Accounting, Organizations and Society 47 (2015) 25e42
Contents lists available at ScienceDirect
Accounting, Organizations and Society
journal homepage: www.elsevier.com/locate/aos
Regulation and the interdependent roles of managers, auditors, and
directors in earnings management and accounting choice*
Robert Libby a, *, Kristina M. Rennekamp a, Nicholas Seybert b
a
b
Samuel Curtis Johnson Graduate School of Management, Cornell University, Ithaca, NY 14853, United States
Robert H. Smith School of Business, University of Maryland, College Park, MD 20742, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 23 June 2015
Received in revised form
3 September 2015
Accepted 11 September 2015
In this paper, we examine the growing number of behavioral studies of how financial reporting, auditing,
and other corporate governance regulations affect earnings management and accounting choice-related
decisions of managers, auditors, and directors. We first describe how experimental and survey studies
can add unique insights into our understanding of earnings management and accounting choice. We
then organize our review of the literature by the type of regulation (financial reporting, auditing, or
corporate governance) and secondarily by which of the three parties are affected. Finally, we point out
useful directions for future research and discuss key methodological choices faced by those who will
conduct that future research.
© 2015 Elsevier Ltd. All rights reserved.
Keywords:
Earnings management
Earnings quality
Accounting choice
Financial reporting
Auditing
Corporate governance
Experimental design
Surveys
Regulation
1. Introduction
This paper examines recent experimental and survey studies of
managers', auditors', and directors' (or audit committee members')
decisions that influence earnings management and accounting
choice, and how these decisions are affected by financial reporting,
auditing, and other corporate governance regulations. We evaluate
what we have learned from these studies, point out useful directions for future research, and discuss key methodological
choices faced by researchers in this area. Like our earlier review
(Libby & Seybert, 2009), we employ a broad definition of earnings
management and accounting choice to include: (1) choices of accounting methods; (2) implementation decisions related to estimates, classifications, levels of detail, and display format used in
mandatory disclosures; (3) the frequency, timing, and content of
*
The authors wish to acknowledge the helpful comments of Scott Asay, Robert
Bloomfield, Scott Emett, Mark Nelson, Ken Trotman and participants at the Accounting, Organizations, and Society's 40th Anniversary Event.
* Corresponding author.
E-mail
addresses:
rl54@cornell.edu
(R.
Libby),
kmr52@cornell.edu
(K.M. Rennekamp), nseybert@rhsmith.umd.edu (N. Seybert).
http://dx.doi.org/10.1016/j.aos.2015.09.003
0361-3682/© 2015 Elsevier Ltd. All rights reserved.
voluntary disclosures; and (4) investment, financing, and operating
choices based on their accounting (rather than economic) consequences (Libby & Seybert, 2009, p. 291). The experimental and
survey studies that we focus on examine the determinants of accounting choice and not their consequences for users and market
prices.1
Since our earlier review, there has been an uptick in experimental and survey studies of the determinants of earnings management and accounting choice. While the majority of the
literature on earnings management and accounting choice uses
archival methods to draw inferences, experiments and surveys
have different strengths and weaknesses that make them particularly useful for studying certain aspects of accounting choice. First,
as Francis (2001, p. 310) notes, most of the earlier accounting choice
literature does not include decision makers other than the manager. The need to study the effects of other parties to the process,
1
Different portions of the consequences literature have been reviewed recently
by Dechow et al. (2010) and Libby and Emett (2014). Studies which examine
managers', auditors', and directors' beliefs about the consequences of their actions
are included where relevant.
26
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
such as auditors and audit committees (or directors), serves as the
basis for many recent experimental and survey studies. Indeed,
experiments and surveys have a comparative advantage in their
ability to tease out the unique contributions of each party. Second,
as Libby and Seybert (2009, p. 293) suggest, archival studies are
limited to examining the effects of existing regulatory regimes,
which makes it difficult to determine which specific elements of
the regulatory regime impact the observed accounting choices. In
experiments, specific elements of the regulatory regime can be
independently manipulated to disentangle their effects on the
parties' actions. Experimental and survey researchers can also
investigate the effects of regulations that do not currently exist. And
third, intermediate process measures are often captured in experiments and surveys, which allow assessment of the impact of
specific motives, beliefs, and cognitive processes of the parties
involved and how they interact with elements of regulation.
On the other hand, experiments, and to a lesser extent surveys,
have limited ability to representatively sample decisions, settings,
and actors. This limits their ability to estimate the magnitude or
importance of effects. Also, experiments that rely on manipulation
of independent variables can only focus on a small number of effects. Furthermore, many variables are held constant, which can
limit the generalizability of results or hide important interactions.
Surveys are limited in their ability to illuminate non-conscious effects and are subject to a number of forms of response bias (see
Nelson & Skinner, 2013, for a detailed discussion). In summary,
different methods are useful for addressing different parts of
research questions related to accounting choice, and a multimethod approach is often warranted.
The earnings management and accounting choice literature
generally views managers' choices as being motivated by managerial self-interest and maximization of current shareholders' interests (Fields, Lys, & Vincent, 2001; Francis, 2001). Tests of the
effects of managerial self-interest rely mostly on differences in
aspects of compensation contracts. Tests of maximization of current shareholders' interests rely mostly on differences in capital
market pressures and differences in the importance of the liquidity
benefits of transparency versus loss of competitive advantage (e.g.,
public vs. private ownership, the need for additional equity or debt
financing, and industry competitiveness).
The broader accounting quality literature (see Dechow, Ge, &
Schrand, 2010, for a recent review) also recognizes the importance of auditors and directors as potential monitors that may
constrain earnings management and accounting choice. In the
auditing literature, auditors are often portrayed as balancing their
wish to satisfy client management with their wish to avoid both
out-of-pocket costs of litigation and regulatory enforcement, as
well as the longer-term costs of reputation damage (e.g.,
Hackenbrack & Nelson, 1996; Watts & Zimmerman, 1978). Similarly, in the corporate governance literature, directors or audit
committee members are portrayed as balancing their wish to
satisfy management with their wish to avoid litigation and regulatory enforcement costs, including longer-term costs to reputation
(e.g., Fama, 1980; Hermalin & Weisbach, 1998).
Consistent with the above description of the various parties'
motives, financial reporting, auditing, and corporate governance
standards and regulations can be viewed as limits set on the effects
of manager, auditor, and director motives. These limits operate by
specifying required and prohibited types of behavior. Violation of
the limits can be sanctioned through the courts or regulatory
processes and the regulations also specify the type and magnitude
of the potential sanctions. Many of the standards and regulations
still leave room for a good deal of discretion (or judgment) on the
part of all three parties involved in accounting choices. And, at the
time of their issuance, there is uncertainty surrounding the exact
manner in which enforcement agencies will interpret the standards
and regulations and impose sanctions for infractions. Enforcement
actions and speeches by regulators fill in many of these missing
details over time. They also allow the regulators to efficiently react
to the changing business environment.
One feature distinguishing the experimental and survey literature is that it places a more significant emphasis on cognitive factors that may affect the manner in which human managers,
auditors, and directors form their beliefs and preferences, which
determine their choices (e.g., Baker & Wurgler, 2013; Koonce &
Mercer, 2005; Koonce, Seybert, & Smith, 2011). These cognitive
factors include self-serving attribution bias, different forms of
overconfidence, anchoring on regulations formerly in force, manager/auditor/director personality traits, weighting of sunk costs,
social identity factors, moral licensing, and others.
There are a number of additional complications in studying the
determinants of accounting choice that have been recognized in the
behavioral literature. One concern is that each regulation can influence the judgments and decisions of any or all of the three parties
involved in the financial reporting process. The limits imposed by
regulations also affect behavior in concert with cross sectional
differences in other attributes of the environment including the
compensation scheme for the managers, auditors, and directors, as
well as the transparency of their actions. Compounding the issue of
cross-sectional differences in the environment, there are reliable
individual differences in the manner in which each of these three
parties respond to regulations and environmental attributes.
Finally, the effects of regulations may also be dependent on other
accounting choices that have been made in the current or prior
periods. All of these complicating factors suggest the possibility of
interesting interactions, and it is an emphasis on these interactions
that differentiates much of the behavioral literature.
The remainder of the paper is organized as follows. In Section 2,
we review the existing literature and derive the key conclusions
from each stream. We organize the literature first based on the type
of regulation, and then by the parties affected. In Section 3, we
discuss directions for future research. In this section we focus on
further research into the aforementioned interactions. Our discussion of future research opportunities also provides some guidance
for a greater focus on understanding causal mechanisms and
evaluating reporting outcomes, and points out the benefits of taking a broader view of regulation. Section 4 examines key research
choices that determine the effectiveness and efficiency of experimental and survey research on accounting choice. These choices
include the realism of stimuli, choice of accounting setting, and
selection of participants. We then make recommendations concerning different approaches to examining decision processes. Key
issues in this regard include the preeminence of clever experimental design, best practices in mediation analysis, and methods to
examine non-conscious processes. Section 5 concludes. Our review
focuses mainly on papers published in the 2008 through 2014
volumes of Accounting, Organizations, and Society; Contemporary
Accounting Research; Journal of Accounting Research; and The Accounting Review. We also include several working papers from
SSRN, and discuss selected older papers that provide the motivation
for the more recent papers.
2. Effects of regulation
As in Libby and Seybert (2009), we discuss how (1) financial
reporting regulations, (2) auditing regulations, and (3) other
corporate governance regulations affect managers', auditors', and
directors' judgments and decisions with respect to earnings management. A key to this organization is recognizing that each regulation can influence the judgments and decisions of any or all of the
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
three parties involved in the financial reporting process (see Fig. 1).
For example, rules on audit committee independence imposed by
the Sarbanes-Oxley Act (SOX) may affect directors by inducing
greater scrutiny of reported financial numbers (e.g., Cohen, Hayes,
Krishnamoorthy, Monroe, & Wright, 2013), but may also affect
auditors by increasing the support that they can expect to receive
from the audit committee in challenging management (e.g., Cohen,
Krishnamoorthy, & Wright, 2010).
2.1. Financial reporting regulation
Financial reporting regulations provide the ground-level
framework for conveying economically relevant and reliable information to investors, analysts, and other firm stakeholders. They
specify both required disclosures and some limits to voluntary
disclosures, and, as noted earlier, the requirements and limits still
leave room for a good deal of discretion. The effects of auditing and
corporate governance regulation depend on the nature of the
financial reporting regulations, as auditors and directors often must
decide whether the firm's financial reports are consistent with the
reporting regulations. The primary responsibility for complying
with financial reporting regulations rests with the CFO and CEO of
the firm, as codified by SOX, where the signatures of these executives assert the accuracy and fair presentation of the reports. For
this reason, much of the financial reporting regulation research
focuses on the effects on managers. However, we also discuss
recent research examining the effects on auditors and directors.
Most of the recent research addresses three broad issues: (1) the
effects of rules- versus principles-based standards on reporting
choices, (2) the effects of information quantity or information
location on reporting choices, and (3) the effects of reporting frequency and accruals timing on investment choices.
2.1.1. Rules versus principles
The first and most rapidly growing area of research in financial
reporting regulation is motivated by what is often described as a
fundamental difference between IFRS and U.S. GAAP accounting e
rules- versus principles-based standards. As suggested by Schipper
(2003), however, principles versus rules can represent a continuum
in any set of accounting standards. An accounting standard is
almost always based on a broad underlying principle, and increased
Managers
REGULATION
Financial ReporƟng
(SEC and FASB)
AudiƟng
(PCAOB)
Auditors
Other Corporate
Governance
(SEC and Stock
Exchanges)
Directors
Fig. 1. Effects of regulation on earnings management and accounting choice.
27
detail and precision are then layered on top in order to specify
criteria for inclusion, exclusion, and application of the principle in
specific circumstances. Nelson, Elliott, and Tarpley (2002) provide
an important early foundation on which the subsequent research is
based, as the authors shed light on how managers attempt to
navigate imprecise accounting standards by altering assumptions
and precise standards by structuring transactions. Unlike the broad
survey evidence in Nelson et al. (2002), recent research often utilizes an experimental design that necessitates focus on a narrow
accounting topic such as revenue or lease recognition. For this
reason, choosing a proxy for principles versus rules that generalizes
to an entire set of standards can be difficult.
Two recent studies investigate the effects of principles versus
rules on managers' operating lease classifications. Jamal and Tan
(2010) test how lease accounting standards impact experienced
financial managers' attempts to avoid balance sheet reporting of
lease liabilities. They manipulate whether standards are principlesor rules-based and also whether the auditor is principles-, rules-, or
client-oriented. The type of standard does not have a substantial
effect on overall aggressive reporting. However, a principles-based
standard coupled with a principles-oriented auditor leads to
increased reporting of lease liabilities on the balance sheet. Agoglia,
Doupnik, and Tsakumis (2011) conduct a similar experiment using
CFO participants to test whether principles or rules lead to more
aggressive operating lease reporting, and how audit committee
strength affects this propensity. CFOs choose the operating lease
treatment more frequently under the rules-based standard than
principles-based standard. A strong audit-committee attenuates
this effect. Results of a follow-up experiment reveal that both the
probability of regulators second-guessing the lease treatment and a
desire to report the economic substance of the transaction underlie
the more conservative lease treatment choice under principlesbased standards.
McEnroe and Sullivan (2013) provide recent survey evidence on
a broad set of accounting issues likely to be impacted by principlesversus rules-based standards. The authors survey CFOs and auditors to ascertain their opinions on whether removing specific accounting rules and replacing them with more general principles
would improve the qualitative characteristics of financial reporting.
