Management accounting in the laboratory and in social context

Description
Experimental research in management accounting (MA) has
changed substantially over the last forty years, and Accounting,
Organizations, and Society has played a significant role in this
change. The present paper explores four basic contrasts in MA
experimental research between the beginning and end of this
forty-year period. The first two of these contrasts illustrate ways
in which the field has expanded and enriched its approaches to
MA and thus added to our understanding, while the other two
contrasts highlight ways in which the field has narrowed and thus
has left us with important unanswered questions.

Management accounting in the laboratory and in social context: Four
contrasts, 1975–2014
Joan Luft
Michigan State University, United States
a r t i c l e i n f o
Article history:
Received 10 July 2015
Accepted 7 August 2015
Available online xxxx
a b s t r a c t
Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction
Experimental research in management accounting (MA) has
changed substantially over the last forty years, and Accounting,
Organizations, and Society has played a signi?cant role in this
change. The present paper explores four basic contrasts in MA
experimental research between the beginning and end of this
forty-year period. The ?rst two of these contrasts illustrate ways
in which the ?eld has expanded and enriched its approaches to
MA and thus added to our understanding, while the other two
contrasts highlight ways in which the ?eld has narrowed and thus
has left us with important unanswered questions.
The ?rst way in which experimental approaches to MA have
been enriched is a set of changes in how experiments represent
the people who use accounting. These users now appear as social
beings with complex and somewhat changeable motivations, not
as isolated operators of stable (probably pro?t-maximizing)
decision models. The second—and related—set of changes has to
do with the way in which MA experiments represent the roles of
accounting in organizations. Many early experiments treated
accounting narrowly as ‘‘an answer machine,” in Burchell, Clubb,
Hopwood, Hughes, and Nahapiet’s (1980) terms: that is, MA pro-
vided numbers—variable values—to slot into pre-existing decision
models, which would then provide managers with answers to
questions about how to price products, make capital investments,
and so on. More recent experiments have documented more
diverse roles for MA: for example, it can help to shape preferences,
to structure people’s mental representations of their work and
environment, and to support or hinder the formation of social
identities (Luft & Shields, 2009).
This enrichment and broadening of experimental research in
some directions has been accompanied by a narrowing of focus
in other directions—perhaps unsurprisingly, as the attention of
researchers has limited scope at any one time. Fundamental
research questions about MA that were addressed relatively fre-
quently by experiments thirty or forty years ago hardly appear at
all in this literature now, although it is not self-evident that these
questions are either unimportant or unaddressable or already
answered.
The ?rst of these un?nished business areas has to do with
Demski and Feltham’s (1976) distinction between decision-facili-
tating and decision-in?uencing roles of accounting (see Sprinkle
(2003) for de?nitions and examples of these roles in MA experi-
ments). In the late 1970s and early 1980s, a majority of the MA
experiments in major journals addressed decision-facilitating roles
of MA. In contrast, in more recent years, the situation has reversed:
a relatively large, robust, and coherent experimental literature
addresses decision-in?uencing roles, while the recent experimen-
tal literature on decision-facilitating uses has been relatively small
and fragmented.
Second, although experiments have always been more likely to
address the effects of MA than its causes, the gap between these
two foci of research has widened in recent years. Most experiments
try to answer questions about what will happen as a result of using
one type of MA rather than another (MA is an independent
variable). How it comes about—by deliberate design or spontaneous
processes—that organizations use one type of MA rather than
another (MA as a dependent variable) is not a question that
experiments in recent years have been very likely to investigate.
To examine these four contrasts in more detail, the rest of this
article proceeds as follows. As an initial overview, the next section
provides a graphic illustration of the growth of MA experiments
between the early and late years of this forty-year period in the
research communities that have clustered around different major
journals. The two following sections present in more detail the
increase in diversity and richness in experiments’ representations
of the users and roles of MA. The subsequent section examines
shifts in experimental research from decision-facilitating to
decision-in?uencing roles of MA and from causes to effects of
MA; it also identi?es some possible approaches for addressinghttp://dx.doi.org/10.1016/j.aos.2015.08.001
0361-3682/Ó 2015 Elsevier Ltd. All rights reserved.
E-mail addresses: [email protected], [email protected]
Accounting, Organizations and Society xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Accounting, Organizations and Society
j our nal homepage: www. el sevi er . com/ l ocat e/ aos
Please cite this article in press as: Luft, J. Management accounting in the laboratory and in social context: Four contrasts, 1975–2014. Accounting, Organiza-
tions and Society (2015),http://dx.doi.org/10.1016/j.aos.2015.08.001
the un?nished business that these changes have left behind. The
last section provides a brief summary and conclusion.
2. Publication of MA experiments: Contrasts between 1976–
1985 and 2005–2014
Some of the changes in MA experiments in the last forty years
are readily visible in a graph of publication patterns in major
journals at the beginning and end of this period. I begin by simply
counting the number of MA experiments that appeared in the three
highest-impact English-language journals that have regularly
published such experiments in the last forty years: Accounting,
Organizations, and Society, of Accounting Research, and The
Accounting Review.
1
For purposes of this count, I de?ne MA as
accounting in organizations, including both accounting information
as such and the organizational processes in which it is involved
(e.g., budgeting, performance evaluation), and including the man-
agement of close inter-organizational relations in joint ventures
and supply chains. The count does not include experiments that
address purely methodological issues relevant to MA but without
any actual MA content (e.g., a comparison of methods for eliciting
subjective probabilities in abstract settings). The count also does
not include experiments that examine how ?nancial-market
participants such as investment bankers or ?nancial analysts might
use MA information if they have it.
Fig. 1 presents the results of these counts for four ?ve-year
periods: 1976–1980, 1981–1985, 2005–2009, and 2010–2014. I
sum the counts by ?ve-year periods to avoid the noise from
small-sample ?uctuations that would appear in year-by-year
counts, and I omit the middle years of the forty-year period in
order to make the contrast more visible between early and recent
experimental studies of MA.
As Fig. 1 shows, trends in the publication of MA experiments are
quite different in the three journals. The Journal of Accounting
Research was the primary venue for such studies in the late
1970s and early 1980s but has rarely published them in recent
years, while an opposite trend appears for The Accounting Review.
In the absence of further information, we might suppose that
Fig. 1 represents a migration of a particular type of studies from
one US journal to another, perhaps as editorial tastes change, while
AOS holds a steady course in the middle, less subject to large
?uctuations in taste.
A closer examination of the studies represented in these counts,
however, will reveal that this is not the case. Rather, the particular
type of MA experiments that JAR stopped publishing were not pub-
lished anywhere else either after the mid 1980s. The type of MA
experiments that took off around the turn of the millennium were
very different, in ways this article intends to explore.
The role of AOS in these changes was twofold. First, it provided a
regular publication venue for a broad variety of research, including
MA experiments of various types. It was a seedbed: it offered a
place where people could continually try out different ways of
doing MA experimentation, so that if one approach came to a dead
end, there were at least preliminary examples of alternative
approaches available that might be developed. Second, and equally
important, AOS in the 1980s published a number of much-cited
qualitative theoretical articles (e.g., Burchell et al., 1980; Cooper,
Hayes, & Wolf, 1981; Hopwood, 1983; March, 1987; Roberts &
Scapens, 1985) that took issue with the rather narrow view of
MA and its users that was prevalent in much of the early
experimental literature.
As the next sections will illustrate in more detail, the kind of
experimentation that took off around the turn of the millennium
had a view of MA and its users that was in many ways more like
the view represented in these much-cited AOS articles from the
1980s than it was like the view represented in earlier MA experi-
ments. As I argue in the following sections, taking this different
view enabled researchers to make a variety of interesting and
valuable contributions.
3. Changing representations of the users of MA
The ?rst of the two major (related) changes that have enriched
MA experiments over the last forty years is a change in the way
that the users of MA are represented. Experiments are typically
close-up, sharp-focus pictures of a small segment of human activ-
ity, not comprehensive overviews, and early MA experiments tend
to bracket out many questions about the nature of users—that is, to
set these questions aside unanswered—in order to focus attention
on other issues.
What is in the focus of experimenters in the ?rst ten years of our
forty-year period, besides basic experimental methods,
2
is typically
users’ cognition, conceptualized as the more or less skilled use of
decision models, and in particular, the use of accounting inputs to
these models. For example, early experiments examine product costs
calculated by different accounting methods as inputs to product
pricing models (Ashton, 1976, 1981; Bloom et al., 1984; Dyckman
et al., 1982), opportunity costs as inputs to investment models
(Neumann & Friedman, 1978, 1980; Hoskin, 1983), cost information
as inputs to variance-investigation models (optimizing or heuristic:
Brown, 1981, 1983; Jacobs, 1978; Lewis, Shields, & Young, 1983;
0
2
4
6
8
10
12
14
16
1976-80 1981-85 2005-09 2010-14
AOS
JAR
TAR
Fig. 1. Number of MA experimental studies published in three major journals,
1976–1985 and 2005–2014. The data points represent the total number of
experimental studies on MA published during the designated time periods in each
of three journals: Accounting, Organizations, and Society, Journal of Accounting
Research, and The Accounting Review. Broken lines between 1985 and 2005 highlight
the fact that no data is presented for this interim period.
1
This count includes all laboratory and ?eld experiments in MA that were
published as research articles or research reports; publications in separate ‘‘Notes” or
‘‘Capsules and Comments” sections of the journals are omitted. Two other journals
that are commonly regarded as internationally signi?cant, Contemporary Accounting
Research and the Journal of Accounting and Economics, are omitted from these counts.
