The effects of monetary incentives on effort and task performance

Description
The purpose of this paper is to review theories and evidence regarding the effects of (performance-contingent)
monetary incentives on individual effort and task performance. We provide a framework for understandingthese
effects in numerous contexts of interest to accountingresearchers and focus particularly on how salient features of
accounting settings may affect the incentives-effort and effort-performance relations. Our compilation and integration
of theories and evidence across a wide variety of disciplines reveals significant implications for accounting research and
practice. Based on the framework, theories, and prior evidence, we develop and discuss numerous directions for future
research in accountingthat could provide important insights into the efficacy of monetary reward systems

The e?ects of monetary incentives on e?ort and
task performance: theories, evidence, and a framework
for research
Sarah E. Bonner
a
, Geo?rey B. Sprinkle
b,
*
a
Leventhal School of Accounting, Marshall School of Business, University of Southern California, 3660 Trousdale Parkway,
Los Angeles, CA 90089-0441, USA
b
Kelley School of Business, Indiana University, 1309 East Tenth Street, Bloomington, IN 47405-1701, USA
Abstract
The purpose of this paper is to review theories and evidence regarding the e?ects of (performance-contingent)
monetary incentives on individual e?ort and task performance. We provide a framework for understanding these
e?ects in numerous contexts of interest to accounting researchers and focus particularly on how salient features of
accounting settings may a?ect the incentives-e?ort and e?ort-performance relations. Our compilation and integration
of theories and evidence across a wide variety of disciplines reveals signi?cant implications for accounting research and
practice. Based on the framework, theories, and prior evidence, we develop and discuss numerous directions for future
research in accounting that could provide important insights into the e?cacy of monetary reward systems. # 2002
Elsevier Science Ltd. All rights reserved.
1. Introduction
Monetary incentives frequently are suggested as
a method for motivating and improving the per-
formance of persons who use and are a?ected by
accounting information (e.g. Atkinson, Banker,
Kaplan, Young, 2001; Horngren, Foster, & Datar,
2000; Zimmerman, 2000), and their use in organi-
zations is increasing (Wall Street Journal, 1999).
Further, researchers have been encouraged to
employ incentives in experimental studies so that
subjects are su?ciently motivated and participate
in a meaningful fashion (e.g. Davis & Holt, 1993;
Friedman & Sunder, 1994; Roth, 1995; Smith,
1982, 1991). Anecdotal and empirical evidence,
however, indicates that monetary incentives have
widely varying e?ects on e?ort and, consequently,
oftentimes do not improve performance (Bonner
et al., 2000; Camerer & Hogarth, 1999; Gerhart &
Milkovich, 1992; Jenkins, 1986; Jenkins, Mitra,
Gupta, & Shaw, 1998; Kohn, 1993; Young &
Lewis, 1995). Consistent with this, accounting
studies examining the e?ects of incentives on indi-
vidual performance ?nd mixed results with regard
to their e?ectiveness (e.g. Ashton, 1990; Awasthi
& Pratt, 1990; Libby & Lipe, 1992; Tuttle & Bur-
ton, 1999; Sprinkle, 2000). If monetary incentives
have disparate e?ects on e?ort and performance,
0361-3682/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PI I : S0361- 3682( 01) 00052- 6
Accounting, Organizations and Society 27 (2002) 303–345
www.elsevier.com/locate/aos
* Corresponding author. Tel.: +1-812-855-3514; fax: +1-812-855-4985.
E-mail address: [email protected] (G.B. Sprinkle).
then suggestions for their use in either the ?eld
or the laboratory should be informed by an
understanding of the factors that moderate their
e?ectiveness.
We have four objectives in this paper. Our ?rst
objective is to provide a conceptual framework for
understanding the e?ects of (performance-con-
tingent) monetary incentives on individual e?ort
and performance and also to discuss theories that
suggest mediators of the incentives-e?ort relation.
Here, our focus is on explicating the motivational
and cognitive mechanisms by which monetary
incentives are presumed to increase performance;
understanding these mechanisms is critical for
determining how to maximize the e?ectiveness of
monetary incentives. Theoretically, monetary
incentives work by increasing e?ort which, in turn,
leads to increases in performance. Given these
relations, we ?rst provide a detailed discussion of
the various components of the e?ort construct:
direction, duration, intensity, and strategy devel-
opment. We then describe theories that detail the
mechanisms through which monetary incentives
are presumed to lead to increases in e?ort. These
theories are expectancy theory, agency theory (via
expected utility theory), goal-setting theory, and
social-cognitive (self-e?cacy) theory.
Our second objective is to enumerate and cate-
gorize important accounting-related variables that
may combine with monetary incentives in a?ecting
task performance. To do this, we express the
monetary incentives-e?ort and e?ort-performance
relations as a function of person variables, task
variables, environmental variables, and incentive
scheme variables. This conceptualization allows
for a full, yet parsimonious, categorization of the
numerous accounting-related variables that may
a?ect these relations, thereby facilitating an
understanding of the e?ects of monetary incen-
tives in numerous contexts of interest to account-
ing researchers.
Our third objective is to review evidence
regarding the e?ects of the combination of these
important accounting-related variables and mone-
tary incentives on individual e?ort and perfor-
mance. Here, we choose one speci?c variable from
each of the person, task, environmental, and
incentive scheme categories within our framework
Fig. 1. Conceptual framework for the e?ects of performance-contingent monetary incentives on e?ort and task performance.
304 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
and discuss its e?ects on the incentives–e?ort and
e?ort–performance relations. We also discuss the
importance of each variable in accounting settings
as well as the theoretical and practical importance
of examining the variable in conjunction with
monetary incentives. We then review studies from
a wide variety of disciplines to discuss the empiri-
cal e?ects of these variables on the incentives–
e?ort and e?ort–performance relations. Next, we
provide insights regarding how the results from
our compilation of studies may have signi?cant
implications for accounting research and practice.
Finally, we brie?y discuss the theoretical and
empirical relations between monetary incentives
and many other important accounting-related
person, task, environmental, and incentive scheme
variables.
Our ?nal objective is to identify and discuss
numerous directions for future research in account-
ing that would help ?ll gaps in our knowledge
regarding the e?cacy of monetary reward systems.
Thus, for each of the person, task, environmental,
and incentive scheme categories within our frame-
work, we enumerate many important questions
about the e?ects of monetary incentives on e?ort
and task performance. We believe it is essential to
address these questions given the important role
that accountants and accounting information play
in compensation practice and the design of per-
formance-measurement and reward systems.
The remainder of this paper is organized as fol-
lows. In Section 2, we introduce our conceptual
framework and discuss two important elements of
the framework. First, we discuss the general e?ects
of monetary incentives on e?ort and task perfor-
mance and explicate the e?ort construct. Second,
we discuss theories that suggest mediators of the
incentives-e?ort relation. In Section 3, we com-
plete our discussion of the conceptual framework;
in particular, we discuss important accounting-
related variables that may moderate the e?ects of
monetary incentives on e?ort and the e?ects of
e?ort on task performance. In discussing these
moderators, Section 3 also provides detailed evi-
dence regarding the e?ects of monetary incentives
on e?ort and task performance under various
situations, the implications of this evidence for
accounting research and practice, and numerous
directions for future accounting research. In Sec-
tion 4, we summarize our main points and o?er
concluding comments.
2. Theories about the e?ects of monetary incen-
tives on e?ort and task performance
The general hypothesis regarding the e?ects of
monetary incentives on e?ort and performance is
that incentives lead to greater e?ort than would
have been the case in their absence.
1,2
This basic
idea, however, does not explain how monetary
incentives lead to increases in e?ort. Accordingly,
theories about mediators of the incentives-e?ort
relation deserve further attention, and we discuss
these theories after explication of the e?ort con-
struct.
3
In turn, increased e?ort is thought to lead
1
A few theories predict that monetary incentives may lead
to decreased e?ort and performance. For example, cognitive-
evaluation theory (e.g. Deci et al., 1981; Deci & Ryan, 1985)
suggests that monetary incentives, by focusing attention on the
external reward related to a task, decrease intrinsic motivation
and, thus, can decrease e?ort and task performance. Addition-
ally, arousal theory (Broadbent, 1971; Easterbrook, 1959;
Eysenck 1982, 1986; Humphreys & Revelle, 1984; Yerkes &
Dodson, 1908) posits an inverted-U relationship between
arousal and e?ort and, consequently, between e?ort and per-
formance. That is, e?ort and performance initially increase as
arousal increases, but then begins to decrease once arousal
increases beyond a moderate level, thereby inducing anxiety.
Since arousal and anxiety are posited to be created, in part, by
motivational devices such as monetary incentives, arousal the-
ory predicts that monetary incentives may either increase or
decrease e?ort and task performance.
2
While e?ort typically is discussed as the key intervening
variable between monetary incentives and performance, some
researchers have focused on other variables such as a?ect
(Stone & Ziebart, 1995) and stress (Shields, Deng, & Kato,
2000). For example, Stone and Ziebart (1995) propose that
monetary incentives increase negative a?ect and, in turn,
increases in negative a?ect directly decrease performance.
However, these researchers also note that variables such as
a?ect and stress likely are important in explaining the relation
between incentives and performance because they can mediate
the incentives-e?ort relation rather than directly intervene
between incentives and performance.
3
The e?ort–performance relation is not connected to
monetary incentives per se, so we do not discuss theories that
explain this relation. For discussions of the e?ort-performance
relation (Bandura, 1997; Kahneman, 1973; Kanfer, 1987;
Locke and Latham, 1990; Navon and Gopher, 1979; Payne et
al., 1990).
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 305
to an improvement in the rewarded dimension of
task performance.
4
Fig. 1 presents a conceptual
framework for the e?ects of monetary incentives
on e?ort and task performance.
5
In the remainder
of the paper, we discuss the various relations
depicted by our framework.
First, we discuss the e?ort construct. Greater
e?ort refers either to e?ort directed toward current
performance of the task, which is thought to lead
to immediate performance increases, or e?ort
directed toward learning, which is thought to lead
to delayed performance increases (improved per-
formance on later trials). Increases in e?ort direc-
ted toward current performance are classi?ed as
changes in e?ort direction, e?ort duration, and
e?ort intensity, whereas e?ort directed toward
learning is characterized as strategy development
(Bettman, Johnson, & Payne, 1990; Kahneman,
1973; Kanfer, 1990; Locke & Latham, 1990).
E?ort direction refers to the task or activity in
which the individual chooses to engage (i.e. what
an individual does). As long as the expected bene-
?ts provided by monetary incentives outweigh the
costs of doing a task or activity, incentives tied to
performance theoretically should lead to e?ort
being directed toward the rewarded task or activ-
ity. In the ?eld, the e?ects of incentives on the
direction of e?ort can be observed with such
measures as absenteeism and task choice (Kanfer,
1990). Laboratory experiments usually constrain
the direction of e?ort to a large degree in that they
o?er only one task to subjects and require subjects
to remain present to receive incentive payments.
However, if subjects focus on a particular dimen-
sion of a laboratory task as opposed to other
dimensions, this is similar to making a choice
among tasks. For example, subjects paid a piece
rate for each completed toy assembly likely would
focus on creating as many assemblies as possible
rather than focusing on the quality of individual
assemblies. Furthermore, subjects can choose to
do the task or daydream. In these ways, monetary
incentives may have e?ects on e?ort direction in
the laboratory.
E?ort duration refers to the length of time an
individual devotes cognitive and physical resour-
ces to a particular task or activity (i.e. how long a
person works). In the ?eld, incentive contracts
typically are based on relatively long periods of
time, such as a year, and on performance measures
that attempt (at least partially) to measure sus-
tained e?ort over those periods. It seems fairly
intuitive that monetary incentives can increase
e?ort duration in these settings (e.g. employees
may take fewer breaks or work overtime). The
e?ects of monetary incentives on e?ort duration,
however, also can occur in laboratory experiments.
Speci?cally, these e?ects can appear in longer
laboratory studies as well as studies in which sub-
jects work at their own pace and control the time
taken to complete the activity or task (i.e. subjects
can leave the experiment at di?erent times).
Finally, increases in e?ort directed toward cur-
rent performance can come in the form of increa-
ses in e?ort intensity, which refers to the amount
of attention an individual devotes to a task or
activity during a ?xed period of time (i.e. how
hard a person works). As Kanfer (1990) notes,
e?ort intensity essentially captures how much of
one’s total cognitive resources are directed toward
a particular task or activity. In both the ?eld and
the laboratory, e?ort intensity may be measured
by assessing performance on timed tasks or tasks
involving explicit (?xed) time limits (assuming
e?ort direction is constrained). Similar to the
e?ects on e?ort direction and duration, monetary
incentives theoretically have positive e?ects on
e?ort intensity if people believe that short-term
increases in cognitive resources deployed toward
4
We discuss the e?ects of monetary incentives on non-
rewarded dimensions of performance in Section 3.4.
5
In our framework, we portray cognitive and motivational
mechanisms as mediating the relation between monetary incen-
tives and e?ort, and person variables, task variables, environ-
mental variables, and incentive scheme variables as moderating
the monetary incentives-e?ort relation and/or the e?ort-task
performance relation. As discussed in Baron and Kenny (1986,
p. 1174) a moderating variable ‘‘a?ects the direction and/or
strength of the relation between the independent variable and a
dependent variable’’. Baron and Kenny (1986, p. 1176) state
that a mediating variable ‘‘explains how external physical
events take on internal psychological processes. Whereas mod-
erator variables specify when certain e?ects will hold, media-
tors speak to how or why such e?ects occur’’ (also see Reber,
1995).
306 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
the task will lead to increases in the performance
measure for which they are being rewarded.
6
Monetary incentives also may motivate people
to invest e?ort to acquire the skills needed to per-
form a task so that future performance and
rewards will be higher than they otherwise would
be (i.e. learn). This notion of increased e?ort is
referred to as strategy development and consists of
conscious problem solving, planning, or innova-
tion on the part of the person performing the task.
Here, individuals may not be working on the task
or activity per se. Compared to increases in e?ort
direction, intensity, and duration, increases in
e?ort directed toward strategy development are
less automatic and also are likely to have a nega-
tive e?ect on performance in the short run, but a
positive e?ect on performance in the long run.
Given this, incentives are thought to promote
e?ort directed toward strategy development when
more automatic mechanisms are not su?cient to
attain desired performance and reward levels
(Locke & Latham, 1990).
Next, we discuss the proposed cognitive
mechanisms by which monetary incentives in?u-
ence the various dimensions of e?ort. Under-
standing these mechanisms is critical for
determining how to maximize the e?ectiveness of
monetary incentives (Bonner, 1999). For example,
organizations may restructure incentive schemes
in an attempt to enhance performance, but if the
restructured elements of the incentives do not tar-
get the key cognitive processes that lead incentives
to a?ect e?ort, then the restructuring will not be
e?ective. Moreover, changes in incentive plans are
costly (Gerhart & Milkovich, 1992), and under-
standing the cognitive processes a?ected by
monetary incentives and setting up compensation
plans that target these processes can reduce these
costs. Such plans likely will have the most positive
e?ects on e?ort and performance.
Although several theories explaining the e?ects
of incentives on e?ort have been o?ered, we dis-
cuss only four. These four theories represent the
predominant explanations o?ered for the e?ects of
monetary incentives on e?ort direction, duration,
and intensity; there is very little information about
the mediators of the incentives-strategy develop-
ment e?ort relation. The theories are expectancy
theory, agency theory (via expected utility theory),
goal-setting theory, and social-cognitive (self-e?-
cacy) theory.
7
We ?rst discuss the theories sepa-
rately, and then present recent conceptualizations
that combine elements of many of them.
Expectancy theory (e.g. Vroom, 1964) proposes
that people act to maximize expected satisfaction
with outcomes. Expectancy theory posits that an
individual’s motivation in a particular situation is
a function of two factors: (1) the expectancy about
the relationship between e?ort and a particular out-
come (e.g. a certain level of pay for a certain level of
performance), referred to as the ‘‘e?ort-outcome
expectancy’’ and (2) the valence (attractiveness) of
the outcome.
8
The motivation created by these
two factors leads people to choose a level of e?ort
that they believe will lead to the desired outcome.
6
Some researchers consider ‘‘arousal’’ to be the same con-
struct as e?ort intensity (e.g. Locke & Latham, 1990; also see
Humphreys & Revelle, 1984). Other researchers and arousal
theory, though, suggest that arousal (or stress) can a?ect the
various dimensions of e?ort (e.g. Ashton, 1990; Eysenck 1982,
1986; Shields et al., 2000) and posit that arousal (or stress) is an
important mediator in the monetary incentives–e?ort relation.
This helps explain how monetary incentives may lead to
increases, and particularly decreases, in e?ort. Because of this,
we refer to arousal and stress directly in later sections of the
paper when they might be viewed as constructs that are sepa-
rate from e?ort intensity.
7
Other theories of motivation discuss factors that, in addi-
tion to monetary incentives, can a?ect expectancies, utility,
goals, and self-e?cacy. In other words, these factors also a?ect
the key processes through which monetary incentives are pre-
sumed to operate. These theories include Maslow’s hierarchy of
needs (1943), achievement theory (McClelland, Atkinson,
Clark, & Lowell, 1953; Weiner, 1972), and reinforcement the-
ory (Hamner, 1974; Skinner, 1953). For overviews of several
motivation theories, see Kanfer (1990), Locke (1991), and
Miner (1980).
8
The e?ort–outcome expectancy can be broken down into
an e?ort–performance expectancy, a performance–evaluation
expectancy, and an evaluation–outcome expectancy to re?ect
factors that operate in the workplace such as imperfect evalua-
tion processes and uncertainty about outcomes (Naylor,
Pritchard, & Ilgen, 1980). These last two expectancies probably
are minimized in laboratory studies of incentive e?ects because
the performance criteria usually are clear and speci?ed in
advance, individuals are not being ‘‘evaluated’’ per se, and pay
likely is the primary outcome of interest. Thus, the e?ort–out-
come expectancy likely reduces to the e?ort–performance
expectancy in the laboratory.
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 307
The e?ect of monetary incentives on e?ort in an
expectancy-theory conceptualization is twofold.
First, the outcome of interest is the ?nancial
reward. Money can have valence for a variety of
reasons. Vroom’s initial conception of the valence
of money is that money is instrumental in obtain-
ing things people desire such as material goods. In
addition, money has symbolic value due to its
perceived relationship to prestige, status, and
other factors (Furnham & Argyle, 1998; Zelizer,
1994). Monetary incentives clearly have higher
valence than no pay (if expected pay is greater
than zero) and also may have higher valence than
noncontingent incentives, depending on the rela-
tive payment schedules.
Second, expectancies also should be, and have
been found to be, higher under monetary incentives
than under no pay or noncontingent incentives due
to the stronger links among e?ort, performance,
and pay (e.g. Jorgenson, Dunnette, & Pritchard,
1973; Locke & Latham, 1990; Pritchard, Leonard,
Von Bergen, & Kirk, 1976). Therefore, according
to expectancy theory, an individual’s motivation
and subsequent e?ort likely are signi?cantly higher
when compensation is based on performance, due
to both an increased expectancy about the e?ort–
outcome relationship and an increased (or at least
no change in the) valence of the outcome.
Agency theory (e.g. Baiman, 1982, 1990; Eisen-
hardt, 1989), via its assumption that individuals
are expected utility maximizers, adds further
structure in explaining the e?ects of monetary
incentives on e?ort. Speci?cally, a fundamental
assumption of agency theory is that individuals
are fully rational and have well-de?ned pre-
ferences that conform to the axioms of expected
utility theory. Further, individuals are presumed
to be motivated solely by self-interest, where self-
interest is described by a utility function that con-
tains two arguments: wealth and leisure. Individuals
are presumed to have preferences for increases in
wealth and increases in leisure (reductions in e?ort).
Agency theory (and most models of economic
behavior) therefore posits that individuals will
shirk (i.e. exert no e?ort) on a task unless it
somehow contributes to their own economic
well-being. Incentives that are not contingent on
performance generally do not satisfy this criter-
ion.
9
Thus, similar to expectancy theory, agency
theory suggests that incentives play a fundamental
role in motivation and the control of performance
because individuals have utility for increases in
wealth. Additionally, agency models typically
assume that individuals (employees) are strictly
risk-averse and, therefore, also must be paid a
risk-premium when monetary incentives are based
on imperfect surrogates of behavior (e.g. output
that is a function of both e?ort and some random
state of nature). Thus, monetary incentives can
lead to ine?cient risk-sharing, although the moti-
vational (e?ort) bene?ts associated with linking
pay to performance are presumed to exceed this
loss in e?ciency. Monetary incentives must there-
fore appropriately balance the need for providing
motivation (to increase e?ort) against the need for
risk-sharing (Holmstrom, 1989).
In essence, both expectancy theory and agency
theory suggest that monetary incentives a?ect the
attractiveness/utility of various outcomes, and
that e?ort a?ects the probability of achieving
these outcomes. Thus, monetary incentives
increase an individual’s desire to increase perfor-
mance and concomitant pay. In turn, this desire
motivates individuals to exert costly e?ort because
increases in e?ort are presumed to directly lead to
increases in expected performance. However, nei-
ther expectancy theory nor agency theory provides
much information about the cognitive mechan-
isms whereby the motivation created by monetary
incentives leads to changes in e?ort. Goal-setting
theory and social-cognitive theory add further
richness to these fundamental ideas.
Goal-setting theory (Locke & Latham, 1990)
proposes that personal goals are the primary
determinant of, and immediate precursor to,
e?ort. In other words, personal goals are the sti-
mulant of the incentive-induced e?ort increases
described above.
10
In particular, research indicates
9
With a su?ciently high level of monitoring and penalties,
noncontingent pay could be optimal. Additionally, Fama
(1980) has argued that the e?ects of reputation on one’s market
wage may reduce (or even eliminate) the need for explicit per-
formance-based incentives.
10
Personal goals are those chosen by individuals and, as
such, may or may not be the same as the goals assigned by an
organization or experimenter.
308 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
that speci?c and challenging personal goals lead to
greater e?ort than goals that are vague or easy, or
no goals at all. Challenging goals lead to greater
e?ort than easy goals simply because people must
exert more e?ort to attain the goal. While goal-
setting theory allows for expectancies to a?ect
personal goals, evidence shows that assigned goals
have a much larger e?ect on personal goals than
do expectancies. Further, expectancies and perso-
nal goals have separate e?ects on e?ort and per-
formance, indicating they capture di?erent
cognitive processes (Locke & Latham, 1990, p. 72).