Ten specific rules are presented, including issues such as the choice
to account for investments using the equity method, the choice
between operating and capital lease treatment, the decision to
expense R&D, and the decision to report retail land sales on an
accrual basis using the percentage of completion method. For eight
of ten issues, participants believe that removing the detailed
reporting rules would actually harm reporting quality. Lease accounting represents the sole area in which participants moderately
agree that a shift from rules to principles could improve reporting.
This result is critical to interpreting the two previously discussed
managerial experiments as well as the subsequent auditor experiments where a primary focus is lease accounting e any results from
this literature should be generalized with caution as the lease issue
may represent a special case in the principles versus rules debate.
While the previous studies examined the behaviors and beliefs
of managers, most recent research on principles versus rules focuses on auditors. Broadly speaking, this line of research addresses
how the precision of the accounting standards upon which auditors
must provide their opinions affects their willingness to accept
aggressive client-preferred accounting methods. Accounting quality is reduced and earnings management is more likely if auditors
are more willing to accept aggressive client-preferred treatments
rather than pushing for methods that better reflect the economic
substance of a transaction. Many have argued that rules-based
standards lead to more aggressive accounting treatments than
principles-based standards because they offer a “bright line” for
28
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
preparers to aim for in structuring transactions. Experimental evidence in recent years is generally consistent with Nelson et al.'s
(2002) survey finding that audit managers and partners say that
they would be less likely to challenge a client's aggressive reporting
when the client structures a transaction to meet precise rulesbased standards.
Segovia, Arnold, and Sutton (2009) provide some of the first
experimental evidence concerning how auditors react to
competing pressures from clients versus from the SEC to constrain
aggressive reporting. They find support for the idea that auditors
are less likely to constrain aggressive accounting under rules-based
than under principles-based standards. Further, they find that auditors' experience drives variation in their responsiveness to client
vs. SEC pressure. More experienced auditors respond primarily to
greater client pressure by requiring greater adjustment to reported
numbers (reducing aggressive reporting), while less experienced
auditors respond primarily to greater SEC pressure by requiring
greater adjustment to reported numbers. It is important to note
that aggressive reporting under the rules-based standard involves
impairing a long-term asset that has experienced no change in
historical performance, while aggressive reporting under the
principles-based standard involves expensing supplies that have
not yet been used. Thus, the effect of principles versus rules may be
attributable to the relatively greater uncertainty about future performance of an impaired asset compared to the relatively lower
uncertainty about past use of supplies.
Recent research also suggests a specific reason why the greater
uncertainty surrounding principles-based standards might
improve accounting quality. Cohen, Krishnamoorthy, Peytcheva,
and Wright (2013) argue that principles-based standards will
improve audit quality because the greater uncertainty created by
principles will induce greater personal accountability for the decision process on the part of auditors. Greater process accountability will in turn cause auditors to seek out more audit evidence
and discourage aggressive reporting. As with much prior research,
the study utilizes a lease scenario where the rules-based standard
indicates operating lease treatment is appropriate, but where the
principles-based standard suggests the economic substance of the
transaction merits capital lease treatment. Auditors are more likely
to require capital lease treatment under the principles-based
standard than under the rules-based standard, and this holds
regardless of the strength of the regulatory regime. Peytcheva,
Wright, and Majoor (2014) find that the increased process
accountability induced by principles may also have detrimental
effects, as auditors request both more diagnostic and nondiagnostic evidence items under principles-based standards.
It is important to note an additional caveat concerning the
manner in which the leasing standard is operationalized in the
previously discussed manager and auditor studies. Typically, the
lease just barely meets the criteria for an operating lease under the
rules-based standard, and thus the aggressive reporting choice is
sanctioned under the standard. The principles-based standard, on
the other hand, does not make an explicit allowance for off-balance
sheet reporting. Because the principles-based standard is thus
more stringent, participants may view aggressive reporting as more
difficult in this case. The exception is Agoglia et al. (2011), who
conduct a robustness test constrained to participants who view the
less precise principles-based standard as conveying similar numerical guidance to the more precise rules-based standard, which
does not alter their primary findings.
Quick (2013) provides an opposing view on the principles versus
rules debate by demonstrating that principles-based standards can
allow more flexibility to report aggressively when aggressive
reporting is warranted. Using audit seniors and managers, her
experiment manipulates the level of legal liability and the precision
of the accounting standard regarding revenue recognition in a
scenario where aggressive revenue recognition represents the
economic substance of the transaction. Results indicate that auditors are more likely to allow aggressive revenue recognition under a
principles-based standard and when they have limited legal
liability.
Two recent papers studying the effects of principles versus rules
on auditors provide the most nuanced views on this topic. Backof,
Bamber, and Carpenter (2014) find that rules-based standards can
facilitate or constrain aggressive reporting compared to principlesbased standards, and that the directional effect depends on the
stringency of the rules-based standard. They find that rules-based
standards lead auditors to allow more aggressive reporting than
principles-based standards for a lease transaction (where the rules
indicate an operating lease is appropriate), but principles-based
standards lead auditors to allow more aggressive reporting than
rules-based standards for a transaction involving revenue recognition for a bundled good (where the rules indicate bundling is
inappropriate). This paper thus directly manipulates a key
confound in many previously discussed studies e the relative
stringency of the rules-based standard as compared to the
principles-based standard, and reveals that it is likely this stringency rather than the precision of the standard that determines
reporting aggressiveness. Kadous and Mercer (2012) highlight a
similar issue regarding the precision of accounting standards.
Investigating jury judgments against auditors, they manipulate
whether client reporting violates (complies with) a precise standard and find that juries return fewer (more) verdicts against auditors when the precise standard is replaced with an imprecise
standard. Backof et al. (2014) also find that providing auditors with
a judgment framework that encourages them to take high-level,
big-picture perspective of the accounting issue rather than a
detailed transaction view can help curb aggressive reporting.
Messier Jr., Quick, and Vandervelde (2014) also find that the
benefits of principles-based standards for constraining aggressive
reporting are not straightforward. In a setting related to the treatment of R&D expenses, they manipulate whether the prior-year
standard is IFRS-based and requires capitalization, IFRS-based and
requires expensing, or GAAP-based and requires expensing, and
find that both US and Norwegian auditor participants anchor on the
prior-year standard in making a decision about a current year accounting treatment. Because of this, some of the purported benefits
of principles-based standards may be reduced by auditors
anchoring on old rules-based standards, which are often still
allowed even when principles-based standards are introduced.
However, they also find that anchoring on old standards may be
reduced if auditors' accountability for the decision process is
increased.
Taken together, the evidence indicates that the ability of a
principles-based standard to constrain aggressive reporting is
probably very limited. However, it seems clear that the joint goals
of reducing the complexity of standards while also constraining
aggressive reporting will only be met where a less-stringent set of
rules is replaced with a more-stringent general principle (as in the
case of leases). Framing the comparison of principles versus rules in
the context of IFRS versus GAAP also seems not to be productive. As
Schipper (2003) and Nelson (2003) note, all standards have the
potential to include relatively more detail and precision, shifting
from a broader principle to a more specific set of rules. The degree
of precision is a debate that continues in standard setting around
the world, and has the potential to affect all stakeholders in the
economy. Unfortunately, few studies have identified the key attributes of principles versus rules that facilitate or constrain aggressive reporting, focusing instead on specific accounting issues such
as leases or revenue recognition. Future research might attempt to
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
develop a broader conceptual framework that aids in operationalizing these key attributes as opposed to selecting a specific accounting standard in an attempt to more directly address a current
reporting controversy. If future research does choose to focus on
comparisons of U.S. versus international accounting standards, the
uncertain status of convergence towards a global set of accounting
standards suggests that more research may be needed on the effects of a world with incomplete convergence, as was considered by
Asay, Brown, Nelson, and Wilks (2013).
While the bulk of research on financial reporting regulation and
reporting decisions clearly falls within the principles versus rules
debate, several other financial reporting regulations have received
attention in the literature. The two primary areas involve the
quantity of information and location of information that must be
presented. These issues are important as they represent other
fundamental aspects of financial reporting regulations that differ
both within and across regimes.
2.1.2. Effects of the required amount and location of information on
reporting choices
A number of recent experimental papers investigate how the
required amount or location of information impact reporting
choices. We first discuss several papers that focus on the amount of
information that the firm must disclose, followed by papers that
focus on the location of the information that the firm must disclose.
Majors (2014) examines whether a mandatory range disclosure
of estimate reasonableness alters aggressive reporting behavior,
particularly for individuals high in Dark Triad personality traits
(Machiavellianism, narcissism, and psychopathy). The study utilizes the experimental economics paradigm with student participants, testing whether incentive-consistent aggressive accounting
estimates will be attenuated by a mandated disclosure of a range of
reasonable estimates. When range is not disclosed, managers high
(compared to low) in Dark Triad personality traits provide more
aggressive point estimates for reporting to investors. When range is
disclosed, this tendency is attenuated such that all managers report
less aggressive estimates. The findings demonstrate that personality interacts with reporting requirements such that the negative
impact of manager psychology can be targeted with reporting
regulation.
Griffin (2014) considers how auditors may also be affected by
regulation requiring increased disclosure surrounding fair value
measurements. Drawing on moral licensing theory, he shows that
supplemental disclosure might actually increase misstatements in
fair value reporting. Griffin (2014) considers both input subjectivity
(i.e., the reliability of the inputs to fair value estimates) and
outcome imprecision (i.e., the degree of variability in future potential outcomes) in his study. He finds an interaction where more
subjective Level 3 estimates are more likely to require adjustment
from auditors only when they lead to a more imprecise range of
possible outcomes. When the range of possible outcomes is precise,
auditors' ratings of the likelihood that they would require an
adjustment do not differ depending on whether inputs are more or
less subjective. However, this interaction goes away when supplemental disclosure is provided to explain the subjectivity in the fair
value estimates, as well as the range of potential alternative outcomes. Combined, the results support the idea that including a
supplemental footnote disclosure makes auditors feel less responsible and therefore less likely to require adjustment. Further, Griffin
(2014) shows that, following auditing standards, auditors' required
adjustments tend to anchor on the lower bound of the range of
possible outcomes rather than the midpoint, such that the required
dollar amount of adjustments is actually lower when estimates of
future outcomes are imprecise and thus have a greater range of
possible outcomes.
29
The final paper on the amount of information reported considers a specific context in which the firm either may or may not
report potentially useful information to stakeholders e the reversal
of a large asset impairment. Under IFRS, impairments should be
reversed when value increases, but under GAAP this treatment is
not permitted. Trottier (2013) focuses on the effect of reversibility
on the initial willingness to record the impairment loss. Trottier
finds that managers are less likely to record impairments that are
not reversible, particularly when they have current bonus incentives. The results demonstrate that managers are worried about
foregoing current bonuses that cannot be regained in the future
should the asset's underlying value increase.
The remaining papers investigate the location in which information must be reported and how this affects reporting outcomes.
The determinants and consequences of information location and
other presentation effects has been the subject of both archival and
experimental research over the last two decades (see Libby &
Emett, 2014, for a recent review). Clor-Proell and Maines (2014)
investigate an important element of reporting regulation, recognition versus disclosure, to test whether cognitive and process
effort by managers contributes to differences in the reliability and
bias in disclosed and recognized numbers. Using controller and CFO
participants, the authors provide three optional methods for estimating a contingent liability. The methods increase in required
effort but decrease in resulting bias such that participants must
trade-off reporting accuracy with the effort they are willing to
exert. The authors predict that capital market pressure on public
company executives will magnify effort and reduce bias in recognized (as compared to disclosed) numbers. The results show that
public company executives utilize lower effort for disclosed liabilities and thus report more biased numbers, whereas private company executives exert equal effort regardless of whether the
liability is recognized or disclosed. These findings suggest that
regulation governing mandatory recognition and disclosure will
interact with capital market pressures to determine the ultimate
effort and bias underlying information provided to investors.
Only one recent paper has examined the effects of earnings
presentation regulations on auditor behavior. IFRS currently requires a higher level of disaggregation of expenses than does US
GAAP. Libby and Brown (2013) find that U.S. audit managers require
correction of smaller misstatements as disaggregation increases on
the face of the financial statements. This is consistent with the idea
that auditors believe that disaggregated numbers would be subject
to greater SEC scrutiny, and that disaggregated line items represent
materiality benchmarks. However, Libby and Brown (2013) find
that the effect goes away when the disaggregation is instead reported in the footnotes (as is allowed under IFRS), presumably
because the footnotes are expected to receive less scrutiny. Interestingly, they also find a substantial amount of disagreement
among auditors on the issue of how disaggregation affects materiality. While disaggregation on the face of the financial statements
reduces the average amount of misstatement that auditors report
they will tolerate, it also increases the variance in their responses.
Follow-up questions further highlight this disagreement e 58% say
disaggregation changes the materiality of the reported expense
amount, but 42% say that it does not. Overall, their study suggests
that disaggregation can increase the reliability of reported financial
statement numbers because smaller misstatements will likely be
corrected at the insistence of auditors, but that there is substantial
disagreement among auditors on the issue. Their results suggest
that better guidance is needed if auditors are expected to make
consistent decisions about materiality.
Taken together, research on the amount and location of reported
information generally suggests that when more detailed information about estimates or economic events must be reported,
30
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
managers will engage in a lower level of misreporting and auditors
will scrutinize the numbers to a greater extent. However, the effect
is dampened when information appears in a footnote disclosure
rather than on the face of the financial statements e managers and
auditors are less concerned about the accurate presentation of
disclosed information. However, it is important to note that there
may be exceptions to this general intuition, as Griffin (2014) shows
that auditors may abdicate their responsibility to require corrections of imprecise estimates if supplemental disclosure makes
them feel as though users have received sufficient warning about
subjectivity in the reported numbers.