CAR did not begin publication until 1984, and thus cannot provide comparative data
for the ?rst ten years of the forty-year period; and JAE has published almost no
experiments. Experiments published in CAR are included in the discussion of recent
research in later sections of this article.
2
Some of the dif?culties of establishing credible and usable experimental
approaches to MA research can be seen in the debates over methods in research on
functional ?xation (Ashton, 1976; Bloom, Elgers, & Murray, 1984; Dyckman, Hoskin, &
Swieringa, 1982; Libby, 1976; Wilner & Birnberg, 1986) and the use of opportunity
costs (Becker, Ronen, & Sorter, 1974; Neumann & Friedman, 1978, 1980; Hoskin,
1983).
2 J. Luft / Accounting, Organizations and Society xxx (2015) xxx–xxx
Please cite this article in press as: Luft, J. Management accounting in the laboratory and in social context: Four contrasts, 1975–2014. Accounting, Organiza-
tions and Society (2015),http://dx.doi.org/10.1016/j.aos.2015.08.001
Magee & Dickhaut, 1978), or sample draws of balls from urns–ab-
stract representations of uncertainty-reducing MA information—as
inputs to generic decision models (Hilton, Swieringa, & Hoskin,
1981; Hilton & Swieringa, 1981; Uecker, 1978, 1980, 1982).
What is outside the focus of experimenters during this period,
for the most part, is either the social (including organizational)
context or the motivation of the users of MA. These users are rep-
resented as isolated operators of decision models, who have no rel-
evant personal relations with anyone else in their organizations.
Insofar as their own objectives are considered in early MA
experiments, these objectives are often assumed to be adequately
represented by the decision models: that is, they are assumed to be
clear, stable, and identical with organizational objectives such as
pro?t-making or compliance with speci?c organizational policies.
A number of AOS articles in the 1980s critiqued such represen-
tations of MA users, arguing that individuals’ objectives are
sometimes uncertain (Burchell et al., 1980), and their preferences
can be vague and changeable, in?uenced by as well as in?uencing
decision processes (March, 1987). Hence accounting can play a role
in ‘‘the construction of organizational participants’ views of the
desirable” (Hopwood, 1983, p. 291). These articles also argue that
‘‘views of the desirable” are constructed in a social process, rather
than being essentially a matter of individual tastes.
More recent MA experiments characterize the users of MA in
ways that are more consistent with this critique than with the
early laboratory experiments. The MA users in more recent exper-
iments are represented as individuals with sometimes uncertain
and changeable preferences, whose ‘‘views of the desirable” can
be in?uenced by accounting. MA often acts on individuals’ views
through processes such as social comparison, the formation of
social identities, and the invocation of social norms. The remainder
of this section provides more detail on this contrast between early
and recent experimentations. It also illustrates some of the gains in
understanding MA that have arisen from combining this theoreti-
cally richer view of MA users with the close-up, sharp-focus tool
of experimentation.
3.1. Developing an interest in MA users’ motivation
Early MA experiments typically did not ask the question, ‘‘What
do users of MA want—what motivates their actions?”
3
One way of
not asking the question consisted of giving participants a decision
task and either telling them that their organization had a particular
policy for making the decision (e.g., Ashton, 1976; Bloom et al., 1984)
or expecting that they would use the policy in place for making sim-
ilar decisions in their own organization (e.g., Harrell & Klick, 1980).
The assumption was that individuals would want to follow the pol-
icy as best they could, and that the effects of other wants were not a
major concern for the experimenter. A second way of not asking
questions about motivation, which appears in later studies during
this initial period, was to provide a monetary incentive that was
intended to be dominant, following the precepts of experimental
economics (e.g., Schepanski & Uecker, 1983; Waller & Chow, 1985).
The objective of using these monetary incentives was typically not
to study motivation as such, but to set aside motivation as a ques-
tion—to control for it so that researchers could focus their attention
elsewhere.
Motivation gradually became a more important question for MA
experimenters, however, for multiple reasons. First, the growth of
principal-agent models as a theoretical approach to MApersistently
called attention to con?icting motivations within organizations as
an in?uence on MA, making it dif?cult to justify experiments that
assumed employees automatically followed organizational policies
and objectives. Second, questions about motivation resisted being
set aside: monetary incentives that were intended to dominate in
experiments sometimes failed to do so, and failed in interesting
ways. By the end of the ?rst ten years of the forty-year period under
consideration, experiments began to document the effects of other
factors that limited the dominance of performance-based pay: for
example, the effects of social pressure in budgeting (Young, 1985)
and the construction of preferences in context, using control-
system elements as cues (Tiller, 1983).
Third, by the last ten years of the forty-year period under
consideration, theoretical and methodological tools were readily
available from behavioral economics, where skilled experimenters
had already devoted considerable attention to addressing ques-
tions about the effects of diverse motivations (e.g., social norms,
social comparisons) in experimental settings that seemed readily
adaptable to MA questions.
4
And ?nally, as the MA experimental
literature developed further, it became evident that motivation
issues and standard MA issues like measurement quality and
control-system design were so deeply entangled with each other—
interactions were so frequent and substantial—that there would be
bene?ts from researching the two sets of issues together.
From the end of the 1980s onward, there was substantial
growth in experimental investigations of how MA in?uences and
is in?uenced by diverse motivations. This literature began simply
by identifying and documenting MA-related consequences of what
behavioral economists call social preferences: individuals’ willing-
ness to incur costs in order to make advantageous social compar-
isons of themselves with others (e.g., Frederickson, 1992), to
avoid what they see as unfair or dishonest behavior (e.g., Luft &
Libby, 1997; Evans, Hannan, Krishnan, & Moser, 2001), or to
maintain their social identity (Towry, 2003; Rowe, 2004).
Luft and Shields (2009) provide a more detailed presentation of
the growth of this literature up to the mid-2000s; in the present
article I focus on a selection of more recent studies
5
that cluster
around two themes.
First, knowing more about what people want besides money
helps us better understand the effects of money—that is, of orga-
nizations’ widespread use of accounting-based monetary incen-
tives. In some cases other motivations (e.g., social comparisons,
reciprocity, autonomy) leverage the effects of monetary incen-
tives, making them a more powerful motivator than would be
expected based on the consumption utility of money alone. In
other cases, however, social preferences mute or distort the
intended effects of accounting-based monetary incentives.
Which of these outcomes actually occur can depend on MAprop-
erties such as measurement quality and elements of control-sys-
tem design.
Second, social institutions (e.g., organizations, work teams,
labor markets with speci?c characteristics) help to determine
3
A few experiments in the ?rst ten-year period investigate motivation as such,
inspired by survey research on motivational effects of participative budgeting.
Brownell (1981) ?nds that participation is motivating for individuals with an internal
locus of control but demotivating for those with an external locus of control; and
Tiller (1983) ?nds that participation can increase commitment and performance via
cognitive dissonance effects.
4
For example, behavioral-economics literature on fairness concerns in pricing and
negotiation (e.g., Kahneman, Knetsch, & Thaler, 1986; Thompson & Loewenstein,
1992) was adapted for laboratory investigation of the detailed working of fairness
concerns in transfer pricing negotiations (Chang, Cheng, & Trotman, 2008;
Kachelmeier & Towry, 2002; Luft & Libby, 1997). Field research had reported the
existence of strong fairness concerns in transfer pricing (Eccles, 1985) but not the
speci?c judgment and bargaining mechanisms through which these concerns
in?uenced prices.
5
Studies published before 2009 are not described except insofar as needed to
explain the foundations of later work. The descriptions of more recent studies that
appear below are intended as representative examples of the ‘‘four contrasts” theme
of this article, not as a comprehensive account of recent research.
J. Luft / Accounting, Organizations and Society xxx (2015) xxx–xxx 3
Please cite this article in press as: Luft, J. Management accounting in the laboratory and in social context: Four contrasts, 1975–2014. Accounting, Organiza-
tions and Society (2015),http://dx.doi.org/10.1016/j.aos.2015.08.001
the effects of incentives by in?uencing individuals’ interpreta-
tions of events and accounting information. Moreover, because
incentive plans are social creations rather than automatic
mechanisms, how they work depends on the desires and beliefs
of the superiors who operate these systems as well as the
desires and beliefs of the subordinates whom they are intended
to in?uence.
The remainder of the section elaborates on these themes, pro-
viding examples of how the up-close, sharp-focus approach of
experimentation enables researchers to tease out complex effects
of multiple motivations—speci?cally, how monetary incentives
combine with preferences for favorable social comparisons,
reciprocity, and autonomy to in?uence individual decisions.
3.2. Social comparison as a source of motivation
Performance, as measured and recognized by the organization,
is likely to be important to individuals not only because it determi-
nes their monetary payoffs but also because it affects their self-
identity. Monetary compensation is unlikely to be the only reason
that managers care about answers to the question, ‘‘Am I a good
manager or not?” This question is often answered by social com-
parisons: it becomes, in effect, ‘‘Am I at least as good as the average
of my peers?” or ‘‘Am I better than most of them?” The desire to
perform better than others, and to be seen as performing better
than others (or at least not conspicuously worse) can be a signi?-
cant motivator even when relative performance has no effect on
monetary payoffs.
This does not mean, however, that social comparisons and mon-
etary payoffs act independently of each other, as an experiment by
Tafkov (2013) neatly illustrates. Participants in his experiment
perform a repeated problem-solving task; all receive periodic
information on how well they are performing absolutely, and some
also receive periodic relative performance information (RPI), i.e.,
how their performance ranks compared with the four other
participants in the group to which they are assigned. Although
the presence of this information has no impact on participants’
monetary payoffs, it does prompt more effort—as we might expect,
given prior evidence on the effort effects of social comparison
(e.g., Frederickson, 1992).