The manner in which monetary incentives a?ect
e?ort in a goal-setting conceptualization is not
completely clear, but several processes have been
proposed. In particular, Locke, Shaw, Saari, and
Latham (1981) proposed three possible ways in
which incentives can a?ect e?ort via goal setting.
First, monetary incentives may cause people to set
goals when they otherwise would not. Such an
e?ect of monetary incentives is not captured by
expectancy theory or most economic models of
behavior since individuals’ ‘‘goals’’ are presumed
to be well-speci?ed in advance. Second, monetary
incentives might cause people to set more challen-
ging goals than they otherwise would; these goals
in turn lead to higher e?ort. One can view this as
being captured by expectancy theory and expected
utility (agency) theory in the sense that people are
simply choosing outcomes that require higher
levels of e?ort because the attractiveness asso-
ciated with their performance outcomes is
increased by incentives. Finally, monetary incen-
tives may result in higher goal commitment (and
thus greater e?ort) than noncontingent incentives
or no incentives. This proposed e?ect of incentives
on e?ort typically would not appear in expectancy
theory or expected utility theory conceptualiza-
tions because commitment is presumed to be
invariant. Consequently, goal-setting theory pro-
vides a description of the e?ect of incentives on
e?ort that goes beyond their e?ects on expectan-
cies and outcomes (probabilities and values).
Social-cognitive (or self-e?cacy) theory (Ban-
dura, 1986, 1991, 1997) proposes self-regulatory
cognitive mechanisms that relate to e?ort. Self-
e?cacy theory e?ectively expands upon both
expectancy theory and goal-setting theory by fur-
ther explicating the cognitive factors that a?ect
e?ort and, consequently, the possible cognitive
mechanisms by which monetary incentives can
a?ect e?ort. Speci?cally, self-e?cacy, or an indi-
vidual’s belief about whether he or she can execute
the actions needed to attain a speci?c level of per-
formance in a given task, is posited to be an
important determinant of e?ort.
11
Self-e?cacy is
thought to help people regulate their e?ort and,
consequently, it can a?ect e?ort direction, e?ort
duration, e?ort intensity, and strategy develop-
ment. In other words, self-e?cacy is thought to be
another variable (in addition to goals) that a?ects
the key dimensions of e?ort. Self-e?cacy also is
posited to a?ect e?ort indirectly through its
impact on goal levels and goal commitment. For
example, people with high self-e?cacy like to take
on challenges by setting high personal goals and
being strongly committed to achieving those goals.
However, while self-e?cacy focuses on goal set-
ting as a principal means of regulating one’s
behavior, it allows for other factors to come into
play. In particular, self-e?cacy is thought to a?ect
e?ort through several cognitive, motivational,
a?ective, and task mechanisms.
12
Cognitive and
motivational mechanisms include goal setting,
expectancies, and the increased use of high-quality
problem-solving strategies. Self-e?cacy also can
have positive e?ects on initial emotional states (those
prior to task performance) and can alleviate aversive
emotional states that arise during task performance.
Finally, self-e?cacy a?ects the initial selection of
tasks in that higher self-e?cacy leads to the choice
of more challenging tasks to perform.
11
Bandura (1997) notes that self-e?cacy is not to be con-
fused with self-esteem. Self-e?cacy is a belief about one’s abil-
ity to perform a speci?c task, whereas self-esteem is a global
evaluation of self-worth.
12
Additionally, Bandura (1997) claims that the self-e?cacy
construct goes beyond the e?ort–performance expectancy idea,
thus expanding on expectancy theory in that it re?ects all fac-
tors (not just e?ort) that a person believes can a?ect his or her
performance on a particular task or activity. Others (e.g. Locke
& Latham, 1990) question this idea from an operational
standpoint. They note that, while in theory, self-e?cacy is a
broader construct than the e?ort–performance expectancy,
measurements of the two likely produce similar results because
people typically are not asked to limit themselves to consider-
ing the e?ects of e?ort on performance when their expectancies
are measured.
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 309
Because self-e?cacy a?ects many factors, the roles
of incentives in a?ecting e?ort in self-e?cacy theory
likely are more numerous than those speci?ed by
expectancy, agency, and goal-setting theories. The
general relation between monetary incentives and
self-e?cacy is as follows (Bandura, 1997). Incentives
lead to increased task interest and, consequently, to
increased e?ort. In turn, increased e?ort generally
leads to improved performance, greater skill on the
task (if the person has the ability to increase skill),
and increased self-e?cacy. The increase in self-
e?cacy due to incentives can then ?ow through to
e?ort through the various goal mechanisms or
through other cognitive, motivational, a?ective, or
selective mechanisms described above.
13
Recent discussions of factors that mediate the
incentives-current e?ort relation appear to incor-
porate elements of many of these theories (e.g.
Klein & Wright, 1994; Lee, Locke, & Phan, 1997;
Locke & Latham, 1990; Riedel, Nebeker, &
Cooper, 1988; Wright, 1989, 1990, 1992; Wright &
Kacmar, 1995). For example, Wright and Kacmar
(1995) propose that performance-contingent
incentives a?ect self-e?cacy (a broadened expec-
tancy) and attractiveness (valence) of goal attain-
ment, which a?ects personal goal level and goal
commitment. Similarly, Riedel et al. (1988) suggest
that incentives a?ect valence and expectancies,
which can lead to spontaneous goal setting and
higher levels of goals and goal commitment. And,
as noted above, self-e?cacy theory speci?cally
includes personal goals as one of the more impor-
tant choices caused by self-regulatory behavior.
In summary, the fundamental hypothesis that pre-
dicts a positive overall relation between the presence
of monetary incentives and task performance is that
incentives increase e?ort and increased e?ort leads
to improvements in performance (either in the
short run or the long run). Furthermore, a number
of mechanisms have been proposed for explicating
the incentives-e?ort link, including expectancies,
self-interest, goal setting, and self-e?cacy.
In contrast to this fundamental hypothesis,
empirical evidence indicates that monetary incen-
tives frequently are not associated with increased
e?ort and improved performance. For example, in
reviewing laboratory studies of incentives, Bonner
et al. (2000) found that incentives lead to sig-
ni?cant performance improvements in no more
than half the studies (also see Camerer & Hogarth,
1999; Jenkins et al., 1998). In addition, Guzzo,
Jette, and Katzell’s (1985) meta-analysis of ?eld
studies of various motivating techniques, includ-
ing ?nancial incentives, indicated that ?nancial
incentives had widely varying e?ects and a mean
e?ect that was not signi?cantly di?erent from zero
(also see Prendergast, 1999). Empirical studies
that examine the e?ect of incentives on mediating
factors also ?nd mixed results. For example,
incentives sometimes lead to higher goals, greater
commitment, and/or enhanced self-e?cacy, and
sometimes do not (Lee et al., 1997; Wright, 1989,
1990, 1992; Wright & Kacmar, 1995).
Studies examining the e?ects of incentives on
performance, as well as studies examining media-
tors of the incentives–e?ort relation note that
there must be factors that moderate these rela-
tions, thereby causing incentive e?ects to not
always be positive (and to not always be consistent
with proposed mediating forces). To date, how-
ever, reviews have discussed relatively few such
factors and, as such, little is known about vari-
ables that interact with incentives in a?ecting task
performance. Again, it is important to identify
factors that moderate the e?ectiveness of incen-
tives so that researchers and organizations can
have better information about the use of monetary
incentives in either the ?eld or the laboratory.
13
Note that this explanation can pertain only to multiple-
trial situations in which subjects perform the task and receive
feedback. In single-trial situations (as is the case in many
experiments) a positive incentives–self-e?cacy relation is less
likely because, as Bandura (1991) notes, the motivation that
comes from self-e?cacy is not likely to be activated unless
people know how well they are performing. Thus, any relation
between incentives and self-e?cacy in these settings would have
to be predicated on an expectancy-theory conceptualization or
a reframing of self-e?cacy as a goal. For example, an expec-
tancy-theory view might be that people believe incentives will
propel them toward greater e?ort in order to attain the reward;
this belief then leads to increased self-e?cacy. Alternatively, as
Baker and Kirsch (1991) discuss, if self-e?cacy represents one’s
beliefs about his or her skills, incentives are unlikely to a?ect
self-e?cacy in the short run because people know they cannot
improve their skills in a short time. Rather, for incentives to
have an e?ect on self-e?cacy, beliefs must represent one’s
intentions (goals).
310 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
The remainder of this paper discusses salient
accounting-related variables that may moderate
the positive e?ects of monetary incentives on task
performance. For each variable presented, we dis-
cuss its importance in accounting settings as well
as the theoretical and practical importance of
examining the variable in conjunction with mone-
tary incentives. We then summarize the prior
research examining the joint e?ects of monetary
incentives and the particular variable. In these
summaries, we discuss the general ?ndings,
attempt to tie these ?ndings back to the theories
and underlying cognitive mechanisms previously
discussed, and then discuss the potential implica-
tions for accounting research and practice. Fol-
lowing this, we highlight numerous open issues
regarding the e?cacy of monetary incentives in
improving task performance and provide sugges-
tions for how future accounting research could
help ?ll these gaps in our knowledge. We attempt
to accomplish these objectives within a framework
(Fig. 1) that we feel is helpful for understanding
the e?ects of monetary incentives on performance.
In the next section, we further elaborate on our
framework and present the empirical evidence.
3. Evidence regarding the e?ects of monetary
incentives on e?ort and performance and a fra-
mework for research
In discussing accounting-related variables that
may combine with monetary incentives to a?ect
e?ort and task performance, we employ and add
to Bonner’s (1999) three broad categories of vari-
ables that determine performance (also see Payne,
Bettman, & Johnson, 1990). Speci?cally, Bonner
(1999) sets performance=f (person variables, task
variables, environmental variables). Such a model
allows for full, yet parsimonious, consideration of
the factors that may a?ect performance.
14
We
modify Bonner’s (1999) model for use in our fra-
mework since we focus speci?cally on the mone-
tary incentives-e?ort and e?ort-performance
relations (Fig. 1). Speci?cally, we suggest that
these relations=f (person variables, task variables,
environmental variables, incentive scheme vari-
ables). Person variables are those that relate to the
individual performing the task; they are char-
acteristics the person brings to the task such as
motivation, personality, and abilities.
15
Task vari-
ables are those that relate to the task itself; a
‘‘task’’ can be de?ned as a piece of work assigned
to or demanded of a person. Task characteristics
can vary within tasks. For example, a bankruptcy
prediction task can be framed as predicting the
probability the company will fail or it can be
framed as predicting the probability the company
will survive. Some task characteristics, like com-
plexity, also vary across tasks. For example, pro-
blem-solving tasks generally are more complex
than other tasks.
Environmental variables include all the condi-
tions, circumstances, and in?uences surrounding a
person who is doing a speci?c task. In other
words, these variables do not relate to a particular
task or person but can surround all tasks and
persons in a given setting. Monetary incentives
typically are considered an environmental vari-
able, along with factors like time pressure,
accountability requirements, and assigned goals
(and Libby and Luft (1993) and Bonner (1999)
discuss them in this way). Because our paper
focuses on the e?ects of performance-contingent
incentives (vis-a` -vis no incentives or non-
contingent incentives), we examine environmental
variables that interact with incentives. Finally, we
consider elements of incentive schemes per se that
could alter the relations between (the presence of)
monetary incentives and e?ort, as well as between
e?ort and performance, such as what dimension(s)
of performance the incentive scheme rewards.
14
This model is similar in spirit to the one employed by
Libby and Luft (1993) in that it serves to enumerate and cate-
gorize variables that may in?uence performance. These cate-
gories di?er from Libby and Luft’s (1993), though, in two
ways. First, Libby and Luft (1993) discuss three speci?c person
variables—ability, knowledge, and motivation—rather than
discussing the more general category that includes other
important characteristics of people. Second, Libby and Luft
(1993) do not include task variables in their formulation other
than noting that the relative e?ects of the other variables may
di?er across types of tasks (i.e. they do not speci?cally include
the main e?ects of various task factors).
15
The term ‘‘ability’’ refers to traits that are formed by the
time one is an adult, i.e. traits that are in?uenced mostly by
genetic factors and early childhood experiences (Carroll, 1993).
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 311
Evidence regarding the e?ects of monetary
incentives on e?ort and performance comes from
a large body of literature in accounting, econom-
ics, ?nance, management, and psychology. Given
the enormity of these literatures, we restrict our
attention to studies that employ laboratory
experiments or highly controlled ?eld experiments.
Additionally, we only consider studies that report
the e?ects of monetary incentives on individual
e?ort and performance and for which there is
some normative performance standard (i.e. the
criterion for high performance is clear). This
means that we do not consider the literature on
incentive e?ects in games or markets (multi-person
settings). While these clearly are of interest in
accounting, we chose to restrict our attention to
the individual. We also do not examine tasks
involving the choice between a certainty equiva-
lent and a gamble, for example, as there is no
normative performance criterion for such tasks.
We ?rst reviewed the 85 experimental studies
covered by the Bonner et al. (2000) review. Sec-
ond, we read a large number of additional studies
related to the e?cacy of incentives that did not
meet the Bonner et al. (2000) criteria but were of
relevance in developing our theoretical arguments.
Third, we read summaries of the literature on the
theoretical mediators of incentive e?ects, including
articles related to expectancies, self-interest
(agency relationships), arousal (stress), self-e?-
cacy, and goal-setting, as well as numerous other
individual empirical and theoretical articles.
Fourth, we read several review papers from
accounting, economics, management, and psy-
chology regarding the e?ects of incentives.
16
Finally, we read numerous papers related to the
person, task, environmental, and incentive scheme
variables we examine and their independent e?ects
on e?ort and performance.
In developing our arguments about the e?ects of
various factors on the incentives–e?ort and e?ort–
performance relations, we discuss papers and the-
ories as necessary. In other words, our paper is not
meant to provide an exhaustive, detailed review of
studies of incentive e?ects. Rather, our goal is to
integrate diverse ?ndings and theories regarding
the mechanisms by which incentive e?ects occur
and/or can be altered. To the extent previous
reviews have investigated the variables we discuss
here, we cite their ?ndings. Finally, to the extent
possible, we discuss whether the variables we con-
sider moderate the incentives–e?ort relation or the
e?ort–performance relation (or both).
3.1. Person variables
In this section, we discuss how person variables
may a?ect the relation between monetary incen-
tives and e?ort and e?ort and task performance.
Person variables include attributes that a person
possesses prior to performing a task, such as
knowledge content, knowledge organization, abil-
ities, con?dence, cognitive style, intrinsic motiva-
tion, cultural values, and risk preferences. These
person variables (like other variables) can a?ect
performance through various cognitive processes
that the person brings to bear while performing a
task, such as memory retrieval, information
search, problem representation, hypothesis gen-
eration, and hypothesis evaluation.
Person variables play an important role in the
performance of many accounting-related tasks.
For example, prior research documents that indi-
vidual factors such as knowledge content (e.g.
Bonner & Lewis, 1990; Bonner & Walker, 1994;
Bonner, Davis, & Jackson, 1992; Cloyd, 1997;
Dearman & Shields, 2001; Hunton, Wier, &
Stone, 2000; Vera-Mun˜ oz, 1998) and knowledge
organization (e.g. Dearman & Shields, 2001; Fre-
derick, 1991; Nelson, Libby, & Bonner, 1995) can
signi?cantly a?ect performance in a wide variety
of accounting tasks. Prior accounting research
also informs us that various abilities, such as ana-
lytical reasoning ability, can a?ect the task per-
formance of accountants as well as those who use
and are a?ected by accounting information
(Awasthi & Pratt, 1990; Bonner et al., 1992; Bon-
16
We also considered the executive compensation literature
(see Pavlik, Scott, & Tiessen, 1993 for a review). Because this
literature examines the relation between compensation (incen-
tives) and ?rm performance, however, it is di?cult to compare
it to the literature we review (that examining individual perfor-
mance). For example, as Pavlik et al. note, the direction of the
relation between executives’ individual performance and com-
pensation is unclear (i.e. it is not clear that the incentives lead
to the performance). As a result, we do not discuss the ?ndings
from the executive compensation literature.
312 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
ner & Lewis, 1990; Hunton et al., 2000; Tan &
Libby, 1997). Furthermore, accounting research
has shown that numerous other person variables,
including con?dence (Bloom?eld et al., 1999; Cote
& Sanders, 1997), cognitive style (Bernardi, 1994;
Johnson, Kaplan, & Reckers, 1998; Mills, 1996;
Pincus, 1990), intrinsic motivation (Becker, 1997),
cultural values (Harrison, Chow, Wu, & Harrell,
1999), and risk preferences (Shields, Chow, &
Whittington, 1989; Young, 1985) also can a?ect
performance in accounting settings.
While there are numerous person variables that
could be studied in conjunction with monetary
incentives, we devote our primary attention to the
role of variables that are included under the rubric
‘‘skill’’. We do so for three reasons. First, skill,
broadly de?ned, subsumes many of the person
variables previously discussed, including knowl-
edge content, knowledge organization, and the
various abilities that are relevant to performance
in a task. Second, skill plays a crucial role in the
performance of numerous accounting-related
tasks (Bonner & Lewis, 1990; Libby & Luft, 1993).
Third, in suggesting solutions for improving task
performance, it often is important to understand
exactly what skills the person brings (or does not
bring) to the task (Bonner, 1999). Since monetary
incentives frequently are suggested as a mechan-
ism for improving performance, it is important to
understand how a person’s skill a?ects the relation
between monetary incentives and performance.
3.1.1. E?ects of skill on the incentives–e?ort–
performance relation: direct role of skill
Skill can alter the e?ects of monetary incentives
on performance because of its important e?ects on
performance via several cognitive processes. For
example, skill includes knowledge (content) of
factual information that, when retrieved from
memory, can enhance task performance. Skill also
includes the organization of knowledge around
meaningful concepts, and appropriate knowledge
organization can facilitate the search for pertinent
information, the initial setup of problems (pro-
blem representation) and the generation of initial
hypotheses. All of these cognitive processes have
substantial e?ects on performance. In a similar
vein, mental and physical abilities of various sorts
aid in various cognitive and physical processes
that in?uence performance on many tasks, so that
the lack of requisite ability can severely constrain
performance. For example, problem-solving abil-
ity can help auditors diagnose errors when using
analytical procedures (Bonner & Lewis, 1990).
The direct e?ects of skill on performance sug-
gest that, despite the perfect rationality assump-
tion governing most economic models (Conlisk,
1996; Simon, 1986), skill may a?ect the incentives-
performance relation by attenuating the positive
e?ects of incentive-induced e?ort on performance.
Speci?cally, individuals may try harder in the pre-
sence of incentives (e.g. exhibit higher e?ort
intensity or higher e?ort duration) but, if they lack
the skill needed for a given task, their performance
will be invariant to increases in e?ort (e.g. Arkes,
1991; Bonner et al., 2000; Camerer, 1995; Kanfer,
1987; Smith & Walker, 1993). Although skill has
been discussed extensively as having an attenuat-
ing e?ect on the e?ort-performance relation, there
are few empirical studies that present direct evi-
dence regarding this issue.
17
Awasthi and Pratt (1990) found that subjects
working under performance-contingent incentives
exhibited higher e?ort duration than subjects
working under ?xed pay, irrespective of skill.
18
However, subjects with incentives did not perform
better than those working under ?xed pay unless
they possessed a high degree of skill. These ?nd-
ings are consistent with the proposed role of skill
in attenuating the e?ort–performance relation.
That is, while monetary incentives motivated sub-
jects to increase e?ort duration, they only
increased performance for those subjects with high
task-relevant skill.
Qualitatively similar ?ndings are reported in
Bonner, Hastie, Young, Hesford, and Gigone
(2001), who examined the e?ects of several incen-
17
There are a number of incentives studies that measure skill
but do not examine whether it moderates the incentives-e?ort
or e?ort-performance relations. Instead, they either adjust per-
formance measures for initial skill or ensure that mean skill
does not di?er across incentive treatments (e.g. Hogarth et al.,
1991; McGraw & McCullers, 1979; Toppen, 1965a, 1965b,
1966).
18
Awasthi and Pratt (1990) did not measure other dimen-
sions of e?ort, such as e?ort intensity or strategy development.
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 313
tive schemes on subjects’ performance in a mental
multiplication task. Although mental multi-
plication is a task that the subjects understood
how to perform, they varied dramatically in their
skill at the task. The authors also allowed time for
further learning by examining subjects’ perfor-
mance in very lengthy experiments. Speci?cally, in
their second experiment, which lasted 90 h over 12
weeks, monetary incentives increased e?ort dura-
tion for all subjects but only increased e?ort
intensity for high-skill subjects. Further, monetary
incentives increased two measures of performance
for high-skill subjects, but only one measure of
performance for low-skill subjects. Collectively,
these ?ndings are somewhat consistent with a lack
of skill attenuating the e?ort–performance relation,
although low-skill subjects’ e?ort only increased as
to duration under monetary incentives. That is, skill
a?ected both the incentives–e?ort and e?ort–per-
formance relations in this study.
Other studies present ?ndings that are sugges-
tive of the role of skill in attenuating the incentive-
induced e?ort–performance relation. These studies
employ a multi-period approach and ?nd that: (1)
incentive-related di?erences in performance in
skill-sensitive tasks increase over time (e.g. Huber,
1985; London & Oldham, 1977; Sprinkle, 2000),
or (2) incentive-related di?erences in performance
of simple tasks for which subjects possess skill do
not change over time (e.g. Bailey, Brown, &
Cocco, 1998; Harley, 1965a, 1965b; Pollack &
Kna?, 1958). The ?rst type of ?nding suggests that
the increased e?ort induced by incentives has a
greater e?ect when subjects’ skill on the task
increases. The second type of ?nding suggests that,
when subjects possess skill, incentive-induced
e?ort increases can ?ow through to performance,
i.e. a positive e?ort–performance relation remains
intact.
Given the small amount of evidence that directly
addresses the role skill plays in attenuating the
e?ort–performance relation, the appropriate
implications for accounting research and practice
are unclear. Under what conditions a lack of skill
means that incentive-induced e?ort will not
improve performance remains a rather complex
open issue. First, in order for skill to attenuate the
incentives-induced e?ort–performance relation,
incentives must lead to increases in e?ort. Thus,
incentives must meet subjects’ reservation wages
or some minimum level of symbolic value. In
other words, the expected utility from the incen-
tives must exceed the disutility from working on
the task.
19
Additionally, there must not be other
person, task, environmental, or incentive scheme
variables that substantially reduce the e?ect of
incentives on e?ort.