While the broader takeaways from the literature may largely
confirm readers' prior beliefs, some of the more interesting inferences stem from the subtler interaction effects. For example,
increasing the quantity of information reported may be particularly
helpful in curbing aggressive behaviors for managers with certain
personality traits (Majors, 2014), requiring more prominent
disclosure of numbers will lead to higher quality information when
capital market pressure is particularly high (Clor-Proell & Maines,
2014), and quantity of information interacts with location of information to determine whether reporting quality improves or
does not (Libby & Brown, 2013). These nuances leave a number of
important considerations for regulators and stakeholders as they
interpret the previous effects of financial reporting regulation and
contemplate potential intended and unintended consequences of
proposed future regulation.
2.1.3. Effects of reporting frequency and accruals timing on
operational decisions
The previously discussed financial reporting regulation studies
investigated the reporting and disclosure decisions brought about
by those regulations. However, reporting regulation may also affect
the real decisions that firms make and thus may have unintended
economic consequences. Over the last few decades, a number of
archival studies investigated the possibility that companies would
increase or decrease investment in accordance with the effects of
these investments on the financial statements (e.g. Baber, Fairfield,
& Haggard, 1991; Roychowdhury, 2006). The Graham, Harvey, and
Rajgopal (2005) survey paper, in which experienced executives
explicitly indicated their willingness to manage earnings through
real (operational) methods, spawned a resurgence of experimental
studies in this area. These experiments are particularly helpful
because they eliminate the endogeneity concerns that severely
limited the breadth of archival studies that could previously be
implemented. The archival research relied on very carefully chosen
settings and cleverly designed comparisons, where it was still not
completely clear that the operational decision could be separated
from or attributed to the accrual effects.
Wang and Tan (2013) provide evidence on the issue of reporting
frequency by examining how the frequency of a voluntary disclosure can alter managers' incentives to sacrifice firm growth for
predictable earnings streams. Utilizing MBA student participants,
the authors manipulate firm policy on the frequency of quarterly
earnings guidance and firm goals regarding that earnings guidance.
The manager's task is to choose between two strategic firm projects, one of which will provide higher economic benefits but lower
earnings predictability from quarter to quarter, and the other of
which will provide lower economic benefits but higher earnings
predictability. When the firm's policy is to guide infrequently,
participants with accuracy guidance goals prefer the project with
lower economic benefits and higher predictability. Participants
with meet/beat guidance goals prefer the project with higher
benefits and lower predictability. However, when the firm's policy
is to guide frequently, participants prefer the lower benefit, higher
predictability project regardless of guidance goal. These results
suggest that guidance frequency and guidance accuracy goals can
each contribute to a short-term focus on earnings predictability
that sacrifices long-term operational profitability.
Several of the papers in this area involve the timing of expenses.
Jackson (2008) examines a depreciation context and finds that
managers are more likely to sell an asset that has experienced more
depreciation under the accelerated depreciation method (as
compared to the straight-line method), holding economic characteristics constant. Additional measures of participant beliefs reveal
that the effect of straight line versus accelerated depreciation on
the asset disposal decision operates primarily through perceptions
of the value that the asset previously provided. Thus, the more
“used up” an asset appears to be based on accrual accounting, the
more sense it makes to a manager to dispose of it, regardless of the
true underlying economic circumstance. Jackson, Keune, and
Salzsieder (2013) demonstrate that source of financing can have
similar effects on asset disposal decisions. Managers are more
hesitant to dispose of debt-financed assets than equity-financed
assets, and this is especially true when the unpaid principal on
the debt is higher. Similar to Jackson (2008), belief measures indicate that an unpaid principal causes participants to perceive that
the asset has provided less value to date. Seybert (2010) explores
another expense timing difference, the capitalization versus
expensing of research and development, to show that managers are
more hesitant to abandon a failing project when the expenditures
have been capitalized and they were responsible for initiating the
project. This tendency is stronger for high self-monitors (those
more concerned about their reputation), and thus the results suggest that reputation concerns will drive investment decisions that
can increase reported earnings.
Brink, Gouldman, and Rose (2014) extend Seybert (2010) in an
experiment with MBA participants who assume the role of middle
managers who have knowledge of their superiors' executive
compensation incentives. The authors hypothesize that mere
knowledge of the executive incentives will induce subordinates to
alter investment decisions to optimize the financial reporting
outcomes. Results indicate that R&D capitalization (i.e., impairment
in the event of project abandonment) increases subordinates'
overinvestment in continuing projects when superiors have shortterm unrestricted stock compensation but decreases overinvestment when superiors have long-term restricted stock
compensation. These findings suggest that R&D reporting format
can negatively impact even middle managers when knowledge of
executive compensation exists. The authors also highlight how this
phenomenon could be magnified by Dodd-Frank requirements that
executive compensation packages receive increased transparency
and disclosure.
Graham, Hanlon, and Shevlin (2011) conduct a survey of tax
executives and multinational corporations to ascertain the effect of
non-cash tax expense on real investment decisions. Under GAAP,
firms that declare their foreign earnings to be permanently reinvested need not accrue tax expense on the future repatriation of
those earnings. This declaration affects the current GAAP tax
expense displayed on the income statement and accrued as a liability but not the future cash taxes paid, as the future cash taxes
depend upon the actual future repatriation decision. Tax executives
are asked to indicate the importance of items such as actual cash
taxes, foreign tax rates, and the GAAP tax expense on their decisions concerning which countries to invest in. Responses indicate
that the U.S. GAAP tax expense is roughly as important (or even
more important for high-R&D firms) as the actual cash taxes and
foreign tax rate in making real investment decisions. The same is
true for earnings repatriation decisions e the GAAP expense is a
very important factor in this real decision. These results suggest
that, contrary to economic theory, accrual basis tax expense alters
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
crucial global investment decisions.
The final paper investigates how impairment reversibility causes managers to make different investment decisions in impaired
divisions. Using student and experienced participants in two experiments, Rennekamp, Rupar, and Seybert (2015) find responsibility for asset impairment in a failing division induces
cognitive dissonance in managers. How managers resolve this
dissonance depends upon whether reporting regulations allow
impairment reversal in the future. When the impairment is
reversible in the event of asset value recovery, managers responsible for the initial impairment (compared to managers not
responsible for the impairment) allocate greater research and
development investment to the impaired division. When the
impairment is irreversible, responsible managers allocate lower
research and development investment to the impaired division.
These differences are attenuated when managers are given an
alternate mode of dissonance reduction through denial of responsibility for the division's poor prior performance. The findings
suggest that reversible impairments instill a sense of hope in
managers who have overseen a failure, while irreversible impairments instill a sense of hopelessness and blame on the failing
division.
Taken together, the results of these studies demonstrate how
reporting differences such as reporting frequency, expense deferral,
and impairment reversibility can alter project choices, investment
location, and investment magnitude. These decisions may be
particularly damaging because they have direct immediate and/or
future real economic consequences and represent a tradeoff between short-term reporting issues and long-term firm value. To
make matters worse, investors and information intermediaries may
have a harder time undoing the effects of these real earnings
management choices to discern the true underlying economic state
of the firm. This is one reason for managers to implement real
earnings management in the first place (e.g., Roychowdhury, 2006).
Evans, Houston, Peters, and Pratt (2015) examine its effects on
auditors' ability to detect real earnings management, which we
discuss in the following section on auditing regulation.
2.2. Auditing regulation
Although managers have a responsibility to refrain from earnings management, auditors play a key role in constraining earnings
management. Because of this, research that examines how auditing
regulation is likely to affect earnings management primarily considers regulations that are intended to increase the likelihood that
auditors constrain aggressive managerial reporting, typically by
shifting auditors' incentives. Auditors are more likely to constrain
aggressive reporting if regulations succeed in either making them
more independent from client management (and thus client pressure) and/or more accountable to non-client constituents (e.g., investors, regulators, etc.). This is consistent with the prior literature
that suggests auditors must balance their desire to satisfy client
management with their desire to avoid litigation, regulatory
enforcement, and reputational damage (Hackenbrack & Nelson,
1996). Both increased independence from clients as well as
increased accountability to non-clients shift auditors' incentives
away from pleasing management and allowing them to report
aggressively. We therefore categorize our discussion based on
research that focuses primarily on either auditor independence
(from client management) or accountability (to non-clients).
2.2.1. Auditor independence
Increasing auditors' independence from client management
may increase auditors' willingness to constrain managers' aggressive accounting decisions, often by reducing the quasi-rents that
31
result from incumbency. The most common mechanisms for
increasing auditor independence that have been examined in the
recent literature are firm or partner rotation. Winn (2014) finds that
mandatory partner rotation reduces planned effort in the final year
of the audit before rotation, and that rotation increases the time
spent on documentation relative to other activities that may be
more likely to improve audit quality. However, Winn (2014) does
not find that rotation affects independence (positively or negatively), where independence is captured by the magnitude of
adjustment proposed by auditors. In her study, a larger proposed
adjustment goes against client wishes, and reflects a more independent mindset.
Bauer (2014) also finds that rotation does not enhance independence, but considers a specific theoretical reason. Drawing on
social identity theory, he predicts that auditors can form very
strong bonds with new clients quickly, particularly when they share
more overlap with the client's norms and values. The results suggest that auditors form a relatively strong relationship with the
client after very short tenure, which would tend to reduce the
effectiveness of rotation for enhancing independence.
More broadly, Daugherty, Dickins, Hatfield, and Higgs (2012)
survey audit partners' opinions on how partner rotation affects
audit quality. The partners in their survey believe that rotation
improves independence, contrary to the evidence in the experimental papers discussed above. However, respondents also suggest
that rotation reduces audit quality because it reduces client-specific
knowledge and makes audit partners more likely to switch to a new
industry rather than relocating their families when it is time to
rotate. As a result, they argue that partner rotation leads to broader,
but shallower, industry expertise.
Efforts that affect auditor independence may also affect the
behavior of the managers who are subject to auditors' monitoring
decisions. Evans et al. (2015) conduct an international experiment
using managers domiciled in the U.S. versus Europe and Asia.
Managers choose how much to manage accruals versus real operations to help meet a target. The authors predict and find that
stronger audit enforcement in the U.S. induces managers to prefer
the less detectable real earnings management to accruals earnings
management, both when they are managing earnings upward and
downward. Under less stringent international audit enforcement,
managers prefer to use accruals for downward but real decisions for
upward earnings management. These results suggest that the differences in audit regulation and enforcement across regimes can
induce managers to engage in different forms of earnings
manipulation.
2.2.2. Auditor accountability
The second stream of research investigating the effects of audit
regulations focuses on increasing auditor accountability to nonclients (e.g., investors, regulators, etc.), although many results in
this stream suggest that these regulations may not work as intended.2 An early example is Kadous, Kennedy, and Peecher (2003)
who show that an SEC mandate to assess the quality of a client's
preferred accounting method, as well as alternative methods,
might actually increase auditors' commitment to a client-preferred
method. This presumably occurs because motivated reasoning
leads them to process information in a way that supports the client's preferences, rather than processing information objectively.
2
Although not specifically focused on the effects of regulation, Trotman et al.
(2015) review literature that examines another source of auditor accountability e
the hierarchical review process. Their discussion provides an overview of research
on how different operationalizations of the review process can affect preparers'
accountability and, ultimately, audit outcomes and audit quality.
32
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
More recent research builds on the idea that increased auditor
accountability may not be effective, or may have unintended consequences that result from cognitive biases.
For example, Piercey (2011) looks at how documentation requirements affect the leniency of auditors' judgments. He finds that
documentation can actually make auditors more lenient with clients on a misstatement task if they document their risk assessments qualitatively rather than quantitatively. Piercey (2011)
suggests that qualitative risk assessments allow for more
wordsmithing and ex-post flexibility in defining what a term
means, so that lenient risk assessments are more justifiable should
something go wrong.
Winn (2014) manipulates the likelihood of a PCAOB inspection,
which should also affect accountability to regulators. While she
finds that the expectation of a PCAOB inspection increases planned
audit effort, she does not find that it increases the extent to which
auditors require correction of a potential misstatement. Thus, one
important takeaway from Winn's (2014) paper may be that
increased audit effort does not necessarily always translate to an
improved audit outcome (in her case, misstatement correction),
which could have implications for whether we view increased
audit effort as either a positive signal regarding the audit process or
instead an indicator of reduced audit efficiency.
At a broader level, Westermann, Cohen, and Trompeter (2014)
survey experienced auditors to understand how different sources
of accountability affect professional skepticism. While accountability that arises from PCAOB inspections and documentation requirements generally increase professional skepticism, competing
sources of accountability (e.g., for engagement profitability) may
reduce professional skepticism. Further, responses also suggest that
the enhanced professional skepticism that arises from regulation
may induce a “check-the-box” mentality, which could actually
harm audit quality. In fact, this type of “check-the-box” mentality
may help account for Winn's (2014) finding that accountability
increases audit effort but does not improve misstatement correction, in that auditors may be primarily concerned with the
appearance of providing a thorough audit, rather than with actually
improving audit processes. Combined, research on the effects of
increasing auditor accountability via regulation suggests that it
may be insufficient for constraining earnings management,
particularly if it cannot overcome auditors' cognitive biases and
competing sources of accountability.
Still other papers examine how regulations that affect auditor
accountability influence the decisions that audit committee
members make in carrying out their oversight responsibilities. If
auditing regulations that increase accountability induce too much
compliance-focused behavior, one potential countermeasure
would be to introduce an Audit Judgment Rule (AJR) that prohibits
second-guessing of auditor's estimates when they are made in
“good faith and in a rigorous, thoughtful, and deliberate manner”
(Peecher, Solomon, & Trotman, 2013). Kang, Trotman, and Trotman
(2014) find that, in the presence of such an AJR, audit committee
members feel greater accountability to ensure the reasonableness
of the financial statements, although, on average, it does not make
them more skeptical or increase the number of probing questions
that they ask of the auditor.