What is novel in Tafkov’s (2013) study is the way in which the
effort-increasing effects of social comparison are leveraged by two
control-system features that make comparisons of performance
more important in participants’ minds. When the control system
attaches monetary rewards to (absolute) measured performance
rather than paying a ?xed wage, and when the system discloses
RPI to more people in the organization, it makes RPI matter more
to individuals’ self-image, and thus the motivational power of
social comparison increases.
6
Brown, Fisher, Sooy, and Sprinkle (2014) provide another exam-
ple, in a budgeting setting, of how MA can alter the shape and force
of social comparisons in in?uencing behavior. In this experiment,
participants have private information and can earn extra compen-
sation for themselves by misrepresenting this information to
secure a larger budget. If they receive information about how their
compensation ranks relative to others, then compensation, as the
basis of social comparison, becomes a more important dimension
of performance in their minds. Hence they misrepresent their pri-
vate information more, compared to a baseline condition in which
the monetary payoffs for dishonesty are the same but are not rein-
forced by compensation rankings. In contrast, if participants
receive rankings of ?rm pro?t, which increases with honest report-
ing, then pro?t is the basis of social comparison; hence it is more
important dimension of performance in participants’ minds, and
their mean level of honesty is substantially higher.
7
Tafkov (2013) and Brown et al. (2014) present social compar-
ison as part of an organizational control-system toolkit that can
be used to in?uence employee behavior in desired ways. Hannan,
McPhree, Newman, and Tafkov (2013) provide a useful reminder
that individuals will also choose their own social comparisons
and that these will not necessarily foster the smooth functioning
of organizational control systems. In this experiment, participants
allocate effort between two tasks; they are likely to be better at
one than the other, due to basic ability and interest differences.
They are told that the organization would like them to allocate
effort equally between the two tasks, and accordingly their com-
pensation is highest if they do so. However, when they—and
others—know how their performance ranks on each task relative
to others in the organization, they shift their effort allocations
somewhat away from equality toward the task at which they are
better performers and thus have a better chance of outranking
peers. This result presumably contributes to their positive self-
image but not to their cash rewards or to the organizational goal.
8
3.3. Reciprocity as source of motivation
Another important motivator of action in both laboratory and
natural environments, in addition to social comparison, is the norm
of reciprocity (see Luft and Shields (2009) for relevant MA labora-
tory experiments, and Falk (2007) for evidence from a ?eld exper-
iment). Recent MA experiments examining this source of
motivation have emphasized the fact that contract terms must be
interpreted by the contracting parties: an expected payoff of $X in
return for a certain set of actions, for example, can mean different
things in different social-institutional contexts, and thus can
prompt different behavior. In recent experiments, the labor market
provides an important social context for the interpretation of con-
tract terms.
Kuang and Moser (2009) provide an example that investigates
the effects of two contrasting employment contracts. One contract
is a theoretically optimal agency contract, which offers the agent
the minimum payoff required to make him better off than his next
best alternative. This contract is not expected to motivate much
effort because—for reasons beyond either the agent’s or the princi-
pal’s control—the agent’s actions cannot be very effectively moni-
tored in the setting under consideration. The other contract is a
‘‘gift exchange” contract in which the principal offers a higher
wage and the agent may (or may not) observe the social norm of
reciprocity by offering more than the minimum enforceable effort.
How agents act and how much principals earn under each of
these contracts depends not only on the formal provisions of the
contract but also on the labor-market context. When each contract
is the only kind of contract available in the market, principals earn
signi?cantly more under the optimal than the gift-exchange con-
tract. This is not the case, however, when both types of contract
can be offered in the same market. The monetary bene?ts and costs
of effort under the optimal contract are the same for agents in both
single-contract and mixed-contract markets; but in mixed-
contract markets the ungenerous nature of the optimal contract
6
The two factors interact, such that the effect of performance-based pay is stronger
when RPI is public.
7
Interestingly, if rankings of both ?rm pro?t and compensation are provided,
honesty levels are similar to those in the condition with only ?rm pro?t rankings. The
authors speculate that this occurs because the two rankings together make the
‘‘antisocial” quality of individual compensation-maximizing more salient; however,
the experiment is not designed to provide explicit evidence on this point.
8
Similar to Tafkov (2013), the distortionary effect occurs even when RPI is
private—when only the individual will know whether he or she performs better than
peers—but is stronger when RPI is public.
4 J. Luft / Accounting, Organizations and Society xxx (2015) xxx–xxx
Please cite this article in press as: Luft, J. Management accounting in the laboratory and in social context: Four contrasts, 1975–2014. Accounting, Organiza-
tions and Society (2015),http://dx.doi.org/10.1016/j.aos.2015.08.001
is more salient to agents, as is the principal’s freedom to choose
more or less generous contracts. Thus in the mixed-contract mar-
ket, the principal’s offer of an optimal contract is more likely to
be interpreted by the agent as a hostile act, and agents more read-
ily interpret their own low effort choices as justi?ed negative
reciprocity. In consequence, they are more likely to make loweffort
choices, thus reducing principals’ payoffs as well as their own.
9
3.4. Autonomy as a source of motivation
Another motivational force investigated in MA experiments cre-
ates a fundamental paradox of organizational control systems:
they aim to motivate employees, but ‘‘control” itself can be demo-
tivating and autonomy can be energizing. For example, an experi-
ment by Williamson (2008) provides initial evidence of employees’
making more costly contributions to ?rm value when they have at
least partial control over the choice of the activities they perform.
In more recent studies, Christ, Sedatole, and Towry (2012) and
Christ, Emmett, Summers, and Wood (2012) investigate ways in
which speci?c control system elements can be more and less ‘‘con-
trolling” in this negative sense. As in the reciprocity studies
described above, the effect of controls in Christ et al.’s experiments
depends on the social meaning ascribed to controls as well as to
their formal contractual properties (e.g., the payoffs they provide
and the probability of receiving these payoffs conditional on
various actions).
In Christ, Sedatole, et al. (2012), participants’ experiences of
working under economically identical incentive contracts labeled
as bonuses or penalties in?uence their effort choices on later tasks
for the same organization that are not governed by the incentive
contract. Effort is lower on the later tasks after experience with
the penalty contracts, because these are interpreted as intrusive
and autonomy-reducing. The bonus contracts, in contrast, are
interpreted as conveying organizational trust; hence they create
more reciprocal trust and prompt more (organizationally
bene?cial) later effort.
In a related study, Christ, Emmett, et al. (2012) investigate a
task in which one dimension of performance is controlled through
a performance-based monetary incentive and the other through a
formal control—either a preventive control that keeps undesired
behavior from happening in the ?rst place, or a detective control
that detects and labels undesired behavior ex post but does not
impose any further consequences. Although the preventive control
might be expected to be unambiguously more useful, since it
actually prevents undesired outcomes from happening, it has
signi?cant negative side effects. Because it is perceived as more
autonomy-reducing than the detective control, it lowers intrinsic
motivation; and lower intrinsic motivation in turn reduces perfor-
mance on both compensated and non-compensated dimensions of
the task.
3.5. Two questions for the future: Do I know how motivated I am?
What motivates the motivators?
Although recent experimental MA research has generated rich
results, it has also raised important questions about motivation
that largely remain unanswered. First, the economic models that
implicitly or explicitly underlie many MA experiments assume that
motivation is conscious—that is, that choices of costly actions are
the result of deliberate decisions to incur these costs with the
aim of receiving desired rewards. Not all motivation is like this,
however: an experiment by Hecht, Tafkov, and Towry (2012)
provides a striking example to the contrary. This study investigates
a multi-task setting in which performance-based rewards are pro-
vided for only one task (a stylized form of the situations in which
some important tasks or performance dimensions are rewarded
less because of measurement limitations). As expected, dispropor-
tionate rewards for one task lead to disproportionate effort alloca-
tions to that task, compared to effort allocations in a ?xed-wage
setting.
Another motivational force mitigates this effect considerably,
however, when rewarded and unrewarded tasks are performed
in close temporal proximity to each other. The incentive for the
rewarded task energizes participants: that is, it increases their
arousal or activation levels. Some of this energy automatically
spills over onto the unrewarded task, increasing effort and
performance on this task without a corresponding decrease for
the rewarded task.
It is unlikely that participants’ additional effort in the close-
proximity condition is the result of a well-deliberated conscious
choice. This raises questions about the MA motivation literature
in general: How conscious is the motivation prompted by elements
of organizational control systems? For example, do individuals
realize how much additional cost they are incurring in order to
compare themselves favorably with others? If the magnitude of
the additional effort was salient to them and they deliberated
about it, then would they make the same choices? The answer
probably depends on their interpretation of their situation: for
example, on whether they believe that their employers are
prompting social comparisons or providing autonomy simply in
order to motivate more effort without paying higher compensa-
tion. Although studies of non-monetary motivations have
occasionally presented their effects as behavior that could be
exploited by organizations as low-cost ways of motivating more
effort, more honesty, etc., it is questionable how effective such
deliberately exploitive tactics would be if employees understood
their purpose.
A second set of motivation questions that are only beginning to
be addressed in MA experiments are about the individuals who
design and operate organizational incentive systems. All of the
recent studies described above are concerned with the motivations
of the individuals at the receiving end of organizational incentive
systems. When recent experimental literature has investigated
higher-level members of organizations, it has usually focused on
cognitive processes, especially with respect to reducing biases in
performance evaluation (e.g., Bol & Smith, 2011; Cardinaels &
van Veen-Dirks, 2010; Bailey, Hecht, & Towry, 2011; Ding &
Beaulieu, 2011), but has paid less attention to the potential
diversity of motivations among these individuals. Some gift-
exchange contract studies (Hannan, 2005; Kuang & Moser, 2009)
include superior-decision components, but these studies are not
primarily interested in superiors’ motivation. They assume superi-
ors want to increase their own payoffs, and the studies examine
how well superiors can do so by identifying and responding to
the complexities of subordinates’ motivation.