Second, in order for a lack of skill to attenuate
the incentives-induced e?ort–performance rela-
tion, skill and e?ort, like numerous other factors
of production, must be complements to some
extent (i.e. the marginal rate of technical substitu-
tion between skill and e?ort must not be constant;
see, e.g. Jehle & Reny, 2001). Thus, increases in
e?ort cannot completely substitute for a lack of
skill. Assuming that there is a continuum describ-
ing the relation between skill and e?ort in their
e?ects on performance (with the endpoints being
skill and e?ort as complete complements versus
skill and e?ort as complete substitutes), the ques-
tion arises as to whether skill and e?ort act more
like complements or more like substitutes. Our
belief is that most accounting-related tasks require
some skill (knowledge and/or ability), but that
possessing skill is not su?cient to guarantee high
levels of task performance. That is, individuals
must exert some e?ort to bring their skill to bear
in most tasks. The tasks that may be exceptions
are tasks that involve relatively automatic cogni-
tive processes such as frequency learning and esti-
mation (Libby & Lipe, 1992). For such tasks, lack
of skill is less likely to attenuate the positive
e?ort–performance relation because this relation
is of much smaller magnitude when tasks require
very little e?ort.
20
Given the small number of tasks examined by
prior research, but the wide variety of accounting-
related tasks, future research directed toward
19
In situations where incentives do not meet reservation
wages or a minimum level of symbolic value, as may occur in
some experiments, there may be no e?ect of incentives on
e?ort, in which case the role of skill in the e?ort-performance
relation becomes moot.
20
The reverse phenomenon would occur when tasks require
virtually no skill, although it is unclear that there are many
accounting-related tasks that require little or no skill.
314 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
understanding the relations among skill, e?ort,
and performance in a broader range of tasks
seems warranted. Such research could provide
useful insights regarding whether and when skill
and e?ort act more like complements or sub-
stitutes and, thus, the relative importance of skill
and e?ort in a?ecting performance in accounting
settings. This research also could examine whether
the relative importance of skill and e?ort depend
on inherent characteristics of the task or, instead,
whether this relationship can be altered by other
factors.
For example, all auditing ?rms are bound by
professional standards to conduct an audit in
accordance with Generally Accepted Auditing
Standards (GAAS). However, ?rms choose to
meet the requirements for a GAAS audit in sub-
stantially di?erent ways. Some ?rms employ
structured audit approaches for gathering evi-
dence; these structured approaches make use of a
number of decision aids and templates. Other
?rms employ very unstructured approaches that
allow auditors to determine the types of evidence
to gather for a particular audit. These di?erences
in structure a?ect the experience level of auditors
assigned to a given task—?rms with structured
approaches assign relatively less experienced audi-
tors than do ?rms with unstructured approaches
(Prawitt, 1995). This might suggest that ?rms
employing structured approaches believe e?ort is
relatively more important (vis-a` -vis skill) in audit-
ing tasks than do ?rms employing unstructured
approaches. Consequently, research that informs
us about the relative importance of skill and e?ort
in key accounting tasks and whether this relative
importance is embedded within the task (versus
created by audit technology) could inform audit
?rms about the potential e?ectiveness of incentives
and the circumstances under which they will be
most e?ective.
We also know very little about whether a lack of
skill attenuates the relation between all dimensions
of e?ort and performance or just that between
some dimensions of e?ort and performance. For
example, while a lack of skill in the short run can
attenuate the positive relation between e?ort and
performance, monetary incentives may lead to
strategy development, which over the long run can
allow individuals to acquire the necessary skill
they need, thereby ultimately restoring a positive
relation between incentives and performance (e.g.
Sprinkle, 2000). This raises questions regarding
the types of skills that can be acquired by exerting
e?ort under incentives, and how long it takes for
e?ort directed toward skill acquisition to ‘‘pay
o?’’. Since many of the skills that have been
examined in accounting research are ‘‘innate’’
abilities and, thus, relatively ?xed by the time
people are adults (Carroll, 1993), it is possible that
for many accounting-related tasks, skill de?-
ciencies created by a lack of ability will attenuate
the e?ort–performance relation over the long run
as well as the short run. Additional research is
needed to address this question.
3.1.2. E?ects of skill on the incentives–e?ort–
performance relation: indirect (self-selection) role
of skill
The second role of skill as it relates to perfor-
mance and, more speci?cally, the e?ect of incen-
tives on performance is the indirect screening or
self-selection role. Skill is a factor that people
consider when assessing their self-e?cacy (Ban-
dura, 1997). Again, self-e?cacy is a person’s
judgment of his or her capability for performing a
speci?c task (Stajkovic & Luthans, 1998). Self-
e?cacy plays an important role in a person’s
choice to perform a task or job, or even work for a
particular ?rm. In other words, self-e?cacy a?ects
initial e?ort direction. In addition, because self-
e?cacy positively a?ects the goals people set, self-
e?cacy also can a?ect e?ort duration and inten-
sity, as well as strategy development. Overall,
then, skill has an indirect e?ect on performance
because skill is positively related to self-e?cacy
and self-e?cacy a?ects the selection of tasks, as
well as consequent e?ort and performance in those
tasks. Consequently, on average, we would expect
that individuals with appropriate levels of skill
would select tasks requiring those skills and
choose to exert high levels of e?ort.
By the same token, ?rms use hiring and promo-
tion processes to assign employees to tasks and
jobs, and such task assignments naturally re?ect a
consideration of individuals’ skill, among other
factors. Thus, ?rms’ perceptions of skill indirectly
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 315
a?ect task performance to the extent these per-
ceptions result in a sample of individuals with
higher actual skill.
21
In contrast, when people are
assigned to tasks in experimental settings, the
experimenter typically does not consider subjects’
skill. Thus, he or she may be asking subjects to
perform tasks for which they lack skill and, con-
sequently, for which they have low self-e?cacy,
low goals, and thus, low e?ort (i.e. subjects ‘‘give
up’’).
The self-selection role of skill in improving per-
formance suggests another role for skill in a?ect-
ing the e?ects of incentives on performance.
Speci?cally, if an experimenter or an employer
assigns people to tasks for which they do not have
the necessary skill, people may know that they
lack the requisite skill and, as a result, the lack of
skill may attenuate a positive incentives–e?ort
relation. In other words, people who lack skill
may ‘‘give up’’ (not increase their e?ort due to
lowered self-e?cacy and lowered goals) under
incentives if they believe that e?ort increases will
not lead to performance increases and consequent
rewards. This giving-up phenomenon could be
particularly prevalent under incentive schemes like
tournaments where individuals may believe there
is a low probability of receiving compensation
(Bull, Schotter, & Weigelt, 1987; Dye, 1984).
By contrast, when individuals are allowed to
select their own incentive contracts for a particular
task, we would expect that persons lacking skill,
on average, would choose contracts that do not
include performance-based incentives (i.e. they
would choose ?xed pay contracts), whereas per-
sons who perceive that they have adequate skill
would choose performance-based contracts. That
is, economic theory suggests that monetary incen-
tives may serve as an important mechanism for
sorting individuals based on skill (e.g. Demski &
Feltham, 1978; Riley, 1979; Spence, 1973).
Because individuals with lower skill choose non-
contingent pay, skill likely does not decrease the
e?ect of incentives on e?ort for those choosing
performance-based incentives contracts. In other
words, incentives can ‘‘work’’ because individuals
choosing performance-contingent rewards have
appropriate skill. Further, it also is possible that
having the opportunity to choose one’s contract
can actually increase the positive e?ect of incen-
tives on e?ort because choosing a performance-
contingent contract also commits one to exerting
high levels of e?ort and, thus, may lead to higher
goals as well as greater commitment to achieving
high performance.
A few empirical studies have examined whether
having the opportunity to choose an incentive
contract a?ects performance. For example, Chow
(1983) found that subjects who selected a budget-
based compensation scheme had higher skill (and
performance) than subjects who selected a ?at rate
scheme. Further, subjects who selected the budget-
based contract outperformed those who were
assigned the same budget-based contract, and
there were no di?erences between subjects who
selected ?at rate schemes and those assigned to
these schemes. The e?ects of selecting the budget-
based contract on performance were due to initial
skill for a group of subjects who had a di?cult
goal, but were due to the combination of initial
skill and the opportunity to choose in a group of
subjects who had a moderate goal. In a follow-up
study, Waller and Chow (1985) also found that
subjects selecting a pure ?xed pay contract had
lower skill and performance than subjects selecting a
pure quota contract (also see Shields & Waller,
1988). Additionally, Waller and Chow (1985) found
that, after controlling for skill, the type of incentive
contract chosen had no e?ect on performance.
Similar ?ndings regarding the relation between
skill and choice of incentive contracts were repor-
ted by Dillard and Fisher (1990); however, they
found no di?erences in performance between sub-
jects who self-selected or were assigned incentives.
Farh, Gri?th, and Balkin (1991) found the self-
selection e?ect both with regard to initial skill and
task performance. Additionally, Farh et al. (1991)
found the same pattern of di?erences between
21
Such perceptions are not always accurate. Prior research
shows that individuals within ?rms who make job assignments
may make errors in their assessments of others’ skill. For
example, superiors judge subordinates’ skill based on a con-
sideration of their own skill (e.g. Kennedy & Peecher, 1997).
Other factors also may contribute to errors in superiors’ judg-
ments of subordinates’ skill such as prior experience with a
particular performance evaluation scheme (Frederickson et al.,
1999). See Hunt (1995) for a review of this literature.
316 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
assigned and self-selecting groups with regard to
goal levels, i.e. self-selecting groups had higher
goals. Here, the authors posited that the opportu-
nity to choose an incentive contract positively
a?ects commitment to that contract which, in
turn, positively a?ects goal levels and task perfor-
mance, although they did not measure the com-
mitment to the contract or directly examine
whether goal levels were a mediator of the choice-
versus-assignment e?ect on performance. An
alternative explanation is that higher levels of self-
e?cacy (due to higher levels of initial skill) led to
the higher levels of goals.
Collectively, the ?ndings from studies that have
examined the interaction of the self-selection of
incentive contracts and skill are consistent with
the notion that high-skill individuals tend to
choose performance-based incentive contracts
while low-skill individuals tend to choose ?at-rate
contracts. Further, subjects who choose perfor-
mance-based contracts often outperform those
who choose ?at-rate contracts, although not
always. In one instance, there is an e?ect for
choice of contracts that is not attributable to
initial skill di?erences; however, most studies ?nd
that task performance di?erences relate to di?er-
ences in initial skill. Finally, subjects who choose
performance-based contracts often outperform
those who are assigned the same contracts, but
subjects who choose ?at-rate contracts only
sometimes underperform those who are assigned
to such contracts. In short, these ?ndings are mostly
consistent with the posited interaction between
incentives and skill in skill’s self-selection role.
One implication of this research for accounting
is that incentive contracts tied to standards can
facilitate the attraction of individuals with higher
skill. This can help ?rms reduce adverse selection
problems which, in turn, can improve task assign-
ment and overall ?rm performance. For example,
job ladders within ?rms, whereby incentives (pro-
motion, etc.) are tied to ful?lling certain job stan-
dards, are believed to greatly facilitate the sorting
of employees based on their skills (Milgrom &
Roberts, 1992). In this way, monetary incentives
linked to standard attainment help cull managers
and executives with the greatest potential to
increase ?rm value. That said, the evidence to date
also raises a number of questions. First, while
prior research reports fairly consistent ?ndings
that performance-based incentives attract high-
skill individuals, the types of contracts examined
have been fairly limited. Most studies have used
budget-based schemes with moderate goals (where
the goal is to exceed average pretest performance).
Thus, it is not clear whether budget-based schemes
with very easy goals or very di?cult goals will
yield the same bene?ts. Further, it is unclear whe-
ther other forms of incentives, such as tournament
contracts, will perform better or worse than bud-
get-based contracts at attracting skilled indivi-
duals. For example, while tournaments may
demotivate the average person to whom they are
assigned (because they believe the probability of
winning is low), if self-selection is allowed, they
may attract very skilled individuals who believe
they have an excellent chance of winning (e.g.
Prendergast, 1999). Thus, when both the e?ort
and selection e?ects are considered, it is unclear
whether budget-based contracts will perform bet-
ter than tournament contracts.
Second, other person factors such as risk pre-
ferences have been proposed to moderate the
relation between skill and the selection of perfor-
mance-based incentive contracts (Chow, 1983).
Further, because people use their perceptions of
skill in making this choice, it is important to
examine factors besides actual skill that a?ect
perceptions of skill. For example, research indi-
cates that men tend to be overcon?dent about
their skill while women tend to be undercon?dent
(Estes & Hosseini, 1988; Lundeberg et al., 1994);
this might suggest that more men than is ‘‘war-
ranted’’ based on actual skill would select perfor-
mance-based pay, while fewer women than is
‘‘warranted’’ would select performance-based pay.
Finally, people may be less able to determine
whether they have appropriate skill for complex
tasks than for simple tasks (Gist & Mitchell,
1992). Being aware of factors that a?ect percep-
tions of skill and moderate the relation between
skill and selection of performance-based incentives
is important not only to experimentalists who
want to consider potential confounds and sources
of noise in ?ndings, but also to employers whose
workforce has many individual di?erences and
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 317
who are assigned to tasks that vary on many
characteristics.
Finally, we know very little about the mechan-
isms that cause individuals to give up (reduce
e?ort) when they are assigned tasks for which they
lack skill and also are assigned performance-based
incentives. People may have lowered self-e?cacy,
lowered self-set goals, or some combination
thereof. In turn, lowered self-e?cacy or lowered
goals may decrease e?ort direction, e?ort inten-
sity, or e?ort duration. Further, people may not
engage in appropriate amounts of strategy devel-
opment. Understanding the mechanisms that
cause reduced e?ort could facilitate ?nding an
appropriate remedy (such as the assignment of
goals) for the giving-up phenomenon.
3.1.3. E?ects of other person variables on the
incentives–e?ort–performance relation
As mentioned earlier, there are a number of
person variables that in?uence performance in
accounting tasks. Many of these variables also
may interact with incentives in a?ecting perfor-
mance. For example, the e?ort and performance
of high need-for-achievement (or intrinsically
motivated) individuals likely is a?ected less posi-
tively by the presence of monetary incentives (e.g.
Mawhinney, 1979). In other words, the positive
incentives–e?ort relation could be reduced for
high need-for-achievement individuals since their
e?ort is likely to be high irrespective of the situa-
tion. Consistent with this, Atkinson and Reitman
(1956) found that incentives had a positive e?ect
on the performance of low need-for-achievement
subjects, but a negative e?ect on the performance
of high need-for-achievement subjects. Further,
Vecchio (1982) found a positive e?ect for incen-
tives for low need-for-achievement subjects and no
e?ect for high need-for-achievement subjects. It is
unclear, though, whether need-for-achievement
(intrinsic motivation) a?ects the bene?ts that
accrue from the self-selection role of contracts.
Research is needed to examine this issue and,
more generally, how incentives and intrinsic moti-
vation combine to a?ect performance. Such
research could have important implications for the
design of accounting-based performance measure-
ment and reward systems and may ultimately
suggest, for example, that tasks and jobs that
attract highly intrinsically motivated individuals
do not require performance-based pay.
Future research also might investigate how cul-
tural background (and its attendant values) a?ects
the e?cacy of monetary incentives. Several studies
in accounting report that individuals from di?er-
ent cultures vary on dimensions such as the degree
of individualism (versus collectivism), power dis-
tance, femininity (versus masculinity), uncertainty
avoidance, and Confucian dynamism (Chow,
Shields, & Wu, 1999; Harrison & McKinnon,
1999). However, prior research has not examined
whether di?erences in these attributes actually
lead individuals to respond di?erentially to mone-
tary incentives. Such research is particularly
important since some (shared) dimensions of cul-
ture could yield opposing predictions regarding
the e?ects of monetary incentives, and it is theo-
retically unclear whether di?erences in cultural
background will lead to di?erential e?ort respon-
ses under monetary incentives (Chow et al., 1999).
Knowledge of such di?erences, though, could
facilitate the design of (and employees’ preferences
for) management control systems in companies
that employ culturally diverse workforces.
In summary, there are several person variables
that could interact with monetary incentives to
a?ect e?ort and task performance. We focus on
skill because it is a key variable related to perfor-
mance in accounting-related tasks. Moreover, skill
can play two roles in interacting with monetary
incentives—one role can attenuate a positive
e?ort–performance relation and one role can
attenuate a positive incentives–e?ort relation.
While much has been written about the important
role of skill in understanding the e?ects of mone-
tary incentives, surprisingly little empirical work
exists. The existing evidence suggests that skill
sometimes reduces the positive e?ects of incentive-
induced e?ort on performance and that highly
skilled individuals frequently choose performance-
based incentives when given the opportunity to do
so. However, there are a number of open questions
that research needs to examine to make useful sug-
gestions for the consideration of skill (and other
person variables) in choosing incentive contracts or
when using incentives in laboratory settings.
318 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
3.2. Task variables
In this section, we discuss how task variables
may a?ect the relations between monetary incen-
tives and e?ort and e?ort and performance. Task
variables include factors that vary both within and
across tasks, such as complexity, e?ort-sensitivity,
and framing (e.g. whether the situation is described
as a gain or a loss). Task variables also include
presentation format (e.g. whether accounting
information is presented in the balance sheet or in
footnotes), processing mode (e.g. whether people
are asked to process information simultaneously
or sequentially), and response mode (e.g. whether
people are asked to respond to a question in terms
of probabilities or frequencies). Finally, tasks are
thought to vary as to their ‘‘attractiveness’’, or
how interesting or fun they are perceived to be.
The importance of understanding the e?ects of
task characteristics on performance cannot be
understated. In a decision-making context, Hogarth
(1993, p. 411) notes: ‘‘To understand decision mak-
ing, understanding the task is more important than
understanding the people.’’ Hogarth’s comment
re?ects several important issues. First, much research
has noted that human decision-making strategies
‘‘evolve’’ to adapt to task demands (Anderson,
1990; Gigerenzer et al., 1999; Newell & Simon,
1972; Payne et al., 1990). Second, as Hogarth notes,
research has documented surprisingly frequently
that task variables explain more performance varia-
tion than key person variables. Because accounting
tasks may di?er dramatically from those used in
psychology research due to professional stan-
dards, regulations, and other factors, it is critical
to understand their characteristics and examine
how they a?ect accountants. Consequently,
numerous researchers have called for more work
related to analyzing these task characteristics (Ash-
ton & Ashton, 1995; Bonner, 1999; Gibbins &
Jamal, 1993; Hogarth, 1993; Peters, 1993). Prior
research in accounting has examined the e?ects of a
number of task characteristics on task perfor-
mance, including complexity (Asare & McDaniel,
1996; Simnett, 1996), framing (Kida, 1984; Lipe,
1993), order of information (Ashton & Ashton,
1988), presentation format (Maines & McDaniel,
2000; Vera-Mun˜ oz, Kinney, & Bonner, 2001), pro-
cessing mode (Ashton & Ashton, 1988; Libby &
Tan, 1999), and task attractiveness (Fessler, 2000).
While there are numerous task variables that
could be studied in conjunction with monetary
incentives, we focus on task complexity. We do so
for the following reasons. First, tasks in account-
ing settings can vary dramatically in complexity,
and complexity has been posited to be one of the
most important determinants of performance in
accounting settings (Bonner, 1994; Hogarth,
1993). Second, recent work in accounting portrays
and ?nds evidence consistent with task complexity
being a type of incentive that accounting profes-
sionals can face, thereby suggesting the impor-
tance of studying task complexity in conjunction
with monetary incentives (Das, Levine, & Sivar-
amakrishnan, 1998; Young, 2001). Finally, task
complexity sometimes has been confused with
e?ort-sensitivity, which is a separate characteristic
of tasks that has clear importance when consider-
ing the e?ects of monetary incentives on perfor-
mance. Consequently, we discuss the di?erences
between these two constructs.
3.2.1. E?ects of task complexity on the incentives–
e?ort–performance relation
Broadly de?ned, task complexity refers to the
amount of attention or processing a task requires
as well as the amount of structure and clarity the
task provides. Thus, task complexity increases as
the required amount of processing increases and
as the level of structure decreases (Campbell, 1988;
Wood, 1986). Task complexity therefore subsumes
constructs such as ‘‘task di?culty’’ and ‘‘task
structure’’ in addition to the algorithmic/heuristic
solution dimension of tasks discussed by Ashton
(1990) and McGraw (1978) since these constructs
relate to the amount and/or clarity of processing
involved in a task. Given this de?nition, there are
three roles task complexity can play in a?ecting
task performance.
First, task complexity can decrease current
e?ort duration and e?ort intensity, which can lead
to decreases in performance. Second, task com-
plexity can increase (or decrease) e?ort directed
toward strategy development, which also can lead
to decreases in short-run (or long-run) performance.
Third, task complexity can attenuate the e?ects of
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 319
e?ort on performance because increases in task
complexity lead to increases in skill requirements.
Thus, if skill is held constant, as is the case in many
experimental or other short-term situations, the
gap between subjects’ skill and tasks’ skill require-
ments increases as task complexity increases,
thereby making it less likely that e?ort will positively
in?uence performance. Overall, then, task complex-
ity can a?ect performance by decreasing current
e?ort duration and e?ort intensity or increasing
(or decreasing) e?ort directed toward strategy devel-
opment, all of which can lead to reductions in short-
run (and long-run) task performance. Additionally,
task complexity can attenuate the relationship
between e?ort and performance because individuals
are more likely to lack skill for complex tasks.
By de?nition, increases in task complexity lead
to increases in the e?ort requirements for a task
(Campbell, 1988; Wood, 1986). Ceteris paribus,
when a task’s e?ort requirements increase, people
may respond by exerting less absolute e?ort than
they would for a simpler task. There are a number
of possible explanations for this phenomenon.
First, standard expected utility theory (and adap-
tive decision-making theory, e.g. Payne et al.,
1990) suggests that, before performing a task,
individuals consider the costs and bene?ts related
to that task. Thus, persons weigh the bene?ts
associated with increasing performance against the
e?ort costs necessary to achieve higher perfor-
mance. If the costs outweigh the bene?ts, then
people will trade o? a reduction in performance
for reductions in e?ort. This may entail using
simpli?ed strategies or heuristics in addition to
exerting less e?ort in terms of duration and inten-
sity. Assuming that the bene?ts in terms of per-
formance are roughly equal under simple and
complex tasks, the e?ort costs would be more
likely to outweigh the bene?ts in complex tasks.
22
Second, task complexity is positively related to
arousal, as are incentives. Since theories posit an
inverted-U relationship between arousal and
e?ort/performance (Eysenck, 1986), the combina-
tion of incentives and a complex task could lead to
a less optimal level of arousal than the combina-
tion of incentives and a simple task.