While an AJR might decrease auditors' accountability, a separate
proposal that could instead increase auditors' accountability is the
requirement of additional audit report disclosure of critical audit
matters (i.e., areas that include significant management judgments
and estimates, or areas with significant measurement uncertainty)
(IAASB, 2013; PCAOB, 2013). This type of mandated disclosure
arguably increases auditors' accountability by drawing attention to
areas requiring greater attention in the audit. Kang (2014) manipulates (1) whether such a disclosure is required, as well as (2)
whether the investor audience is sophisticated, and finds that audit
committee members are more likely to challenge management's
estimates in the absence of the mandated disclosure, but only when
the investor base is relatively sophisticated. When the additional
audit disclosure is mandated, audit committee members are less
likely to question management's estimates, suggesting that, while
the additional proposed disclosure in the audit report may increase
auditors' accountability, it may simultaneously reduce the amount
of oversight by audit committee members. Additional analyses
suggest that the decline in audit committee oversight may be
driven by greater perceived accountability to management than to
investors, suggesting that the outcome may be different if responsibility towards investors is particularly salient to the audit
committee.
Overall, the research that examines how auditing regulations
affect earnings management suggests that the effects of regulations
are more complicated than they may at first appear to be. Results in
a variety of settings demonstrate that both increased auditor independence and increased auditor accountability can fail to produce desired outcomes and, in some cases, can even lead to
unanticipated consequences. Some research suggests that
increased independence from client management may not improve
outcomes, particularly if auditors are quick to form bonds with new
clients such that independence is compromised. Other research on
accountability to non-clients suggests that it increases effort and/or
documentation, but may not actually improve outcomes. Still other
research suggests that increasing the accountability of one group
reduces perceived accountability of another group and may lower
overall reporting quality. Thus, it is still largely an open question as
to which types of regulation might successfully shift auditors' incentives away from pleasing client management and towards
pleasing other non-client constituents. We now turn our discussion
to research on the effects of corporate governance regulations.
2.3. Corporate governance regulation
Audit committee members may be the last line of defense
constraining earnings management. Because of this, much of the
research on the effects of corporate governance regulation has
examined how regulations might help the audit committee to be
more effective in carrying out its oversight duties. Like auditors,
audit committee members must balance their desire to please
management with their desire to constrain management and
please others. Also like auditors, audit committee members are
more likely to constrain aggressive reporting by management as
they become more independent from management and/or more
accountable to constituents other than management. We therefore
structure our discussion of corporate governance regulation based
on research that focuses primarily on either audit committee independence (from management) or accountability (to nonmanagement constituents). We conclude the section with a discussion of research that investigates how auditors react to changes
in audit committee members' independence and accountability.
2.3.1. Audit committee independence
To facilitate oversight of financial reporting, one specific
requirement of SOX is that the audit committee consist of independent directors, presumably to improve the flow of information
with auditors and to reduce the potential for influence by managers. In interviews with 22 experienced directors, Cohen,
Krishnamoorthy, et al. (2013) provide evidence on how this may
affect financial reporting quality. Consistent with the idea that SOX
gives audit committees greater authority over the audit process, the
vast majority of respondents (86%) indicate that the audit committee has the most say over hiring and firing decisions with
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
respect to the external auditor, although they also agree, on
average, that management still has some influence in the process.
Directors also generally agree that the frequency of meetings with
external auditors has increased post-SOX, and that these meetings
involve more back-and-forth exchange of substantive questions
and ideas.
SOX also requires that the audit committee include at least one
“financial expert”.3 Cohen, Krishnamoorthy, et al. (2013) also provide some evidence on the effects of this requirement. While the
majority of directors agree that there is sufficient financial expertise on audit committees (post-SOX), some expressed concern that
those who qualify as “financial experts” tend to be more rulesoriented and less knowledgeable about the broader business of
the firm. As a result, some believed that financial experts were less
likely to understand and identify the business risks of the firm,
albeit were more likely to understand and identify risks related to
financial reporting quality.
While the structure of the board can directly affect its independence, Magilke, Mayhew, and Pike (2009) provide evidence on
a specific institutional feature e stock-based compensation e that
may further affect audit committee members' independence and
ability or willingness to constrain aggressive financial reporting.
Using an abstract experimental market with student participants,
they find that audit committee members are most objective in the
absence of stock-based compensation.4 In the presence of stockbased compensation, they find that audit committee members
are biased towards upward earnings management when they
receive stock-based compensation tied to the earnings of current
shareholders, but more conservative, downward, earnings management when they receive stock-based compensation tied to the
earnings of future shareholders. The results of the experiment
suggest that independence requirements imposed by SOX may be
compromised by the form of compensation that is provided to audit
committee members, in that stock-based compensation may
reduce willingness to constrain aggressive earnings management.
Persellin (2013) uses an experiment with MBA student participants serving as proxies for audit committee members to examine
the effects of both compensation and likelihood of PCAOB inspection. Consistent with prior research on the effects of stock option
compensation (e.g., Magilke et al., 2009), she finds that participants
are more likely to side with management (in this case agreeing that
an income-decreasing adjustment is not necessary) when
compensation is based on stock options rather than cash. However,
this only occurs when the likelihood of a PCAOB inspection is low.
The effect is moderated when the likelihood of PCAOB inspection is
high, where participants are instead more likely to agree with auditors that an income-decreasing adjustment is necessary, regardless of the form of compensation. Thus, while PCAOB inspection
may be primarily intended to increase the accountability of auditors, Persellin (2013) finds that the threat of inspection may also
increase the willingness of audit committee members to act independently of management's wishes.
2.3.2. Audit committee accountability
In cases where audit committee independence may be
compromised, one potential solution is to increase disclosure about
audit committee members' compensation and/or relationships
with management. This has the potential to increase audit committee members' accountability to non-management constituents,
3
Alternatively, if no financial expert serves on the audit committee the firm must
disclose why that is the case.
4
Section 4 includes additional discussion on when non-experts are most likely to
be a reasonable proxy for experienced professionals in an experiment.
33
by drawing attention to potential areas of concern. Two recent
experiments using active directors as participants look at how
transparent disclosures can increase accountability and alleviate or
exacerbate problems associated with a lack of audit committee
independence. Rose, Mazza, Norman, and Rose (2013) examine the
interactive effects of director stock ownership and transparency of
director decisions. They show that stock-owning director participants in their experiment are more likely to oppose earnings
management attempts by management if their board discussions
are transparently disclosed. This indicates that the often suggested
policy of required stock ownership for directors should be tied to
greater transparency of director reporting if the audit committee is
to be an effective check against earnings management.
Rose, Rose, Norman, and Mazza (2014) examine whether
disclosure of any relationship between a CEO and board members
can increase director accountability to non-management and help
to mitigate problems associated with a lack of independence from
management. They first demonstrate that friendship ties increase
directors' willingness to accept proposed R&D cuts to meet the
CEO's bonus target. These cuts improve reported short-term performance to benefit management, but at the expense of long-term
performance of the firm. Further, the authors find that disclosure of
friendship ties may backfire, and actually increase directors' propensity to approve the cuts proposed by the CEO. This counterintuitive result is presumably driven by the fact that disclosure of
friendship ties provides directors with a subconscious psychological “moral license” to engage in more selfish behavior. In other
words, the fact that the relationship is disclosed makes directors
feel as though they have fulfilled their professional responsibilities,
giving directors more psychological freedom to act in the best interest of their friend (the CEO). Finally, Rose et al. (2014) find that
investors are more likely to agree with the directors' decisions
when friendship ties are disclosed, consistent with the idea that
investors view the disclosures as a signal of directors' transparency
and objectivity. Combined, the results of this study suggest the
importance of independence between directors and CEO's, and that
simply disclosing any relationship between the two is not sufficient
for increasing directors' perceived accountability and constraining
earnings management.
2.3.3. Effects of corporate governance regulations on auditors
Although corporate governance regulations are aimed primarily
at changing the behavior of management and audit committee
members, research has examined how auditors might also be
affected through the amount of support they expect to receive from
audit committee members in constraining aggressive reporting.
The recent research builds on earlier studies discussed in more
detail by Libby and Seybert (2009). For example, Libby and Kinney's
(2000) experiment suggests that, pre-SAB99 and pre-SOX, boards
were not a particularly effective control on earnings management
to meet earnings benchmarks. They find that, even in the presence
of a requirement to report uncorrected misstatements to the audit
committee, auditors still allow managers to engage in earnings
management, and avoid full correction of quantitatively immaterial
misstatements, if correction would lead to a missed analyst
consensus forecast.
Further supporting the importance of a strong audit committee,
Ng and Tan (2003) similarly find that audit managers, on average,
expect a smaller adjustment of a material misstatement in the
absence of a strong audit committee and authoritative guidance on
the audit issue. DeZoort and Salterio's (2001) earlier survey of
experienced directors finds that auditors can expect more support
from audit committee members that have either more audit
knowledge or more experience on independent corporate boards.
More recent papers further support the idea that auditors can
34
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
expect more support and involvement from the audit committee in
the post-SOX environment. Beasley, Carcello, Hermanson, and Neal
(2009) find that audit committee members appointed post-SOX are
more likely to question auditors about methods and potential alternatives. Cohen et al. (2010) interview experienced auditors, who
indicate that their planning phase has become much more
dependent on perceived strength of corporate governance in the
firm. Whereas meetings with the audit committee pre-SOX were
fairly passive, auditors' perception is that meetings are now much
more active and involve more give-and-take. Providing complementary evidence, Cohen, Krishnamoorthy, et al. (2013) conduct
detailed interviews of experienced directors to get their opinions
on the effects of SOX. Consistent with Cohen et al.'s (2010) interview results, Cohen, Krishnamoorthy, et al. (2013) find that directors believe the meetings between auditors and the audit
committee have become more frequent and more involved, and
that the issues discussed are more substantive. However, they also
find that directors perceive that, post-SOX, auditors have become
more frustrated, more overworked and much more rules-oriented
or rigid, presumably out of greater concern about facing litigation
and PCAOB inspection. Directors express the belief that auditors
now act more like “policemen” rather than collaborators with
management and the audit committee. While directors' responses
in Cohen, Krishnamoorthy, et al. (2013) suggest that SOX may have
had some negative effects on auditors' attitudes, they also suggest
that auditors may be taking their monitoring role even more seriously to crack down on aggressive accounting. This is also consistent with responses provided by CFOs in the survey by Dichev,
Graham, Harvey, and Rajgopal (2013).
Cohen, Gaynor, Krishnamoorthy, and Wright (2011) examine
audit partners' and managers' decisions in light of the strength of
audit committee independence. They predict and find an ordinal
interaction. Auditors waive the largest amount of the proposed
adjustment, regardless of audit committee independence, when
managers have low incentives to manage earnings, presumably
because the risk of misstatement is low. When managers have
strong incentives to manage earnings (i.e., to meet/beat a forecast),
auditors waive more of a proposed adjustment when the audit
committee is less independent. Their results provide additional
support for the idea that audit committee independence may increase the extent to which auditors believe that they will have
backing from the audit committee in pushing for an adjustment. A
big takeaway from Cohen et al. (2011) is that auditors appear to be
relatively lenient when their enforcement is needed the most e
that is, when incentives to manipulate are high and the audit
committee is not very independent and therefore unlikely to
challenge the CEO.
Not all studies support the idea that increased audit committee
independence has been beneficial for auditors. McEnroe (2007)
surveys CFOs and audit partners and finds that, while they agree
that SOX reduces egregious earnings management that violates
GAAP, they believe it has limited impact on within-GAAP manipulations. Gibbins, McCracken, and Salterio (2007) survey CFOs and
find that they think audit committee independence as well as who
serves as the chair of the committee are unimportant in resolving
auditor-manager disputes. Respondents in their survey indicate
that the audit committee does not help to resolve conflicts, and
instead wants information brought to their attention only after a
resolution has been reached. Only half of the respondents say that
the audit committee plays an important role in resolving conflicts
with management.
Combined, the evidence in the research on corporate governance regulations suggests that audit committee involvement has
improved in recent years, and more knowledgeable and independent directors have been more likely to support correction of
discovered errors and constrain earnings management. But, in
some cases it appears that directors are not as effective at helping
auditors to actually reign in within-GAAP earnings management.
3. Future research
From our review of the literature, some broad themes emerge
about where future research has the potential to make the greatest
contributions. We structure our review of research along the dimensions of financial reporting regulations, auditing regulations
and corporate governance regulations. This is, in part, driven by the
fact that most research examines the effects of only a single given
regulation, and mostly on a single decision maker. In the real world,
however, multiple regulations have the potential to influence
behavior, often in competing ways.
3.1. Important interactions
3.1.1. Multiple actors
There is room for more research on how a given group is affected
by regulations that may be primarily intended to affect another
group. For example, much of the research on financial reporting
regulations looks at effects on managers or on auditors, but pays
less attention to the effects on audit committee members who are
not the primary target of those regulations. Additional research on
the effects of other aspects of financial reporting regulation on
audit committees would be useful. Both Libby and Kinney (2000)
and Nelson et al. (2002) rely on auditor participants' perceptions
to draw inferences about the effects of regulation on audit committee members and managers, respectively. This indirect
approach may also be useful in other areas, especially given the
difficulty in recruiting experienced directors for experiments and
surveys.
