You are on page 1of 52

If You Want My Advice:

Status Motives and Audit Consultations about Accounting Estimates

W. Robert Knechel

University of Florida

Justin Leiby*

University of Georgia

June 2016

*Corresponding Author
255 Brooks Hall
310 Herty Drive
Athens, GA 30602
706.369.4012
jleiby@uga.edu

We are thankful for suggestions and comments from Denny Beresford, Joe Brazel, Tim Bauer,
Tim Brown, Frank Buckless, Jackie Hammersley, Robin Pennington, Linda Quick, Bridget
Stomberg, Eileen Taylor, and workshop participants at North Carolina State University, the
University of Georgia, and the 2015 International Symposium on Auditing Research (ISAR).
Justin Leiby is grateful for support from a University of Florida Fisher School of Accounting
Summer Research Grant.
If You Want My Advice:
Status Motives and Audit Consultations about Accounting Estimates

Abstract

Effective consultation is critical for improving the audit of estimates. In an experiment where audit
managers acted as consultants to other auditors, we examine conditions in which consultants either
recommend estimates that differ substantially from advice-seekers assessments (contrariness) or
recommend narrower reasonable ranges of estimates (precision). Psychology theory argues that
both of these attributes can improve estimates. We examine whether these attributes depend on
consultants' status motives, i.e., the desire to gain respect from or power over others. We find that
active status motives lead consultants with higher specialized knowledge to provide
recommendations that are less contrary, but more precise. However, consultants increase precision
by tightening range bounds in a manner not counter to managements preference and thus unlikely
to prompt the audit team to challenge the estimate. Further, we find that higher consultant decision
authority also constrains precision. Our findings suggest limits to consultations potential
effectiveness in improving estimates. For instance, our findings suggest that firms and standard
setters emphasize consultant scrutiny on the range bound that is most likely to constrain
management opportunism.

1
1. Introduction

Financial reports contain many complex accounting estimates that require significant

auditor judgment and can increase the risk of material misstatements. Auditors often struggle to

maintain the requisite knowledge and questioning mindset necessary to effectively assess these

estimates (PCAOB [2015a; 2015b, 2015c; 2015d]; Griffith et al. [2015]). Due to these difficulties,

auditors routinely engage in consultation with other professionals to improve their own judgments

in these areas (Kennedy et al. [1997]; Salterio and Denham [1997]; AICPA [2002]; IAASB [2005];

Barrows et al. [2010]; Ranzilla et al. [2011] AICPA [2012]). Consultation refers to an auditor

receiving advice from another knowledgeable individual (consultant). Such advice can range

from formal input from a designated expert to informal input from a co-worker or colleague.1

Auditors often rely heavily on input from consultations when evaluating estimates (Griffith et al.

[2015]), and psychology research suggests that this input can be particularly helpful in improving

uncertain estimates (Yaniv [2004]). Yet, little is known about how the efficacy of consultation and

the conditions under which consultants provide advice might improve the audit of estimates.

Consultation in auditing is pervasive and has substantial potential benefits, thus the availability of

useful advice is often a necessary condition to improve the audit of estimates.

In this study, we examine two properties of advice that are likely to help improve auditor

judgment on accounting estimates: contrariness and precision. Contrariness refers to the degree

to which a consultants advice differs from the advice-seekers own opinion. Many of the

presumed benefits of consultation lie in the potential to expose auditors to new information and

alternative opinions (Barrows et al. [2010]; Ranzilla et al. [2011]), and there is ample evidence

1
The consultants advice could be opinions or suggestions for or against an option, about the best process to address
an issue, or additional information about an issue. We use advice-seeker and auditor interchangeably to refer to
the intended recipient of advice.

2
that exposure to contrary opinions improves novice and expert judgment (Tetlock [2005]; Larrick

and Soll [2006]). Precision refers to the narrowness of the range of options presented by a

consultant to an advice seeker, i.e., reducing the range of possible outcomes to be considered.

Because auditing standards explicitly require adjustments only to estimates that fall outside a

reasonable range (AICPA [2006]), recommending narrower ranges for estimates or their inputs

can help constrain managements leeway to use estimates opportunistically.

In spite of these potential benefits, the inherent frictions of an audit setting can make

positions of contrariness and precision costly to auditors or an audit firm. Audit engagements face

intense budget pressure as well as scrutiny by both internal quality reviews and external PCAOB

inspections. These conditions may pressure an audit team to prematurely reach an artificial level

of consensus on contentious issues and to focus on confirming evidence (e.g., AICPA [2012]) so

that the audit appears to be efficient and effective. By its nature, contrary advice can inhibit and

delay generating consensus. In addition, audit clients may prefer imprecision that allows leeway

for estimates that reflect managements incentives, rather than underlying fundamentals (Bamber

et al. [2010]; Dechow et al. [2010]), and can use their superior knowledge about their business to

justify this leeway. Thus, increasingly precise recommendations require that consultants possess

the necessary expertise to justify constraining clients discretion over estimates. Such conditions

may work against consultants providing contrary or precise recommendations.

We argue in this paper that contrariness and precision may increase when consultants are

motivated to signal their own self-confidence, as higher contrariness and precision are both

interpreted as indicators of confidence (Kanodia et al. [1989]; Levy [2004]; Moore and Healy

[2008]). Based on evolutionary psychology, contrariness and precision depend on a consultants

status motives, i.e., the desire to gain respect from, or power over, others (Griskevicius et al.