23
For these
reasons, then, task complexity may attenuate a
positive incentives-e?ort duration relation and a
positive incentives–e?ort intensity relation.
Whether task complexity attenuates the incen-
tives–e?ort relation also depends on the relative
weights an individual places on good performance
(expected bene?ts) and e?ort (expected costs).
Incentives for good performance should increase
the relative weight placed on good performance.
However, whether expected bene?ts then outweigh
expected costs for complex tasks likely depends on
the individual’s belief that he or she can perform
well by exerting additional e?ort. In other words,
the individual’s perception of his or her skill or,
more generally, his or her self-e?cacy likely in?u-
ences the bene?ts a person expects from good
performance. As skill and self-e?cacy increase,
people are more likely to believe that they can
attain the expected bene?t (the incentive payment)
for a complex task, thereby increasing the expec-
ted (value of the) bene?t and the likelihood that
task complexity will not attenuate the incentives–
e?ort relation.
However, task complexity also can a?ect self-
e?cacy. Task complexity can make self-e?cacy
more di?cult to assess, making it more variable
than it would be with a simple task (Gist &
Mitchell, 1992). Further, because some individuals
will recognize that task complexity decreases per-
formance capabilities, ceteris paribus, self-e?cacy
may decrease. In other words, as skill increases,
self-e?cacy will increase and the expected bene?ts
for good performance in a complex task will out-
weigh the expected costs of obtaining good per-
formance. Consequently, the attenuating e?ect of
task complexity on the incentives–e?ort relation-
23
An alternative conceptualization is that, as task complex-
ity increases, the gap between performance capability and task
demands rise, giving rise to stress, which is presumed to have a
negative e?ect on e?ort and performance.
22
This is an important assumption that may not hold in
many settings as the rewards for performing well at complex
tasks often exceed those for performing well at simple tasks.
For example, the relative magnitude of CEO compensation to
average employee compensation (Crystal, 1991) likely re?ects,
among other things, di?erential complexity of tasks performed
by these individuals. At some point, then, higher rewards for
complex tasks (vis-a` -vis simple tasks) outweigh the higher cost
of exerting e?ort in these tasks. Consequently, the incentives–
e?ort relation would not be attenuated.
320 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
ship ultimately will decrease. However, this
reduction may be o?set by decreases in self-e?-
cacy due to task complexity itself.
Overall, then, increases in task complexity are
posited to attenuate the positive incentives–e?ort
relation, unless subjects have high self-e?cacy
(which should, at least partially be based on their
actual skill). We expect to observe this attenuation
in most experimental or other short-run settings.
Experimenters typically recruit college student
subjects whose skills are relatively constant; these
subjects do not have the opportunity to acquire
skills during the course of the experiment. In other
short-run settings, people may not have the oppor-
tunity to acquire many skills simply because of time
constraints. Since skill requirements (along with
e?ort requirements) increase as tasks become more
complex (Bonner et al., 2000; Campbell, 1988;
Locke & Latham, 1990; Wood, 1986), having
subjects or employees whose skills are e?ectively
constant decreases the probability people have
requisite skills when complexity increases. In turn,
this means their self-e?cacy should remain low.
24
A second response to increases in task complex-
ity could be to engage in more strategy develop-
ment than would be the case for simple tasks. This
could occur because people recognize that com-
plex tasks require complicated strategies for good
performance (Locke & Latham, 1990). However,
such e?ort directed toward strategy development
could decrease performance in the short run
because people change strategies quite frequently
in order to ?nd an appropriate strategy, thereby
often employing strategies that are not appro-
priate (e.g. Naylor & Clark, 1968; Naylor &
Dickinson, 1969; Naylor & Schenk, 1968). In the
long run, e?ort directed toward strategy develop-
ment could enhance performance because it ulti-
mately creates greater knowledge (e.g. Campbell &
Ilgen, 1976; Creyer et al. 1990; Sprinkle, 2000). On
the other hand, if the rewards from successful
performance are held constant, people may engage
in less strategy development as task complexity
increases because the costs of engaging in strategy
development increase as complexity increases. If
individuals respond to task complexity in this way,
their performance likely will su?er in both the
short run and the long run.
The third possible e?ect of task complexity on
performance occurs because increases in task
complexity lead to increases in skill requirements
in addition to increases in e?ort requirements.
Consequently, in settings like laboratory experi-
ments, subjects are less likely to have the skills
needed for complex tasks than for simple tasks. If
subjects are less likely to have the skills necessary
for good performance in complex tasks, even if
monetary incentives increase current e?ort, then
increases in e?ort may not translate into perfor-
mance increases. Thus, for example, if subjects
have high self-e?cacy because they have di?culty
determining that they actually lack skills for com-
plex tasks (Gist & Mitchell, 1992), incentive-
induced e?ort likely will not have a positive e?ect
on performance. So, the gap between the skills
required by complex tasks and the skills people
have may reduce the e?ects of incentives on per-
formance either by reducing self-e?cacy and, thus,
e?ort, or by reducing the e?ects of e?ort on per-
formance. Because self-e?cacy is a?ected by a
number of factors (Bandura, 1997), it is quite
conceivable that subjects lacking skills for com-
plex tasks could have widely varying levels of self-
e?cacy and e?ort.
Note that the typical prediction of all the the-
ories above is that increases in task complexity will
decrease a positive e?ect of incentives on perfor-
mance, either by attenuating the incentives–e?ort
relation or by attenuating the e?ort–performance
relation. The exception to this prediction is when
both self-e?cacy and actual skill (including exist-
ing strategies) are at a high enough level to overturn
these negative e?ects. Findings from laboratory
studies that use both across-task and within-task
de?nitions of task complexity are consistent with
this typical prediction. For example, in a study
that de?ned complexity across broad categories of
tasks, Bonner et al. (2000) found that the prob-
ability that incentives positively a?ect performance
decreases as complexity increases. The results of
24
We discuss empirical ?ndings after discussing theories
about task complexity–incentives interactions as the studies in
this area have not directly addressed the speci?c mechanisms by
which task complexity interacts with incentives. Thus, their
?ndings are compatible with more than one of the theories
advanced here.
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 321
studies manipulating task complexity within the
context of a single task are similar. Speci?cally,
Glucksberg (1962) manipulated task complexity in
two experiments. In each experiment, incentives
had a positive e?ect on performance in the easy
version of the task but a negative e?ect on per-
formance in the complex version. Similarly, Pel-
ham and Neter (1995) found that subjects with the
easy version of a task performed better with
incentives, while those subjects with the complex
version did not. Finally, Wright and Aboul-Ezz
(1988) found that incentives had a greater positive
e?ect in simple tasks than in more complex tasks.
None of these studies measured skill, self-e?-
cacy, goals, arousal (stress), or any dimensions of
e?ort, although we do know that meta-analyses of
the e?ect of task complexity in the positive rela-
tions between, respectively, self-e?cacy and goal
setting and performance ?nd that the e?ects of
these factors decrease as task complexity increases
(Stajkovic & Luthans, 1998; Wood et al., 1987).
Further, theory and much empirical evidence sug-
gests that the positive e?ects of arousal decrease as
task complexity increases (Eysenck, 1986). Over-
all, then, it appears that incentives are less likely to
positively a?ect performance as task complexity
increases, at least in short-duration studies in
which subjects have little to no experience with the
task and thus lack requisite skills.
The two accounting studies that appear to bear
on the interaction of task complexity and incen-
tives, however, ?nd results that seem contradictory
to those above. We believe that this is because
these studies address variables other than task
complexity. First, Libby and Lipe (1992) exam-
ined the e?ect of incentives on recall and recogni-
tion of internal controls and found that incentives
had a greater e?ect on recall performance than on
recognition performance. They hypothesized that
this occurred because recall is more sensitive to
e?ort than is recognition, but also noted that
recall is a less structured (more complex) task,
implicitly suggesting that incentives have a greater
positive e?ect in more complex tasks. We suggest
that the recall versus recognition manipulation is a
manipulation of the e?ort–sensitivity of the task,
and that it does not speak to task complexity
issues. Both the recall and recognition tasks
employed by Libby and Lipe were relatively sim-
ple memory tasks for which subjects possessed
some prior knowledge (and, to the extent they did
not, incentives had a lessened e?ect). Task com-
plexity and e?ort sensitivity may be related when
subjects do not possess skill or when tasks are far
more sensitive to skill variations than to e?ort var-
iations. In these situations, complex tasks would be
less e?ort sensitive because subjects simply cannot
do the task regardless of the level of e?ort they
exert. When subjects possess skill or when tasks
are sensitive to both e?ort and skill variations (as
in Libby & Lipe), however, the relation of com-
plexity to e?ort sensitivity is not clear.
The second accounting study that could be
interpreted as providing contradictory ?ndings
about the role of task complexity in the incentives-
e?ort or e?ort-performance relations is Ashton
(1990). In this study, auditors (who likely had little
skill with regard to the task—see Heiman, 1990),
provided bond ratings either with or without a
decision aid and with or without incentives. The
aid can be construed to be a manipulation of task
di?culty, which is an element of complexity, with
the presence of the aid making the task more dif-
?cult (Heiman, 1990). Using this interpretation,
the incentive e?ect is consistent with the results
described above—incentives had a positive e?ect
in the simpler version of the task (that without the
aid) and no e?ect in the more di?cult version of
the task (that with the aid). However, this inter-
pretation is problematic because the aid had a
positive e?ect on performance; this would suggest
that the aid made the task less di?cult rather than
more di?cult. An alternative interpretation, dis-
cussed by Heiman (1990), which seems more
likely, is that the aid increased arousal in that it
provided a very di?cult performance goal that
subjects tried to surpass when faced with incen-
tives. This additional arousal (beyond that pro-
vided by incentives) may have led to the failure of
incentives to increase performance with the aid.
We discuss this interpretation further in the envir-
onmental variables section.
While the evidence to date is fairly consistent in
suggesting that task complexity reduces the e?ect
of incentives on performance, we know very little
about how and under what conditions this occurs.
322 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
For example, the studies that have examined this
issue have been very short in duration. It is not
clear what will happen when studies increase in
length because, for example, while subjects may
have the opportunity to acquire skill, task com-
plexity itself may reduce self-e?cacy, so that sub-
jects still are not willing to exert additional e?ort
under incentives. Further, longer-duration settings
common to most accounting-related jobs may lead
to more arousal or stress, so that the combination
of complex tasks and incentives could make per-
formance worse in the long run while subjects
engage in ‘‘too much’’ strategy development.
Additionally, other distinct features of accounting
settings such as the presence of accountability may
create additional arousal or stress (Ashton, 1990).
On the other hand, increases in self-e?cacy
through the acquisition of skill could mitigate
decreases in self-e?cacy from task complexity
and/or arousal/stress due to the task complexity-
incentives combination, so that incentives can
positively in?uence e?ort. Further, distinct fea-
tures of accounting settings such as the review
process also may a?ect key intervening variables
such as self-e?cacy by providing clear perfor-
mance capability information. Additionally, while
task complexity may reduce e?ort, it could in fact
lead to strategies that improve performance for
many individuals. For example, research shows
that inexperienced ?nancial analysts are more
likely to conform to a consensus earnings forecast
than more experienced analysts (Hong, Kubik, &
Solomon, 2000). This ?nding is consistent with
their believing that increased e?ort may not lead
to better performance and simply reducing their
e?ort to make an earnings forecast similar to that
of others. More importantly, research also has
found that this behavior is consistent with self-
interest. Inexperienced analysts who make fore-
casts far from the consensus are more likely to lose
their jobs than more experienced analysts who
make such ‘‘bold’’ forecasts (Hong et al., 2000),
suggesting that at least ?rms view their conformity
as ‘‘better performance.’’
One assertion that seems relatively clear is that,
when considering the e?ects of task complexity on
the incentives–e?ort and e?ort–performance rela-
tions, one also must consider issues related to skill.
Consequently, many of the questions we raised
related to skill and incentives also are pertinent
here. In particular, when designing reward systems
for employees who perform complex tasks, orga-
nizations need to consider whether rewards linked
to performance in the short run are wise if they
wish to encourage learning over time. This may be
a particular problem in situations where account-
ing data are used to measure performance.
Another question that arises in this area is how
e?ort-sensitivity and task complexity di?er, and
how these di?erences a?ect predictions about the
e?ects of tasks on the incentives–performance
relation. Tasks in accounting vary both as to e?ort
sensitivity and task complexity, so further work
explicating these constructs could be very useful.
Again, as is the case with many other variables, we
know virtually nothing about the processes by
which task complexity can a?ect the incentives–
performance relation. Understanding these pro-
cesses is critical to suggesting solutions for poor
performance. For example, mechanisms that tend
to increase self-e?cacy such as persuasion could
be used to o?set the negative e?ects of task com-
plexity on self-e?cacy (if this is indeed the
mechanism by which task complexity operates to
reduce the e?ectiveness of incentives).
Another interesting issue to pursue is the idea
that task complexity may actually serve as an
incentive itself in some accounting-related ?elds.
Das et al. (1998) argue that, because companies
with low earnings predictability are di?cult to
make earnings forecasts for, analysts have an
incentive to make optimistic forecasts (forecasts
that are too high) in order to curry favor with
management. In turn, management will provide
them with more information than analysts issuing
less optimistic forecasts and their forecasts will be
more accurate (despite the bias). The incentive
argument derives from the following assumptions:
(1) analysts’ performance is at least partially
judged on their accuracy, (2) getting more infor-
mation from management will lead to greater
accuracy, and (3) management prefers optimistic
forecasts. While there is some evidence to support
the ?rst assumption (Mikhail, Walther, & Willis,
1999), there is little evidence supporting the second.
Further, there is some evidence that goes contrary
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 323
to the third (Brown, 1999). Nevertheless, the idea
that task complexity can serve as an incentive in
this situation deserves further attention.
3.2.2. E?ects of other task variables on the
incentives–e?ort–performance relation
Task complexity, when broadly de?ned, sub-
sumes two key dimensions of accounting infor-
mation: amount and clarity. Another dimension
on which accounting information can vary dra-
matically is the manner in which it is presented,
including the framing of items. This issue is impor-
tant in many accounting contexts. For example, one
of the principal tasks of the Financial Accounting
Standards Board is to determine the presentation
of ?nancial information. In management account-
ing and auditing, there are several situations in
which information can be framed di?erently such
as variance investigation (e.g. Lipe, 1993) or
going-concern judgments (Kida, 1984).
While we are not aware of any studies examin-
ing the interaction of incentives and presentation
format or framing, the literature on preference
reversals o?ers some clues as to predictions (see
Thaler, 1992 for a review). Preference reversals
occur when people’s expressed preferences for one
item of two (typically a hypothetical monetary
gamble) reverse when the problem is presented in
a di?erent way. Prospect theory (Kahenman &
Tversky, 1979) suggests that these reversals and
other similar phenomena re?ect the fact that peo-
ple think about ?nancial outcomes in terms of
gains or losses vis-a` -vis a reference point as
opposed to ultimate wealth positions. In turn,
problems that elicit thinking in terms of gains
versus losses lead to di?erent responses. This way
of thinking appears to be relatively ‘‘hardwired.’’
Consequently, we would expect that framing or
presentation of information could attenuate the
positive relation between incentives-induced e?ort
and performance, speci?cally because while people
may try harder, their way of thinking likely will
not change. Consistent with this, attempts to
reduce preference reversals with incentives have
not been e?ective (e.g. Grether & Plott, 1979).
Other presentation format or framing e?ects
may occur because, similarly, they tap into rela-
tively ‘‘hardwired’’ (automatic) cognitive pro-
cesses. For example, Hopkins (1996) ?nds
presentation format e?ects that are due to di?er-
ential categorization of accounting information.
Categorization is one of the most fundamental
cognitive processes and people appear to be natu-
rally inclined to categorize items in certain ways,
such as by cause rather than e?ect (Lien & Cheng,
1990; Nelson et al., 1995). However, many cate-
gories in accounting settings are arti?cial and are
learned through training and experience. This
suggests that, over the long run, incentives might
be e?ective in eliminating task performance pro-
blems related to framing or presentation format
variation. The key issue here is determining which
cognitive process the presentation format taps into
and to what extent this cognitive process is amen-
able to changes that can come about through
incentive-induced e?ort.
Finally, tasks of interest to accounting
researchers and organizations naturally vary, for
example, from more aversive factory work to
more interesting strategic cost management or
resource allocation decisions. Deci and his collea-
gues (Deci, Betley, Kahle, Abrams, & Porac, 1981;
Deci & Ryan, 1985), as well as others (e.g. Kohn,
1993), have proposed that ?nancial incentives can
harm performance in tasks that are inherently
attractive (or interesting) rather than enhancing
performance as would likely be the case with
aversive (or boring) tasks. This hypothesis is based
on the notion that extrinsic sources of motivation
such as ?nancial incentives rob subjects of the
intrinsic motivation they initially have for these
tasks. Since intrinsically motivated behavior is
posited to result in more creativity and ?exibility
in decision-making than extrinsically motivated
behavior, extrinsic incentives may actually
degrade e?ort and performance.
Most of the studies in this paradigm have
focused on changes in intrinsic motivation by
examining the time subjects spend on the task
during a ‘‘free choice’’ period, which occurs after
subjects have performed the task under incentives.
This behavior is then compared to their pre-
incentives behavior or to the behavior of a control
group with no incentives. While some studies
document negative e?ects of incentives on intrinsic
motivation, other studies show opposite results
324 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
(e.g. Farr, 1976; Scott, Farr, & Podsako?, 1988;
Wimperis & Farr, 1979). More importantly, recent
reviews of this literature indicate that incentives
do not have di?erential e?ects on task perfor-
mance when the task is interesting versus boring
(Cameron & Pierce, 1994; Jenkins et al., 1998;
Rummel & Feinberg, 1988; Tang & Hall, 1995;
Wiersma, 1992). That said, recent research in
accounting suggests that monetary incentives may
reduce intrinsic motivation (e?ort) and perfor-
mance on tasks viewed as attractive and that
require some level of creativity or innovation
(Fessler, 2000). More research is clearly needed,
though, to sort out the relations among incentives,
task attractiveness, intrinsic motivation, e?ort,
and performance.
To summarize, there are many task variables
that could interact with monetary incentives to
a?ect e?ort and task performance. We concentrate
on task complexity because it is a key variable
related to performance in accounting-related
tasks, and it appears to decrease the e?ectiveness
of incentives. Moreover, task complexity can play
multiple roles in interacting with monetary incen-
tives, although very few studies have addressed
these roles speci?cally. Further, the speci?c
mechanisms by which task complexity interacts
with incentives and under what circumstances this
occurs remain largely unexplored.
3.3. Environmental variables
Environmental variables include all the condi-
tions, circumstances, and in?uences surrounding a
person who is performing a particular task (Bon-
ner, 1999). These variables include factors such as
time pressure, accountability relationships, assigned
goals, and feedback. A ?rm’s accounting system
also can be viewed as an environmental variable
and, to this end, much research in accounting
focuses on whether and how the environmental
variables associated with accounting settings a?ect
task performance. For example, accounting
researchers have examined, either in isolation or in
conjunction with other person, task and environ-
mental variables, how factors such as time pressure
(McDaniel, 1990; Spilker, 1995), accountability
(Kennedy, 1993, 1995; Peecher, 1996), assigned
goals (Chow, 1983; Hirst & Yetton, 1999), fea-
tures of the regulatory environment (Hronsky &
Houghton, 2001), and feedback (Briers, Chow,
Hwang, & Luckett, 1999; Frederickson et al.,
1999; Jermias, 2001) a?ect task performance.
Monetary incentives not only are an important
part of management control systems but also can
be viewed as an environmental variable. Thus,
their e?cacy in motivating e?ort and performance
has been studied extensively in accounting and
other disciplines. However, monetary incentives
are only one of many environmental variables that
may enhance (or detract from) motivation and
performance. Thus, it is important to examine
whether there are dependencies among these vari-
ables, and whether salient accounting-related
environmental variables serve as complements to
(or substitutes for) incentive compensation.
There are numerous environmental variables
that may interact with monetary incentives in
a?ecting task performance. Similar to the person
and task sections, we primarily devote our atten-
tion to environmental variables that are important
in accounting settings and that have been studied
in combination with monetary incentives. In par-
ticular, we focus on assigned goals. Goals are
important to accountants because, similar to
incentive compensation, they are thought to be an
important element of an organization’s control
system (Merchant, 1998). Speci?cally, organiza-
tions continuously develop and revise perfor-
mance targets and employ both short-term and
long-term goals to reach these targets (Locke &
Latham, 1990; Merchant, 1998; Shields, 2001).
For example, many ?rms develop ?nancial budgets
that contain explicit return-on-investment goals,
sales-revenue goals, and production-cost goals.
The goals (standards) contained in these budgets
frequently are used as benchmarks in evaluating
the performance of, and therefore to motivate,
employees (Merchant, 1998; Shields, 2001).
3.3.1. E?ects of assigned goals on the incentives–
e?ort–performance relation
Several studies have manipulated assigned goals
alone or in conjunction with monetary incentives.
Similar to monetary incentives, assigned goals and
performance targets are thought to positively
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 325
in?uence e?ort direction, e?ort duration, and
e?ort intensity and, as a result, improve perfor-
mance (Earley & Lituchi, 1991; Locke & Latham,
1990; Meyer, Schacht-Cole, & Gellatly, 1988).
Assigned goals have been shown to positively
in?uence e?ort through two mechanisms. First,
assigned goals a?ect the level of personal (self-set)
goals, which in turn, positively in?uences the var-
ious dimensions of e?ort. Second, assigned goals
positively a?ect self-e?cacy and, in turn, self-e?-
cacy has (as discussed earlier) a positive in?uence
on both personal goal levels and the various
dimensions of e?ort.
25
Locke and Latham (1990)
note that these e?ects of assigned goals on e?ort
direction, duration, and intensity are relatively
automatic once individuals accept goals. They
further suggest that, under some circumstances,
goals can positively a?ect e?ort directed toward
strategy development.
Moreover, in contrast to predictions from neo-
classical economic (agency) theory that goals per
se do not a?ect e?ort and performance, there are a
large number of empirical studies and meta-ana-
lyses indicating two key ?ndings about the e?ects
of goals (Latham & Locke, 1991; Locke &
Latham, 1990; Tubbs, 1986). First, the level of
di?culty of the assigned goal is positively related
to performance, until goals become excessively
di?cult, at which point performance levels o?.