Studies of audit regulations look primarily at the effects on auditors. Some notable exceptions are Persellin (2013), Kang (2014),
and Kang et al. (2014), which all examine how existing or proposed
auditing regulations affect the behavior of audit committee members. These studies provide some evidence on how regulations
affecting auditors' accountability and/or independence affect the
stance that audit committee members take in their interactions
with auditors. Trotman, Bauer, and Humphreys (2015) call for
additional research that both manipulates audit committee member behavior and observes resulting auditor behavior or, alternatively, manipulates auditor behavior and observes the resulting
behavior of audit committee members (e.g., with respect to their
questioning of auditors). Future work could also examine how
auditing regulations affect managers' actions and interactions with
auditors. Evans et al. (2015), which examines the effects of stronger
audit enforcement on managers' choices between accruals and real
earnings management, is a recent example examining this category
of issues. Finally, although corporate governance regulations are
often aimed at both directors and managers, less work is being
done in the experimental literature on how corporate governance
regulations affect managers' behavior. One exception is Ugrin and
Odom (2010), which investigates the likely impact of increased
sanctions imposed by SOX on executive attitudes toward fraudulent
reporting. They find that increasing jail time from one to ten years
significantly dampens attitudes towards misreporting, whereas
further increases from ten to twenty years have no significant
impact. Further analyses indicate that the primary costs of jail time
to executives come in the form of career opportunities and social
costs, such that even relatively small criminal sanctions represent a
strong disincentive to engage in fraudulent reporting. We suggest
that future work should do more to examine the effects of corporate governance regulations on managers, particularly since they
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
are often targeted by such regulations. To date, most of the evidence in this area has been provided via archival, rather than
experimental, research (see, e.g., Beyer, Cohen, Lys, & Walther,
2010; Cohen, Dey, & Lys, 2008).
In a similar vein, there is also still a lack of research examining
the effects of a single regulation on multiple actors (managers,
auditors, and directors). Much of the research focuses on how a
given regulation, along with other factors, influences the judgments and decisions of a given group with respect to earnings
management decisions. While this provides useful information, it
would also be useful to better understand how regulations influence the behavior of a given group in light of how another group is
influenced. For example, Cohen, Krishnamoorthy, et al. (2013)
provides evidence on auditors' perceptions of how board members' activities are affected by corporate governance regulations,
and how board members' interactions with auditors have changed
in response to auditors' reactions to corporate governance
regulations.
Although survey research has provided some evidence on how
different groups react to regulations in light of the reactions of
other groups, we propose that future experimental research could
provide additional insights to the literature. This is particularly
important given than an individual regulation may push one group
in a direction that is offset by another group. For example, imposing
an AJR that limits second-guessing of auditors' judgments may
reduce auditors' accountability to non-clients, but may be offset by
increased feelings of responsibility for oversight by directors (Kang
et al., 2014). Similarly, auditors' responses to changes in the auditors' report (e.g., disclosure of “key” or “critical” audit matters;
IAASB 2015) may be seen as relieving either managers or audit
committees of some of their responsibilities.
3.1.2. Multiple regulations
We also suggest that future research should do more to consider
that regulations do not operate in a vacuum, and that individuals
may be affected by multiple regulatory efforts at once. One paper
that bridges this gap is Persellin (2013), which examines audit
committee members' behavior in light of both their own
compensation (which is impacted by corporate governance regulations) but also PCAOB inspection (an auditing regulation).
Another is Griffin (2014) whose effects result from a combination of
financial reporting regulations and auditing standards related to
fair value measurement. We believe that these joint effects are a
particularly important area for future research.
3.1.3. Interactions among accounting choices
Related to the need to examine how the interaction of regulations or groups affect reporting outcomes, there is also a lack of
literature examining interactions among accounting choices and
how they are affected by context and the regulatory environment.
The tradeoff between real and accruals earnings management
represents one such broad category. Very few papers (including
archival papers) over the last decade have examined this issue. One
exception is the Evans et al. (2015) study suggesting that managers
in the U.S. GAAP regulatory environment exhibit a greater preference for real over accruals earnings management than managers in
an international IFRS environment. However, the specific drivers of
this difference are open for future research. Numerous financial
reporting, auditing, and corporate governance regulations would
likely impact managers' assessments of the tradeoff between these
two earnings management methods and/or auditors' and audit
committee members' ability to detect them. Another broad category of accounting choice interactions would be numbers reported
in the financial statements versus voluntary disclosures of quantitative or qualitative performance information. While studies have
35
examined the effects of disclosure regulation itself (e.g., Clor-Proell
& Maines, 2014; Majors, 2014), it remains unknown how regulations targeting qualitative disclosure would simultaneously affect
the reporting of other quantitative information (or vice versa) as
either complements or substitutes. In the real world, a large set of
reporting and disclosure choices are interdependent and managers
must choose the optimal combination, whatever their objective
function may be.
3.2. Understanding causal mechanisms and outcomes
As noted in Section 1, key elements of the experimentalists'
comparative advantage in the study of earnings management and
accounting choice are the ability to study effects on each party to
the process and to separate the effects of specific elements of a
regulatory regime. Where a finer understanding of the process is
important, there is a need for future experimental or survey
research to tackle issues that have so far been addressed primarily
by archival research. For example, the PCAOB is considering
whether to require the engagement partner on an audit to sign the
audit report. Carcello and Li (2013) use archival data to examine
whether such a requirement will affect audit quality. Their study
compares U.K. firms (that currently have a similar signature
requirement) to a matched sample of U.S. firms (where no such
requirement exists), and finds evidence consistent with the idea
that a signature requirement improves audit quality. While a
matched sample helps mitigate concerns about differences in U.K.
and U.S. firms, other cultural, regulatory, and enforcement differences between the two countries are likely to remain and may
affect their analyses. Again, an experiment can hold constant differences across firms and environments to provide convergent
evidence on the potential effects of these types of proposed regulations, as well as provide insights into the precise causal factors
though which these effects occur.
3.2.1. Understanding causal mechanisms
Understanding causal mechanisms requires a careful conceptual
structuring of the earnings management process and an understanding of which parts are targeted for change by an existing or
proposed regulation. We believe that research can benefit from
structuring at three levels: (1) elements of the earnings management process, (2) economic drivers, and (3) psychological drivers.
3.2.1.1. Elements
of
the
earnings
management
process.
Researchers should carefully consider whether they expect a
regulation to improve prevention, detection, and/or correction of
earnings management. For example, increases in the magnitude of
potential sanctions against management and the client firm for
GAAP violations, or within-GAAP biases in reporting, presumably
are aimed at preventing or discouraging earnings management
attempts. Holding constant auditors' and audit committee members' willingness to correct misstatements, regulations that are
meant to increase effort likely constrain earnings management
through better detection. Holding constant auditors' and audit
committee members' detection of misstatements, regulations that
are meant to increase independence from management likely
constrain earnings management through increased correction of
detected misstatements.
Some regulations may also affect more than one element of the
earnings management process. As Nelson et al. (2002) demonstrate, the level of precision in a financial reporting standard can
affect both the likelihood of an earnings management attempt and
auditors' correction decisions given detection. Furthermore, some
regulations may be designed to improve both detection and
correction.
36
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
In designing an investigation of the effects of a regulation, it is
useful for researchers to first think carefully about what the regulation is meant to achieve. If a regulation is meant to operate
through improved detection, the setting must be such that variation in detection can be observed, and the dependent variable must
likewise relate to detection of a misstatement rather than correction (and vice versa if the regulation is meant to improve correction). Furthermore, carefully delineating detection effects from
correction effects and choosing the appropriate dependent variables may change our perceptions of whether a given regulation
appears to be effective. For example, if a given regulation is
designed to improve detection, one should not conclude that the
regulation is ineffective if it does not lead to differences in
dependent variables that capture correction of misstatements.
3.2.1.2. Economic drivers. In Section 1, we discussed the fact that
managers' choices are often modeled as balancing managerial selfinterest and current shareholders' interests. Some reporting regulations such as those increasing transparency seem aimed at
bringing those interests into closer alignment. Reporting regulations that affect timing of recognition of revenues and expenses and
timing of reports appear to operate simply by changing what real
actions are in the manager's self-interest or the current shareholders' interests. Auditing regulations and corporate governance
regulations seem to be aimed either at increasing independence by
decreasing the value of satisfying client management, or increasing
accountability by increasing the costs of litigation, regulatory
enforcement, and long-term reputation loss. For example, mandatory audit firm rotation or consulting prohibitions should increase
independence by decreasing the value of satisfying management.
Inspection requirements increase the likelihood of sanctions on
auditors, and other aspects of the Sarbanes-Oxley Act increase the
magnitudes of sanctions, both of which increase the expected costs
of litigation, regulatory enforcement, and long-term reputation loss
to auditors. Similarly, regulation of the compensation of directors
can increase independence by affecting the value of satisfying client
management, and regulation of disclosures about board actions can
increase accountability by increasing the expected costs of litigation, regulatory enforcement, and reputation loss. Future research
would benefit from more precise specification of the economic
drivers underlying hypothesized changes in behavior.
3.2.1.3. Psychological drivers. More fundamentally, a number of
recent studies have focused on the effects of underlying cognitive
processes on reporting choices. For example, recent papers have
examined the effects of the related concepts of mindset and construal level on auditors' evaluations of accounting choices (Griffith,
Hammersley, Kadous, & Young, 2015; Rasso, 2015). Both studies
have implications for audit guidance and the organization of
workpapers provided within audit firms. An understanding of
these effects may also have implications for understanding the effects of auditing regulations that operate through the same processes. Similarly, Asay (2014) and Brown (2014) show how horizoninduced optimism and cognitive dissonance can bias both managers' probability estimates and utilities for outcomes that are key
determinants of earnings management attempts. These effects may
also provide a basis for designing and predicting the effects of
specific regulations aimed at reducing these effects in management
choices. Recent experimental and archival studies (Guggenmos,
2015; Pacelli, 2015) also demonstrate how broad elements of
corporate culture can affect the aggressiveness of financial
reporting choices and the accuracy of analysts' estimates. These
types of carryover effects suggest that corporate culture may
moderate the effects of a regulatory change on actions by the
parties involved in earnings management and accounting choice
(see Tarullo, 2014, for a discussion in the context of banking regulation). These issues are ripe for experimental and survey
explorations.
3.2.2. Evaluating reporting outcomes
Takeaways from the research on the effects of regulation would
be much clearer if researchers considered three issues related to
financial reporting outcomes. First, it is clear that there are
important contingenciesdunder some conditions the effect of a
regulation can be positive and under some conditions it will be
negative. Second, some regulatory changes have multiple effects
(e.g., they provide a benefit but increase a cost). And third, some
effects on process do not lead to reporting improvements.
Considering these three possibilities will help researchers explain
the possible implications of their research for practice, as well as
consider related directions for future research. For example, some
research on the effects of audit regulation appears to treat
increased audit effort as being synonymous with increased audit
quality. Similarly, some research on the effects of corporate
governance regulations appears to treat increased meetings and
interaction between auditors and the audit committee as a signal of
improved audit committee oversight. However, in both cases it
could instead be argued that the regulations are increasing the
appearance of audit quality and audit committee oversight, but that
actual outcomes are not improving.
3.3. A broader view of regulation
Finally, we encourage researchers to expand their definition of
what constitutes research on “regulations”, since variation in
enforcement can occur in the absence of formal regulatory changes.
The breadth and importance of these less formal regulatory channels can be seen from the issues discussed at the 2014 AICPA National Conference on Current SEC and PCAOB Developments (see
PwC, 2014 for a review). At that meeting, representatives of the
SEC discussed current regulations that were under review, indicated specific concerns about how factors used to determine
operating segments following current standards were being
weighted, indicated that the SEC would increase its focus on adequacy of income tax expense disclosures for 2015, reported on
changes in enforcement practices requiring that more companies
admit wrongdoing rather than “neither admit nor deny” responsibility, and discussed the future of IFRS for domestic companies as well as many other topics. Similarly, on the auditing side,
PCAOB representatives discussed a variety of current standardsetting initiatives, results and findings from current inspections,
and recent enforcement actions. FASB and IASB staff also discussed
current projects (e.g., leases) and implementation efforts related to
new standards (e.g., revenue recognition). Note that these topics
include more detailed interpretations of existing standards and
regulations, warnings in advance about areas of increasing scrutiny
which could lead to more detailed interpretations or enforcement
actions in the future, and even changes in the magnitude of sanctions imposed for certain infractions. None of these regulatory or
enforcement changes require a change in existing laws, standards,
or regulations.
4. Methodological considerations
Some methodological best practices for experimental research
in accounting are universal. For a discussion of these we refer
readers to discussions in Libby, Bloomfield, and Nelson (2002),
Kachelmeier and King (2002), Nelson and Tan (2005), and Bonner
(2008). In this section of the paper, we instead focus on methodological suggestions that are particularly relevant for research on
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
the effects of regulations on earnings management and accounting
choice.
4.1. Realism of stimuli
A key advantage of using experimental research to address
questions about the effects of regulations is that experiments are
able to create settings that do not yet exist. This allows for ex ante
research on accounting standards, which may help regulators understand the consequences (both intended and unintended) of
what is being proposed before costly regulations are put in place.