3
[2009a; 2009b]; Kenrick et al. [2009]). Status motives are one of a group of specialized decision

rules that evolved to help early humans cope with survival dilemmas they may have confronted

(Cosmides and Tooby [1994]; Kenrick et al. [2003]).2 Status motives can activate in the presence

of cues of dominance, prestige, or competition, including accomplishments or rivalries, and are

inactive in the absence of such cues. For example, status motives could activate in the presence of

a competitor, a highly successful mentor or superior, luxurious surroundings or products, or when

explicitly considering career progression or professional accomplishments.

In an audit consultation, the presence or absence of cues to activate status motives is likely

to vary naturally. For instance, consultations necessarily involve high-profile decisions and clients,

and interactions with high-level personnel such as experts and successful partners. When status

motives are activated, people are likely to derive significant utility from actions that cultivate

prestige or respect in others eyes. One such action is signaling self-confidence (Anderson et al.

[2006]; Radzevik and Moore [2011]. As a result, status motives may stimulate incrementally more

contrary and/or more precise recommendations in a consultation setting.

Further, we examine whether the effect of status motives also depends on two key factors

in an auditing context. The first is the tendency of auditors to specialize by subject matter or

industry, which leads auditors to process information differently as their specialized knowledge

increases (Solomon et al. [1999]; Taylor [2000]; Hammersley [2006]). The second is the

consultants decision authority, i.e., the degree to which the advice-seeker is expected to rely on

2
Kenrick et al. [2003; 2009] identify a small number of these specialized sets of decision rules, i.e., evolutionary
motives, specifically, status, self-protection, coalition formation, mate acquisition, mate retention, and kin care, all of
which serve to solve particular survival problems that confronted people in hunter-gatherer times. Status motives
address the dilemma of competing with others for scarce resources. In this framework, preferences and resource
allocations are utility-maximizing functions that depend on which evolved set of decision rules is active. For example,
when the status system is active, people derive utility from actions that increase prestige or dominance, while people
derive utility from actions that increase safety when the self-protection system is active. Note that we are not interested
in how possessing status affects consultant recommendations, but rather in the motivation to gain or maintain status.

4
the advice (e.g., Gold et al. [2012]). These frictions of an audit setting may influence the

willingness of consultants with active status motives to be contrary or precise.

We conduct an experiment with 102 audit managers and senior managers from a U.S. Big

Four accounting firm. Following theory that status motives are context-specific adaptations

activated in specific situations, we manipulate status motives by priming auditors with a brief story

prior to the task (Griskevicius et al. [2009b]). Participants then read a case in which they act as a

consultant to another auditor regarding the discount rate a client uses to estimate the fair value of

a securitized asset. The auditor is inclined to allow an aggressive discount rate that is lower than

that of industry leaders and increases reported income. We manipulate decision authority as high

or low and measure participants securitization knowledge as our proxy of specialized knowledge.

Auditors then provide their estimate of the most appropriate discount rate for calculating the

underlying accounting estimate and assess the range of reasonable discount rates that the client

could use. Our measure of contrariness is the signed difference between the consultants discount

rate estimate and the advice-seekers best estimate. Our measure of precision is based on the

absolute size of each consultants assessed range of reasonable discount rates.

We find that a consultants recommendations are influenced by specialized knowledge and

decision authority conditional on their status motives. More specifically, we find that when status

motives are active, consultants with higher specialized knowledge are more precise but less

contrary than those with lower knowledge. Consultants willingness to challenge an auditors

relatively pro-client judgment decreases as the consultants expertise increases. By contrast,

recommending narrower reasonable ranges could help constrain management opportunism

depending on which end point is affected by the consultation. However, we find that consultants

5
recommend changes to the end of the range that is further from managements preference, thus

this increased precision is unlikely to influence the evaluation and ultimate quality of the estimate.

These insights are relevant to ongoing debates over standards on estimates and the use of

specialists. Given that experienced, higher knowledge consultants (a reasonable proxy for

valuation specialists) are more prone to increase the precision of a range rather than the

contrariness of a point estimate, firm policies or standards could explicitly advocate precision as a

desired property of recommendations from consultants and specialists. This could include

guidance for consultants to focus scrutiny on the end of the range that is nearest to managements

likely preference, as tightening this end of a range is most likely to influence the audit teams

evaluation in a manner that affects the ultimate, audited estimate.

Moreover, we find that precision decreases with higher decision authority and increases

only when decision authority is lower and status motives are active. This is likely opposite of what

is desired from consultations. Our finding that higher decision authority can have negative audit

quality implications is relevant to audit firm policies, which often vest substantial authority in

consultants, and to the ongoing debate over standards for the use of specialists. In tandem with

findings that auditors rely heavily on specialists who often have little formal authority in practice

(Griffith 2015; Griffith et al. 2015), our findings suggest it may be beneficial to advocate lower

decision authority. Unlike knowledge-driven increases in precision, we do not find that increased

precision under low decision authority involves disproportionately tightening one end of the range

more than another, which suggests heretofore unacknowledged benefits of low decision authority.

We discuss further implications in the conclusions section.

Finally, if the goal of consultation is to promote self-critical, more effective professional

judgment (Bell et al. [2005]), then audit quality may also benefit from firms and regulators

6
acknowledging the value of consultation with non-specialist but experienced peers. This could

include explicitly encouraging more informal consultations with peers. We find that this group is

most willing to challenge acquiescent assessments of management estimates, and there is evidence

that experienced professionals exhibit similar judgment quality outside, as opposed to inside their

specialties under many circumstances (Tetlock [2005]). Standards on brainstorming related to

fraud recognize the benefits of such peer-to-peer consultations (AICPA [2002]). It may be valuable

to extend this approach to other areas of the audit.