The explanation for this goal-di?culty e?ect is
that goal levels are positively correlated with e?ort
intensity and e?ort duration (e?ort direction is
?xed once a goal has been accepted). Here, Locke
and Latham (1990) note that more di?cult goals
may set a higher standard for people’s satisfaction
with their performance; this standard may con-
tribute to or explain the e?ects of di?cult goals on
e?ort intensity and duration. Second, speci?c
(quantitative) di?cult goals lead to higher e?ort
and performance than vague di?cult goals such as
‘‘do your best’’ or than no assigned goals, which
often are assumed to be implicit ‘‘do your best’’
goals.
For this latter e?ect, Locke and Latham (1990)
o?er two explanations. First, speci?c goals likely
positively in?uence e?ort direction. Second, they
posit that there is substantial variation in the per-
formance levels at which people will be satis?ed
with their performance under ‘‘do your best’’
goals versus speci?c, di?cult goals. Consequently,
there also will be substantial variation in e?ort
duration or intensity among individuals with ‘‘do
your best’’ goals, and there will be lower variation
in e?ort among people with speci?c, di?cult goals
who are attempting to attain a high level of per-
formance. Increased variation among the ‘‘do
your best’’ group implies a lower mean of e?ort in
this group because the e?ort level of people with
speci?c, di?cult goals is uniformly high. Con-
sistent with this, many studies have shown that the
e?ects of speci?c, challenging goals lead to greater
e?ort duration and/or e?ort intensity and greater
performance than ‘‘do your best’’ goals (Locke &
Latham, 1990; Tubbs, 1986).
Like assigned goals, monetary incentives can
a?ect e?ort direction, duration, and intensity, as
well as strategy development. In other words, both
monetary incentives and goals are motivational
techniques. A simple hypothesis about the joint
e?ects of monetary incentives and goals, then,
would be that assigned goals have additive posi-
tive e?ects on e?ort and performance over mone-
tary incentives and not interactive e?ects (e.g.
Locke, Shaw, Saari, & Latham, 1981).
Empirical results tend to support this hypoth-
esis—numerous studies show that monetary
incentives and assigned goals generally have addi-
tive e?ects on performance (e.g. Campbell, 1984;
Latham, Mitchell, & Dossett, 1978; Locke, Bryan,
& Kendall, 1968; London & Oldham, 1976;
Pritchard & Curtis, 1973; Terborg & Miller, 1978).
In other words, on average, assigned goals and
monetary incentives have independent, positive
e?ects on performance. The broad implication of
this research for accounting is that goals and
incentives are not substitutes, therefore suggesting
that organizations should employ performance
targets in conjunction with ?nancial incentives to
best motivate their employees.
25
The positive e?ect of assigned goals on self-e?cacy occurs
because assigned goals (vis-a` -vis no goals) provide normative
information about the level of performance people can be
expected to reach in a particular setting. This normative infor-
mation positively a?ects self-e?cacy (Locke & Latham, 1990;
Meyer & Gellatly, 1988).
326 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
The ?ndings of recent research, however, raise
an important question regarding a possible inter-
action between goals and incentives. Speci?cally,
what are the e?ects on e?ort and performance
when monetary incentives are linked to goal
(standard) attainment versus when incentives and
goals are kept as separate motivating mechanisms?
Some incentive schemes, such as budget-based
(quota) schemes, explicitly include assigned goals
and link compensation to achieving these goals.
Such compensation schemes typically pay indivi-
duals a ?at wage up to some targeted level of per-
formance, then either a bonus for reaching the
target or a piece rate for each additional unit of
output above the target. Alternatively, other
incentive schemes, such as piece-rate or pro?t-
sharing schemes, need not (and frequently do not)
include either an explicit or implicit goal because
compensation is linked to each unit of output.
However, independent performance goals can be
assigned to employees when they are working
under these types of schemes.
A recent review of incentives experiments found
that incentive schemes that include an explicit
assigned goal tend to lead to higher performance
than incentive schemes that do not include an
explicit goal (Bonner et al., 2000). This result may
occur because budget-based (quota) incentive
schemes include both an explicit goal and an expli-
cit link between pay and performance, whereas
other incentive schemes usually only include the
pay-for-performance link. What this review does
not tell us, however, is whether incentive schemes
that embed goals (e.g. budget-based schemes) are
superior to situations in which employees are pro-
vided with explicit goals and performance-con-
tingent incentives that are not explicitly linked to
achieving these goals. Further, this review does not
address the dimensions of assigned goals that may
create or alter any observed interactions between
goals and incentives, such as their level of di?culty.
Finally, it does not address the cognitive and moti-
vational mechanisms by which this result occurs.
A few empirical studies have attempted to
address these issues. For example, Fatseas and
Hirst (1992) and Lee et al. (1997) found that sub-
jects working under a quota scheme performed
better than subjects working under piece-rate or
?at-rate schemes when goals were easy or moder-
ate.
26
However, when the goal became very di?-
cult, both Fatseas and Hirst (1992) and Lee et al.
(1997) found that the quota scheme resulted in
signi?cantly lower performance than piece-rate or
?at-rate schemes. Lee et al. (1997) found that these
results were accounted for by personal goals and
self-e?cacy, but not by goal commitment.
Other studies have examined additional issues
related to a possible goal-monetary incentives
interaction. For example, Wright (1992) found a
three-way interaction among incentive schemes,
level of pay, and goals. In particular, subjects with
the high level of the quota scheme outperformed
subjects with the high piece-rate scheme at both
easy and moderate goals; piece-rate subjects did
better at the di?cult goal level. Subjects with the
low quota scheme, however, performed worse
than low piece-rate subjects at all goal levels. This
study also found that the results were partially
explained by goal commitment, and other similar
studies have found that incentives tied to goals
(quota schemes) can result in lower personal goals
and lower self-e?cacy than do incentives not tied
to goals (piece-rate schemes) (Wright, 1989;
Wright & Kacmar, 1995).
Ashton (1990) examined the interaction between
a decision aid that implied a di?cult goal and
monetary incentives. When subjects did not have
the decision aid, performance with incentives was
26
In the goal-setting literature, goal levels frequently are
de?ned by reference to a pilot test of similar subjects. For
example, in Fatseas and Hirst (1992) the ‘‘low’’, ‘‘moderate’’,
and ‘‘di?cult’’ goals were set at a level such that, a priori, sub-
jects had an 80% (20th percentile on pilot test), 50% (50th
percentile on pilot test), or 20% (80th percentile on pilot test)
chance, respectively, of achieving them. Goal levels also have
been de?ned by reference to each individual subject’s perfor-
mance in a practice session. For example, in Erez et al. (1990)
the ‘‘easy’’, ‘‘moderate’’, and ‘‘di?cult’’ goal levels were set by
having each subject perform at the 20th, 50th, and 80th per-
centiles of their own distribution of performance in the practice
session, respectively. Moreover, a typical (but by no means
widely accepted) de?nition of low, moderate, and di?cult goals
seems to be around the 20th, 50th, and 80th percentile of per-
formance. Additionally, when linking compensation to goal
attainment, as under a quota scheme, these studies do not
increase the level of rewards as goal di?culty increases. Thus,
raising the goal reduces the expected payo? associated with a
given level of e?ort under a quota scheme.
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 327
better than performance without incentives; when
they had the decision aid with the di?cult goal,
subjects with incentives performed no better than
those without incentives. Erez, Gopher, and Arzi
(1990) examined the interaction between the pre-
sence of a goal-based incentive and whether goals
were assigned or self-selected. Results indicated an
interaction due to subjects with incentives per-
forming the same when they had assigned or self-
set goals, but subjects without incentives perform-
ing better with self-set goals. This interaction
occurred only at the moderate goal level, however.
At the easy goal level, neither incentives nor
assignment of goals had an e?ect on performance.
At the high goal level, these factors had main
e?ects, but no interactive e?ects.
Overall, it appears that while early studies indi-
cated no interaction between goals and incentives,
more recent studies that examine various dimen-
sions of goals and/or incentives ?nd such interac-
tions. The strongest evidence suggests an
interaction between goal di?culty (when goals are
speci?c) and the type of performance-contingent
incentive scheme. Speci?cally, while performance
tends to increase as speci?c goals increase under
piece-rate schemes, performance initially increases,
but then decreases, as goals increase under quota
schemes. This interaction typically results in per-
formance that is better under piece-rate schemes
than quota schemes when goals are very di?cult,
but performance that is better under quota
schemes (than piece-rate schemes) when goals are
moderate. There also may be an interaction
between goal level and incentive magnitude, as
well as a three-way interaction involving these
factors and type of scheme. Finally, there may be
interactions that involve whether goals are
assigned or self-set. Evidence on the mechanisms
by which these interactions a?ect performance is
limited to personal goals, goal commitment, and
self-e?cacy, and these results are mixed. Further,
none of the studies has examined the various
dimensions of e?ort.
The results described above are quite limited as
to the dimensions of goals and incentives exam-
ined and, further, the ?ndings are not uniform.
Consequently, drawing conclusions about the best
combination of goals and incentives would be
premature. For example, Lee et al. (1997) note
that concluding that the best incentive system
would be a quota scheme embedded with moder-
ate goals is dangerous from the standpoint that it
is di?cult to maintain goals as ‘‘moderate’’ over
time due to learning and other factors.
Future research directed toward examining how
goals and incentives should be employed in tan-
dem to achieve maximal e?ort and performance
from individuals could provide several useful
insights on a number of issues of practical and
theoretical importance to accounting. First, such
research could inform us of whether compensation
should be tied to meeting performance targets or
whether goals and incentives should be indepen-
dent. Second, research could inform us about
important nonlinearities in such goal-incentive
combinations (such as those related to goal level
and pay level) and, thus, where the bene?ts accru-
ing to these combinations start, stop, diminish, or
change sign. As discussed by Luft and Shields
(2001a), there can be important theoretical and
practical reasons for examining whether nonlinear
relations exist. Such research also is valuable
because the exact level of goal di?culty that max-
imizes performance is ambiguous (Hirst & Yetton,
1999; Shields, 2001). Along with this, prior
research has not examined possible interactions
among incentives and non-speci?c goals, such as
‘‘do your best’’ goals. Because of the possible
problems with keeping speci?c goals at certain
levels of di?culty, such research could have sub-
stantial practical signi?cance.
Finally, research could provide more insight
into the mechanisms by which incentives and goals
combine to in?uence performance since the results
to date focus mostly on their e?ects on personal
goals and goal commitment. For example, if the
combination of di?cult assigned goals and quota
schemes leads to lowered personal goals, does this
translate into people ‘‘giving up’’ (e.g. decreasing
e?ort intensity, direction, or duration)? Alter-
natively, do people who have di?cult assigned
goals and incentives tied to those goals maintain
high personal goals and goal commitment, thus
not ‘‘giving up,’’ but instead engage in ine?ective
strategy development under the high pressure to
perform well? It is important to understand these
328 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
mediators of goal-incentive interactions because
there may be very simple solutions to problems
caused by particular combinations. For example,
some research indicates that incentives can
increase goal commitment (see Locke & Latham,
1990 for a review), so there may be some level of
incentives at which the negative e?ects of di?cult
goals on personal goals (and self-e?cacy and
e?ort) can be eliminated.
3.3.2. E?ects of other environmental variables on
the incentives–e?ort–performance relation
As previously discussed, there are numerous
other environmental variables that can in?uence
performance in accounting-related tasks. For
example, feedback (information provided to a per-
son regarding some aspect of his or her task per-
formance) is an integral component of accounting
because a fundamental role of accounting infor-
mation is to facilitate individual and organiza-
tional learning (Atkinson et al., 2001; Sprinkle,
2000, 2001). Further, accounting methods can
drastically alter both the amount and type of
feedback users of accounting information receive
and, as a result, feedback can vary greatly across
accounting settings (e.g. Luft & Shields, 2001b).
While monetary incentives and assigned goals
theoretically are posited to have only motivational
e?ects, feedback has been posited to have both
cognitive e?ects (learning) and motivational e?ects
(e.g. Kessler & Ashton, 1981; Nelson, 1993). Speci-
?cally, in their meta-analysis, Kluger and DeNisi
(1996) note that feedback has been shown to
positively a?ect motivation and learning as well as
other factors such as self-e?cacy. The most typical
predicted e?ect of feedback is that it tends to
increase the various dimensions of e?ort, the ability
to learn and, consequently, enhances performance
(e.g. Ashford & Cummings, 1983; Pritchard, Jones,
Roth, Stuebing, & Ekeberg, 1988). In many cases,
however, feedback has no apparent e?ect on per-
formance (e.g. Kluger & DeNisi, 1996; Locke, 1967)
and, in some cases, even debilitates performance
(e.g. Jacoby, Mazursky, Troutman, & Kuss, 1984;
Kluger & DeNisi, 1996). Moreover, the motiva-
tional and cognitive e?ects of feedback can interact.
Thus, it is unclear whether feedback has additive or
interactive e?ects with monetary incentives.
Most empirical evidence tends to show that
feedback does not interact with incentives in
a?ecting task performance (e.g. Arkes, Dawes, &
Christensen, 1986; Chung & Vickery, 1976;
Hogarth, Gibbs, McKenzie, & Marquis, 1991;
Montague & Webber, 1965; Phillips & Lord, 1980;
Sipowicz, Ware, & Baker, 1962; Weiner, 1966;
Wiener, 1969; Weiner & Mander, 1978). Further,
Kluger and DeNisi’s (1996) meta-analysis found
that monetary incentives do not moderate the
e?ect of feedback. In short, prior research suggests
that incentive and feedback e?ects, when they
exist, typically are independent and additive so
that there is no simple two-way interaction.
In these prior studies, however, feedback was
automatically provided to experimental partici-
pants and, thus, there was no cost to acquiring
feedback. Recent research in accounting shows
that monetary incentives can motivate individuals
to both acquire and use feedback to improve long-
run task performance (Sprinkle, 2000). One impli-
cation of this research for accounting is that
monetary incentives and feedback can be comple-
ments. Organizations not only need to provide
information that is valuable for decision making,
but also need to employ monetary incentives to
ensure that people actually use this information
(feedback) to enhance learning and improve orga-
nizational performance.
Much future research is needed, though, to fully
understand how feedback and monetary incentives
combine to a?ect e?ort and performance. For
example, there are various types of feedback in
accounting settings, including outcome feedback,
cognitive feedback, and task-properties feedback
(Kessler & Ashton, 1981). Further, such feedback
can have di?erential e?ects on individuals’ abil-
ities to learn and, presumably, their motivation/
e?ort (Balzer, Doherty, & O’Connor, 1989; Bon-
ner & Pennington, 1991; Bonner & Walker, 1994;
Kluger & DeNisi, 1996). To date, though, studies
examining the e?ects of monetary incentives and
feedback have employed only outcome feedback
and, consequently, we know little about how cog-
nitive feedback or task-properties feedback a?ect
the relations between incentives and e?ort and
e?ort and performance. Additionally, since feed-
back can increase the degree of information
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 329
asymmetry between employees and employers, it
may provide a means for employees to shirk more
e?ectively (Baiman & Sivaramakrishnan, 1991)
and monetary incentives can exacerbate this moti-
vation. Moreover, since a primary goal of
accounting is to provide information (feedback)
for decision-making, it is important to understand
when and how incentives motivate individuals to
use feedback from accounting systems to the ben-
e?t or detriment of the organization.
In addition to assigned goals and feedback,
there are other important accounting-related envir-
onmental variables that should be studied in con-
junction with performance-contingent monetary
incentives. Such variables include training, account-
ability, the assignment of decision rights, and time
pressure. For example, training typically is intended
to increase the skill level of individuals (Anderson,
1995), and much of accounting education and
instruction is directed toward increasing the skill level
of individuals or helping individuals better deploy
their existing skill (Bonner & Pennington, 1991;
Bonner & Walker, 1994). If training enhances skill
(and skill and e?ort are complements to some extent),
then we might expect that training would interact
with incentives in the same manner that skill is pro-
posed to interact with incentives. That is, the posi-
tive e?ect of monetary incentives on performance
would increase as skill-related training increases.
Compared to goals and feedback, far fewer
studies have examined the combined e?ects of
training and performance-contingent monetary
incentives. Further, existing empirical ?ndings (e.g.
Arkes et al., 1986; Baker & Kirsch, 1991; Giger-
enzer, Hofrage, & Kleinbolting, 1991) almost uni-
formly show that training does not interact with
incentives to a?ect task performance.
27
In our
view, though, the prior research in this area does
not provide a good test of whether training inter-
acts with incentives. In particular, the studies by
Arkes et al. (1986) and Gigerenzer et al. (1991)
provide no evidence that the training actually
enhanced skill. Further, the task employed by
Baker and Kirsch (1991), immersing one’s hand in
ice water for as long as possible, appears to be far
more sensitive to e?ort than to skill, so that train-
ing likely did not have much of an e?ect on per-
formance (Baker & Kirsch, 1991, p. 508). In short,
we feel that much additional research is needed to
understand how training and monetary incentives
combine to a?ect performance, especially since
training is thought to be a signi?cant determinant
of performance in many accounting-related tasks
(Libby & Luft, 1993).
There are several other key environmental vari-
ables such as accountability and the assignment of
decision rights that have not been studied in con-
junction with ?nancial incentives. This is unfortu-
nate since, similar to performance-contingent
monetary incentives, these variables can be viewed
as motivating mechanisms, and it is unclear whe-
ther they will have interactive or additive e?ects
with incentives (see also Pelham & Neter, 1995 for
other variables). Future research is clearly needed
to document the exact nature of these relations as
well as the underlying motivational and cognitive
processes governing these relations. Given that
signi?cant dependencies may exist among
accounting-related environmental variables, it is
important to understand how they combine to
a?ect task performance (Libby & Luft, 1993).
Finally, as articulated by Libby and Luft (1993),
often the key to understanding these dependencies
is to understand the mechanisms that determine
task performance.
3.4. Incentive scheme variables
In this section, we discuss how various dimen-
sions of the incentive scheme per se may a?ect the
relations between (the presence of) monetary
incentives and e?ort and e?ort and task perfor-
mance. Incentive scheme variables include, for
example, the timing of the incentive, whether it
embodies competition, what dimension(s) of
27
One could consider the results from binary outcome pre-
diction studies by Siegel (1961) and Tversky and Edwards
(1966) to be the exceptions. Speci?cally, in a task where sub-
jects are asked to predict which one of two lights will illuminate
over a series of hundreds of trials, subjects receiving perfor-
mance–contingent incentives are, once trained regarding the
Bernoulli process which govern the lights, more likely to
employ the optimal rule of choosing the signal with a higher
probability more frequently rather than perform probability
matching. Thus, subjects receiving incentives are more likely to
use a decision rule to enhance performance than subjects not
receiving incentives.
330 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
performance the incentive rewards, and payo?
magnitude. Incentive scheme variables also
include whether the incentive contract is assigned
or self-selected and whether the incentive contract
incorporates assigned goals. These latter elements
were discussed previously under the person vari-
ables and environmental variables sections,
respectively.
Several studies in accounting have focused on
whether and how dimensions of the incentive
scheme itself a?ect task performance and the
underlying cognitive and motivational factors by
which these dimensions determine task perfor-
mance. For example, researchers in accounting
have examined the e?ects on performance of: (1)
the type of incentive scheme (Bonner et al., 2000;
Frederickson, 1992), (2) the timing of the incentive
(Libby & Lipe, 1992), (3) the framing of the
incentive contract in bonus or penalty terms (Luft,
1994), (4) the magnitude of the incentive (Hannan,
2001), and (5) the assignment versus self-selection
of the incentive scheme (Chow, 1983).
Such research is important in accounting
because accountants not only play a major role in
designing compensation plans but also in deter-
mining the speci?c attributes of these plans (e.g.
Atkinson et al., 2001; Indjejikian, 1999). Thus,
accounting research directed toward under-
standing how properties of incentive schemes
a?ect e?ort and performance can help uncover the
characteristics of incentive schemes that best align
the employees’ interests with those of the organi-
zation and, therefore, can help determine the most
e?ective compensation arrangements. Further, in
designing e?ective incentive schemes, it is bene-
?cial to understand the relations among incentive
scheme variables and the cognitive and motiva-
tional processes that a?ect task performance. For
example, research may show that while a parti-
cular incentive scheme increases e?ort duration
and task performance, the cost to the ?rm asso-
ciated with achieving this e?ort increase exceeds
the bene?t from improved performance (also see
Luft, 1994).
There are numerous incentive scheme variables
that could be studied. Similar to prior sections, we
restrict our primary focus to one of these variables
and discuss others brie?y. Speci?cally, we discuss
the performance dimension that the incentive
scheme rewards. We do so because employees
usually perform several di?erent tasks as part of
their jobs or a single task with several dimensions
of performance (e.g. Feltham & Xie, 1994; Hem-
mer, 1996; Holmstrom & Milgrom, 1991). For
instance, production employees frequently are
responsible for both the quantity and quality of
output. In such settings, a fundamental role of
accounting is to design a set of performance mea-
sures that best re?ect ?rm value and employees’
contributions to ?rm value. The accounting per-
formance measurement and reward system also
needs to consider how the measured dimensions of
performance will objectively (or subjectively) be
used in evaluating employees and, thus, whether
an employee’s ?nancial compensation will be
linked to each performance measure. Finally, the
compensation weights assigned to each perfor-
mance measure need to be speci?ed.
3.4.1. E?ects of the rewarded dimension of
performance on the incentives–e?ort–performance
relation
The potential role of monetary incentives in a
situation where multiple dimensions of perfor-
mance (or multiple tasks) exist is twofold. First,
similar to a one-dimensional setting, incentives
presumably serve a role in motivating high levels
of e?ort (Holmstrom & Milgrom, 1991; Merchant,
1998). Second, incentives are posited to serve an
informational role and, thus, are thought to be
important in directing employees’ e?ort toward
their various responsibilities (Holmstrom & Mil-
grom, 1991; Merchant, 1998). In both roles, eco-
nomic theory informs us that appropriate
incentives need to be provided on each task or
dimension thereof in order to induce employees to
optimally allocate their e?ort among their various
responsibilities (e.g. Prendergast, 1999).
It frequently is very di?cult, however, to
measure all dimensions of performance with equal
precision and, consequently, theory suggests that,
given employees’ aversion to risk, it can become
exceedingly costly for ?rms to achieve the desired
allocation of e?ort using ?nancial incentives.
Ceteris paribus, as the di?culty of measuring per-
formance on any one activity increases, economic
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 331
theory indicates that the desirability of providing
?nancial incentives decreases, so much so that
some have posited that a ?at-wage contract may
be optimal in multidimensional task situations
(Holmstrom & Milgrom, 1991). Theoretically, this
occurs for two reasons. First, linking monetary
incentives to one dimension of performance but
not the other(s) severely attenuates the incentives–
e?ort relation on the unrewarded dimension and,
thus, reduces overall task performance. Second,
individuals derive some utility from work activities
and, thus, even in the absence of performance-
contingent incentives, will exert non-trivial e?ort
on tasks. Further, since pay is not contingent on
performance, employees will allocate their e?orts
according to the ?rm’s wishes.