10
Two recent studies, however, focus more directly on the mixed
nature of superiors’ motivation and present interesting unsolved
puzzles for future research. Maas, van Rinsum, and Towry (2012)
examine superiors’ motives in making discretionary allocations of
pre-established bonus pools among the members of teams of
9
See Choi (2014) for a related example of labor-market in?uences on the
interpretation of contract terms and the consequent effectiveness of gift-exchange
contracts.
10
Hannan (2005) includes no hypotheses about the behavior of superiors, although
she expects and ?nds that they are willing to offer high wages in the expectation of
receiving (unenforceable) high effort in return. Kuang and Moser (2009, p. 1685)
include one hypothesis about superiors’ choices, based on the assumption that
superiors want to increase their own payoffs and that they will ‘‘learn from
experience that employees react negatively [to the optimal contract] and therefore
switch to offering the gift-exchange contract.”
J. Luft / Accounting, Organizations and Society xxx (2015) xxx–xxx 5
Please cite this article in press as: Luft, J. Management accounting in the laboratory and in social context: Four contrasts, 1975–2014. Accounting, Organiza-
tions and Society (2015),http://dx.doi.org/10.1016/j.aos.2015.08.001
subordinates. They ?nd that fairness, trust, and reciprocity consid-
erations motivate superiors to sacri?ce a portion of their own
payoffs in order to acquire the necessary information to make
bonus allocations that are proportional to subordinates’ contribu-
tions. (In the short-term conditions of the experiment, superiors
know that this payoff sacri?ce cannot bene?t them by prompting
more cooperation from subordinates in the future.)
In contrast, an experiment by Hales and Williamson (2010) sug-
gests limitations on the success of fairness, trust, and reciprocity—
and even that often powerful factor, reputation—in supporting
pro?table cooperation when it cannot be contractually enforced.
Hales and Williamson (2010) employ a three-level hierarchy, in
which an owner chooses whether or not to invest in a ?rm, a
worker chooses howproductively to use the asset if the investment
is made, and a manager, after observing the worker’s productivity,
decides how much of the resulting surplus is paid to the worker
and how much remains as pro?t for the owner. Owners will invest
if they believe that workers will be productive and that managers
will allocate a suf?cient share of the surplus to owners; and work-
ers will be productive if they believe that managers will reward
their efforts. In the experiment setting, however, there is no way
of contractually forcing either productivity or rewards for it;
reciprocity and reputation are the only possible mechanisms.
Reciprocity alone is not effective in this setting: in conditions
where the participants are rematched in each period and thus
cannot develop reputations through repeated interactions, earn-
ings of all participants are low. Reputation in repeated-interaction
settings is helpful—investments and productivity are signi?cantly
higher—but only if managers’ pay is not very sensitive to owners’
pro?ts. If cutting workers’ pay and delivering high pro?ts to owners
results in substantially higher pay for managers, then managers
make myopically large allocations of surplus to owners, workers
reduce productivity (apparently much more than managers antic-
ipate), and the payoffs to all three parties, including managers,
decline: the welfare of all three parties is lower when managers’
pay depends heavily on owners’ pro?ts.
Part of the motivation for managers to reduce workers’ pay in
this way is that their own bonus for giving the owner a larger share
is certain, and future payoffs from high worker productivity are
uncertain. But Hales and Williamson (2010) argue that no plausible
magnitude of risk aversion can account for the observed magni-
tude of value-destroying, own-payoff-destroying behavior by
managers. Thus an interesting puzzle presents itself: why are the
managers in this setting so myopic? Why does reputation fail here?
And why are trust and reciprocity mechanisms so powerless in this
setting? It is possible that the three-level hierarchy presents a
more complex decision problem for individuals than the two-
person superior-subordinate setting in other experiments, and
grabbing for immediate, certain payoffs is a simplifying response
to this complexity.
11
It is also possible that the high reward
managers receive for high current pro?t makes current pro?t a much
more salient and/or legitimate-seeming goal, and developing
longer-term pro?table relations with employees loses attention
and/or desirability.
The contrast with Maas et al. (2012) also raises questions about
the social meaning of different kinds of decisions and the possible
implications for superior motivation and outcomes. In both Maas
et al. (2012), which uses random rematching to exclude reputation
effects, and the random-rematching condition in Hales and
Williamson (2010), superiors’ decisions to incur costs to reward
subordinates fairly are formally similar: they are reductions in the
superiors’ own payoffs in order to reciprocate subordinates’ trusting
effort choices. But it is possible that these decisions do not feel the
same to the individuals involved. Perhaps the action, ‘‘Acquire infor-
mationabout subordinates’ actions” (Maas et al., 2012) has positive-
value overtones for managers that ‘‘Reduce pro?ts by paying subor-
dinates more” (Hales & Williamson, 2010) does not. Or perhaps
there is a clear fairness benchmark in Maas et al. (2012)—pay more
to the team member who contributes more—and this benchmark
cannot be attained without buying information. In Hales and
Williamson (2012), however, there is not such a clear benchmark
for how the common surplus should be divided among owners,
managers, and workers, given their varying contributions; this
unclarity gives managers more leeway for opportunistic behavior.
Future research on both the motivation and cognition of superiors
might offer further insights on these puzzles.
4. Changes in the representation of the roles of MA
A second major set of changes in experimentation in the last
forty years is that the roles of MA in organizations have come to
be conceptualized differently in experimental research. The role
of MA in most early experiments is quite simple. It provides
numerical values for the variables in individuals’ decision
models—for example, values for product costs in product-pricing
decision models (Ashton, 1976, 1981; Bloom et al., 1984;
Dyckman et al., 1982), values for realized costs in cost-variance
investigation models (Ansari, 1976; Brown, 1981, 1983, 1985;
Jacobs, 1978; Lewis et al., 1983; Magee & Dickhaut, 1978) or values
for measured performance in incentive-contracting models (Chow,
1983; Waller & Chow, 1985).
The critique of this approach in AOS articles of the 1980s argued,
in contrast, that accounting is not just a provider of numerical val-
ues to ?ll into pre-given decision models. For example, MA is
involved in ‘‘the construction of organizational participants’ views
of the desirable” and the ‘‘construction of organizational reality”
(Cooper et al., 1981; Hopwood, 1983). It can in?uence ‘‘individuals’
perceptions of the world, the way they ‘bracket’ or ‘punctuate’ their
stream of experiences” (Cooper et al., 1981); it can also be used to
rationalize existing situations and to express and create meaning
(Burchell et al., 1980; Cooper et al., 1981; Roberts & Scapens, 1985).
Some examples of these additional uses of MA have already
appeared in the previous section. For example, when MA-related
control elements in?uence the dimensions and/or importance of
social comparisons, they are in?uencing individuals’ ‘‘views of
the desirable.” In another recent experiment that documents MA
in?uence on views of the desirable, Bloom?eld and Tayler (2011,
p. 753) provide evidence that formal controls ‘‘directly in?uence
people’s sense of what is appropriate in the setting (personal
norms) and indirectly alter people’s tendency to conform to the
behavior of those around them (descriptive norms).” These effects
persist even after the control system has changed, since norms are,
so to speak, sticky: they can evolve over time, but organizations
cannot alter them discontinuously from one period to the next,
the way they can alter features of the formal control system like
the size of the bonus pool or the performance measures that
determine its allocation.
Recent experiments have also explored other ways in which MA
contributes to ‘‘individuals’ perception of the world” and their
‘‘construction of organizational reality.” Two important ways in
which MA plays these roles can be distinguished in the recent
literature:
First, imperfect accounting measures of constructs like ef?-
ciency, quality, and cost are sometimes treated as perfectly
equivalent to the constructs themselves, and the extent to
which this happens is in?uenced by features of control systems.
11
Brüggen and Luft (2016) present evidence for more myopic behavior in three-
person than two-person settings in a very different task (capital budgeting).
6 J. Luft / Accounting, Organizations and Society xxx (2015) xxx–xxx
Please cite this article in press as: Luft, J. Management accounting in the laboratory and in social context: Four contrasts, 1975–2014. Accounting, Organiza-
tions and Society (2015),http://dx.doi.org/10.1016/j.aos.2015.08.001
Second, MA can in?uence how people conceptualize others in
an organization: are the others ‘‘they” or are they part of ‘‘we”?
The ?rst theme deals with a core issue in accounting: the fact
that accounting measures are typically incomplete and noisy
indicators. This is also true of the non?nancial measures that are
often bundled together with traditional accounting measures in
performance-measurement systems: the customer satisfaction or
quality measures reported are not necessarily identical to actual
customer satisfaction or quality. Users of MA sometimes ignore
these differences between measures and constructs: for example,
they might believe (or at least to act as if) the organization’s actual
service quality has become worse in the latest period when in fact
only the measure has deteriorated, perhaps due to small-sample
variation in a sample-based measure.
In a pair of papers, Choi, Hecht, and Tayler (2012, 2013)
investigate in?uences on what they call surrogation, that is, the
tendency for individuals to treat a measure as if it fully captured
the underlying construct, even when at some level they know that
it does not. In the experiments, surrogation is greater—that is,
accounting has more of a tendency to become reality—when peo-
ple have been paid based on the measures in the past, even though
they are not thus paid in the present and hence there is no extrinsic
motivation for them now to treat the measures as reality. This
effect is larger when there is only one measure per construct rather
than multiple measures, since different measures of the same con-
struct make it more evident that none of the measures actually is
the construct (Choi et al., 2012).