In sum, our study identifies conditions affecting the degree to which consultant

recommendations are likely to help remedy deficiencies in the audit of estimates. This is an

important step, as prior literature on advice focuses primarily on the advice-seeker and treats the

properties of consultant recommendations as exogenous or fixed (Yaniv [2004]; Ng and Shankar

[2010]; Kadous et al. [2013]). Our findings can inform audit firm policies that require consultation

with knowledgeable persons, as well as standard-setters and regulators whose responsibility is to

provide guidance on using consultation to conduct more effective audits of financial statement

estimates. Auditing standards that recommend consultation but provide few specifics and limited

guidance may presume too much about its overall effectiveness.

Finally, this is the first study of which we are aware to show that adaptive needs for status

can shape how professionals respond to features of their environment. Consultation fails frequently

in professional settings (Lee [1997]; Missal [2008]), sometimes with catastrophic consequences

(Vaughan [1997]), and our study is a first step towards understanding how the adaptive need for

prestige or respect can affect a consultants tendency to provide useful advice that promotes a

self-critical, point-counterpoint style of thinking that benefits judgment (Tetlock [2005, 21]).

Thus, our contributions potentially extend to other professions beyond auditing and accounting. In

7
general settings like social and consumer psychology, active status motives lead people to show

off by challenging others and acting distinctively (Griskevicius [2009b]). By contrast, in a

professional auditing setting that emphasizes consensus, we find limited evidence of this tendency

as high knowledge consultants avoid challenging others and instead show off by increasing

precision. Consultants in other professions that encourage consultation and prize consensus, such

as engineering or medicine, may similarly satisfy status motives through means other than

challenging others.

In the remainder of the paper, Section 2 develops our theory and hypotheses, Section 3

describes our experiment, Section 4 reports results, and Section 5 discusses conclusions and future

directions.

2. Theory and Hypothesis Development

2.1 CONSULTATION IN AUDITING

Consultation is integral to an audit firms quality control system (Salterio and Denham

[1997]). Practitioners, scholars, and regulators note the potential for consultation to improve

auditor judgment by exposing auditors to new opinions and information (Barrows et al. [2010];

Ranzilla et al. [2011]; PCAOB [2012]; Glover and Prawitt [2014]).3 Accounting estimates are an

area in which consultation has the potential to be particularly useful because auditors often struggle

to maintain an independent perspective and sufficient knowledge when evaluating accounting

estimates (Griffith et al. [2015]). The usefulness of consultation in auditing estimates (Glover and

Prawitt [2014]) is supported by evidence in psychology that advice is effective in improving

judgments, especially those that involve estimates (e.g., Yaniv [2004]; Vul and Pashler [2008];

3
There is no standard definition of consultation offered by regulators or practitioners. Our definition of consultation
is relatively broad and encompasses a range of consultative interactions that vary in their degree of formality. In the
conclusions section, we discuss the implications of our definition on our theorys generalizability.

8
Herzog and Hertwig [2009]; also see Bonaccio and Dalal [2006] for a review). As a result,

understanding the conditions that affect a consultants provision of useful advice provides insight

into a necessary condition for improving the audit of estimates. We argue that two attributes of

advice could be particularly important in this context: (1) the direction of the advice (confirming

vs contrary) and (2) the specificity (precision) of the advice.

In general, advice from a consultant can either be contrary to a decision-makers attitudes

or it can confirm those attitudes. In many domains, such as medicine and organizational behavior,

advice is considered more decision relevant when it is not contrary (Kray [2000]; Jonas and Frey

[2003]; Zikmund-Fischer et al. [2006]). However, because auditing emphasizes a questioning

(skeptical) mindset (PCAOB AU 230 [2006]; Nelson [2009]; Barrows et al. [2010]; KPMG

[2011]), contrary advice may be more decision relevant to an auditor. Even when evidence

suggests that a clients assertions are reasonable, which is true in many audit contexts, professional

skepticism requires the auditor to challenge the reasonableness of the assertions.4

Tetlock [2005] alludes to the value of contrary opinions in an extensive analysis of expert

judgments by concluding that the greatest deficit in expert judgment is the vice of closed-

mindedness, of dismissing dissonant possibilities too quickly. In general psychology research,

increasing the contrast between a consultants and an advice-seekers best estimates increases the

chance that the two will bracket the best answer. In such conditions, estimates will improve as

long as the advice-seeker moves their judgment closer to that of the consultant (Larrick and Soll

[2006]; Vul and Pashler [2008]; Herzog and Hertwig [2009]). We refine this logic for an auditing

4
For this project we also conducted multiple informal interviews with senior auditing practitioners with national
office experience. Among other things, these practitioners note that it is important for consultants to be contrary
even towards conclusions that appear to be well-justified and correct, as the costs of not challenging a conclusion
that turns out to be wrong far exceed the costs of challenging a conclusion that turns out to be right.

9
setting, in which estimates are often biased in a direction consistent with managements incentives.

In auditing, contrariness is helpful when it moves the estimate towards the most reasonable value.