However, this also creates some challenges from an experimental
design perspective when it comes to determining the appropriate
level of realism for experimental stimuli. The regulatory process
can take many years, and proposals often go through many iterations (e.g., see the Libby & Kinney, 2000, study of the effects of the
proposed SAS 89). Researchers who map their stimuli too closely to
a current proposal may be unable to provide much insight should a
proposed regulation be changed down the road. From a design
perspective, researchers should choose a level of realism in their
materials that appropriately captures the constructs of interest
rather than the specifics of a proposal at a given time. We
encourage researchers to think about the conceptual aspects of a
given proposal that are most interesting. This allows us to learn
something even if (or, more likely, when) a proposal changes or is
never adopted. For example, regardless of whether the United
States ever actually adopts IFRS, the SEC has expressed interest in
moving away from rules-based standards and towards more “objectives-based” standards (SEC, 2012). This suggests that, regardless
of whether IFRS are ever adopted in the U.S., research is still
important if it can inform us on the effects of differences in the
specificity, use of examples, and other attributes of “rules-based”
vs. “principles-based” standards. As another example, recent
research has started to investigate the effects of discussing Key or
Critical Audit Matters (CAMs) in the audit report, both in terms of
how it will affect auditors (Kang, 2014; Kang et al., 2014) and investors/jurors (Backof et al., 2014; Brasel, Doxey, Grenier, & Reffett,
2014; Brown, Majors, & Peecher, 2015; Gimbar, Hansen, & Ozlanski,
2014; Kachelmeier, Schmidt, & Valentine, 2014). While there is still
some uncertainty about what a CAM disclosure might look like or
include, we can nevertheless learn from these papers so long as
their materials successfully capture important elements of the
construct of increased disclosure about areas that posed the most
difficulty for auditors or required the most subjective and complex
auditor judgments. Future research can also examine which elements of CAM disclosures cause any discovered effects.
4.2. Choice of accounting setting
Related to the idea that an appropriate level of realism should be
chosen, researchers must also be careful to select an appropriate
experimental setting. From our review of the literature, it is clear
that there is a lot of interest in understanding the effects of principles- vs. rules-based standards. Many of these studies use lease
accounting settings to contrast the two types of standards, perhaps
because it is a convenient one since there is general agreement that
current lease accounting standards are rules-based.
However, McEnroe and Sullivan (2013) survey Fortune 1000
CFOs and experienced auditors about principles-based vs. rulesbased standards, and find that lease settings may differ from
most others. They give respondents ten settings where GAAP incorporates rules, and ask whether elimination of the rule would
lead to accounting that better achieves the conceptual framework's
definition of useful financial information. In almost all settings, the
majority of respondents agree that financial reporting is improved
37
by keeping the rule rather than switching to more principles-based
standards. Importantly, the one setting where they believe the rule
should be eliminated is with operating leases. This suggests that a
lease setting may be quite different from many other settings,
which could affect the generalizability of the results in prior
research. At the very least, we encourage researchers and standard
setters to be clearer about the meaning of rules-based and
principles-based, and researchers to expand the number of settings
that they examine in the future when contrasting the effects of
elements of principles- vs. rules-based standards.
4.3. Selection of participants
Libby et al. (2002) suggest that the choice of participants should
match the goal(s) of an experiment, and that researchers should not
use more sophisticated participants than is necessary to achieve
those goals. Sophisticated participants are appropriate when the
goal of an experiment is to investigate the effects of a regulation in a
setting that requires a strong understanding of institutional features (i.e., studies that “peer into the minds of experts”). For
example, experienced auditors are likely necessary for research
that investigates how auditors' behavior changes in response to
regulations that affect the style and content of the audit report.
Alternatively, less sophisticated participants are appropriate when
the goal of an experiment is to investigate fundamental psychological biases or economic behaviors (Libby et al., 2002). Student
subject pools may be a useful source of less sophisticated participants (see, e.g., Elliott, Hodge, Kennedy, & Pronk, 2007, for a discussion of when MBA students are likely to be appropriate).
Researchers are also increasingly using Amazon's Mechanical Turk
(AMT) platform to recruit participants (see, e.g., Farrell, Grenier, &
Leiby, 2014; Krische, 2014; Rennekamp, 2012).
Regardless of their source, the use of less sophisticated participants allows researchers to address many more unanswered
questions, because the availability of less sophisticated participants
is much greater. A number of studies have used less sophisticated
MBA student participants as proxies for managers to investigate
regulatory effects (Brink et al., 2014; Jackson, 2008; Seybert, 2010;
Wang & Tan, 2013). However, very few studies have used less sophisticated participants to examine the effects of regulation on
earnings management and accounting choice on auditors or directors (see Persellin, 2013, for an exception). While we believe the
studies we review in this paper have generally made appropriate
use of sophisticated auditor and director participants, we do not
know how many interesting research questions have been abandoned because researchers assumed that they could not be
answered using less sophisticated participants. Some of this may be
due to entirely appropriate concerns that those involved in the
review process will be dismissive of “auditor” and “audit committee” studies that do not use actual auditors and directors as participants. Thus we not only encourage researchers to use less
sophisticated participants (again, where appropriate), we also
encourage journal editors and reviewers to be open-minded to
their use, and to actively consider that less sophisticated participants might be appropriate for answering certain questions related
to fundamental biases and behaviors. At the same time, if the goal is
to “peer into the minds of experts,” more costly experienced participants are necessary. Here we encourage researchers to be
entrepreneurial in their attempts to obtain appropriate participants. We also encourage editors to recognize that the costs of
additional experimentation are much higher in such studies.
4.4. Examining decision processes
As we note throughout the paper, an important part of the
38
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
comparative advantage of experiments and surveys is the ability to
disentangle the effects of correlated variables and gather intermediate process measures that allow assessment of the impact of
specific motives, beliefs, and cognitive processes of the parties
involved. There are many ways that experiments and surveys can
be designed to provide a more detailed description of decision
processes. We believe that the current literature would benefit
from a broader view of decision process analysis. We discuss this
issue under three headings: 1) the preeminence of clever experimental design, 2) process measures and mediation, and 3) nonconscious processes.
4.4.1. The preeminence of clever experimental design
Certainly, the most influential decision researchers of the last 50
years are Daniel Kahneman and Amos Tversky.5 Every decision
researcher should read and reread their work. It is important to
note that their early work was the subject of intense criticism from
many who had more orthodox views of decision processes. They
very successfully fended off and converted many of their critics by
their ability to design clever experiments that were remarkably
simple in concept: they constructed two settings which their theory suggests would result in different effects, and then randomly
assigned subjects to those settings. When alternative explanations
were raised, they would either point out how those theories could
not explain the results, or design another simple experiment whose
results could not be explained by those alternative theories. We
note that they rarely, if ever, gathered intermediate process measures (we discuss their use below).
It is important to note that simple two-celled experiments
cannot always be designed to eliminate alternative theories and
hypotheses. This has led to the use of more complex designs where
well-specified interactions provide much of the discrimination
among the alternatives. The power of interactions to eliminate
alternative explanations was recognized first in the experimental
accounting literature in studies of expertise (see Libby & Luft, 1993,
for a review) and much more recently in the archival literature in
the focus on “diff in diff” designs based on natural experiments
(e.g., Hanlon, Maydew, & Shevlin, 2008). Studies that demonstrate a
two-way interaction help to rule out alternative explanations in
that, while many alternatives can explain an observed main effect of
independent variables, it is much less likely that alternatives can
explain the observed interaction. If they can, researchers should
more carefully consider whether different manipulation choices
would help to rule out any alternatives. Three-way interactions
with personality traits where two-way interactions are stronger for
individuals who should be more prone to an effect are particularly
powerful methods for supporting process explanations (e.g.,
Seybert, 2010). The big take-away here is that any time you rely on
correlations of measures, you are giving up a major part of the
experimentalist's comparative advantage. Design (manipulation)
just about always trumps measurement. Unfortunately, we are always limited in the number of independent variables that can be
manipulated.
4.4.2. Process measures and mediation
Papers using experimental methodology increasingly rely on
“process measures” to help tell their story. These measures are
typically captured on 7-point (or 9- or 11-point) scales, and are then
used to test for mediation using the approach popularized by Baron
and Kenny (1986) or a related approach. If a process measure
“mediates” the relationship between an independent and dependent variable, then researchers often conclude that their
5
See Tversky and Kahneman (1974) and many other papers.
underlying story is supported and the process measure explains the
observed effects on the dependent variable. We believe that applications of this approach have provided important process insights (e.g., Hodge, Martin, & Pratt, 2006; Koonce & Lipe, 2010).
However, future research in accounting should consider the
strengths and weaknesses of different versions of mediation analysis to guide its application.
Some of the most important pitfalls in many accounting applications relate to the choice of the mediators themselves. Sometimes the “mediators” are simply different ways of eliciting the
same dependent variable. In such a case, the mediation analysis is
not informative. Also, researchers must always choose whether to
include the mediator scales before or after the dependent variable
scale, or the order can be manipulated. In the first two cases, specific types of carryover effects are a concern (i.e., the first scale may
contaminate the second). When mediator scales are placed before
the dependent variable scale, there is a significant risk that they
will be reactive and cause participants to respond to the dependent
variable scale in a certain way (e.g., the participants may infer that
the researcher believes the mediator variables should be weighted
heavily in the subsequent decision or choice). This is often called a
demand effect. If the dependent variable question is asked first,
subsequent responses to the mediator scales may reflect participants' attempts to answer in a fashion consistent with or to justify
their choices. When order is varied, often times the tests for order
effect interactions have little statistical power, and when they are
significant, they are often difficult to explain. Any significant
interaction effects with order can result from differences in the
carryover effects. Manipulating order is preferred when sample
sizes are very large. When they are not, we recommend that when
the main focus is on the effects on the decision or choice, then the
decision or choice scales should be administered first. When the
focus is on the judgments concerning the mediators, they should go
first. Finally, many papers ask many more debriefing questions than
they report as mediators, raising a cherry-picking problem. Such
analyses should be described as exploratory in nature.
This leaves us with an important question: What makes a great
mediator? Our general answer is that good process measures are
(1) as unobtrusive as possible (see Webb, Campbell, Schwartz, &
Sechrest, 1966), (2) distinct from the dependent variable, (3) not
simply comprehension checks, and (4) at the least are not leading.
Some of the best of these have traditionally been thought of as
“real” process measures. Examples include time spent, information
accessed, and a variety of physiological measures.6 Important examples of creative use of measures in accounting include Rasso
(2015), Griffith et al. (2015), and Elliott, Rennekamp, and White
(2015). Although not always discussed explicitly as process measures, these papers use unobtrusive measures that are unrelated to
the primary dependent variables in order to provide evidence
without leading participants to a certain response.
For example, Rasso (2015) asks participants to complete an
unrelated task developed by Fujita, Trope, Liberman, and Levin-Sagi
(2006). In the Fujita et al. (2006) task, participants are given a variety of ideas to consider (e.g., “sweeping the floor”) and asked to
choose one of two interpretations to describe it (e.g., “moving a
broom”, which represents a low-level construal focused on how the
item is done, or “being clean” which represents a high-level construal focused on why the item is done). The combined responses to
the different items provide an unobtrusive measure of whether
participants are thinking primarily at a high- or low-level of
6
Most of the physiological measures (skin conductance, respiration, heart rate,
and neural activity) are not available in accounting studies. Although see Farrell,
Goh, and White (2014) for an exception using fMRI.
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
construal. Griffith et al. (2015) ask their auditor participants to
make reasonableness assessments of a client's fair value estimate,
and then to recommend next steps to take. Auditor participants are
then asked to provide open-ended explanations for their responses.
These open-ended responses are coded for whether they mention
four errors seeded in the case, which serves as an unobtrusive
measure of the extent to which auditors employed a deliberative
mindset (i.e., identified those seeded errors). Finally, Elliott et al.
(2015) demonstrate that concrete disclosures help to reduce investors' feelings of psychological distance, which increases their
willingness to invest in a firm. As a supplemental measure of psychological distance, they ask participants to estimate the distance,
in miles, between their current location and the target firm that is
being evaluated. This unobtrusive measure is captured at the end of
the study, and allows Elliott et al. (2015) to show that more concrete
disclosures do, in fact, reduce psychological distance (reflected in
reduced estimates of geographical distance), without influencing
responses on the dependent variables related to investment
decisions.
Recent papers point out other issues that accounting researchers should consider when using mediation analysis. Zhao,
Lynch, and Chen (2010) suggest that “many research projects
have been terminated early in a research program or later in the
review process because the data did not conform to Baron and
Kenny's criteria, impeding theoretical development.” Specifically,
Zhao et al. (2010) point out that independent variables may affect
dependent variables indirectly via process measures, even in the
absence of significant direct effects. This suggests that process
measures may be one key to understanding “failed” experiments
where no significant differences are observed in dependent variables. Furthermore, this could be especially useful in research on
regulations where the initial conclusion is that a given proposal
does not have its intended effect.
For example, suppose that a given proposal is intended to
reduce earnings management by increasing auditors' independence from client management. Further suppose that researchers
test the proposal, but find no significant effect on auditors' willingness to correct a material misstatement. Without a significant
direct effect to report, the authors may choose to abandon the
project, despite the fact that their (lack of) results may not present
the complete picture of the proposal's effects. One possibility suggested by Zhao et al. (2010) is that there are multiple indirect effects
with opposite signs, and that they are canceling out any observation of a direct effect. Returning to the example, suppose the proposal does increase auditors' feelings of independence, which has a
positive indirect effect on willingness to correct the misstatement.
At the same time, the proposal may increase auditors' concerns
about building rapport and commitment, which could have a
negative indirect effect on willingness to correct. Combined, the
positive and negative indirect effects may cancel out, leading the
researchers to erroneously conclude that there is no effect of the
proposal on willingness to correct the misstatement.