Surrogation of strategic measures is also more likely when indi-
viduals are assigned a strategy to implement rather than choosing
it themselves. Interestingly, Choi, Hecht, and Tayler (2013) are able
to establish that this difference occurs, not because participants
have deliberated about the strategy more when they choose it,
but because they personally identify more with the strategic
constructs. Thus the constructs are more accessible in their minds
and less likely to be substituted by the limited measures (Choi
et al., 2013).
In the Choi et al. (2012, 2013) experiments, the ‘‘organizational
reality” that is shaped by MA is the characteristics of products,
which are seen as more or less successful depending on incomplete
measures of their success. In other studies, the reality shaped by
MA is the social landscape of the organization. A number of recent
experiments provide evidence of the in?uence of MA on individu-
als’ tendency to think of themselves and co-workers as either ‘‘we,”
or as ‘‘I and they,” with signi?cant consequences for behavior.
Towry (2003) and Rowe (2004) provide the foundation for this line
of research, with experimental investigations of how MA in?u-
ences and is in?uenced by social identity—i.e., the tendency to ‘‘de-
personalize the self-concept” (Brewer & Schneider, 1990, p. 170), to
think of oneself more as part of a group and less as a separate indi-
vidual. Social identity has both cognitive and motivational aspects,
because thinking ‘‘we” not only means putting a higher value on
others’ welfare but also means shifts in attention and information
processing. Stronger social identity means that individuals become
‘‘more attuned to the interrelatedness of their actions, focusing on
the ways in which they can jointly affect outcomes.” (Towry, 2003,
p. 1079)
Chen, Williamson, and Zhou (2012) provide an interesting
recent example, in which teams create business proposals under
one of four different incentives: individual piece-rate, group
piece-rate, individual tournament, or intergroup tournament. All
incentive payoffs are based on the creativity of the proposals as
judged by a panel of peers. Although the group piece-rate and
intergroup tournament provide equal mean payoffs and the payoff
structure in the tournament is riskier, the intergroup tournament
is more effective in generating proposals that are judged as highly
creative, because intergroup competition (‘‘us versus them”)
promotes intragroup cohesiveness—a stronger social identity as
‘‘us”—and this in turn facilitates more effective collaborative
processes.
12
Chang, Cheng, and Trotman (2013) provide another example of
how features of control systems affect organizational processes by
de?ning the identity of other players primarily as opponents or
collaborators. In this experiment, participants negotiate a supply-
chain contract and are paid partly for the outcome of the negotia-
tion (how much pro?t it provides for their own business unit) and
partly for the quality of their reports on either the outcome or the
process. The report-topic manipulation changes negotiators’ atten-
tion to and evaluation of different elements of their experiences
and thus, in effect, change these experiences. If negotiators must
later explain the outcome, they focus more on distributive issues
(getting a larger share of the surplus at the expense of the other
party), employ more distributive tactics in negotiation, and exhibit
a larger ‘‘?xed-pie bias,” i.e., the incorrect belief that the total sur-
plus is ?xed and cannot be increased through integrative tradeoffs.
In contrast, when they must report later on the negotiation process,
they focus more on their common experience within the negotia-
tion; and in consequence, integrative tactics are more common
and the ?xed-pie bias is smaller.
These examples illustrate how experiments can provide
close-up, detailed evidence of the microprocesses by which MA
contributes to the shaping of individuals’ experience and hence
to the actions that arise out of that experience. Although this
recent literature is rich in some respects, it is somewhat narrow
in other respects: it is heavily focused on effects of the
performance measurement and reward roles of MA, not on other
roles or on the causes of MA. The remainder of this article contrasts
this narrowness of scope with the broader scope of earlier MA
experiments. It also identi?es some recent studies that provide
examples of the bene?ts of beginning to expand the scope of MA
experimentation again.
5. Narrowing the scope of MA experiments: Two contrasts and
the future
5.1. Decision-facilitating and decision-in?uencing roles of MA
A common way of identifying roles for accounting, particularly
in economics-based research, is to distinguish between decision-
facilitating and decision-in?uencing roles of accounting (Demski
& Feltham, 1976) or (in alternative terms) decision and control
problems (e.g., Arya, Glover, & Sivaramakrishnan, 1997). In this
view, MA facilitates decisions by reducing pre-decision uncertainty,
while it in?uences decisions through its ex post use in reward sys-
tems that motivate people to do what they would otherwise prefer
not to do.
The diversity of roles for MA that appear in the literature ana-
lyzed above suggest that the traditional descriptions of decision-
facilitating and decision-in?uencing do not fully describe the roles
of accounting in organizations. Studies described in the previous
section have illustrated how MA can in?uence decisions not only
by providing a basis for monetary rewards for subordinates, but
also (up to a point) by in?uencing what people (both subordinates
and superiors) value and how they conceptualize their business
problems, their setting, and their relations with others in
the organization (see also Luft & Shields, 2009). Similarly, the
decision-facilitating roles of MA are likely to go beyond the
12
Consistent with the social comparison literature described earlier, the individual
tournament results in more individual effort than individual piece-rate pay; but in the
absence of the intensi?ed social identity prompted by the intergroup tournament,
these increased individual efforts do not result in better group output.
J. Luft / Accounting, Organizations and Society xxx (2015) xxx–xxx 7
Please cite this article in press as: Luft, J. Management accounting in the laboratory and in social context: Four contrasts, 1975–2014. Accounting, Organiza-
tions and Society (2015),http://dx.doi.org/10.1016/j.aos.2015.08.001
traditional role of reducing uncertainty about variables (and some-
times parameters) in individuals’ decision models.
We have less information about diverse roles of MA in decision
facilitation, however, because the last forty years have seen a
signi?cant shift of experimental researchers’ attention from deci-
sion-facilitating to decision-in?uencing roles of MA. Fig. 2 shows
the number of experiments published in AOS, JAR, and TAR that
examine these two roles of MA (broadly de?ned) in the ?rst and
last ten years of the forty-year period. In the ?rst ten years, deci-
sion facilitation is a more common subject of experiments:
researchers in this period are interested in knowing how account-
ing is used in decisions about resource allocations, product prices,
and variance investigations. In contrast, in the last ten years (as in
the preceding decade; see Sprinkle, 2003) the situation has
reversed. Most experiments address the role of control systems
in motivating people to exert more effort or better-allocated effort
or to share private information honestly; and even in decision-
facilitation experiments, the decision addressed is often the evalu-
ation of subordinates’ performance, which is an activity closely
related to the decision-in?uencing role.
This shift of interest has had multiple causes. Decision-facilitating
roles of accounting have been regarded skeptically in some
quarters: for example, Zimmerman (2001, p. 424) conjectured that
an ‘‘emphasis on decision making, not control” in some areas of
empirical MA research was inappropriate because control is a more
important role for accounting in organizations. Perhaps more
importantly, interest in the decision-in?uencing role has been
consistent with experimentalists’ increased interest in motivation
and has been consistent with the principal-agent models that have
tended to dominate analytic research in MA in recent decades.
These models have often provided useful underpinnings for
experiments—for example, suggesting elements of task structure,
important variables, and baseline predictions of behavior.
Acceptable analytic models for decision-facilitating roles of MA
have not always been so readily available.
MA does play decision-facilitating roles, however, and there is
no evident reason why experimentation cannot address these
roles. New studies are likely to take a different approach than
the decision-facilitation experiments of the 1970s and early
1980s, however; and in recent experiments it is possible to see
some promising approaches toward making this kind of research
successful.
Early experiments on the decision-facilitating role of account-
ing tended to focus on the statistical properties of accounting infor-
mation and to ask whether the users of MA understood the
importance of sample size or could judge correlations accurately
from a set of data. These are by no means irrelevant questions,
but the ability of early studies to provide useful answers was often
limited by two design features of the experiments. First, they
tended to assume or provide decision models without adequate
assurance that these models represented the way people actually
were likely to think, or should optimally think, about tasks like
product pricing or strategy evaluation. Second, many early studies
represented accounting at a high level of abstraction—for example,
as larger or smaller samples of balls from urns. These design
choices were limiting because actual uses of MA information are
likely to depend on how people conceptualize their business
problems (perhaps differently from the standard decision-theory
models); and actual uses of MA information are also likely to
depend not only on abstract problem structure but also on
concrete features of MA systems that aid users in interpreting
and using information.
Although recent studies of the decision-facilitating role of
accounting are not numerous, they provide some evidence of the
potential of experimental research in this area when it moves away
from the approaches used in early studies. For example, Jackson,
Rodgers, and Tuttle (2010) provide an interesting alternative
approach to the question, frequently addressed in the 1970s and
80s, of how accounting-method differences in?uence asset pricing.
In ?ve experiments, using diverse market institutions and partici-
pants ranging from accounting undergraduates to executives with
over thirty years of experience, participants choose selling prices
for used machines, given book values and in some conditions mar-
ket prices as well. The variation in book value under different
depreciation methods has a strong effect on participants’ selling
prices, but not for any of the reasons put forward in the older
‘‘functional ?xation” studies.
Jackson et al. (2010) verify that when individuals are willing to
take lower prices for machines given accelerated instead of
straight-line depreciation, it is not because they fail to understand
depreciation, or because they believe lower book value means the
machines are in worse condition or have lower market value. It is
also not because the participants are concerned about book losses
or because they have no incentive to pay attention to the problem.
Rather, it is because of the role that depreciation plays in their
basic conceptualization of the asset pricing task.