At the same time, because estimates and many other consultation situations in auditing

involve subjectivity, increasingly precise advice can reduce the uncertainty surrounding a specific

accounting judgment. By eliminating less justifiable values, a consultant can help the decision

maker to narrow the range of reasonable values under consideration. Narrower ranges generally

reduce the latitude for manager opportunism (Bamber et al. [2010]; Griffin [2014]). This suggests

that the precision of a recommendation is also an important attribute in audit consultations.

In brief, both consultant precision and contrariness are consistent with adopting a

questioning mindset towards management assertions and increase the potential for consultation to

improve auditor judgment. Our study examines conditions under which consultants advice will

be relatively more contrary or relatively more precise.

2.2 CONSULTANTS MOTIVATIONS

To understand consultation, it is critical to examine the motivations of consultants. In some

cases, consultants are motivated to reach the best possible conclusion, but in other cases they may

be motivated to either agree or disagree with the advice-seeker regardless of the underlying best

answer (Jonas et al. [2005]). However, consultants are only likely to pursue directional goals

within the constraints of reason, e.g., if doing so does not threaten their reputation or career, or

their self-perception as objective professionals (Kunda [1990]; Kadous et al. [2003]).

In auditing, multiple pressures may motivate the consultant to agree with an advice-seeker,

including motivations to minimize a consultants own responsibility for the decision (e.g., on high

risk engagements), to minimize audit team effort (e.g., when they face a tight deadline), or to

satisfy third parties whose reviews might cause an auditor to emphasize the consistency of

10
documented conclusions (e.g., PCAOB inspectors and internal quality reviewers).5 Consultants

may also be motivated to avoid increased precision, as increasing precision constrains

managements discretion in financial reporting (Bamber et al. [2010]) and could strain client

relations for the engagement team. Thus, it is questionable whether consultants really adopt the

questioning mindset to the extent desired by regulators and audit firms. We focus our examination

on one motivation that varies naturally in an auditing (or any professional) context and could

influence a consultants recommendations: maintaining or improving ones status.

People often achieve status through observable actions that communicate their capabilities

to others. These actions include confronting or challenging others and making choices that are

either costly or distinctive (Griskevicius et al. [2009a; 2009b; 2010]). In auditing, status-motivated

consultants may exhibit a tendency to signal their self-confidence, as confidence is often perceived

as status by observers (Anderson et al. [2006]). In particular, consultants can signal confidence

through the increased contrariness and/or precision of their recommendations. Contrariness

implicitly signals that the consultant values their own abilities and beliefs more than those of the

advice-seeker (Kanodia et al. [1989]; Levy [2004]). Thus, contrariness acts as a possible signal of

the underlying capability of the consultant. Alternatively, precision is a signal of the consultants

degree of certainty that they know the true state of the world, which is commonly viewed as an

indicator of a persons confidence (Moore and Healy [2008]). Thus, increasing the precision of a

recommendation is an alternative means for a consultant to signal their confidence.

We specifically examine how a consultants status motives, i.e., the desire to gain prestige,

respect, or power over others (Griskevicius et al. [2009b]; Kenrick et al. [2009]), affect the

5
As an example of the pressures to maintain agreement, the AICPAs (2012) standard QC 10.4 requires that the
engagement team document disagreements that arise during consultation and delay the release of the audit report
until the disagreements are resolved.

11
contrariness and precision of recommendations. We draw the concept of status motives from

evolutionary theory, which argues that status motives are among a small group of specialized

problem solving systems, each of which evolved to cope with a particular survival dilemma that

confronted early humans (Cosmides and Tooby [1994]; Kenrick et al. [2003]). In this framework,

a persons preferences and resource allocations are based on a combination of specialized sets of

decision rules, rather than a single rational decision system (Gintis et al. [2001]; Kenrick et al.

[2009]; Griskevicius and Kenrick [2013]). Status motives relate to the survival dilemma of

obtaining desirable resources while competing with others for those resources (Kenrick et al.

[2003]). Consequently, when status motives are active, people derive utility from actions that

increase prestige or power. Note that status motives do not refer to the possession of status or to a

personality trait that values status, but rather to a temporary motivation to attain status that can

exist in any person when cues activate the motivation.

Status motives are likely to activate in the presence of situational cues of dominance,

prestige, or competition, such as accomplishments, rivalries, or highly-regarded products or

people (Griskevicius and Kenrick [2013], 378). Because auditors routinely confront ill-structured

information that is rich with evidential, task, economic, and social cues (Bell et al. [2005]; Ranzilla

et al. [2011]), cues that activate status motives are likely to arise and vary naturally in an audit

setting. Such cues tend to represent either potential threats to ones status or opportunities to gain

status. In an auditing setting, possible examples include the presence on an engagement of either

a competitor or highly successful coworker; thinking about or receiving praise for ones

accomplishments (or seeing someone else praised); working on a prestigious client; or witnessing

formal or informal rewards such as promotions, raises, desirable job assignments, and training

opportunities. Because consultation is often a high-profile assignment, involves interaction with

12
important decision makers or peers, and can impact a professionals career progression, it is

reasonable to assume that consultation contexts will activate status motives.

As discussed above, features of the auditing setting may constrain increased contrariness

or increased precision such that either action is potentially costly to the consultant. As a result,

consultants with active status motives may be constrained in their ability to satisfy status motives

through these actions. While active status motives may increase the consultants willingness to

incur the costs of increased contrariness or precision, these motives alone may not be sufficient to

influence a consultants behavior. Rather, we expect that consultants with active status motives are

likely to increase contrariness or precision only in conditions when such actions seem reasonable,

i.e., when their position is consistent with their self-image as an objective, responsible professional

(Kunda [1990]; Kadous et al. [2003]). What a consultant deems reasonable depends on their level

of knowledge and the constraints placed on their decision making by the audit process and

environment. In the following sections, we focus on two conditions that are likely to influence how

a consultants status motives impact their willingness to be contrary or precise: (1) the consultants

level of specialized knowledge and (2) the consultants level of decision authority.