In contemplating the usefulness of this approach, consider the
benefits of the study in the above example to regulators and researchers under two hypothetical outcomes. First, assume the
study collects no process measures but does, in fact, show that the
proposed regulation directly increases auditors' willingness to
correct a misstatement. Regulators now have evidence that
implementing the regulation will have a certain directional effect,
but they do not know what is driving the improvement, or how
altering the regulation might impact its efficacy. Researchers have
gained practical knowledge concerning the effects of the regulation, but the theoretical implications and conclusions to be drawn
are likely to be limited to the specific regulation involved. Second,
assume the study collects the aforementioned process measures
39
but obtains no direct effect. Regulators now know that implementing the regulation may not have the desired effect, but they
also know that it has both intended and unintended effects. This
enables them to adjust the regulation to combat any potential
negative effects. Further, knowledge of the unintended consequences of the regulation should also aid in identifying such
problems in future proposed regulations. Insights into the positive
and negative impacts of the regulation at a conceptual level should
also aid in theory development and the design of future research
studies on the effects of regulation.
4.4.3. Non-conscious processes and the comparison of between- vs.
within-subjects responses
While the above suggests that process measures can certainly be
useful (and perhaps in some cases even more useful than researchers or regulators previously realized), we do not believe that
they are necessary for telling an interesting and informative story.
This is particularly true in cases where effects are driven by nonconscious processes. By definition, participants cannot
consciously provide an explanation for their judgments in these
cases, and we would not expect typical n-point scale measures to
capture these processes. Even in cases where effects are driven by
conscious processes, participants' perceptions may not correlate
with responses that can be sufficiently captured on an n-point
scale. We encourage both researchers and reviewers of accounting
research not to become too reliant on process measures, or to place
too much emphasis on the importance of demonstrating “successful” process measures. Instead, we suggest that researchers
may need to use alternative methods to provide evidence on underlying processes. For example, researchers may use two or more
different operationalizations of a construct to provide convergent
evidence in different settings. Finding convergent evidence in
multiple settings provides more support for the authors' proposed
underlying story, and makes it less likely that there is a parsimonious alternative explanation for the observed results.
One underutilized method for determining whether an effect is
non-conscious in the first place involves the use of a combination
between- vs. within-subjects design. Kahneman and Tversky
(1996) argue that a between-subjects design allows for a clean
test of participants' natural reasoning, whereas a within-subjects
design draws attention to the manipulated independent variables
and allows participants to correct for any unintentional errors or
biases in their responses. Using a combination of a between- and
within-subjects design therefore allows for both a clean test of
participants' natural reactions to the independent variables, as well
as a test of whether those reactions appear to be intentional (Libby
et al., 2002). If participants change their original (between-subjects) responses when presented with within-subjects materials, it
suggests that their initial reactions were unintentional, or nonconscious.
Kang (2014) uses this method to show that, in response to
within-subjects questions, audit committee members do not
believe that investor sophistication would affect their propensity to
ask probing questions about management's significant accounting
estimates, despite the fact that investor sophistication does have a
significant effect on between-subjects responses. Libby and Brown
(2013) also use this method, but their study confirms that betweensubjects responses were intentional. Both the between-subjects
and within-subjects results show that presenting disaggregated
information on the face of the financial statements increases the
perceived materiality of errors in the disaggregated accounts,
although both sets of results also demonstrate that there is substantial disagreement among experienced auditors.
We encourage researchers to expand their use of the betweenvs. within-subjects comparison design. This type of design may be
40
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
particularly valuable in studies using experienced managers, auditors, and directors. Given these individuals' strong institutional
knowledge, it is especially worthwhile to understand whether their
reactions to regulations are intentional or not. Like studies using npoint scales as mediators, such analyses are subject to carryover
effects.
4.5. Replication
A fundamental tenet of the scientific method is the need for
replication of research results. Recent articles in the popular press
reiterate the need for replication, while acknowledging that it is
often neglected (Lehrer, 2010; Zimmer, 2011). This perhaps comes
as no surprise to researchers, who understand that a successful
publication outcome often depends on, at the very minimum,
finding support for at least some of their hypotheses. Research
projects that fail to document significant results are typically
modified or abandoned. This could be especially costly when it
comes to research on the effects of regulations. If a proposed
regulation has no effect on earnings management and accounting
choice, then that is important information for regulators to
consider before implementing costly requirements.
A promising development in accounting research to address this
problem is the recent call by Rick Hatfield, Editor of the journal
Behavioral Research in Accounting calling for submissions of (1)
replications of prior work and (2) studies with non-significant results (BRIA, 2014). The stated goal of this call is “to inform the
literature, which is often biased against publishing these types of
studies, and to aid researchers working in the field of behavioral
accounting.” Replication requires that researchers provide complete documentation of their materials, preferably in the appendices to a paper. We encourage researchers to make this
information available and editors to require inclusion of the information in published papers (or online supplements) in order to
facilitate replication and vetting of ideas in the accounting
literature.
5. Conclusion
This paper discusses what we have learned since the review by
Libby and Seybert (2009) about how financial reporting, auditing,
and corporate governance regulations influence the earnings
management and accounting choice decisions of managers, auditors and directors. For managers, our review focuses on how regulations change their ability as well as their incentives to engage in
earnings management, with an emphasis on (1) rules- vs.
principles-based standards, (2) standards that affect the amount
and location of additional information, and (3) standards that affect
the frequency and timing of reporting. For auditors and directors,
whose primary role is to constrain earnings management, our review distinguishes between regulations that affect their behavior
by increasing their independence (from management) and/or
accountability (to non-management).
Despite the recent uptick in experimental studies of accounting
regulation, we also recognize that there is substantial room for
future work. Specifically, we argue that future work can do more to
examine the interactions of the different parties that are affected by
regulatory changes, or the interactions between different regulations on a given party. We also argue that future work can do more
to understand both (1) the causal mechanisms underlying the effects of regulatory changes (whether they are primarily economic
or psychological in nature) and (2) whether a given effect, or
outcome, is actually desirable. For example, much of the research
treats increased audit effort as a desirable outcome, without
considering whether actual audit quality improves as a result of
that effort. Lastly, we encourage researchers to broaden their
definition of what constitutes a “regulation” when deciding which
research questions to tackle. While changes in formal regulation
may be an obvious choice for research, changes in the interpretation or enforcement of existing regulations may be just as
important.
To help guide future research on the topics that we propose, we
also discuss methodological considerations that are likely to impact
the effectiveness and efficiency of experimental research. We
suggest that researchers who study the effects of regulation should
be especially cognizant of issues surrounding realism of stimuli
(since proposed standards can change often), the selection of accounting setting (since the setting needs to represent the broad
constructs of interest), and the selection of participants (since
experienced participants are particularly difficult to obtain). We
also make recommendations about different possibilities for
examining underlying processes that drive participants' reactions.
While process is important to research on regulation, our discussion of process evidence relates more broadly to all experimental
research in accounting.
Our discussion reflects our belief that experiments and surveys
are uniquely suited to investigating the effects of both existing and
proposed regulation. For existing regulation, experiments can
isolate specific components of a regulation to disentangle the underlying factors that drive accounting choice. Furthermore, experiments can examine how regulations differentially effect individual
parties (e.g., managers, auditors and directors) as well as how those
parties react to the expected or observed behavior of other parties.
For proposed regulation, experiments and surveys have the added
benefit of being able to investigate settings that do not yet exist,
allowing results to inform regulators before potentially costly and/
or detrimental requirements are put in place.
References
Agoglia, C. P., Doupnik, T. S., & Tsakumis, G. T. (2011). Principles-based versus rulesbased accounting standards: the influence of standard precision and audit
committee strength on financial reporting decisions. The Accounting Review,
86(3), 747e767.
Asay, H. S. (2014). Horizon-induced optimism as a gateway to earnings management.
Working paper. University of Iowa.
Asay, H. S., Brown, T., Nelson, M. N., & Wilks, T. J. (2013). Contrast effects under
incomplete convergence of accounting standards. Working paper. University of
Iowa, University of Illinois, Cornell University, and Brigham Young University.
Baber, W., Fairfield, P., & Haggard, J. (1991). The effect of concern about reported
income on discretionary spending decisions. The Accounting Review, 66,
818e829.
Backof, A., Bamber, L., & Carpenter, T. (2014). Do auditor judgment frameworks help in
constraining aggressive reporting? Evidence under more precise and less precise
accounting standards. Working paper. University of Virginia and University of
Georgia.
Baker, M., & Wurgler, J. (2013). Behavioral corporate finance: an updated survey. In
G. M. Constantinides, M. Harris, & R. M. Stulz (Eds.), Handbook of the economics
of finance (Vol. 2, Part A, pp. 357e424). Elsevier Press.
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in
social-psychological research: conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173e1182.
Bauer, T. (2014). The effects of client identity strength and professional identity
salience on auditor judgments. The Accounting Review, 90(1), 95e114.
Beasley, M. S., Carcello, J. V., Hermanson, D. R., & Neal, T. L. (2009). The audit
committee oversight process. Contemporary Accounting Research, 26(1), 65e122.
Behavioral Research in Accounting (BRIA). (2014). A call for research notes: Replications and non-significant results studies. Available at http://aaahq.org/abo/bria/
Bria%20Call%20for%20replications.pdf.
Beyer, A., Cohen, D. A., Lys, T. Z., & Walther, B. R. (2010). The financial reporting
environment: review of the recent literature. Journal of Accounting and Economics, 50(2e3), 296e343.
Bonner, S. E. (2008). Judgment and decision making in accounting. Prentice Hall.
Brasel, K., Doxey, M., Grenier, J. H., & Reffett, A. (2014). Risk disclosure preceding
negative outcomes: The effects of reporting critical audit matters on judgments of
auditor liability. Working paper. Ball State University, University of Alabama e
Culverhouse, Miami University, Miami University of Ohio.
Brink, A. G., Gouldman, A., & Rose, J. (2014). Effects of managerial knowledge of executive compensation structure and R&D reporting method on real earnings
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
management. Working paper. Virginia Commonwealth University.
Brown, T. (2014). Advantageous comparison and rationalization of earnings management. Journal of Accounting Research, 52(4), 849e876.
Brown, T., Majors, T. M., & Peecher, M. E. (2015). The impact of a judgment rule and
critical audit matters on assessments of auditor legal liability e The moderating
role of legal knowledge. Working paper. University of Illinois at UrbanaChampaign.
Carcello, J. V., & Li, C. (2013). Costs and benefits of requiring an engagement partner
signature: recent experience in the United Kingdom. The Accounting Review,
88(5), 1511e1546.
Clor-Proell, S. M., & Maines, L. A. (2014). The impact of recognition versus disclosure
on financial information: a preparer's perspective. Journal of Accounting
Research, 52(3), 671e701.
Cohen, D. A., Dey, A., & Lys, T. Z. (2008). Real and accrual-based earnings management in the pre- and post-Sarbanes Oxley periods. The Accounting Review,
82(3), 757e787.
Cohen, J., Gaynor, L., Krishnamoorthy, G., & Wright, A. (2011). The impact on auditor
judgments of CEO influence on audit committee independence. Auditing: A
Journal of Practice & Theory, 30(4), 129e147.
Cohen, J., Hayes, C., Krishnamoorthy, G., Monroe, S., & Wright, A. (2013). The
effectiveness of SOX regulation: an interview study of corporate directors.
Behavioral Research in Accounting, 25(1), 61e87.
Cohen, J., Krishnamoorthy, G., Peytcheva, M., & Wright, A. (2013). How does the
strength of the financial regulatory regime influence auditors' judgments to
constrain aggressive reporting in a principles-based versus rules-based accounting environment? Accounting Horizons, 27(3), 579e601.
Cohen, J., Krishnamoorthy, G., & Wright, A. (2010). Corporate governance in the
post-Sarbanes-Oxley era: auditors' experiences. Contemporary Accounting
Research, 27(3), 751e786.
Daugherty, B., Dickins, D., Hatfield, R., & Higgs, J. (2012). An examination of partner
perceptions of partner rotation: direct and indirect consequences to audit
quality. Auditing: A Journal of Practice & Theory, 31(1), 97e114.
Dechow, P., Ge, W., & Schrand, C. (2010). Understanding earnings quality: a review
of the proxies, their determinants and their consequences. Journal of Accounting
and Economics, 50(2e3), 344e401.
DeZoort, F., & Salterio, S. (2001). The effects of corporate governance experience and
financial-reporting and audit knowledge on audit committee members' judgments. Auditing: A Journal of Practice & Theory, 20(2), 31e47.
Dichev, I., Graham, J., Harvey, C., & Rajgopal, S. (2013). Earnings quality: evidence
from the field. Journal of Accounting and Economics, 56(1), 1e33.
Elliott, W. B., Hodge, F., Kennedy, S. J., & Pronk, M. (2007). Are MBA students a good
proxy for non-professional investors? The Accounting Review, 82(1), 139e168.
Elliott, W. B., Rennekamp, K. M., & White, B. J. (2015). Does concrete language in
disclosures increase willingness to invest? Review of Accounting Studies, 20(2),
839e865.
Evans, M., Houston, R., Peters, M., & Pratt, J. (2015). Reporting regulatory environments and earnings management: U.S. and non-U.S. firms using U.S. GAAP or
IFRS. The Accounting Review, 90(5), 1969e1994.
Fama, E. F. (1980). Agency problems and the theory of the firm. Journal of Political
Economy, 88(2), 288e307.
Farrell, A. M., Goh, J. O., & White, B. J. (2014). The effect of performance-based
incentive contracts on system 1 and system 2 processing in affective decision
contexts: fMRI and behavioral evidence. The Accounting Review, 89(6),
1979e2010.
Farrell, A. M., Grenier, J. H., & Leiby, J. (2014). Scoundrels or stars? Theory and evidence
on the quality of workers in online labor markets. Working paper. Miami University of Ohio, Miami University, and University of Florida.
Fields, T., Lys, T., & Vincent, L. (2001). Empirical research on accounting choice.