Consistent with research by Heath and Fennema (1996) on how
asset values are used in consumer decisions, what is on partici-
pants’ minds in setting selling prices is not just getting the highest
price the market will give them, but ‘‘getting their money’s worth”
from the machine, both through its use in generating revenue and
through its salvage value. In the absence of information to the con-
trary, they expect that more depreciation expense probably means
the machine has provided more bene?ts (revenue, reduced pro-
duction costs) to the ?rm already. Thus they have already got
‘‘more of their money’s worth” from the machine and can afford
to sell it for a lower price. This may not be a wealth-maximizing
approach to pricing, but it has some intuitive force; and the results
underline the importance of understanding what might be called
‘‘folk theories” of activities like pricing, resource allocation, and
strategy evaluation in order to understand the role of accounting
in these activities.
Other recent decision-facilitation experiments provide evidence
of how MA can nudge people toward different ways of approaching
and using the information available to them. In these studies, what
matters about MA is not simply the decision-relevance of the infor-
mation it provides, but rather how this information combines with
the expectations created by structural elements of the MA system.
0
5
10
15
20
25
1976-80 1981-85 2005-09 2010-14
Decision facilitating
Decision in?luencing
Fig. 2. Decision facilitating and decision in?uencing roles of accounting in MA
experiments, 1976–1985 and 2005–2014. The data points represent the total
number of studies examining decision facilitating and decision in?uencing roles of
MA in three journals (Accounting, Organizations, and Society; Journal of Accounting
Research; and The Accounting Review) in the four time periods 1976–1980, 1981–
1985, 2005–2009, and 2010–2014. Broken lines between 1985 and 2005 highlight
the fact that no data is presented for this interim period. Total counts in this ?gure
exceed the total number of published studies represented in Fig. 1 because in some
cases a single study examines both roles of MA and therefore is counted twice.
8 J. Luft / Accounting, Organizations and Society xxx (2015) xxx–xxx
Please cite this article in press as: Luft, J. Management accounting in the laboratory and in social context: Four contrasts, 1975–2014. Accounting, Organiza-
tions and Society (2015),http://dx.doi.org/10.1016/j.aos.2015.08.001
For example, Mastilak (2011) has participants examine cost data in
order to evaluate the progress of a quality program—that is, to
answer the question, ‘‘When we have spent more on this program,
have we tended to get higher quality?” The quality-improvement
costs (e.g., supervision) and outcomes (e.g., spoilage costs) could
be in either the same or different cost pools. Holding the strength
of the actual quality-improvement effects constant, participants
identify and estimate the effects more accurately when costs and
outcomes are within-pool than when they are across-pool, because
participants use the pool structure as a cue for where to look ?rst,
or look more closely, for relevant relations. The cue does not, in
most cases, make participants see relations that do not exist (i.e.,
between expenditures on quality and uncorrelated quality out-
comes within the same pool): the pool-structure cue supports
learning but does not substitute for it.
Kelly (2010) provides another example of how individuals are
guided in their use of extensive data by cues in the MA system—
and interestingly, like Mastilak (2011), she ?nds that they are
not misguided when the data are inconsistent with the guiding
cues. In her experiment, participants must learn from experience
which of two lines of business tends to produce higher long-term
returns on R&D expenditures. The information participants receive
either does or does not include an initial judgment about which
line of business probably generates higher returns, and this initial
judgment is either accurate or inaccurate.
13
Participants learn sig-
ni?cantly better from their experience when the information pro-
vided to them includes an initial judgment—but they learn at least
as well when the judgment is wrong as when it is right. Kelly
(2010) argues that the inconsistency between expectations and
experience when the judgment is wrong prompts more elaborate
information processing, which supports better knowledge acquisi-
tion and thus makes up for initial errors caused by the miscue.
The lack of a critical mass of related experiments on the deci-
sion-facilitation role of MA means that we know much less about
the robustness and limitations of the effects in studies like those
described above than we know about the decision-in?uencing
effects of MA documented in experiments. However, these and
other recent studies suggest that decision-facilitation experiments
have potential to add to our understanding of how people can learn
from MA and use it in decision making. These recent studies also
suggest two keys to realizing that potential: ?rst, more attention
to how people actually conceptualize and approach business
problems (which may or may not correspond well with standard
decision models), and second, how structural features of MA sys-
tems cue particular approaches to the information these systems
provide.
5.2. Accounting as independent or dependent variable
A persistent feature of experimental MA research is that
accounting is the independent not the dependent variable in most
studies. (See Hopwood (1983) for criticism of this choice in empir-
ical research in general, and see Luft and Shields (2003), Table 1, for
evidence of the persistence of this choice in literature on budgeting
(Map A) and on control microprocesses and individual decisions
(Maps F and G).) Experiments have provided much more informa-
tion about the effects of MA than about its causes. As Fig. 3
illustrates, the preference for treating MA as the independent not
the dependent variable is noticeably stronger in the last decade
of this forty-year period than in the ?rst.
Three related reasons can be suggested for the decline of this
literature on MA as DV after the mid-1980s. First, an implicit or
explicit functionalist approach sometimes made it appear as if
investigations of the effects of MA were suf?cient for explaining
the causes of MA as well. Such an approach assumes, for example,
that if we provide evidence that one kind of MA supports more ef?-
cient resource allocation than others, then we have also provided a
suf?cient explanation of why that kind of MA actually appears in
organizations.
Such explanations are often not in fact suf?cient, however, and
thus additional investigation of MA as a dependent variable is
required. The MA that an organization actually uses is also in?u-
enced by intra- and inter-organizational con?icts and processes
of negotiation among individuals with different understandings
and different levels of power (e.g., Kurunmäki, 1999). Moreover,
MA can be in?uenced by consultants, who have aims of their
own (e.g., Christensen & Skaerbek, 2010; Cooper & Qu, 2011); it
is also in?uenced by the institutions that train accountants and
regulate accounting, by an organization’s own past, and by
historical events that move on a larger scale than individual
organizations’ concerns (see, for example, Loft (1986) on the role
of World War I in the development of MA).
A second reason for the decline of research on MA as a depen-
dent variable may have been the way that MA was operationalized
in early experiments. In many of these studies, the key decision to
be made was the accuracy of the information provided by the MA
system, and more information could be obtained by purchasing
more of it: drawing a larger sample of balls from urns or turning
over more cards on an information board. Studies like these were
interested in determining whether and under what conditions par-
ticipants could identify the optimal level of accuracy of MA, given
the bene?ts of accuracy and the costs of information; but they set
aside the question of how organizations could actually increase the
accuracy of MA information if they wanted to do so. In the lab the
answer was simple: participants simply purchased more pieces of
data. But in natural environments the answer was not always so
obvious, and by the later 1980s the question about how to improve
accuracy perhaps seemed more urgent than questions of whether
people chose optimal sample sizes or purchased more information
than they could effectively process. Questions about how to
improve accuracy were often framed as questions about the effects
0
5
10
15
20
25
30
1976-80 1981-85 2005-09 2010-14
DV
IV
Fig. 3. Accounting as the dependent or independent variable in MA experiments,
1976–1985 and 2005–2014. The data points represent the number of studies using
MA as the dependent or independent variable in three journals (Accounting,
Organizations, and Society; Journal of Accounting Research; and The Accounting
Review) in the four time periods 1976–1980, 1981–1985, 2005–2009, and 2010–
2014. Broken lines between 1985 and 2005 highlight the fact that no data is
presented for this interim period. Total counts in this ?gure exceed the total
number of published studies represented in Fig. 1 because in some cases MA is both
independent and dependent variable in a study. For example, a change in one MA-
related element of a control system (the independent variable) might cause a
change in another element of the system (the dependent variable).
13
These judgments serve as a reduced, stylized representation of the causal models
included with some performance measurement systems, specifying expected rela-
tions between leading and lagging indicators in the systems.
J. Luft / Accounting, Organizations and Society xxx (2015) xxx–xxx 9
Please cite this article in press as: Luft, J. Management accounting in the laboratory and in social context: Four contrasts, 1975–2014. Accounting, Organiza-
tions and Society (2015),http://dx.doi.org/10.1016/j.aos.2015.08.001
of different MA-system choices, and thus questions about the
causes of these choices received less attention.
A third reason for the decline of MA as a dependent variable in
experiments arose from another limitation of the early studies,
namely, that they represented the overall quality of MA as a choice
made by an individual. The overall quality of MA, however, is the
result of many individual choices, organizational negotiations,
and social in?uences. Laboratory experiments typically investigate
the actions of individuals or small groups, and as such, they are
more likely to add to our understanding if they focus on the
individual moments in MA creation—that is, the speci?c decisions
which individuals and small groups do in fact make, and which
provide speci?c contributions at speci?c points in time to the
overall quality of MA.
The most common (in recent years, almost the only) explana-
tions that experiments have provided for the varying quality of
MA information are explanations of when and why individuals
are more likely to share their private information and thus make
it part of organizational MA in the form of budget standards, cost
reports, etc.
14
This kind of information-sharing can be a signi?cant
in?uence on overall MA quality (see Rowe, Birnberg, and Shields
(2008) for a ?eld-study example), and when the information is in
fact private to individuals, then decisions about how accurately to
transmit this information are indeed ‘‘individual moments” in the
creation of MA that can usefully be addressed in experiments.
However, they hardly exhaust the possibilities for using experiments
to help explain how MA comes to be what it is.
Cardinaels and Labro (2008) provide an excellent but thus far
isolated example of a different kind of individual moment in the
creation of MA. In this experimental study, individuals report on
how they have allocated their time among different work activi-
ties. Such reports are common in organizations as a basis for (more
or less accurate) allocations of costs to these activities. In the
experiment, participants are provided with incentives to report
as accurately as they can; thus, accuracy de?cits are not the result
of deliberate and opportunistic misreporting but rather of obsta-
cles to accurate memory that are created by properties of the MA
system and the structuring of work activities. Cardinaels and
Labro (2008) ?nd that the accuracy of individuals’ reports—and
thus the accuracy of the cost estimates the MA system can pro-
duce—depends on the way individuals’ work is structured (task
ordering more or less consistent with cost-system structure), the
number of activity pools, the type of measure requested (percent
or absolute time estimates), and the timing of the report requests.