2.2.2 Hypotheses 1a and 1b: The Consultants Status Motives and Specialized Knowledge

Auditors begin specializing early in their careers, often by industry (Solomon et al. [1999])

but also by functional area (Boritz et al. [2015]). Within their specialization, they integrate

information differently and better distinguish between relevant and irrelevant information (Taylor

[2000]; Low [2004]; Hammersley [2006]). Auditors with greater specialized knowledge have well-

developed mental models for detecting misstatements and dismissing non-misstatements (Kaplan

and Reckers [1989]; Kaplan et al. [1992]; Bedard and Biggs [1991]; Biggs et al. [1993]).

Specialized knowledge enables auditors to develop and evaluate a larger number of plausible

13
hypotheses for a questionable account balance fluctuation or estimate (Wright and Wright [1997];

Solomon et al. [1999]). As a result, if status motives lead consultants to want to signal self-

confidenceeither through greater contrariness or greater precisionthen it is possible that status

motives affect recommendations differently depending on the consultants degree of knowledge

on the issue.

Typically, consultants possess specialized knowledge acquired through experience or

training, which is why they are in a consultative position. However, their degree of knowledge on

an issue can vary based on factors such as the task (e.g., program planning versus substantive

evidence evaluation) or the direct experience a consultant could have with a specific issue (e.g.,

an unusual type of transaction or estimate for which no precedents exist) (DeZoort and Stanley

[2006]; Meservy et al. [2006]). Thus, consultants are often experienced and knowledgeable, but

this knowledge can vary in practice from situation to situation.

In general, we expect that precision is the preferred way to signal confidence and thereby

to satisfy status motives. An individuals ability to increase the precision of their responses directly

depends on the knowledge they have available to justify their advice. While contrariness can signal

self-confidence, organizations often discourage conflicting opinions and reward yes men even

when they claim to prefer independent thought (Prendergast [1993]). Moreover, advice-seekers in

auditing are likely to expect consultants to justify their recommendations (Kadous et al. [2013]).

As a result, consultants may feel accountable to the advice-seeker and feel pressure to match their

recommendations to the preferences of the advice-seeker (Peecher [1996]). These factors suggest

that an expert can better achieve their status motives using precision rather than contrariness. An

expert is likely to possess the knowledge necessary to suggest a narrower range of options in a

given decision situation. This knowledge may also enable the consultant to more effectively help

14
the engagement team challenge a clients preference for less precision in accounting estimates.

Accordingly, because status motives cause a consultant to signal self-confidence, a consultant is

likely to do so via incremental precision as their knowledge increases.

In contrast, as consultants expertise on an issue decreases, they are less likely to believe

that they possess the necessary knowledge to reasonably justify increased precision. Moreover,

those with less knowledge on an issue are increasingly likely to use effort-reducing heuristics

(Kadous et al. [2013]). Because an evolved response to status motives is to appear distinctive or

argumentative (Grisckevicius [2009a; 2009b]), consultants with less expertise are likely to adopt

a position of contrariness in response to status motives, knowing that they may lack the expertise

in a given context to defend an increase in precision. That is, a less knowledgeable consultant will

offer contrary recommendations that reflect a devils advocate position, rather than the more

cognitively challenging approach of justifying more precise recommendations.

Thus, we expect active status motives to increase precision among consultants with high

specialized knowledge, but not among those with low specialized knowledge. That is, precision

will be highest among consultants with both active status motives and high levels of specialized

knowledge. Further, we expect active status motives to increase contrariness among consultants

with low specialized knowledge, but not among those with high specialized knowledge. That is,

contrariness will be highest among consultants with both active status motives and low levels of

specialized knowledge. This leads to our first two hypotheses:

H1a: Active status motives increase contrariness when evaluating accounting estimates
among consultants with lower specialized knowledge, but not among those with higher
specialized knowledge.

H1b: Active status motives increase precision when evaluating accounting estimates
among consultants with higher specialized knowledge, but not among those with lower
specialized knowledge.

15
2.2.3 Hypotheses 2a and 2b: The Consultants Status Motives and Decision Authority

In auditing, it is common for a consultants decision authority to vary. Low decision

authority suggests that advice-seekers have significant discretion to act on or discount a

consultants advice, whereas high decision authority suggests advice-seekers have limited

discretion. In formalized consultation policies, a consultants decision authority can vary from

situation to situation, even within the same firm (Gold et al. [2012]). To be sure, low consultant

decision authority does not necessarily mean that advice-seekers will ignore the consultant, though

there are noteworthy instances in which an engagement team has ignored expert consultant

recommendations (e.g., Missal [2008], 474 476). In auditing, the consultation process is more of

a give and take process through which consensus emerges. The value of consultation lies in

challenging the advice-seeker to justify their conclusions, and the process of doing so could

involve either the consultant or advice-seeker converging towards the others opinion.

