Nandakumar, P., Datar, S. M., and R. Akella. 1993. Models for Measuring and
Accounting for Cost of Conformance Quality. Management Science. 39 (1): 1-16.

Palmer, K., Oates, W. and P. Portney. 1995. Tightening Environmental Standards: The

Beneﬁt-Cost or the No-Cost Paradigm. Journal of Economic Perspectives 9 (4):
1 l9-l32.

Pashigian, B. P. 1984. The Effect of Environmental Regulation on Optimal Plant Size
and Factor Shares. Journal of Law & Economics 27 (April): 1-28.

Perera, S. Harrison, G. and M. Poole. 1997. Customer-Focused Manufacturing Strategy
and the Use of Operations-Based Non-Financial Performance Measures: A
Research Note. Accounting, Organizations and Society. 22 (6): 557-572.

Ponemon, L. A., Carr, L. P. and J. P. Wendell. 1994. The Validity of Quality
Costs: An Empirical Study of Mills in the Paper and Pulp Industry. Advances in
Management Accounting. 3: 213-236.

Porter, M. E. 1991. America’s Green Strategy. Scientific American (April): 168.

and C. van der Linde. 1995a. Green and Competitive: Ending the Stalemate.
Harvard Business Review (Sept./Oct.): 120-134.

. and C. van der Linde. 1995b. Toward a New Conception of the Environment-
Competitiveness Relationship. Journal of Economic Perspectives 9 (Fall): 97-117.

Prendergast, D. and R. Topel. 1993. Discretion and Bias in Performance Evaluation.
European Economic Review. 37: 355-365.

Russo, M. V. and P. A. Fouts. 1997. A Resource-Based Perspective on Corporate
Environmental Performance and Proﬁtability. Academy of Management Journal
40, No. 3: 534-559.

Sim, K. L. and L. N. Killough. 1998. The Performance Effects of Complementarities
Between Manufacturing Practices and Management Accounting Systems. Journal
of Management Accounting Research. 10: 325-346.

Sterman, J. D., Repenning, N. P. and F. Kofman. 1997. Unanticipated Side Effects of
Successful Quality Programs: Exploring a Paradox of Organizational
Improvement. Management Science. 43 (4): 503-521.

Swilka, D. 2002. On the Use of Nonﬁnancial Performance Measures in Management
Compensation. Journal of Economics and Management Strategy. 11 (3):485-510.

The Conference Board. 1993. Does Quality Work? A Review of Relevant Studies. New
York, NY: The Conference Board, Inc.

92

U. S. Environmental Protection Agency. 2000. 1998 Toxics Release Inventory Public
Data ReleaseReport. Washington, DC: U. S. EPA.
http://www.epa.gov/tri/tri98/pdr

Walley, N. and B. Whitehead. 1994. It’s Not Easy Being Green. Harvard Business
Review (May/June): 46-52.

White, A. L., Savage, D., Brody J. Cavander, D., and L. Lach. 1995. Environmental Cost
Accounting for Capital Budgeting: A Benchmark Survey of Management

Accountants. Boston, MA: Tellus Institute.

Wooldridge, J. M. 2000. Introductory Econometrics. South-Western College Publishing.

Wruck, K. H. and M. C. Jensen. "Science, Speciﬁc Knowledge and Total Quality
Management." Journal of Accounting and Economics (Nov. 1 994): 24 7-8 7.

 

s1\l II ‘ 1‘“

 

llllllllllllllllllllllll
3 1293 025

  

   

HIS

04

 

 

 

"1111111