This discussion raises an important question
regarding the provision of incentives in multi-
dimensional tasks and the extent to which extrin-
sic incentives can lead to an e?cient allocation of
e?ort among an employee’s various responsi-
bilities. Along these lines, several prior studies
have examined whether incentives tied to one
dimension of performance (usually quantity of
output) a?ect another dimension of performance,
such as the quality of output or the learning of
incidental material. The results of two of these
studies (Bahrick, Fitts, & Rankin, 1952; McNa-
mara & Fisch, 1964) show that incentives can have
a negative e?ect on the performance of dimensions
that are not rewarded.
28
However, the results of
most of these studies indicate that incentives have
no e?ect on the performance of dimensions that
are not rewarded (Dornbush, 1965; Hamner &
Foster, 1975; Kausler & Trapp, 1962; Riedel et al.,
1988; Terborg & Miller, 1978; Wimperis & Farr,
1979), and we are not aware of any empirical
studies reporting that incentives have a positive
e?ect on the performance of dimensions that are
not rewarded. Further, almost all of these studies
do not measure the e?ect incentives have on e?ort
direction or e?ort duration regarding the unre-
warded dimension of performance. Finally, most
of these studies show that incentives do enhance
the rewarded dimension of performance (output
quantity). These general conclusions are consistent
with those of Jenkins et al. (1998), who found that
incentives have an average positive e?ect on per-
formance quantity, but no e?ect on performance
quality, when only performance quantity is
rewarded.
The implication of this research for accounting
is that it appears to be desirable to link ?nancial
incentives to one dimension of performance even
if other important dimensions of performance
cannot be measured or contracted on. For exam-
ple, if ?rms can e?ectively measure some impor-
tant dimensions of performance, such as return on
investment, but not e?ectively measure other
important dimensions of performance, such as
customer satisfaction, then it still seems preferable
to at least contract on a limited dimension of per-
formance rather than no dimension at all. In other
words, it does not appear that incentives solely
tied to one dimension of performance lead to an
ine?cient reallocation of e?ort. Thus, it does not
appear that incentives should be muted in multi-
dimensional tasks.
Given the prior research that has been con-
ducted in this area, though, these conclusions are
tentative and there are a number of directions that
future research might take. Most prior research in
this area tends to use output quantity and output
quality as the two dimensions of performance.
However, since subjects typically are rewarded for
each unit of good output, quantity and quality
clearly are linked as any e?ort expended toward
producing output of less than acceptable quality
(but higher quantity) will not be rewarded. Thus,
subjects receiving incentives have a motivation to
maximize output quantity, but only conditioned
on each unit passing the quality threshold.
Future research should conduct a much stronger
test of the Holmstrom and Milgrom (1991) pro-
position that incentives can lead to an inappropri-
ate allocation of e?ort and a reduction in overall
performance and ?rm welfare. In particular,
accounting researchers could employ two separate
and distinct tasks, whereby subjects need to expli-
28
Also see Schwartz (1988), who ?nds that incentives can
have negative transfer (carryover) e?ects when individuals
switch tasks. Here, monetary incentives reinforce a ‘‘mental
set’’ and a pattern of repetition that is di?cult to break when
individuals are asked to perform a new task that is substantially
di?erent (in terms of the requirements and strategies necessary
for good performance) from the previously rewarded task.
332 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
citly decide which task to work on and how much
e?ort to devote to each task. Performance-con-
tingent monetary incentives could then be pro-
vided on one task but not the other, and this could
be compared to subjects’ performance under a
pure ?xed-wage contract. By performing such a
study, researchers could clearly measure the e?ort
direction and e?ort duration expended toward
each task as well as the performance on each task.
In doing so, though, researchers should be careful
to equate the expected payo? between compensa-
tion conditions and hold the total time available
constant so that appropriate conclusions can be
drawn regarding both the e?ectiveness and the
e?ciency of each contract.
At a more fundamental level, the multi-
dimensional task contracting problem frequently
reduces to motivating employees to innovate and
take risks (Holmstrom, 1989). Speci?cally, man-
agers can be exposed to both compensation risk
and human capital risk when the various dimen-
sions of performance are not equally sensitive to
their e?ort (Milgrom & Roberts, 1992). Addition-
ally, even when the dimensions of performance are
equally sensitive to e?ort, managers frequently
must select from a menu of projects that vary
greatly in both risk and expected return. For
example, managers frequently engage in capital
budgeting decisions in which they evaluate and
select among investments that di?er in the timing,
magnitude, and riskiness of cash ?ows. In these
situations, the accounting performance measure-
ment and reward system not only needs to moti-
vate high levels of e?ort from employees, but also
needs to encourage the appropriate level of risk
taking (i.e. encourage employees to maximize
expected performance).
Prior research has not addressed these issues.
Speci?cally, it has not examined which incentive
schemes, or combinations and dimensions thereof,
induce managers to take appropriate levels of risk
(i.e. select projects that maximize expected value)
while concurrently motivating high levels of e?ort.
So, for example, do incentives encourage man-
agers to focus on maximizing low variance, low
return performance measures (projects) over high
variance, high return performance measures (pro-
jects)? We currently do not know the answer to
this question, and future research directed toward
examining this issue would be quite valuable.
Moreover, similar sentiments have been echoed by
Stephen Ross, who recently commented: ‘‘No time
has been spent on asking how incentives a?ect the
willingness of the employee to take on risk. More
time is going to have to be spent on the reaction to
the carrot, not just the stick’’ (Valance, 2001).
Research directed toward understanding whe-
ther and how incentive contracts and their speci?c
properties motivate individuals to focus on certain
activities and dimensions of performance (e?ort
direction, strategy development), how hard they
work on these activities and dimensions (e?ort
duration, e?ort intensity), including selecting pro-
jects of di?ering risk and expected return, would
be valuable for several reasons. First, such
research could facilitate job design and improve
our understanding of how decision rights should
be partitioned in an organization. This has clear
implications for the design of responsibility
accounting systems and whether, for example,
organizations should seek to change an employee’s
opportunity costs by limiting the tasks and activ-
ities they can work on. Second, such research
could facilitate the design and development of
performance measures and how precise they need
to be to motivate the desired levels of e?ort, allo-
cation(s) of e?ort, and risk taking. Finally, such
research could be quite informative about the
appropriate (relative) compensation weights
assigned to each measure of performance (e.g.
Banker & Datar, 1989).
3.4.2. E?ects of other incentive scheme variables
on the incentives–e?ort–performance relation
Incentive schemes vary on a variety of dimen-
sions. Another dimension that may alter the
incentives–e?ort relation is the level of pay. The
issue regarding whether payo? magnitude a?ects
e?ort and performance is of obvious interest to
organizations and researchers. Speci?cally, for
e?ciency reasons, organizations would like to
minimize the cost associated with eliciting desired
levels of e?ort and performance from their
employees (e.g. Merchant, 1998, chapter 11).
Analogously, many accounting researchers con-
ducting laboratory experiments would like to
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 333
stretch limited research dollars as far as possible
yet employ, for motivational purposes, monetary
rewards that are salient and dominant (Smith,
1982).
Theoretically, it is unclear how payo? magni-
tude might a?ect an individual’s e?ort and per-
formance. First, agency theory (via expected
utility theory) suggests that, to elicit any e?ort
from individuals, the expected rewards net of
expected e?ort costs must exceed an individual’s
reservation wage (Baiman, 1982, 1990). Thus,
increasing payo? magnitude could increase per-
formance due to the expected cost-bene?t calcula-
tion that individuals theoretically execute prior to
performing a task. Moreover, as pay increases the
expected bene?ts for good performance increase
and costs stay the same (assuming the task is held
constant). Therefore, as the tradeo? between costs
and bene?ts tips more toward bene?ts, people are
expected to exert more current e?ort and/or
engage in strategy development (also see Smith &
Walker, 1993). However, as mentioned earlier, this
increase in e?ort is dependent upon an individual
having a high enough level of self-e?cacy (and
skill) so that they believe the probability of
attaining the accurate performance is high (along
with the value of attaining this level of perfor-
mance). If the expected probability of performing
accurately is low, raising the value of good per-
formance likely will not increase e?ort. Further,
the e?ect on e?ort of increasing incentives beyond
some point depends on the initial point. If the
initial level of incentives leads to maximal e?ort,
then increases in payo? magnitude will not
increase e?ort and, thus, have no e?ect on perfor-
mance.
Because standard expected utility theory also
presumes that most individuals have diminishing
marginal utility for wealth, though, increasing
rewards may decrease motivation and perfor-
mance over time since savings increase and further
increases in wealth have less value (Lambert,
1983). On the other hand, wealth e?ects arising
from increases in pay likely make individuals more
locally risk-neutral and may reduce risk premiums
as well as engender decisions that are more con-
sistent with expected value maximization (e.g. Ang
& Schwarz, 1985). Finally, work by Akerlof (1982,
1984) suggests that the relation between payo?
magnitude and e?ort and performance is likely to
be positive since increasing pay can lead employ-
ees and organizations to engage in mutual gift
exchange. As compensation increases beyond
employees’ reservation wages, they are posited to
reciprocate by supplying higher levels of e?ort and
performance (c.f. Fehr & Ga¨ chter, 2000).
Empirical evidence tends to re?ect this theore-
tical lack of clarity and indicates that increasing
the level of payments to subjects has mixed e?ects
on performance (e.g. Camerer & Hogarth, 1999, p.
21; also see Libby, Bloom?eld, & Nelson, 2001).
Speci?cally, many studies ?nd that increasing the
level of rewards increases performance (e.g.
Cabanac, 1986; Hannan, 2001; Pritchard & Curtis,
1973; Smith & Walker, 1993; Toppen, 1965a;
Weiner, 1966; Weiner & Walker, 1966). Numerous
other studies, however, ?nd that increasing the
level of rewards has no e?ect on performance (e.g.
Bonem & Crossman, 1988; Craik & Tulving, 1975;
Enzle & Ross, 1978; Farr, Vance, & McIntyre,
1977; Riedel et al., 1988; Toppen, 1965b). Finally,
a couple of studies ?nd that increasing the level of
rewards results in a decrease in performance (e.g.
Pritchard & DeLeo, 1973; Tomporowski, Simp-
son, & Hager, 1993).
Unfortunately, there is not enough information
contained in prior studies that would allow us to
disentangle these e?ects in some meaningful fash-
ion. Moreover, prior research generally does not
report on whether varying payo? magnitude
a?ects the underlying e?ort mechanisms that are
theoretically posited to a?ect task performance.
29
Thus, we know very little about whether increas-
ing or decreasing the level of rewards a?ects e?ort
direction, e?ort duration, e?ort intensity, and
strategy development and, if it does, whether this
occurs through self-e?cacy, goal setting, utility
maximization, or other mechanisms.
Given this and the mixed performance ?ndings,
it is di?cult to assess the implications that prior
research on the e?ects of payo? magnitude has for
organizations, accounting researchers, or the
design of accounting-based reward systems.
29
One exception is Enzle and Ross (1978) who ?nd that
increasing rewards results in lower task interest.
334 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
However, we can highlight several fruitful avenues
for future research. Speci?cally, research needs to
be much more systematic and directed in isolating
the underlying mechanisms that might produce
payo? magnitude e?ects.
For example, future research might examine the
predictive ability of various e?ciency wage the-
ories (e.g. Akerlof 1982, 1984; Weiss, 1990; Yellen,
1984) in a setting where the task being performed
requires physical and mental e?ort. Such research
could speci?cally examine the e?ects that increas-
ing the level of pay has on e?ort direction, e?ort
duration, e?ort intensity and strategy develop-
ment. Additionally, such research could determine
whether the observed e?ects are permanent or
more transitory and, thus, whether issues con-
nected with consumption smoothing ultimately
arise. Future research also could examine the
e?ects that payo? magnitude has on employees’
propensity to take risks and innovate. Speci?cally,
does increasing the level of rewards induce risk-
taking behavior because individuals are locally
risk-neutral? Or, does increasing rewards engender
more risk-averse behavior because individuals feel
more performance pressure? Finally, research
should pay more careful attention to person vari-
ables (skill) and task variables (complexity) that
might interact with payo? magnitude as well (e.g.
Wright, 1992). In short, such research could prove
to be valuable in facilitating cost management and
designing an e?cient reward system, both for
organizations and accounting researchers con-
ducting laboratory experiments.
Monetary incentives also can vary as to the
timing of their introduction. Speci?cally, incen-
tives can be introduced at di?erent stages of pro-
cessing (prior to or after information search), and
timing could vary naturally as the result of di?er-
ent phases of production being under the control
of di?erent departments or supervisors. Consistent
with the general hypothesis regarding how incen-
tives a?ect performance, it seems natural that
incentives will be more e?ective when introduced
prior to the processing stages in which perfor-
mance is most sensitive to e?ort. In support of
this, memory studies examining subjects’ retention
of information (via recall or recognition) have
shown that incentives introduced prior to the
encoding of information are signi?cantly more
e?ective than equivalent incentives introduced
after encoding, but prior to retrieval (Harley,
1968; Libby & Lipe, 1992; Weiner, 1966; Wickens
& Simpson, 1968). Such e?ects occur because
incentives given before trace formation or trace
storage motivate subjects to exert e?ort to better
organize and rehearse the stimuli, whereas incen-
tives given after trace formation or trace storage
are less e?ective because additional e?ort at this
point can do little to in?uence retention.
Other timing issues are less clear. For example,
how should compensation be structured over an
employee’s tenure with an organization? Such a
question relates to career concerns, the dynamics
of contracting, and the provision and structure of
incentives over time. While such issues generally
have not been examined empirically, they are quite
important to organizations and the design of an
accounting-based reward system. Thus, should
organizations employ a constant payment sche-
dule or an increasing payment schedule? Deferring
compensation theoretically is posited to bene?t
organizational performance by reducing turnover
and retaining the most able employees as well as
obtaining high levels of e?ort from an employee
throughout their tenure with the ?rm (e.g. Lazear,
1981; Salop & Salop, 1976). Such payment sche-
dules also may be desirable on equity grounds.
Archival research, though, has had di?culty
interpreting why ?rms defer compensation and
understanding the bene?ts that do or do not
accrue to such a compensation arrangement (Pre-
ndergast, 1999). Experimental research, on the
other hand, could provide useful evidence on the
structure of payment schedules and their corre-
sponding e?ects on e?ort and performance over
time. Experimental methods also could help
answer other timing questions such as whether the
ratio of incentive pay to ?xed pay should be con-
stant, increase, or decrease over time. Again, some
theory and empirical evidence indicates that the
pay-for-performance sensitivity should increase
the longer the employee is with the ?rm (Gibbons
& Murphy, 1992), although it is unclear that such
an arrangement elicits the optimal e?ort levels
from employees compared to other payment
methods.
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 335
Other salient dimensions of incentive schemes
include whether the incentive contract explicitly
embodies competition, whether the incentive
scheme incorporates assigned goals, and whether
pay should be linked to performance at the indi-
vidual unit or more aggregate level (e.g. Bonner et
al. 2000). In addition to the research opportunities
previously discussed, we feel that it is crucial for
researchers to examine combinations of incentive
schemes. For example, tournament schemes fre-
quently are criticized on the grounds that they
induce excessive risk-taking behavior or induce
people to ‘‘give-up’’ because they believe that the
probability of winning the prize is low (Camerer &
Hogarth, 1999; Dye, 1984). On the other hand,
piece-rate and budget-based schemes may be best
at motivating high levels of e?ort duration and
e?ort intensity, but may discourage risk-taking
and e?ort directed toward strategy development
(and innovation). Prior research, though, has
examined incentive schemes in isolation and it is
quite possible that, as in the natural environment,
the optimal compensation arrangement is one
where tournaments are combined with budget-
based compensation and a ?xed salary. Moreover,
such a combination of incentive schemes may best
balance short-term e?ort duration and intensity
with longer-term e?ort directed toward strategy
development as well as motivating choices con-
sistent with expected value maximization.
In summary, there are numerous dimensions of
incentive schemes per se that may a?ect task per-
formance. Similar to person, task, and environ-
mental variables, prior research provides some
guidance regarding how to best relate pay to per-
formance. That said, more research clearly is nee-
ded to understand whether and how the explicit
and implicit features of incentive contracts elicit
the desired levels and types of e?ort and, thus,
align the interests of employees with those of the
organization. At a fundamental level, such
research is important to accounting because it
relates to the dimensions of performance that are
measured, when and how they are measured, and
how such measures are ultimately used in evaluat-
ing and rewarding employees. Such research also
relates to cost management and, therefore, can
help facilitate the design of the most e?ective and
e?cient accounting-based performance measure-
ment and reward system.
4. Conclusions
In this paper, we present theories, evidence, and
a framework for understanding the e?ects of
monetary incentives on e?ort and task perfor-
mance. We ?rst describe the fundamental incen-
tives–e?ort and e?ort–performance relations and
the four dimensions of e?ort that monetary incen-
tives theoretically are posited to a?ect: direction,
duration, intensity, and strategy development. We
then discuss psychological and economic theories
that explicate the incentives-e?ort link. Here, we
detail many of the underlying cognitive and moti-
vational process mechanisms by which monetary
incentives are presumed to lead to increases in
e?ort and, thus, increases in performance.
We also provide a conceptual framework for the
e?ects of monetary incentives on e?ort and task
performance. This framework facilitates a com-
prehensive consideration of the variables that may
combine with monetary incentives in a?ecting
performance. Speci?cally, we formulate the incen-
tives–e?ort and e?ort–performance relations as a
function of person variables, task variables, envir-
onmental variables, and incentive scheme vari-
ables. We then use our conceptual framework to
organize and integrate a large amount of evidence
on the e?cacy of monetary incentives. In this
regard, the framework is employed to focus on
how salient features of accounting settings may
moderate the positive e?ects of monetary incen-
tives and, thus, to understand the e?ects of mone-
tary incentives in numerous contexts of interest to
accounting researchers.
We then choose one speci?c variable from each
of the person, task, environmental, and incentive
scheme categories within the framework and dis-
cuss its relation with monetary incentives. The
four particular variables we examine in-depth are:
skill, task complexity, assigned goals, and the
rewarded dimension of performance. For each of
these variables, we discuss its importance in
accounting settings as well as the theoretical and
practical importance of examining the variable in
336 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
conjunction with monetary incentives. We then
present theoretical predictions and review the
empirical evidence regarding the combination of
these accounting-related variables and monetary
incentives on individual e?ort and performance.
We pay particular attention to the signi?cant
implications that our integration and compilation
of theories and evidence has for accounting
research and practice. Following this, we highlight
numerous directions for future research in
accounting that could provide important insights
into the e?cacy of monetary reward systems.
Finally, while we restrict our primary attention to
four speci?c variables we also brie?y discuss the-
ories, empirical evidence, and directions for future
research for several other person, task, environ-
mental, and incentive scheme variables that are
important in accounting settings.
Our framework and review of the attendant
evidence indicates that there are a number of
accounting-related variables that can alter the
e?ects of incentives on performance. For example,
we ?nd that, on average, explicit performance tar-
gets (assigned goals) have additive positive e?ects
on e?ort and performance over monetary incen-
tives, thereby suggesting that organizations should
employ performance targets in conjunction with
monetary incentives to motivate employees. How-
ever, we also ?nd evidence of an interaction
between the di?culty of the goal and the type of
incentive scheme. Speci?cally, compared to piece-
rate schemes, performance typically is better under
budget-based (quota) schemes when goals are
moderate, but worse when goals are di?cult. This
evidence has implications regarding whether
assigned goals and incentives should be kept as
separate motivating mechanisms or whether
incentives should be linked to goal attainment.
We also ?nd that features of accounting settings
can attenuate the positive e?ects of monetary
incentives on performance by altering either the
e?ect of incentives on e?ort or altering the e?ect
of incentives-induced e?ort on performance. For
example, we ?nd evidence that lack of skill can
attenuate the e?ort–performance relation because,
while monetary incentives may induce higher
levels of e?ort, the performance of individuals
who lack requisite skills is not sensitive to these
e?ort increases. Additionally, lack of skill can
attenuate the incentives–e?ort relation. Speci?-
cally, when individuals are assigned tasks for
which they do not have the necessary skills, they
may not increase their e?ort under monetary
incentives because they believe that e?ort increases
will not lead to performance increases and con-
sequent rewards. Alternatively, when individuals
are allowed to select their own contracts for a
particular task, individuals with high skill are
more likely to choose contingent pay, thereby
restoring a positive incentives–e?ort relation.
Because we delve deeply into the underlying
cognitive and motivational processes, we explore
the multiple roles that a particular variable, such
as skill, can play in a?ecting the incentives–per-
formance relation. These multiple roles are not
inconsequential. For example, recognizing that
task complexity itself can a?ect self-e?cacy and,
in turn, whether and how incentives a?ect the
various dimensions of e?ort, highlights the critical
role that task characteristics may play in deter-
mining short-run and long-run performance under
monetary incentives. Moreover, understanding the
processes by which accounting-related variables
alter the incentives–performance relation and
which part of the relation they alter can be critical
for suggesting solutions that might restore a posi-
tive e?ect of incentives on performance. For
instance, understanding how incentive contracts
and their dimensions (e.g. which dimension(s) of
performance they reward) a?ect employees’ allo-
cations and levels of e?ort has possible, and per-
haps distinct, implications for the structure of
monetary reward systems, how performance is
measured, the development of responsibility
accounting systems, and job design.
Our paper also makes clear that there are many
important research questions that need to be
addressed in accounting—and other disciplines—
before it is appropriate to make strong recom-
mendations to organizations and experimenters
regarding the use and form of monetary incen-
tives. In this vein, for each category within our
framework, we develop and discuss several direc-
tions for future research that we feel would help
?ll gaps in our knowledge regarding the e?ective-
ness of monetary reward systems. Moreover, we
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 337
identify opportunities for future accounting
research that reports on how salient accounting-
related person, task, environmental, and incentive
scheme variables combine with monetary incen-
tives to a?ect individual e?ort and performance.
We feel such future research is vital given the
important role that accountants and accounting
information play in compensation practice and the
design of performance-measurement and reward
systems. In this way, organizations and research-
ers will have better information regarding the cir-
cumstances under which monetary incentives yield
desired levels and types of e?ort and performance
in either the ?eld or the laboratory.