As a consultants decision authority increases, the consultant may feel more responsible

for the decision and believe that subsequent problems with an audit conclusion may be attributed

to the consultant. In turn, consultants likely experience greater internal dissonance in formulating

a recommendation when they have high decision authority. We argue that higher decision authority

likely influences the consultants regulatory focus, i.e., the mindset consultants adopt in processing

information that leads them to more strongly focus either on improving a situation or on not making

it worse (Higgins [1997]). High decision authority may cause the consultant to adopt a more

defensive mindset to process information and make recommendations in manner that avoids bad

outcomes that could be attributed to their advice (Polman [2012b]).6 That is, they focus on not

6
Regulatory focus can be more defensive (prevention focused), which emphasizes safety and responsibilities, or
positive (promotion focused), which emphasizes hopes and improvements. In a representative study that contrasts
prevention and promotion foci, Crowe and Higgins ([1997]) examine the effects of rewarding decisions makers for
performing well on a variety of tasks, as opposed to rewarding them for not performing poorly (instantiating a

16
being the cause of less justifiable conclusions. Conversely, as decision authority decreases, a

consultant may adopt a more positive mindset that is less concerned with avoiding bad outcomes

and more concerned with improving the conclusion.

There is evidence that differences in regulatory focus affect expert, highly confident, and

experienced decision makers, as well as novices. For example, physicians recommend different

treatments to patients than they would select to treat themselves, and recommend different patient

treatment plans to other physicians than they would select if they made the final treatment decision

themselves (Ubel et al. [2011]; Rocke et al. [2014]). Similarly, Jaskiewicz and Luchak [2013] rely

on regulatory focus to reason that a CEO leading a family firm is more focused on preventing

losses or damage to the familys financial interests if he or she is a member of the family, rather

than an outsider. Consistent with this logic, family-owned firms are less successful in rapidly

changing environments when led by a family-member CEO, rather than a non-family CEO.

All else equal, when decision authority is higher, a consultant is likely to be more

concerned with not worsening an audit conclusion or undermining the process of reaching that

conclusion. As discussed above, while contrariness and precision can both improve audit

conclusions, either behavior can also lead to negative side effects such as delayed audit reports,

budget overruns, worsened client relations, or heightened litigation risk. As a result, the consultant

is likely to adopt a defensive mindset to avoid potential negative outcomes by being less contrary

or less precise when decision authority is high.

promotion or prevention focus, respectively). They find that, relative to promotion-focused people, those who are
prevention-focused tend to perform worse after failure on a difficult task and generate fewer decision alternatives.
This concept differs from the concept of accountability. Whereas accountability relates to external pressures to defend
ones views, this theory relates to internal pressures to perform well and/or avoid mistakes. Due to the myriad
accountability pressures that auditors face (Gibbins and Newton [1994]; Peecher et al. [2013]), we expect
accountability to be quite high regardless of decision authority.

17
Accordingly, we argue that the defensive mindset generated by high decision authority is

likely to moderate the effects of active status motives on a consultants recommendations.

Increased contrariness or precision differentiates the consultants position from that of the advice

seeker and increases the chances of a mistake attributable to the consultant. As noted above, active

status motives lead consultants to signal confidence through the means that are most reasonable to

them. If consultants with higher decision authority have a defensive mindset that emphasizes

avoiding potential negative outcomes, then they will be less willing to increase contrariness or

precision to satisfy status motives. Thus, decision authority is likely to weaken the effects of status

motives on the recommendations of consultants, leading to our last two hypotheses:

H2a: Active status motives increase consultants contrariness when evaluating


accounting estimates when decision authority is low but not when it is high.

H2b: Active status motives increase consultants precision when evaluating


accounting estimates when decision authority is low but not when it is high.

3. Method

3.1 PARTICIPANTS

The participants were 102 senior managers and managers from a Big Four accounting firm

in the United States who were asked to participate in the experiment during training sessions

sponsored by their firm. All training session attendees received an email from the firm that asked

them to complete the experiment. The email included a link to the experiment and a brief message

that indicated the firms support for the research. Mean (standard deviation) auditing experience

of the participants was 9.72 (3.18) years and ranged from 5 to 19 years, and 71% of participants

indicated that they had served in a consultative role at some point during their career.

18
Procedures

The experimental case (adapted from Peecher et al. [2010] and Kadous et al. [2013])

instructed participants to assume they were consulting with a colleague at the same rank but on a

different engagement on a valuation issue arising during the audit of Retail Credit Incorporated

(RCI), a client that securitizes credit card receivables. The case provided extensive background on

the securitization in question, evidence collected by the colleague, and industry benchmarks for

comparison. The audit evidence was ambiguous and left open the possibility of intentional

overstatement of gains on securitization. The advice-seekers conclusion was that a discount rate

lower than the rates used by industry leaders would be appropriate, resulting in a higher estimated

fair value and larger gain on securitization. Participants, serving as consultants, recommended the

appropriate discount rate they thought would apply to the case and assessed their confidence in

that recommendation. Participants also indicated the range of reasonable discount rates that could

be used for the estimate. Finally, they answered a series of post-test and demographic questions.

3.2 INDEPENDENT VARIABLES

Our first independent variable was active versus non-active status motives. Prior to reading

the securitization case, all auditors read a 600 word story that we manipulated between-

participants. In the active status motive condition, participants read a story (adopted from

Griskevicius et al. [2009b; 2010]) that asked them to assume that they had just explored the job

market and accepted a new job at a prestigious firm. The story described observing the luxury cars

in the parking lot, visiting the opulent office building, and the prospect of competing with two co-

workers who were starting that day.7

7
We activate status motives with a validated priming manipulation unrelated to the judgment task in order to ensure
the internal validity of our experiment (Griffith et al. [2016]). Otherwise, manipulating status motives by varying some
aspect of the task or context would risk confounding status motives with another source of motivation, e.g., incentives,
task familiarity or knowledge, etc. We are grateful to Vlad Griskevicius for generously sharing his materials.