Cardinaels and Labro (2008) is particularly interesting in its
attention to how individuals actually acquire the private informa-
tion which, in environments with different incentives, they may or
may not choose to share. In most honesty experiments,
participants are simply endowed with private information by the
experimenter (e.g., Evans et al., 2001)
15
—but private acquisition of
information that then becomes part of MA is an individual moment
in the creation of MA that could usefully be investigated further in
the laboratory.
6. Conclusion
The volume of MA experiments published in major journals has
more than doubled between the ?rst and last ?ve years of the
forty-year period considered here. Not only is the experimental lit-
erature larger than it once was; it also exhibits two important signs
of robustness and maturity that were not present thirty years ago.
First, many key ?ndings in the experimental literature have been
replicated in multiple experiments, as indicated by the analysis
of studies of social-preference effects in earlier sections of this arti-
cle. Second, the experimental literature has begun to in?uence
analytical and archival studies. For example, some principal-agent
studies now model endogenous social norms (Fischer & Huddart,
2008) or effects of honesty preferences on collusion (Mittendorf,
2008), and archival studies have provided evidence of reciprocity
(Chen & Sandino, 2012) and fairness effects (Bol, Matsumura,
Keune, & Shin, 2010; Matsumura & Shin, 2006) like the effects
found in experiments.
Although the experimental MA literature shows important
signs of maturity in some areas, other areas of MA that are in prin-
ciple accessible to experimental investigation remain relatively
unexplored. The experimental literature on MA and motivation
has been limited by its near-exclusive focus on the individuals
who are, so to speak, on the receiving end of MA—the subordinates
whose behavior upper-level management wants to in?uence—and
its relative lack of attention to the motivation of many other indi-
viduals who shape MA.
This limitation of the motivation-related literature is part of
two broader limitations of recent experiments: their tendency to
address the effects of MA much more frequently than its causes,
and their tendency to address decision-in?uencing roles of MA
much more frequently than decision-facilitating roles. An excep-
tion to the lack of attention to causes—perhaps an unsurprising
exception, because it addresses the motivation of lower-level indi-
viduals—has been the experimental literature on decisions to con-
tribute private information (sometimes accurate, sometimes
distorted) to MA institutions such as budgets and performance
evaluations. Other individual moments and microprocesses
involved in the shaping of MA and its decision-facilitating roles
remain to be identi?ed and explored in experimental studies.
Acknowledgements
Many thanks to Chris Chapman, Mike Shields, and participants
at the AOS 40th Anniversary conference for helpful comments on
this paper.
References
Ansari, S. L. (1976). Behavioral factors in variance control: Report on a laboratory
experiment. Journal of Accounting Research, 14(2), 189–211.
Arya, A., Glover, J. C., & Sivaramakrishnan, K. (1997). The interaction between
decision and control problems and the value of information. The Accounting
Review, 72(4), 561–574.
Ashton, R. H. (1976). Cognitive changes induced by accounting changes:
Experimental evidence on the functional ?xation hypothesis. Journal of
Accounting Research, 14(Suppl.), 1–17.
Ashton, R. H. (1981). A descriptive study of information evaluation. Journal of
Accounting Research, 19(1), 42–61.
Bailey, W. J., Hecht, G., & Towry, K. L. (2011). Dividing the pie: The in?uence of
managerial discretion extent on bonus pool allocation. Contemporary Accounting
Research, 28(5), 1562–1584.
Becker, S. W., Ronen, J., & Sorter, G. H. (1974). Opportunity costs—An experimental
approach. Journal of Accounting Research, 12(2), 317–329.
Bloom, R., Elgers, P. T., & Murray, D. (1984). Functional ?xation in product pricing: A
comparison of individuals and groups. Accounting, Organizations, and Society, 9
(1), 1–11.
Bloom?eld, R., & Tayler, W. (2011). Norms, conformity, and controls. Journal of
Accounting Research, 49(3), 753–790.
Bol, J. C., Matsumura, E. M., Keune, T., & Shin, J. Y. (2010). Supervisor discretion in
target setting: An empirical investigation. The Accounting Review, 85(6),
1861–1886.
Bol, J. C., & Smith, S. D. (2011). Spillover effects in subjective performance
evaluation: Bias and the asymmetric effect of controllability. The Accounting
Review, 86(4), 1213–1230.
Brewer, M. B., & Schneider, S. K. (1990). Social identity and social dilemmas: A
double-edged sword. In D. Abrams & M. A. Hogg (Eds.), Social identity theory:
Constructive and critical advances. New York: Springer-Verlag.
Brown, C. (1981). Human information processing for decisions to investigate cost
variances. Journal of Accounting Research, 19(1), 62–85.
14
Some recent studies on this topic are described earlier in the present article; see
Luft and Shields (2009) for a more extensive treatment of this literature.
15
See Church, Hannan, and Kuang (2014) for an interesting exception, however.
10 J. Luft / Accounting, Organizations and Society xxx (2015) xxx–xxx
Please cite this article in press as: Luft, J. Management accounting in the laboratory and in social context: Four contrasts, 1975–2014. Accounting, Organiza-
tions and Society (2015),http://dx.doi.org/10.1016/j.aos.2015.08.001
Brown, C. (1983). Effects of dynamic task environment on the learning of standard
cost variance signi?cance. Journal of Accounting Research, 21(2), 413–431.
Brown, C. (1985). Causal reasoning in performance assessment: Effects of cause and
effect temporal order and covariation. Accounting, Organizations, and Society, 10
(3), 255–266.
Brown, J. L., Fisher, J. G., Sooy, M., & Sprinkle, G. B. (2014). The effect of rankings on
honesty in budget reporting. Accounting, Organizations, and Society, 39(4),
237–246.
Brownell, P. (1981). Participation in budgeting, locus of control, and organizational
effectiveness. The Accounting Review, 56(4), 844–860.
Brüggen, A., & Luft, J. (2016). Cost estimates, cost overruns, and project continuation
decisions. The Accounting Review, 91(3) [in press].
Burchell, S., Clubb, C., Hopwood, A. G., Hughes, J., & Nahapiet, J. (1980). The roles of
accounting in organizations and society. Accounting, Organizations and Society, 5
(1), 5–27.
Cardinaels, E., & Labro, E. (2008). On the determinants of measurement error in
time-driven costing. The Accounting Review, 83(3), 735–756.
Cardinaels, E., & van Veen-Dirks, P. M. G. (2010). Financial versus non-?nancial
information: The impact of information organization and presentation in a
balanced scorecard. Accounting, Organizations and Society, 35(6), 565–578.
Chang, L., Cheng, M., & Trotman, K. (2008). The effect of framing and negotiating
partner’s objective on judgments about negotiated transfer prices. Accounting,
Organizations and Society, 33(7–8), 93–107.
Chang, L., Cheng, M., & Trotman, K. (2013). The effect of outcome and process
accountability on customer-supplier negotiations. Accounting, Organizations,
and Society, 38(2), 93–107.
Chen, C. X., & Sandino, T. (2012). Can wages buy honesty? The relationship between
relative wages and employee theft. Journal of Accounting Research, 50(4),
967–1000.
Chen, C. X., Williamson, M. G., & Zhou, F. H. (2012). Reward system design and
group creativity: An experimental investigation. The Accounting Review, 87(6),
1885–1911.
Choi, J. (2014). Can offering a signing bonus motivate effort? Experimental evidence
of the moderating effects of labor competition. The Accounting Review, 89(2),
545–570.
Choi, J., Hecht, G. W., & Tayler, W. B. (2012). Lost in translation: The effects of
incentive compensation on strategy surrogation. The Accounting Review, 87(4),
1135–1163.
Choi, J., Hecht, G. W., & Tayler, W. B. (2013). Strategy selection, surrogation, and
strategic performance measurement systems. Journal of Accounting Research, 51
(1), 105–133.
Chow, C. W. (1983). The effects of job standard tightness and compensation scheme
on performance: An exploration of linkages. The Accounting Review, 58(4),
667–685.
Christ, M. H., Emmett, S. A., Summers, S. L., & Wood, D. A. (2012b). The effects of
preventive and detective controls on employee performance and motivation.
Contemporary Accounting Research, 29(2), 432–452.
Christ, M. H., Sedatole, K. L., & Towry, K. L. (2012a). Sticks and carrots: The effect of
contract frame on effort in incomplete contracts. The Accounting Review, 87(6),
1913–1938.
Christensen, M., & Skaerbek, P. (2010). Consultancy outputs and the puri?cation of
accounting technologies. Accounting, Organizations, and Society, 35(5), 524–545.
Church, B. K., Hannan, R. L., & Kuang, X. (2014). Information acquisition and
opportunistic behavior in managerial reporting. Contemporary Accounting
Research, 31(2), 398–419.
Cooper, D. J., Hayes, D., & Wolf, F. (1981). Accounting in organized anarchies:
Understanding accounting systems in ambiguous situations. Accounting,
Organizations and Society, 6(3), 175–191.
Cooper, D. J., & Qu, S. Q. (2011). The role of inscriptions in producing a balanced
scorecard. Accounting, Organizations, and Society, 36(6), 344–362.
Demski, J. S., & Feltham, G. A. (1976). Cost determination: A conceptual approach.
Ames, IA: Iowa State University Press.