Acknowledgements
We thank Ramji Balakrishnan, Joan Luft, Lau-
reen Maines, Jamie Pratt, Jerry Salamon, Mike
Shields, David Upton, Michael Williamson, Mark
Young, and two anonymous reviewers for their
very helpful comments and suggestions. Sarah
Bonner also thanks BDO Seidman, LLP for their
generous ?nancial support.
References
Akerlof, G. A. (1982). Labor contracts as partial gift exchange.
The Quarterly Journal of Economics, 97, 543–569.
Akerlof, G. A. (1984). Gift exchange and e?ciency-wage the-
ory: four views. The American Economic Review, 74, 79–83.
Anderson, J. R. (1990). The adaptive character of thought.
Hillsdale, NJ: Erlbaum.
Anderson, J. R. (1995). Learning and memory: an integrative
approach. New York, NY: John Wiley.
Ang, J. S., & Schwarz, T. (1985). Risk aversion and informa-
tion structure: an experimental study of price variability in
the securities markets. The Journal of Finance, 40, 825–844.
Arkes, H. R. (1991). Costs and bene?ts of judgment errors:
implications for debiasing. Psychological Bulletin, 110, 486–
498.
Arkes, H. R., Dawes, R. M., & Christensen, C. (1986). Factors
in?uencing the use of a decision rule in a probabilistic task.
Organizational Behavior and Human Decision Processes, 37,
93–110.
Asare, S. K., & McDaniel, L. S. (1996). The e?ect of familiarity
with the preparer and task complexity on the e?ectiveness of
the audit review process. The Accounting Review, 71, 139–160.
Ashford, S. J., & Cummings, L. L. (1983). Feedback as an
individual resource: personal strategies of creating informa-
tion. Organizational Behavior and Human Performance, 32,
370–398.
Ashton, A. H., & Ashton, R. H. (1988). Sequential belief revi-
sion in auditing. The Accounting Review, 63, 623–641.
Ashton, R. H. (1990). Pressure and performance in accounting
decision settings: paradoxical e?ects of incentives, feedback,
and justi?cation. Journal of Accounting Research, 28(Suppl.),
148–180.
Ashton, R. H., & Ashton, A. H. (1995). Judgment and decision-
making research in accounting and auditing. Cambridge, UK:
Cambridge University Press.
Atkinson, A. A., Banker, R., Kaplan, R. S., & Young, S. M.
(2001). Management accounting (3rd ed.). Upper Saddle
River, NJ: Prentice-Hall.
Atkinson, J. W., & Reitman, W. R. (1956). Performance as a
function of motive strength and expectancy of goal-attain-
ment. Journal of Abnormal and Social Psychology, 53, 361–
366.
Awasthi, V., & Pratt, J. (1990). The e?ects of ?nancial incen-
tives on e?ort and decision performance: the role of cognitive
characteristics. The Accounting Review, 65, 797–811.
Bahrick, H. P., Fitts, P. M., & Rankin, R. E. (1952). Incentives
and reactions to peripheral stimuli. Journal of Experimental
Psychology, 44, 400–406.
Bailey, C. B., Brown, L. D., & Cocco, A. F. (1998). The e?ects
of monetary incentives on worker learning and performance
in an assembly task. Journal of Management Accounting
Research, 10, 119–131.
Baiman, S. (1982). Agency research in managerial accounting:
a survey. Journal of Accounting Literature, 1, 154–213.
Baiman, S. (1990). Agency research in managerial accounting:
a second look. Accounting, Organizations and Society, 15,
341–371.
Baiman, S., & Sivaramakrishnan, K. (1991). The value of pri-
vate pre-decision information in a principal-agent context.
The Accounting Review, 66, 747–766.
Baker, S. L., & Kirsch, I. (1991). Cognitive mediators of pain
perception and tolerance. Journal of Personality and Social
Psychology, 61, 504–510.
Balzer, W. K., Doherty, M. E., & O’Connor Jr., R. (1989).
E?ects of cognitive feedback on performance. Psychological
Bulletin, 106, 410–433.
Bandura, A. (1986). Social foundations of thought and action.
Englewood Cli?s, NJ: Prentice-Hall.
Bandura, A. (1991). Social cognitive theory of self-regulation.
Organizational Behavior and Human Decision Processes, 50,
248–287.
Bandura, A. (1997). Self-e?cacy: the exercise of control. New
York, NY: W.H. Freeman and Company.
Banker, R. D., & Datar, S. M. (1989). Sensitivity, precision,
and linear aggregation of signals for performance evaluation.
Journal of Accounting Research, 27, 21–39.
Baron, R. M., & Kenny, D. A. (1986). The moderator-med-
iator variable distinction in social psychological research:
conceptual, strategic, and statistical considerations. Journal
of Personality and Social Psychology, 51, 1173–1182.
Becker, D. A. (1997). The e?ects of choice on auditors’ intrinsic
338 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
motivation and performance. Behavioral Research in
Accounting, 9, 1–19.
Bernardi, R. A. (1994). Fraud detection: the e?ect of client
integrity and competence and auditor cognitive style. Audit-
ing: A Journal of Practice and Theory, 13(Suppl.), 68–84.
Bettman, J. R., Johnson, E. J., & Payne, J. W. (1990). A com-
ponential analysis of cognitive e?ort in choice. Organiza-
tional Behavior and Human Decision Processes, 45, 111–139.
Bloom?eld, R., Libby, R., & Nelson, M. W. (1999). Con?dence
and the welfare of less-informed investors. Accounting,
Organizations and Society, 24, 623–647.
Bonem, M., & Crossman, E. K. (1988). Elucidating the e?ects
of reinforcement magnitude. Psychological Bulletin, 104,
348–362.
Bonner, S. E. (1994). A model of the e?ects of audit task com-
plexity. Accounting, Organizations and Society, 19, 213–234.
Bonner, S. E. (1999). Judgment and decision-making research
in accounting. Accounting Horizons, 13, 385–398.
Bonner, S. E., Davis, J. S., & Jackson, B. R. (1992). Expertise
in corporate tax planning: the issue identi?cation stage.
Journal of Accounting Research, 30(Suppl.), 1–28.
Bonner, S. E., Hastie, R., Sprinkle, G. B., & Young, S. M.
(2000). A review of the e?ects of ?nancial incentives on per-
formance in laboratory tasks: implications for management
accounting. Journal of Management Accounting Research, 13,
19–64.
Bonner, S. E., Hastie, R., Young, S. M., Hesford, J., & Gigone,
D. (2001). E?ects of monetary incentives on the performance
of a cognitive task: the moderating role of skill. Working
paper, University of Southern California.
Bonner, S. E., & Lewis, B. L. (1990). Determinants of auditor
expertise. Journal of Accounting Research, 28(Suppl.), 1–20.
Bonner, S. E., & Pennington, N. (1991). Cognitive processes
and knowledge as determinants of auditor expertise. Journal
of Accounting Literature, 10, 1–50.
Bonner, S. E., & Walker, P. L. (1994). The e?ects of instruction
and experience on the acquisition of auditing knowledge. The
Accounting Review, 69, 157–178.
Briers, M. L., Chow, C. W., Hwang, N. R., & Luckett, P. F.
(1999). The e?ects of alternative types of feedback on pro-
duct-related decision performance: a research note. Journal
of Management Accounting Research, 11, 75–92.
Broadbent, D. E. (1971). Decision and stress. London, UK:
Academic Press.
Brown, L. D. (1999). Managerial behavior and the bias in ana-
lysts’ earnings forecasts. Working paper, Georgia State Uni-
versity.
Bull, C., Schotter, A., & Weigelt, K. (1987). Tournaments and
piece rates: an experimental study. Journal of Political Econ-
omy, 95, 1–33.
Cabanac, M. (1986). Money versus pain: experimental study of
a con?ict in humans. Journal of the Experimental Analysis of
Behavior, 46, 37–44.
Camerer, C. F. (1995). Individual decision making. In
J. H. Kagel, & A. E. Roth (Eds.), The handbook of experi-
mental economics (pp. 587–703). Princeton, NJ: Princeton
University Press.
Camerer, C. F., & Hogarth, R. M. (1999). The e?ects of ?nancial
incentives in experiments: a review and capital-labor-produc-
tion framework. Journal of Risk and Uncertainty, 19, 7–42.
Cameron, J., & Pierce, W. D. (1994). Reinforcement, reward,
and intrinsic motivation: a meta-analysis. Review of Educa-
tional Research, 64, 363–423.
Campbell, D. J. (1984). The e?ects of goal-contingent payment
on the performance of a complex task. Personnel Psychology,
37, 23–40.
Campbell, D. J. (1988). Task complexity: a review and analysis.
Academy of Management Review, 13, 40–52.
Campbell, D. J., & Ilgen, D. R. (1976). Additive e?ects of task
di?culty and goal setting on subsequent task performance.
Journal of Applied Psychology, 61, 319–324.
Carroll, J. B. (1993). Human cognitive abilities. New York, NY:
Cambridge University Press.
Chow, C. W. (1983). The e?ects of job standard tightness and
compensation scheme on performance: an exploration of
linkages. The Accounting Review, 58, 667–685.
Chow, C. W., Shields, M. D., & Wu, A. (1999). The importance
of national culture in the design of and preference for man-
agement controls for multi-national operations. Accounting,
Organizations and Society, 24, 441–461.
Chung, K. H., & Vickery, W. D. (1976). Relative e?ectiveness
and joint e?ects of three selected reinforcements in a repeti-
tive task situation. Organizational Behavior and Human Per-
formance, 16, 114–142.
Cloyd, C. B. (1997). Performance in tax research tasks: the
joint e?ects of knowledge and accountability. The Accounting
Review, 72, 11–131.
Conlisk, J. (1996). Why bounded rationality? Journal of Eco-
nomic Literature, 34, 669–700.
Cote, J. M., & Sanders, D. L. (1997). Herding behavior: expla-
nations and implications. Behavioral Research in Accounting,
9, 20–45.
Craik, F. I. M., & Tulving, E. (1975). Depth of processing and
the retention of words in episodic memory. Journal of
Experimental Psychology: General, 104, 268–294.
Creyer, E. H., Bettman, J. R., & Payne, J. W. (1990). The
impact of accuracy and e?ort feedback and goals on adap-
tive decision behavior. Journal of Behavioral Decision Mak-
ing, 3, 1–16.
Crystal, G. S. (1991). In search of excess: the overcompensation
of American executives. New York, NY: W.W. Norton &
Company.
Das, S., Levine, C. B., & Sivaramakrishnan, K. (1998). Earn-
ings predictability and bias in analysts’ earnings forecasts.
The Accounting Review, 73, 277–294.
Davis, D. D., & Holt, C. A. (1993). Experimental economics.
Princeton, NJ: Princeton University Press.
Dearman, D., & Shields, M. D. (2001). Cost knowledge and
cost-based judgment performance. Journal of Management
Accounting Research, 14 (in press).
Deci, E. L., Betley, G., Kahle, J., Abrams, L., & Porac, J.
(1981). When trying to win: competition and intrinsic moti-
vation. Personality and Social Psychology Bulletin, 7, 79–83.
Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 339
determination in human behavior. New York, NY: Plenum
Press.
Demski, J., & Feltham, G. (1978). Economic incentives in
budgetary control systems. The Accounting Review, 53, 336–
359.
Dillard, J. F., & Fisher, J. G. (1990). Compensation schemes,
skill level and task performance: an experimental examina-
tion. Decision Science, 21, 121–137.
Dornbush, R. L. (1965). Motivation and positional cues in
incidental learning. Perceptual and Motor Skills, 20, 709–714.
Dye, R. A. (1984). The trouble with tournaments. Economic
Inquiry, 22, 147–149.
Earley, P. C., & Lituchi, T. R. (1991). Delineating goal and
e?cacy e?ects: a test of three models. Journal of Applied
Psychology, 76, 81–98.
Easterbrook, J. A. (1959). The e?ect of emotion on cue utili-
zation and the organization of behavior. Psychological
Review, 66, 183–201.
Eisenhardt, K. M. (1989). Agency theory: an assessment and
review. Academy of Management Review, 14, 57–74.
Enzle, M. E., & Ross, J. M. (1978). Increasing and decreasing
intrinsic interest with contingent rewards: a test of cognitive
evaluation theory. Journal of Experimental Social Psychol-
ogy, 14, 588–597.
Erez, M., Gopher, D., & Arzi, N. (1990). E?ects of goal di?-
culty, self-set goals, and ?nancial rewards on dual task per-
formance. Organizational Behavior and Human Decision
Processes, 47, 247–269.
Estes, R., & Hosseini, J. (1988). The gender gap on Wall Street:
an empirical analysis of con?dence in investment decision
making. The Journal of Psychology, 122, 577–590.
Eysenck, M. W. (1982). Attention and arousal: cognition and
performance. Berlin, DE: Springer.
Eysenck, M. W. (1986). A handbook of cognitive psychology.
London, UK: Erlbaum.
Fama, E. F. (1980). Agency problems and the theory of the
?rm. Journal of Political Economy, 88, 288–307.
Farh, J., Gri?th, R. W., & Balkin, D. B. (1991). E?ects of
choice of pay plans on satisfaction, goal setting, and perfor-
mance. Journal of Organizational Behavior, 12, 55–62.
Farr, J. L. (1976). Task characteristics, reward contingency,
and intrinsic motivation. Organizational Behavior and Human
Performance, 16, 294–307.
Farr, J. L., Vance, R. J., & McIntyre, R. M. (1977). Further
examinations of the relationship between reward contingency
and intrinsic motivation. Organizational Behavior and Human
Performance, 20, 31–53.
Fatseas, V. A., & Hirst, M. K. (1992). Incentive e?ects of
assigned goals and compensation schemes on budgetary per-
formance. Accounting and Business Research, 22, 347–355.
Fehr, E., & Ga¨ chter, S. (2000). Fairness and retaliation: the
economics of reciprocity. Journal of Economic Perspectives,
14, 159–181.
Feltham, G. A., & Xie, J. (1994). Performance measure con-
gruity and diversity in multi-task principal/agent relations.
The Accounting Review, 69, 429–453.
Fessler, N. (2000). Incentive compensation contracts: when are
they ine?ective? Working paper, Abilene Christian Uni-
versity.
Frederick, D. M. (1991). Auditors’ representation and retrieval
of internal control knowledge. The Accounting Review, 66,
240–258.
Frederickson, J. R. (1992). Relative performance information:
the e?ects of common uncertainty and contract type of agent
e?ort. The Accounting Review, 67, 647–669.
Frederickson, J. R., Pe?er, S. A., & Pratt, J. (1999). Perfor-
mance evaluation judgments: e?ects of prior experience under
di?erent performance evaluation schemes and feedback fre-
quencies. Journal of Accounting Research, 37, 151–165.
Friedman, D., & Sunder, S. (1994). Experimental methods: a
primer for economists. New York, NY: Cambridge Uni-
versity Press.
Furnham, A., & Argyle, M. (1998). The psychology of money.
New York, NY: Routledge.
Gerhart, B., & Milkovich, G. T. (1992). Employee compensa-
tion: research and practice. In M. D. Dunnette, &
L. M. Hough (Eds.), Handbook of industrial and organiza-
tional psychology (pp. 481–569). Palo Alto, CA: Consulting
Psychologists Press.
Gibbins, M., & Jamal, K. (1993). Problem-centred research and
knowledge-based theory in the professional accounting set-
ting. Accounting, Organizations and Society, 18, 451–466.
Gibbons, R., & Murphy, K. J. (1992). Optimal incentive con-
tracts in the presence of career concerns. Journal of Political
Economy, 100, 468–506.
Gigerenzer, G., Hofrage, U., & Kleinbolting, H. (1991). Prob-
abilistic mental models: a Brunswikian theory of con?dence.
Psychological Review, 98, 506–528.
Gigerenzer, G., Todd, P. M. & the ABC Research Group.
(1999). Simple heuristics that make us smart. New York, NY:
Oxford University Press.
Gist, M. E., & Mitchell, T. R. (1992). Self-e?cacy: a theoretical
analysis of its determinants and malleability. Academy of
Management Review, 17, 183–211.
Glucksberg, S. (1962). The in?uence of strength of drive on
functional ?xedness and perceptual recognition. Journal of
Experimental Psychology, 63, 36–41.
Grether, D. M., & Plott, C. (1979). Economic theory of choice
and the preference reversal phenomenon. The American
Economic Review, 75, 623–638.
Guzzo, R. A., Jette, R. D., & Katzell, R. A. (1985). The e?ects of
psychologically based intervention programs on worker pro-
ductivity: a meta-analysis. Personnel Psychology, 38, 275–291.
Hamner, W. C. (1974). Reinforcement theory and contingency
management in organizational settings. In H. L. Tosi, &
W. C. Hamner (Eds.), Organizational behavior and manage-
ment: a contingency approach. Chicago, IL: St. Clair.
Hamner, W. C., & Foster, L. W. (1975). Are intrinsic and
extrinsic rewards additive: a test of Deci’s cognitive evalua-
tion theory of task motivation. Organizational Behavior and
Human Performance, 14, 398–415.
Hannan, R. L. (2001). The e?ect of ?rm pro?t on fairness per-
ceptions, wages and employee e?ort. Working paper, Georgia
State University.
340 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
Harley Jr., W. F. (1965a). The e?ect of ?nancial incentive in
paired associate learning using a di?erential method. Psy-
chonomic Science, 2, 377–378.
Harley Jr., W. F. (1965b). The e?ect of ?nancial incentive in
paired associate learning using an absolute method. Psycho-
nomic Science, 3, 141–142.
Harley Jr., W. F. (1968). Delay of incentive cues in paired-
associate learning and its e?ects on organizing responses.
Journal of Verbal Learning and Verbal Behavior, 7, 924–
929.
Harrison, G. L., & McKinnon, J. L. (1999). Cross-cultural
research in management control systems design: a review of
the current state. Accounting, Organizations and Society, 24,
483–506.
Harrison, P. D., Chow, C. W., Wu, A., & Harrell, A. M.
(1999). A cross-cultural investigation of managers’ project
evaluation decisions. Behavioral Research in Accounting, 11,
143–160.
Heiman, V. (1990). Discussion of pressure and performance in
accounting decision settings: paradoxical e?ects of incen-
tives, feedback, and justi?cation. Journal of Accounting
Research, 28, 181–186.
Hemmer, T. (1996). On the design and choice of ‘‘modern’’
management accounting measures. Journal of Management
Accounting Research, 8, 87–116.
Hirst, M. K., & Yetton, P. W. (1999). The e?ects of budget
goals and task interdependence on the level of and variance
in performance: a research note. Accounting, Organizations
and Society, 24, 205–216.
Hogarth, R. M. (1993). Accounting for decisions and decisions
for accounting. Accounting, Organizations and Society, 18,
407–424.
Hogarth, R. M., Gibbs, B. J., McKenzie, C. R. M., & Marquis,
M. A. (1991). Learning from feedback: exactingness and
incentives. Journal of Experimental Psychology: Learning,
Memory and Cognition, 17, 734–752.
Holmstrom, B. (1989). Agency costs and innovation. Journal of
Economic Behavior and Organization, 12, 305–327.
Holmstrom, B., & Milgrom, P. (1991). Multitask principal-
agent analyses: incentive contracts, asset ownership, and job
design [special issue]. Journal of Law, Economics, & Organi-
zation, 7, 24–52.
Hong, H., Kubik, J. D., & Solomon, A. (2000). Security ana-
lysts’ career concerns and herding of earnings forecasts.
RAND Journal of Economics, 31, 121–144.
Hopkins, P. E. (1996). The e?ect of ?nancial statement classi-
?cation of hybrid ?nancial instruments on ?nancial analysts’
stock price judgments. Journal of Accounting Research,
34(Suppl.), 33–50.
Horngren, C. T., Foster, G., & Datar, S. M. (2000). Cost
accounting: a managerial emphasis (10th ed.). Upper Saddle
River, NJ: Prentice-Hall.
Hronsky, J. J. F., & Houghton, K. A. (2001). The meaning of a
de?ned accounting concept: regulatory changes and the e?ect
on auditor decision making. Accounting, Organizations and
Society, 26, 123–139.
Huber, V. L. (1985). Comparison of ?nancial reinforcers and
goal setting as learning incentives. Psychological Reports, 56,
223–235.
Humphreys, M. S., & Revelle, W. (1984). Personality, motiva-
tion, and performance: a theory of the relationship between
individual di?erences and information processing. Psycholo-
gical Review, 91, 153–184.
Hunt, S. C. (1995). A review and synthesis of research in per-
formance evaluation in public accounting. Journal of
Accounting Literature, 14, 107–139.
Hunton, J. E., Wier, B., & Stone, D. N. (2000). Succeeding in
managerial accounting. Part 2: a structural equations analy-
sis. Accounting, Organizations and Society, 25, 751–762.
Indjejikian, R. J. (1999). Performance evaluation and compen-
sation research: an agency perspective. Accounting Horizons,
13, 147–157.
Jacoby, J., Mazursky, D., Troutman, T., & Kuss, A. (1984).
When feedback is ignored: disutility of outcome feedback.
Journal of Applied Psychology, 69, 531–545.
Jehle, G. A., & Reny, P. J. (2001). Advanced microeconomic
theory. Boston, MA: Addison-Wesley.
Jenkins, G. D. Jr. (1986). Financial incentives. In E. A. Locke
(Ed.), Generalizing from laboratory to ?eld settings (pp. 167–
180). Lexington, MA: D.C. Hearth and Company.
Jenkins, G. D. Jr., Mitra, A., Gupta, N., & Shaw, J. D. (1998).
Are ?nancial incentives related to performance? A meta-
analytic review of empirical research. Journal of Applied
Psychology, 83, 777–787.
Jermias, J. (2001). Cognitive dissonance and resistance to
change: the in?uence of commitment con?rmation and feed-
back on judgment usefulness of accounting systems.
Accounting, Organizations and Society, 26, 141–160.
Johnson, E. N., Kaplan, S. E., & Reckers, P. M. J. (1998). An
examination of potential gender-based di?erences in audit
managers’ performance evaluation judgments. Behavioral
Research in Accounting, 10, 47–75.
Jorgenson, D., Dunnette, M. D., & Pritchard, R. D. (1973).
E?ects of the manipulation of a performance-reward con-
tingency on behavior in a simulated work setting. Journal of
Applied Psychology, 57, 271–280.
Kahneman, D. (1973). Attention and e?ort. Englewood Cli?s,
NJ: Prentice-Hall.
Kahenman, D., & Tversky, A. (1979). Prospect theory: an
analysis of decision under risk. Econometrica, 47, 263–291.