19
In the non-active status motive condition, the 600 word story was neutral with respect to

status and other evolutionary motives. This story asked participants to imagine that they are

searching for lost concert tickets, the panic of having lost them, and then the exhilaration after

finding them and going to a long-awaited show. To ensure that participants believed that the story

in the priming manipulation was unrelated to the audit judgment case, the initial experimental

instructions asked participants to assume that they were two separate tasks. The instructions

described the story as similar to a memory task and stated that there would be questions about the

story at the end of the experiment. We included three memory questions in the post-test materials.

The second independent variable was decision authority. In the introduction to the audit

case, we varied the instructions that the engagement partner gave to the hypothetical advice-seeker

in this consultation. In the low authority condition, participants were told that [t]he partner

suggested that your colleague should pick your brain on the issue but the colleague is expected

to resolve the issue on his or her own. In the high authority condition, participants were told that

[t]he partner has directed your colleague to solicit your input and to take and document actions

as a result of the advice you provide.

Our third independent variable was specialized knowledge. We measured participants

self-reported knowledge of securitizations on a 0 (very low) to 100 (very high) scale. We

classified participants whose knowledge was above the sample median as having high

specialized knowledge, and those below the median as having low specialized knowledge.8

8
An alternative measure of specialized knowledge is specialization in industries in which securitizations are most
common, such as financial services, insurance, and banking. In our sample, 29 auditors were specialists in these
industries, and 73 were specialists in other industries. However, securitizations are increasingly common in many
industries, and several participants did not list financial industry specializations but assessed relatively high
securitization knowledge. For example, one participant was a health care specialist who reported securitization
knowledge of 90 out of 100, as this participant had a functional specialization in receivables, which are commonly
securitized. Thus, we do not believe that industry is the best proxy for specialized knowledge for our task. Nonetheless,
we run our analyses using industry specialization, and our results do not differ substantively from those using self-
reported knowledge.

20
3.3 DEPENDENT MEASURES

To measure contrariness, we use the difference between the participants recommendation

and the hypothetical colleagues estimate.9 To measure precision, we use the difference between

the highest and lowest values specified in the participants range of reasonable discount rates.

Because precision is presumed to decrease as the size of the range increases, we construct a

measure in which higher values indicate greater precision to facilitate the exposition of our results.

To do so, we subtract the highest range size value in our sample from each auditors range size,

and take the absolute value. Thus, higher values indicate greater precision. For supplemental tests,

we construct a measure of the degree to which advice is likely to affect audit quality.

4. Results

4.1 MANIPULATION CHECKS

Regarding the manipulation check for status motives, we asked participants to rate their

feelings after reading the story. On a nine point Likert scale with anchors of 1 = Not at all and 9

= Very much, participants rated the extent to which they felt competitive, which is our

manipulation check for the status prime. Relative to the non-active status motive prime, the status

prime led to significantly higher competitiveness (6.54 versus 2.97, t(84) = -5.60, p < 0.001). As

intended, the primes did not lead to different levels of positive and negative arousal that could

otherwise influence consultants motivations to be contrary or precise, specifically, enthusiasm,

excitement, anger, and frustration (all p 0.248). The active status prime did lead to higher self-

assessed nervousness than the non-active status motive prime (5.50 versus 4.20, t(81) = -1.97, p =

9
We use the signed difference because auditors beliefs are likely influenced by management (Griffith et al. [2015]),
which is likely to be biased in a manner that is consistent with managements incentives (e.g., Dechow et al. [2010]).
Thus, it is reasonable to assume that greater contrariness translates into more effective auditor performance by
recommending greater belief revision away from managements incentives. We note that eight of our 102
participants recommended a rate lower (i.e., less conservative) than the advice-seekers best estimate. Our findings
are identical when we either exclude these eight participants from our analysis or use the absolute difference
between the recommendation and the advice-seekers best estimate as our dependent measure.

21
0.052).10 These responses indicate that the status prime was successful in arousing the motivation

to feel and appear superior to others among participants.

Regarding our manipulation of decision authority, we ask participants to rate their private

confidence on a 9 point scale with the same anchors as the status manipulation check. This

measures the degree to which the manipulation made them feel more internal pressure about the

recommendation. We also ask participants to rate the extent to which they feel accountable for

their recommendations to ensure that our decision authority manipulation does not also vary

accountability. The results indicate that, relative to the low decision authority condition,

consultants in the high decision authority condition rated their confidence as lower (53.77 versus

64.77, t(95)= 2.64, p = 0.010) while their accountability did not differ (66.95 versus 66.49, t(96)

= 0.10, p = 0.922). We conclude that our manipulation of decision authority manipulates internal

pressure while holding external pressures constant.

<<<<< Insert Table 1 about here >>>>>

4.2 HYPOTHESIS 1: STATUS MOTIVES BY SPECIALIZATION

As the foundation for our analyses, we run a three-way ANOVA for each of our primary

dependent measures (contrariness and precision). Specifically, we first run a 2 (Motive: status

active, status not active) X 2 (Specialized knowledge: higher, lower) X 2 (Decision authority: high,

low) ANOVA with contrariness as the dependent measure. Second, we run a similar ANOVA with

precision as the dependent measure. See Table 2, Panels A and B for ANOVA results for the

contrariness and precision measures, respectively. To ensure that our hypothesis tests are based on

the most efficient partitioning of the error variance, we test our hypotheses with F-test contrasts

from these ANOVA models. See Table 3 for hypothesis tests. We note that the main effect of

10
Our degrees of freedom vary slightly by measure, as some participants did not complete all of the measures. The
pattern of non-responses appears to be random across measures.