Ding, S., & Beaulieu, P. (2011). The role of ?nancial incentive in balanced scorecard-
based performance evaluations: Correcting mood congruency biases. Journal of
Accounting Research, 49(5), 1223–1247.
Dyckman, T. E., Hoskin, R. E., & Swieringa, R. J. (1982). An accounting change and
information processing changes. Accounting, Organization, and Society, 7(1),
1–11.
Eccles, R. (1985). The transfer pricing problem. Lexington, MA: Lexington Books.
Evans, J. H., Hannan, R. L., Krishnan, R., & Moser, D. V. (2001). Honesty in managerial
reporting. The Accounting Review, 72(2), 537–559.
Falk, A. (2007). Gift exchange in the ?eld. Econometrica, 75(5), 1501–1511.
Fischer, P., & Huddart, S. (2008). Optimal contracting with endogenous social norms.
American Economic Review, 98, 1459–1475.
Frederickson, J. R. (1992). Relative performance information: The effects of common
uncertainty and contract type on agent effort. The Accounting Review, 67(4),
647–669.
Hales, J., & Williamson, M. G. (2010). Implicit employment contracts: The limits of
management reputation for promoting ?rm productivity. Journal of Accounting
Research, 48(1), 147–176.
Hannan, R. L. (2005). The combined effect of wages and ?rm pro?t on employee
effort. The Accounting Review, 80(1), 167–188.
Hannan, R. L., McPhree, G. P., Newman, A. H., & Tafkov, I. D. (2013). The effect of
relative performance information on performance and effort allocation in a
multi-task environment. The Accounting Review, 88(2), 553–575.
Harrell, A. M., & Klick, H. D. (1980). Comparing the impact of monetary and
nonmonetary human asset measures on executive decision making. Accounting,
Organizations and Society, 5(4), 393–400.
Heath, C., & Fennema, M. G. (1996). Mental depreciation and marginal decision
making. Organizational Behavior and Human Decision Making, 68(2), 95–108.
Hecht, G., Tafkov, I., & Towry, K. L. (2012). Performance spillover in a multitask
environment. Contemporary Accounting Research, 29(2), 563–589.
Hilton, R. W., & Swieringa, R. J. (1981). Perception of initial uncertainty as a
determinant of information value. Journal of Accounting Research, 19(1),
109–119.
Hilton, R. W., Swieringa, R. J., & Hoskin, R. E. (1981). Perception of accuracy as a
determinant of information value. Journal of Accounting Research, 19(1), 86–108.
Hopwood, A. G. (1983). On trying to study accounting in the contexts in which it
operates. Accounting, Organizations and Society, 8(2–3), 287–305.
Hoskin, R. E. (1983). Opportunity cost and behavior. Journal of Accounting Research,
21(1), 78–95.
Jackson, S. B. T., Rodgers, C., & Tuttle, B. (2010). The effect of depreciation method
choice on asset selling prices. Accounting, Organizations and Society, 35(8),
757–774.
Jacobs, F. H. (1978). An evaluation of the effectiveness of some cost variance
investigation models. Journal of Accounting Research, 16(1), 190–203.
Kachelmeier, S. J., & Towry, K. L. (2002). Negotiated transfer pricing: Is fairness
easier said than done? The Accounting Review, 78(4), 1069–1095.
Kahneman, D., Knetsch, J. L., & Thaler, R. H. (1986). Fairness and the assumptions of
economics. Journal of Business, 59(Suppl.), S285–S300.
Kelly, K. (2010). Accuracy of relative weights on multiple leading performance
measures: Effects on managerial performance and knowledge. Contemporary
Accounting Research, 27(2), 577–608.
Kuang, X., & Moser, D. V. (2009). Reciprocity and the effectiveness of optimal agency
contracts. The Accounting Review, 84(5), 1671–1694.
Kurunmäki, L. (1999). Professional vs. ?nancial capital in the ?eld of health care—
Struggles for the redistribution of power and control. Accounting, Organizations
and Society, 24(2), 94–124.
Lewis, B., Shields, M. D., & Young, S. M. (1983). Evaluating human judgment and
decision aids. Journal of Accounting Research, 21(1), 271–285.
Libby, R. (1976). Discussion of ‘‘Cognitive changes induced by accounting changes:
Experimental evidence on the functional ?xation hypothesis”. Journal of
Accounting Research, 14(Suppl.), 18–24.
Loft, A. (1986). Toward a critical understanding of accounting: The case of cost
accounting in the U.K., 1914–1925. Accounting, Organizations and Society, 11(2),
137–169.
Luft, J., & Libby, R. (1997). Pro?t comparisons, market prices, and managers’
judgments about negotiated transfer prices. The Accounting Review, 72(2),
217–229.
Luft, J., & Shields, M. D. (2003). Mapping management accounting: Graphics and
guidelines for theory-consistent empirical research. Accounting, Organizations
and Society, 28(2–3), 169–249.
Luft, J., & Shields, M. D. (2009). Psychology models of management accounting.
Foundations and Trends in Accounting, 4(3–4), 199–345.
Maas, V. S., van Rinsum, M., & Towry, K. L. (2012). In search of informed discretion:
An experimental investigation of fairness and trust reciprocity. The Accounting
Review, 87(2), 617–644.
Magee, R. P., & Dickhaut, J. W. (1978). Effects of compensation plans on heuristics in
cost variance investigations. Journal of Accounting Research, 16(2), 294–314.
March, J. G. (1987). Ambiguity and accounting: The elusive link between
information and decision making. Accounting, Organizations and Society, 12(2),
153–168.
Mastilak, M. C. (2011). Cost pool classi?cation and judgment performance. The
Accounting Review, 86(5), 1709–1729.
Matsumura, E. M., & Shin, J. Y. (2006). An empirical investigation of an incentive
plan with relative performance measures: Evidence from a postal service. The
Accounting Review, 81(3), 533–566.
Mittendorf, B. (2008). Infectious ethics: How upright employees can ease concerns
of tacit collusion. Journal of Law, Economics, and Organization, 24(2), 356–370.
Neumann, B. R., & Friedman, L. A. (1978). Opportunity costs: Further evidence
through an experimental replication. Journal of Accounting Research, 16(2),
400–410.
Neumann, B. R., & Friedman, L. A. (1980). The effects of opportunity costs on project
investment decisions: A replication and extension. Journal of Accounting
Research, 18(2), 407–419.
Roberts, J., & Scapens, R. (1985). Accounting systems and systems of accountability
– Understanding accounting practices in their organisational contexts.
Accounting, Organizations and Society, 10(4), 443–456.
Rowe, C. (2004). The effect of accounting report structure and team structure on
performance in cross-functional teams. The Accounting Review, 79(4),
1153–1180.
Rowe, C., Birnberg, J. G., & Shields, M. D. (2008). Effects of organizational process
change on responsibility accounting and managers’ revelations of private
knowledge. Accounting, Organizations and Society, 33(2–3), 164–198.
Schepanski, A., & Uecker, W. (1983). Towards a positive theory of information
evaluation. The Accounting Review, 58(2), 259–283.
Sprinkle, G. B. (2003). Perspectives on experimental research in managerial
accounting. Accounting, Organizations and Society, 28, 287–318.
Tafkov, I. D. (2013). Private and public relative performance information under
different compensation contracts. The Accounting Review, 88(1), 327–350.
J. Luft / Accounting, Organizations and Society xxx (2015) xxx–xxx 11
Please cite this article in press as: Luft, J. Management accounting in the laboratory and in social context: Four contrasts, 1975–2014. Accounting, Organiza-
tions and Society (2015),http://dx.doi.org/10.1016/j.aos.2015.08.001
Thompson, L., & Loewenstein, G. F. (1992). Egocentric interpretations of fairness and
interpersonal con?ict. Organizational Behavior and Human Decision Processes, 51,
176–197.
Tiller, M. (1983). The dissonance model of participative budgeting: An empirical
exploration. Journal of Accounting Research, 21(2), 581–595.
Towry, K. L. (2003). Control in a teamwork environment: The impact of social ties
on the effectiveness of mutual monitoring contracts. The Accounting Review, 78
(4), 1069–1095.
Uecker, W. C. (1978). A behavioral study of information choice. Journal of Accounting
Research, 16(1), 169–189.
Uecker, W. C. (1980). The effects of knowledge of the user’s decision model in
simpli?ed information evaluation. Journal of Accounting Research, 18(1),
191–213.
Uecker, W. C. (1982). The quality of group performance in simpli?ed information
evaluation. Journal of Accounting Research, 20(2, Part 1), 388–402.
Waller, W. S., & Chow, C. W. (1985). The self-selection and effort effects of standard-
based employment contracts: A framework and some empirical evidence. The
Accounting Review, 60(3), 458–476.
Williamson, M. G. (2008). The effects of expanding employee decision making on
contributions to ?rm value in an informal reward environment. Contemporary
Accounting Research, 25(4), 1183–1209.
Wilner, N., & Birnberg, J. G. (1986). Methodological problems in functional ?xation
research: Criticism and suggestions. Accounting, Organization and Society, 11(1),
71–80.
Young, S. M. (1985). Participative budgeting: The effects of risk aversion and
asymmetric information on budgetary slack. Journal of Accounting Research, 23
(2), 829–842.
Zimmerman, J. L. (2001). Conjectures regarding empirical management accounting
research. Journal of Accounting and Economics, 32(1–3), 411–427.
12 J. Luft / Accounting, Organizations and Society xxx (2015) xxx–xxx
Please cite this article in press as: Luft, J. Management accounting in the laboratory and in social context: Four contrasts, 1975–2014. Accounting, Organiza-
tions and Society (2015),http://dx.doi.org/10.1016/j.aos.2015.08.001

doc_906030161.pdf