Journal of Accounting and Economics, 31(1), 255e307.
Francis, J. (2001). Discussion of empirical research on accounting choice. Journal of
Accounting and Economics, 31(1), 309e319.
Fujita, K., Trope, Y., Liberman, N., & Levin-Sagi, M. (2006). Construal levels and selfcontrol. Journal of Personality and Social Psychology, 90(3), 351e367.
Gibbins, M., McCracken, S., & Salterio, S. (2007). The chief financial officer's
perspective on auditor-client negotiations. Contemporary Accounting Research,
24(2), 387e422.
Gimbar, C., Hansen, B., & Ozlanski, M. E. (2014). The effects of critical audit matter
paragraphs and accounting standard precision on auditor liability: Can anything
the auditors say be used against them?. Working paper. Virginia Polytechnic
Institute & State University, Virginia Tech University, and University of New
Hampshire.
Graham, J. R., Hanlon, M., & Shevlin, T. (2011). Real effects of accounting rules:
evidence from multinational firms' investment location and profit repatriation
decisions. Journal of Accounting Research, 49(1), 137e185.
Graham, J. R., Harvey, C. R., & Rajgopal, S. (2005). The economic implications of
corporate financial reporting. Journal of Accounting and Economics, 40(1e3),
3e73.
Griffin, J. (2014). The effects of uncertainty and disclosure on auditors' fair value
materiality decisions. Journal of Accounting Research, 52(5), 1165e1193.
Griffith, E. E., Hammersley, J. S., Kadous, K., & Young, D. (2015). Auditor mindsets
and audits of complex estimates. Journal of Accounting Research, 53(1), 49e77.
Guggenmos, R. D. (2015). The effect of creative culture on aggressive financial
reporting. Working paper. Cornell University.
Hackenbrack, K., & Nelson, M. N. (1996). Auditors' incentives and their application
of financial accounting standards. The Accounting Review, 71(1), 43e59.
41
Hanlon, M., Maydew, E., & Shevlin, T. (2008). An unintended consequence of booktax conformity: a loss of earnings informativeness. Journal of Accounting and
Economics, 46(2e3), 294e311.
Hermalin, B. E., & Weisbach, M. S. (1998). Endogenously chosen boards of directors
and their monitoring of the CEO. The American Economic Review, 88, 96e118.
Hodge, F. D., Martin, R. D., & Pratt, J. H. (2006). Audit qualifications of incomedecreasing accounting choices. Contemporary Accounting Research, 23(2),
369e394.
International Accounting and Assurance Standards Board (IAASB). (2013). Exposure
draft: Reporting on audited financial statements: Proposed new and revised international standards on auditing (ISAs). Available at http://www.ifac.org/sites/
default/files/publications/files/All%20Proposed%20ISAs%20Included%20in%
20ED.pdf.
Jackson, S. B. (2008). The effect of firms' depreciation method choice on managers'
capital investment decisions. The Accounting Review, 83(2), 351e376.
Jackson, S. B., Keune, T. M., & Salzsieder, L. (2013). Debt, equity, and capital investment. Journal of Accounting and Economics, 56(2e3), 291e310.
Jamal, K., & Tan, H.-T. (2010). Joint effects of principles-based versus rules-based
standards and auditor type in constraining financial managers' aggressive
reporting. The Accounting Review, 85(4), 1325e1346.
Kachelmeier, S. J., & King, R. R. (2002). Using laboratory experiments to evaluate
accounting policy issues. Accounting Horizons, 16(3), 219e232.
Kachelmeier, S. J., Schmidt, J. J., & Valentine, K. (2014). The disclaimer effect of
disclosing critical audit matters in the auditor's report. Working paper. University
of Texas at Austin.
Kadous, K., Kennedy, J. S., & Peecher, M. E. (2003). The effect of quality assessment
and directional goal commitment on auditors' acceptance of client-preferred
accounting methods. The Accounting Review, 78(3), 759e778.
Kadous, K., & Mercer, M. (2012). Can reporting norms create a safe harbor? Jury
verdicts against auditors under precise and imprecise accounting standards. The
Accounting Review, 87(2), 565e587.
Kahneman, D., & Tversky, A. (1996). On the reality of cognitive illusions: a reply to
Gigerenzer's critique. Psychological Review, 103, 582e591.
Kang, Y. J. (2014). Are audit committees more challenging given a sophisticated investor
base? Does the answer change given anticipation of additional mandatory audit
report disclosure?. Working paper. University of Massachusetts at Amherst.
Kang, Y. J., Trotman, A., & Trotman, K. (2014). The effect of an audit judgment rule on
audit committees' questioning of accounting estimates. Working paper. University
of Massachusetts at Amherst, Northeastern University, and UNSW Australia.
Koonce, L., & Lipe, M. G. (2010). Earnings trend and performance relative to
benchmarks: how consistency influences their joint use. Journal of Accounting
Research, 48(4), 859e884.
Koonce, L., & Mercer, M. (2005). Using psychology theories in archival financial
accounting research. Journal of Accounting Literature, 24, 175e212.
Koonce, L., Seybert, N., & Smith, J. (2011). Causal reasoning in financial reporting and
voluntary disclosure. Accounting, Organizations and Society, 36(4e5), 209e225.
Krische, S. (2014). Who is the average individual investor? Numerical skills and implications for accounting research. Working paper. American University.
Lehrer, J. (2010, December 13). The truth wears off: Is there something wrong with the
scientific method? The New Yorker. Retrieved from http://www.newyorker.com.
Libby, R., Bloomfield, R., & Nelson, M. N. (2002). Experimental research in financial
accounting. Accounting, Organizations and Society, 27(8), 775e810.
Libby, R., & Brown, T. (2013). Financial statement disaggregation decisions and
auditors' tolerance for misstatement. The Accounting Review, 88(2), 641e665.
Libby, R., & Emett, S. A. (2014). Earnings presentation effects on manager reporting
choices and investor decisions. Accounting and Business Research, 44(4),
410e438.
Libby, R., & Kinney, W. (2000). Does mandated audit communication reduce
opportunistic corrections to manage earnings forecasts? The Accounting Review,
75(4), 383e404.
Libby, R., & Luft, J. (1993). Determinants of judgment performance in accounting
settings: ability, knowledge, motivation, and environment. Accounting, Organizations and Society, 18(5), 425e450.
Libby, R., & Seybert, N. (2009). Behavioral studies of the effects of regulation on
earnings management and accounting choice. In C. S. Chapman, D. J. Cooper, &
P. Miller (Eds.), Accounting, organizations, and institutions: Essays in honour of
Anthony Hopwood. Oxford, U.K.: Oxford University Press.
Magilke, M. J., Mayhew, B. W., & Pike, J. E. (2009). Are independent audit committee
members objective? Experimental evidence. The Accounting Review, 84(6),
1959e1981.
Majors, T. M. (2014). Communicating measurement uncertainty: An experimental
study of financial reporting implications for managers and investors. Working
paper. University of Illinois at Urbana-Champaign.
McEnroe, J. E. (2007). Perceptions of the effect of Sarbanes-Oxley on earnings
management practices. Research in Accounting Regulation, 19, 137e157.
McEnroe, J. E., & Sullivan, M. (2013). An examination of the perceptions of auditors
and chief financial officers of various regulations introduced by the Dodd-Frank
financial reform bill. Advances in Public Interest Accounting, 16, 145e185.
Messier, W., Jr., Quick, L., & Vandervelde, S. (2014). The influence of process
accountability and accounting standard type on auditor usage of a status quo
heuristic. Accounting, Organizations and Society, 39(1), 59e74.
Nelson, M. N. (2003). Behavioral evidence on the effects of principles- and rulesbased standards. Accounting Horizons, 17(1), 91e104.
Nelson, M. N., Elliott, J., & Tarpley, R. (2002). Evidence from auditors about managers' and auditors' earnings management decisions. The Accounting Review,
42
R. Libby et al. / Accounting, Organizations and Society 47 (2015) 25e42
77(Suppl.), 175e202.
Nelson, M. N., & Skinner, D. (2013). How should we think about earnings quality? A
discussion of “Earnings quality: evidence from the field”. Journal of Accounting
and Economics, 56(2e3), 34e41.
Nelson, M. N., & Tan, H.-T. (2005). Judgment and decision-making research in
auditing: a judgment, person and interpersonal interaction perspective. Auditing: A Journal of Practice & Theory, 24(1), 41e71.
Ng, T. B.-P., & Tan, H.-T. (2003). Effects of authoritative guidance availability and
audit committee effectiveness on auditors' judgments in an auditor-client
negotiation context. The Accounting Review, 78(3), 801e818.
Pacelli, J. (2015). Corporate culture and analysts forecast quality. Working paper.
Indiana University.
Peecher, M. E., Solomon, I., & Trotman, K. (2013). An accountability framework for
financial statement auditors and related research questions. Accounting, Organizations and Society, 38(8), 596e620.
Persellin, J. S. (2013). The influence of PCAOB inspections on audit committee
members' judgments. Behavioral Research in Accounting, 25(2), 97e114.
Peytcheva, M., Wright, A., & Majoor, B. (2014). The impact of principles-based versus
rules-based accounting standards on auditors' motivations and evidence demands. Behavioral Research in Accounting, 26(2), 51e72.
Piercey, D. (2011). Documentation requirements and quantified versus qualitative
audit risk assessments. Auditing: A Journal of Practice & Theory, 30(4), 223e248.
Public Company Accounting Oversight Board (PCAOB). (2013). PCAOB release no.
2013-005. Washington, DC. Available at http://pcaobus.org/Rules/Rulemaking/
Docket034/Release_2013-005_ARM.pdf.
PwC. (2014). 2014 AICPA national conference on current SEC and PCAOB developments.
In Depth, US2014-09 http://www.pwc.com/en_US/us/cfodirect/assets/pdf/indepth/us2014-09-aicpa-conference.pdf.
Quick, L. (2013). When less is more: The effect of reducing auditor liability on auditor
judgments in principles-based and rules-based environments (Unpublished
dissertation). The University of South Carolina.
Rasso, J. (2015). Construal instructions and professional skepticism in evaluating
complex estimates. Accounting, Organizations and Society (in press).
Rennekamp, K. M. (2012). Processing fluency and investors' reactions to disclosure
readability. Journal of Accounting Research, 50(5), 1319e1354.
Rennekamp, K., Rupar, K., & Seybert, N. (2015). Impaired judgment: the effects of
asset impairment reversibility and cognitive dissonance on future investment.
The Accounting Review, 90(2), 739e759.
Rose, J., Mazza, C., Norman, C., & Rose, A. (2013). The influence of director stock
ownership and board discussion transparency on financial reporting quality.
Accounting, Organizations and Society, 38(5), 397e405.
Rose, J., Rose, A., Norman, C., & Mazza, C. (2014). Will disclosure of friendship ties
between directors and CEOs yield perverse effects? The Accounting Review,
89(4), 1545e1563.
Roychowdhury, S. (2006). Earnings management through real activities manipulation. Journal of Accounting and Economics, 42(3), 335e370.
Schipper, K. (2003). Principles-based accounting standards. Accounting Horizons,
17(1), 61e72.
Securities and Exchange Commission (SEC). (2012). Work plan for the consideration
of incorporating international financial reporting standards into the financial
reporting system for U.S. issuers. Available at http://www.sec.gov/spotlight/
globalaccountingstandards/ifrs-work-plan-final-report.pdf.
Segovia, J., Arnold, V., & Sutton, S. (2009). Do principles- vs. rules-based standards
have a differential impact on U.S. auditors' decisions? Advances in Behavioral
Accounting Research, 12(1), 61e84.
Seybert, N. (2010). R&D capitalization and reputation-driven real earnings management. The Accounting Review, 85(2), 671e693.
Tarullo, D. (2014, October 20). Good Compliance, Not Mere Compliance. Lecture presented at Reforming Culture and Behavior in the Financial Services Industry Conference. New York.
Trotman, K., Bauer, T., & Humphreys, K. (2015). Group judgment and decision
making in auditing: past and future research. Accounting, Organizations and
Society (forthcoming).
Trottier, K. (2013). The effect of reversibility on a manager's decision to record asset
impairments. Accounting Perspectives, 12(1), 1e22.
Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and
biases. Science, 185(4157), 1124e1131.
Ugrin, J. C., & Odom, M. D. (2010). Exploring Sarbanes-Oxley's effect on attitudes,
perceptions of norms, and intentions to commit financial statement fraud from
a general deterrence perspective. Journal of Accounting and Public Policy, 29,
439e458.
Wang, E., & Tan, H.-T. (2013). The effects of guidance frequency and guidance goal
on managerial decisions. Journal of Accounting Research, 51(3), 673e700.
Watts, R., & Zimmerman, J. (1978). Towards a positive theory of the determination
of accounting standards. The Accounting Review, 53(1), 112e134.
Webb, E. J., Campbell, D. T., Schwartz, R. D., & Sechrest, L. (1966). Unobtrusive
measures: Nonreactive research in the social sciences. Chicago: Rand McNally.
Westermann, K., Cohen, J., & Trompeter, G. (2014). The influence of accountability on
professional skepticism. Working paper. Florida International University, Boston
College, and the University of Central Florida.
Winn, A. (2014). Partner rotation and PCAOB inspections: Effects on end-of-term audit
quality. Working paper. The University of Illinois at Urbana-Champaign.
Zhao, X., Lynch, J. G., Jr., & Chen, Q. (2010). Reconsidering Baron and Kenny: myths
and truths about mediation analysis. Journal of Consumer Research, 37(August),
197e206.
Zimmer, C. (2011, June 25). It's science, but not necessarily right. The New York Times.
Retrieved from http://www.nytimes.com.