Kanfer, R. (1987). Task-speci?c motivation: an integrative
approach to issues of measurement, mechanisms, processes,
and determinants. Journal of Social and Clinical Psychology,
5, 237–264.
Kanfer, R. (1990). Motivation theory and industrial and orga-
nizational psychology. In M. D. Dunnette, & L. M. Hough
(Eds.), Handbook of industrial and organizational psychology
(pp. 75–170). Palo Alto, CA: Consulting Psychologists Press.
Kausler, D. H., & Trapp, E. P. (1962). E?ects of incentive set
and task variables on relevant and irrelevant learning in
serial verbal learning. Psychological Reports, 10, 451–457.
Kennedy, J. (1993). Debiasing audit judgment with account-
ability: a framework and experimental results. Journal of
Accounting Research, 31, 231–245.
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 341
Kennedy, J. (1995). Debiasing the curse of knowledge in audit
judgment. The Accounting Review, 70, 249–273.
Kennedy, J., & Peecher, M. E. (1997). Judging auditors’ tech-
nical knowledge. Journal of Accounting Research, 35, 279–
293.
Kessler, L., & Ashton, R. H. (1981). Feedback and prediction
achievement in ?nancial analysis. Journal of Accounting
Research, 19, 146–162.
Kida, T. (1984). The impact of hypothesis-testing strategies on
auditors’ use of judgment data. Journal of Accounting
Research, 22, 332–340.
Klein, H. J., & Wright, P. M. (1994). Antecedents of goal
commitment: an empirical examination of personal and
situational factors. Journal of Applied Social Psychology, 24,
95–114.
Kluger, A. N., & DeNisi, A. (1996). The e?ects of feedback
interventions on performance: a historical review, a meta-
analysis, and a preliminary feedback intervention theory.
Psychological Bulletin, 119, 254–284.
Kohn, A. (1993). Why incentive plans cannot work. Harvard
Business Review, 71, 54–63.
Lambert, R. A. (1983). Long-term contracts and moral hazard.
Bell Journal of Economics, 14, 441–452.
Latham, G., & Locke, E. A. (1991). Self-regulation through
goal-setting. Organizational Behavior and Human Decision
Processes, 50, 212–247.
Latham, G. P., Mitchell, T. R., & Dossett, D. L. (1978).
Importance of participative goal setting and anticipated
rewards on goal di?culty and job performance. Journal of
Applied Psychology, 63, 163–171.
Lazear, E. P. (1981). Agency, earnings pro?les, productivity,
and hours restrictions. The American Economic Review, 71,
606–620.
Lee, T. W., Locke, E. A., & Phan, S. H. (1997). Explaining the
assigned goal-incentive interaction: the role of self-e?cacy
and personal goals. Journal of Management, 23, 541–559.
Libby, R., Bloom?eld, R., & Nelson, M. W. (2001). Experi-
mental research in ?nancial accounting. Accounting, Organi-
zations and Society (in press).
Libby, R., & Lipe, M. G. (1992). Incentive e?ects and the cog-
nitive processes involved in accounting judgments. Journal of
Accounting Research, 30, 249–273.
Libby, R., & Luft, J. (1993). Determinants of judgment per-
formance in accounting settings: ability, knowledge, motiva-
tion, and environment. Accounting, Organizations and
Society, 5, 425–450.
Libby, R., & Tan, H-T. (1999). Analysts’ reactions to warnings
of negative earnings surprises. Journal of Accounting
Research, 37, 415–435.
Lien, Y., & Cheng, P. (1990). A bootstrapping solution in the
problem of di?erentiating spurious from genuine causes. Paper
presented at the 31st Annual Meeting of the Psychonomic
Society.
Lipe, M. G. (1993). Analyzing the variance investigation deci-
sion: the e?ects of outcomes, mental accounting, and fram-
ing. The Accounting Review, 68, 748–764.
Locke, E. A. (1967). Motivational e?ects of knowledge of
results: knowledge or goal setting? Journal of Applied Psy-
chology, 51, 324–329.
Locke, E. A. (1991). The motivation sequence, the motivation
hub, and the motivation core. Organizational Behavior and
Human Decision Processes, 50, 288–299.
Locke, E. A., Bryan, J. F., & Kendall, L. M. (1968). Goals and
intentions as mediators of the e?ects of ?nancial incentives
on behavior. Journal of Applied Psychology, 52, 104–121.
Locke, E. A., Shaw, K., Saari, L., & Latham, G. (1981). Goal-
setting and task performance: 1969–1980. Psychological Bul-
letin, 90, 125–152.
Locke, E. A., & Latham, G. P. (1990). A theory of goal setting
and task performance. Englewood Cli?s, NJ: Prentice-Hall.
London, M., & Oldham, G. R. (1976). E?ects of varying goal
types and incentive systems on performance and satisfaction.
Academy of Management Journal, 19, 537–546.
London, M., & Oldham, G. R. (1977). A comparison of group
and individual incentive plans. Academy of Management
Journal, 20, 34–41.
Luft, J. (1994). Bonus and penalty incentives: contract choice
by employees. Journal of Accounting and Economics, 18, 181–
206.
Luft, J., & Shields, M. D. (2001a). Mapping management
accounting: making structural models from theory-based
empirical research. Accounting, Organizations and Society (in
press).
Luft, J., & Shields, M. D. (2001b). Why does ?xation persist?
Experimental evidence on the judgment performance e?ects
of expensing intangibles. The Accounting Review, 76 (in
press).
Lundeberg, M. A., Fox, P. W., & Punccohar, J. (1994). Highly
con?dent but wrong: gender di?erences and similarities in
con?dence judgments. Journal of Educational Psychology, 86,
114–121.
Maines, L. A., & McDaniel, L. S. (2000). E?ects of compre-
hensive-income characteristics on nonprofessional investors’
judgments: the role of ?nancial-statement presentation for-
mat. The Accounting Review, 75, 179–207.
Maslow, A. (1943). A theory of human motivation. Psycholo-
gical Review, 50, 370–396.
Mawhinney, T. C. (1979). Intrinsic  extrinsic motivation:
perspectives from behaviorism. Organizational Behavior and
Human Performance, 24, 411–440.
McClelland, D. C., Atkinson, J. W., Clark, R. A., & Lowell,
E. L. (1953). The achievement motive. New York, NY:
Appleton-Century-Crofts.
McDaniel, L. S. (1990). The e?ects of time pressure and audit
program structure on audit performance. Journal of
Accounting Research, 28, 267–285.
McGraw, K. O. (1978). The detrimental e?ects of reward on
performance: a literature review and a prediction model. In
M. R. Lepper, & D. Greene (Eds.), The hidden costs of
reward (pp. 33–60). Hillsdale, NJ: Lawrence Erlbaum.
McGraw, K. O., & McCullers, J. C. (1979). Evidence of a det-
rimental e?ect of extrinsic incentives on breaking a mental
set. Journal of Experimental Social Psychology, 15, 285–294.
McNamara, H. J., & Fisch, R. I. (1964). E?ect of high and low
342 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
motivation on two aspects of attention. Perceptual and
Motor Skills, 19, 571–578.
Merchant, K. A. (1998). Modern management control systems.
Upper Saddle River, NJ: Prentice-Hall.
Meyer, J. P., & Gellatly, I. R. (1988). Perceived performance
norm as a mediator in the e?ect of assigned goal on personal
goal and task performance. Journal of Applied Psychology,
73, 410–420.
Meyer, J. P., Schacht-Cole, B., & Gellatly, I. R. (1988). An
examination of the cognitive mechanisms by which assigned
goals a?ect task performance and reactions to performance.
Journal of Applied Social Psychology, 18, 390–408.
Mikhail, M. B., Walther, B. R., & Willis, R. H. (1999). Does
forecast accuracy matter to security analysts?. The Account-
ing Review, 74, 185–200.
Milgrom, P. R., & Roberts, J. (1992). Economics, organization
& management. Upper Saddle River, NJ: Prentice-Hall.
Mills, T. Y. (1996). The e?ect of cognitive style on external
auditors’ reliance decisions on internal audit functions.
Behavioral Research in Accounting, 8, 49–73.
Miner, J. B. (1980). Theories of organizational behavior. Hins-
dale, IL: The Dryden Press.
Montague, W. E., & Webber, C. E. (1965). E?ects of knowl-
edge of results and di?erential ?nancial reward on six unin-
terrupted hours of monitoring. Human Factors, 7, 173–180.
Navon, D., & Gopher, D. (1979). On the economy of the human
processing system. Psychological Review, 86, 214–255.
Naylor, J. C., & Clark, R. D. (1968). Intuitive inference strate-
gies in interval learning tasks as a function of validity mag-
nitude and sign. Organizational Behavior and Human
Performance, 3, 378–399.
Naylor, J. C., & Dickinson, T. L. (1969). Task structure, work
structure, and team performance. Journal of Applied Psy-
chology, 53, 167–177.
Naylor, J. C., Pritchard, R. D., & Ilgen, D. R. (1980). A theory
of behavior in organizations. New York, NY: Academic Press.
Naylor, J. C., & Schenk, E. A. (1968). The in?uence of cue
redundancy upon the human inference process for tasks of
varying degrees of predictability. Organizational Behavior
and Human Decision Processes, 3, 47–61.
Nelson, M. W. (1993). The e?ects of error frequency and
accounting knowledge on error diagnosis in analytical
review. The Accounting Review, 68, 804–824.
Nelson, M. W., Libby, R., & Bonner, S. E. (1995). Knowledge
structure and the estimation of conditional probabilities in
audit planning. The Accounting Review, 70, 221–240.
Newell, A., & Simon, H. A. (1972). Human problem solving.
Englewood Cli?s, NJ: Prentice Hall.
Pavlik, E. L., Scott, T. W., & Tiessen, P. (1993). Executive
compensation: issues and research. Journal of Accounting
Literature, 12, 131–189.
Payne, J. W., Bettman, J. R., & Johnson, E. J. (1990). The
adaptive decision maker. New York, NY: Cambridge Uni-
versity Press.
Peecher, M. (1996). The in?uence of auditors’ justi?cation pro-
cesses on their decisions: a cognitive model and experimental
evidence. Journal of Accounting Research, 34, 125–140.
Pelham, B. W., & Neter, E. (1995). The e?ect of motivation of
judgment depends on the di?culty of the judgment. Journal
of Personality and Social Psychology, 68, 581–594.
Peters, J. M. (1993). Decision making, cognitive science and
accounting: an overview of the intersection. Accounting,
Organizations and Society, 18, 383–405.
Phillips, J. S., & Lord, R. G. (1980). Determinants of intrinsic
motivation: locus of control and competence information as
components of Deci’s cognitive evaluation theory. Journal of
Applied Psychology, 65, 211–218.
Pincus, K. V. (1990). Auditor individual di?erences and fair-
ness of presentation judgments. Auditing: A Journal of Prac-
tice & Theory, 9, 150–166.
Pollack, I., & Kna?, P. R. (1958). Maintenance of alertness by
a loud auditory signal. Journal of the Acoustical Society, 30,
1013–1016.
Prawitt, D. F. (1995). Sta?ng assignments for judgment-orien-
ted audit tasks: the e?ects of structured audit technology and
environment. The Accounting Review, 70, 443–465.
Prendergast, C. (1999). The provision of incentives in ?rms.
Journal of Economic Literature, 37, 7–63.
Pritchard, R. D., & Curtis, M. I. (1973). The in?uence of goal
setting and ?nancial incentives on task performance. Orga-
nizational Behavior and Human Performance, 10, 175–183.
Pritchard, R. D., & DeLeo, P. J. (1973). Experimental test of
the valence–instrumentality relationship in job performance.
Journal of Applied Psychology, 57, 264–270.
Pritchard, R. D., Jones, S. D., Roth, P. L., Stuebing, K. K., &
Ekeberg, S. E. (1988). E?ects of group feedback, goal setting,
and incentives on organizational productivity [Monograph].
Journal of Applied Psychology, 73, 337–358.
Pritchard, R. D., Leonard, D. W., Von Bergen Jr., C. W., &
Kirk, R. J. (1976). The e?ects of varying schedules of rein-
forcement on human task performance. Organizational
Behavior and Human Performance, 16, 205–230.
Reber, A. S. (1995). The Penguin dictionary of psychology.
London, UK: Penguin Books.
Riedel, J. A., Nebeker, D. M., & Cooper, B. L. (1988). The
in?uence of ?nancial incentives on goal choice, goal com-
mitment, and task performance. Organizational Behavior and
Human Decision Processes, 42, 155–180.
Riley, J. G. (1979). Informational equilibrium. Econometrica,
47, 331–360.
Roth, A. E. (1995). Introduction to experimental economics. In
A. E. Roth, & J. H. Kagel (Eds.), The handbook of experi-
mental economics (pp. 3–109). Princeton, NJ: Princeton Uni-
versity Press.
Rummel, A., & Feinberg, R. (1988). Cognitive evaluation the-
ory: a meta-analytic review of the literature. Social Behavior
and Personality, 16, 147–164.
Salop, S., & Salop, J. (1976). Self-selection and turnover in the
labor market. Quarterly Journal of Economics, 90, 619–627.
Schwartz, B. (1988). The experimental synthesis of behavior:
reinforcement, behavioral stereotypy, and problem solving.
The Psychology of Learning and Motivation, 22, 93–138.
Scott Jr., W. E., Farr, J., & Podsako?, P. M. (1988). The e?ects
of ‘intrinsic’ and ‘extrinsic’ reinforcement contingencies on
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 343
task behavior. Organizational Behavior and Human Decision
Processes, 41, 405–425.
Shields, M. D. (2001). Operating budgeting systems. Working
paper, Michigan State University.
Shields, M. D., Chow, C. W., & Whittington, O. R. (1989). The
e?ects of state risk and controllability ?lters on compensa-
tion contract and e?ort choice. Abacus, 25, 39–55.
Shields, M. D., Deng, F. J., & Kato, Y. (2000). The design and
e?ects of control systems: tests of direct- and indirect-e?ects
models. Accounting, Organizations and Society, 25, 185–202.
Shields, M. D., & Waller, W. S. (1988). An experimental labor
market study of accounting variables in employment con-
tracting. Accounting, Organizations and Society, 13, 581–594.
Siegel, S. (1961). Decision making and learning under varying
conditions of reinforcement. Annals of the New York Acad-
emy of Science, 89, 766–783.
Simnett, R. (1996). The e?ect of information selection, informa-
tion processing and task complexity on predictive accuracy of
auditors. Accounting, Organizations and Society, 21, 699–719.
Simon, H. A. (1986). Rationality in psychology and economics
[special issue]. Journal of Business, 59, S209–S224.
Sipowicz, R. R., Ware, J. R., & Baker, R. A. (1962). The e?ects
of reward and knowledge of results on the performance of a
simple vigilance task. Journal of Experimental Psychology,
64, 58–61.
Skinner, B. F. (1953). Science and human behavior. New York,
NY: Macmillan.
Smith, V. L. (1982). Microeconomic systems as an experimental
science. The American Economic Review, 72, 923–955.
Smith, V. L. (1991). Rational choice: the contrast between
economics and psychology. Journal of Political Economy, 99,
877–897.
Smith, V. L., & Walker, M. (1993). Monetary rewards and
decision cost in experimental economics. Economic Inquiry,
31, 245–261.
Spence, A. M. (1973). Job market signaling. Quarterly Journal
of Economics, 87, 355–374.
Spilker, B. C. (1995). The e?ects of time pressure and knowl-
edge on key word selection behavior in tax research. The
Accounting Review, 70, 49–70.
Sprinkle, G. B. (2000). The e?ect of incentive contracts on
learning and performance. The Accounting Review, 75, 299–
326.
Sprinkle, G. B. (2001). Perspectives on experimental research in
managerial accounting. Accounting, Organizations and
Society (in press).
Stajkovic, A. D., & Luthans, F. (1998). Self-e?cacy and work-
related performance: a meta-analysis. Psychological Bulletin,
124, 240–261.
Stone, D. N., & Ziebart, D. A. (1995). A model of ?nancial
incentive e?ects in decision making. Organizational Behavior
and Human Decision Processes, 61, 250–261.
Tan, H., & Libby, R. (1997). Tacit managerial versus technical
knowledge as determinants of audit expertise in the ?eld.
Journal of Accounting Research, 35, 97–113.
Tang, S., & Hall, V. C. (1995). The overjusti?cation e?ect: a
meta-analysis. Applied Cognitive Psychology, 9, 365–404.
Terborg, J. R., & Miller, H. E. (1978). Motivation, behavior,
and performance: a closer examination of goal setting and
?nancial incentives. Journal of Applied Psychology, 63, 29–
39.
Thaler, R. H. (1992). The winner’s curse: paradoxes and
anomalies of economic life. New York, NY: The Free Press.
Tomporowski, P. D., Simpson, R. G., & Hager, L. (1993).
Method of recruiting subjects and performance on cognitive
tests. American Journal of Psychology, 106, 499–521.
Toppen, J. T. (1965a). E?ects of size and frequency of money
reinforcement on human operant (work) behavior. Percep-
tual and Motor Skills, 20, 1193–1199.
Toppen, J. T. (1965b). Money reinforcement and human oper-
ant (work) behavior: II. Within-subjects comparisons. Per-
ceptual and Motor Skills, 21, 907–913.
Toppen, J. T. (1966). Money reinforcement and human operant
(work) behavior: IV. Temporally extended within-S compar-
isons. Perceptual and Motor Skills, 22, 575–581.
Tubbs, M. E. (1986). Goal-setting: a meta-analytic examination
of the empirical evidence. Journal of Applied Psychology, 71,
474–483.
Tuttle, B., & Burton, G. (1999). The e?ects of a modest incen-
tive on information overload in an investment analysis task.
Accounting, Organizations and Society, 24, 673–687.
Tversky, A., & Edwards, W. (1966). Information versus reward
in binary choices. Journal of Experimental Psychology, 71,
680–683.
Valance, N. (2001). Bright minds, big theories. CFO, 17, 64–70.
Vecchio, R. P. (1982). The contingent-noncontingent compen-
sation controversy: an attempt at resolution. Human Rela-
tions, 35, 449–462.
Vera-Mun˜ oz, S. C. (1998). The e?ects of accounting knowledge
and context on the omission of opportunity costs in resource
allocation decisions. The Accounting Review, 73, 47–72.
Vera-Mun˜ oz, S. C., Kinney, W. R., & Bonner, S. E. (2001).
Assuring information relevance: the e?ects of domain-spe-
ci?c knowledge and task presentation format. The Account-
ing Review, 76 (in press).
Vroom, V. (1964). Work and motivation. New York, NY: John
Wiley.
Wall Street Journal. (1999). Linking pay to performance is
becoming a norm in the workplace. 115(April 6), 1.
Waller, W., & Chow, C. W. (1985). The self-selection and e?ort
e?ects of standard-based employment contracts: a frame-
work and some empirical evidence. The Accounting Review,
60, 458–476.
Weiner, B. (1966). Motivation and memory. Psychological
Monographs: General and Applied, 80, 1–22.
Weiner, B. (1972). Theories of motivation: from mechanism to
cognition. Chicago, IL: Markham.
Weiner, B., & Walker, E. L. (1966). Motivational factors in
short-term retention. Journal of Experimental Psychology, 71,
190–193.
Weiner, M. J., & Mander, A. M. (1978). The e?ects of reward
and perception of competency upon intrinsic motivation.
Motivation and Emotion, 2, 67–73.
Weiss, A. W. (1990). E?ciency wages: models of unemployment,
344 S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345
layo?s, and wage dispersion. Princeton, NJ: Princeton Uni-
versity Press.
Wickens, D. D., & Simpson, C. K. (1968). Trace cue position,
motivation, and short-term memory. Journal of Experimental
Psychology, 76, 282–285.
Wiener, E. L. (1969). Money and the monitor. Perceptual and
Motor Skills, 29, 627–634.
Wiersma, U. J. (1992). The e?ects of extrinsic rewards in
intrinsic motivation: a meta-analysis. Journal of Occupational
and Organizational Psychology, 65, 101–114.
Wimperis, B. R., & Farr, J. L. (1979). The e?ects of task content
and reward contingency upon task performance and satisfac-
tion. Journal of Applied Social Psychology, 9, 229–249.
Wood, R. E. (1986). Task complexity: de?nition of the con-
struct. Organizational Behavior and Human Decision Pro-
cesses, 37, 60–82.
Wood, R. E., Mento, A. J., & Locke, E. A. (1987). Task com-
plexity as a moderator of goal e?ects: a meta-analysis. Jour-
nal of Applied Psychology, 72, 416–425.
Wright, P. M. (1989). Test of the moderating role of goals in
the incentive-performance relationship. Journal of Applied
Psychology, 74, 699–705.
Wright, P. M. (1990). Monetary incentives and task experience
as determinants of spontaneous goal setting, strategy devel-
opment, and performance. Human Performance, 3, 237–258.
Wright, P. M. (1992). An examination of the relationships
among monetary incentives, goal level, goal commitment,
and performance. Journal of Management, 18, 677–693.
Wright, P. M., & Kacmar, K. M. (1995). Moderating roles of
self-set goals, goal commitment, self-e?cacy, and attractive-
ness in the incentive-performance relation. Human Perfor-
mance, 8, 263–296.
Wright, W. F., & Aboul-Ezz, M. E. (1988). E?ects of extrinsic
incentives on the quality of frequency assessments. Organi-
zational Behavior and Human Decision Processes, 41, 143–
152.
Yellen, J. L. (1984). E?ciency wage models of unemployment.
The American Economic Review, 74, 200–205.
Yerkes, R. M., & Dodson, J. D. (1908). The relation of
strength of stimulus to rapidity of habit formation. Journal
of Comparative Neurology and Psychology, 18, 459–482.
Young, S. M. (1985). Participative budgeting: the e?ects of risk
aversion and asymmetric information on budgetary slack.
Journal of Accounting Research, 23, 829–842.
Young, S. M. (2001). The e?ects of earnings predictability on
optimism and accuracy in forecasts of ?nancial analysts.
Working paper, Emory University.
Young, S. M., & Lewis, B. L. (1995). Experimental incentive-
contracting research in managerial accounting. In
R. H. Ashton, & A. H. Ashton (Eds.), Judgment and decision
making research in accounting and auditing (pp. 55–75). New
York, NY: Cambridge University Press.
Zelizer, V. (1994). The social meaning of money. New York,
NY: Basic Books.
Zimmerman, J. L. (2000). Accounting for decision making and
control (3rd ed.). Boston, MA: Irwin/McGraw-Hill.
S.E. Bonner, G.B. Sprinkle / Accounting, Organizations and Society 27 (2002) 303–345 345

doc_269326535.pdf