22
status motives is not significant for either of our dependent measures. This is consistent with our

predictions that status motives interact with specialized knowledge and decision authority.

<<<<< Insert Tables 2 and 3 about here >>>>>

H1a (H1b) predicts that active status motives increase consultants contrariness (precision)

only when consultants knowledge is lower (higher). Consistent with our prediction of a status

motives by specialized knowledge interaction, there are significant interactions for both the

contrariness measure (F1,94 =7.72, p = 0.007, Table 2, Panel A) and the precision measure (F1,94 =

7.75, p = 0.006, Table 2, Panel B). We use linear contrasts to formally test H1a and H1b as each

hypothesis predicts an ordinal interaction (Buckless and Ravenscroft [1990]). For H1a, we assign

a weight of +3 for the active status motive, lower knowledge condition and -1 for each of the

remaining status motive and knowledge conditions. For H1b, we assign a weight of +3 for the

active status motive, higher knowledge condition and -1 for each of the remaining conditions. As

shown in Table 3, the contrasts are significant both for contrariness (F1,94 = 9.09, p = 0.003) and

precision (F1,94 = 9.70, p = 0.002).11 See Figure 1 for graphed means.

<<<<< Insert Figure 1 about here >>>>>

Thus, our results support both H1a and H1b, i.e., consultants respond to active status

motives by increasing the precision of their recommendations when their knowledge is higher, and

by increasing their contrariness when their knowledge is lower. The significant contrasts are

consistent with descriptive statistics in Table 1, Panels A and B and partial effects analyses in

Table 3, Panel A. Partial effects reveal that the general evolved response to active status motives,

i.e., to challenge others, only manifests in audit consultation when consultants have lower

11
We report one-tailed p-values for the partial effects, as these predictions are directional. All other reported p-
values are two-tailed.

23
knowledge (2.19 versus 1.08, F1,94 = 8.53, p = 0.002). More knowledgeable consultants respond

by increasing precision as a means to signal confidence (4.50 versus 3.34, F1,94 = 8.71, p = 0.002).

These results suggest that the institutional features of auditing, which often constrain

disagreement, moderate the evolved response to status motives. That is, the decision rules to

enhance status operate differently in audit consultations depending on the level of the consultants

knowledge. This suggests that professional settings change, but do not undo, the effects of

evolutionary motives. To the extent that contrariness is desired by regulators and firms, these

findings are not encouraging as the auditing setting constrains a natural motivation to exhibit

contrariness under conditions in which contrariness would have the most potential to improve

auditor judgment, i.e., when auditors consult with highly knowledgeable consultants.

4.3 HYPOTHESIS 2: STATUS MOTIVES BY DECISION AUTHORITY

H2a and H2b predict that active status motives increase both contrariness and precision

when decision authority is low. As shown in Table 2, Panels A and B, neither of the status motive

by decision authority disordinal interaction terms are significant (both p 0.199). There is a main

effect of decision authority on precision (F1,94 = 4.53, p = 0.036, Table 2 Panel B) such that

consultants are, on-average, less willing to provide precise recommendations when they have high

decision authority. We use linear contrasts to formally test H2a and H2b (Buckless and

Ravenscroft [1990]), which are reported in Table 3, Panel A.

<<<<< Insert Figure 2 about here >>>>>

For H2a, we assign a weight of +3 for the active status motive, low decision authority

condition and -1 for each of the remaining status motive and decision authority conditions. While

this contrast is insignificant for contrariness (F1,94 = 2.15, p = 0.146), Thus, the data do not support

H2a. Recall that we theorize that active status motives increase contrariness among low knowledge

24
consultants, whereas high knowledge consultants seek to avoid contrariness when possible.

Consequently, it is possible that high decision authority may attenuate the effect of active status

motives on contrariness when knowledge is high, as these consultants are most motivated to avoid

contrariness. That is, the effect hypothesized in H2a may manifest only for high knowledge

consultants. In support of this reasoning, the predicted linear contrast is marginally significant

when specialized knowledge is higher (F1,94 = 2.87, p = 0.094, untabulated) but is insignificant

when specialized knowledge is lower (F1,94 = 0.02, p = 0.887, untabulated). This is a discouraging

possibility, as decision authority may constrain contrariness when it is most potentially beneficial,

i.e., from higher knowledge consultants.

For H2b, we assign a weight of +3 for the active status motive, low decision authority

condition and -1 for each of the remaining conditions. This contrast is significant for the precision

measure (F1,94 = 4.73, p = 0.032, Table 3).12 Thus, H2b is supported, i.e., consultants are likely to

increase precision in response to active status motives when decision authority is low, but not when

it is high. This is consistent with the idea that high decision authority increases consultants

avoidance mentality, which reduces their willingness to formulate more precise recommendations.

In sum, our evidence suggests that high decision authority constrains precision on-average,

and weakens the effect of active status motives on precision. We also find some evidence that high

decision authority makes higher knowledge consultants even less willing to be contrary in response

to active status motives. Thus, decision authority constrains each property of consultant

12
As shown in Table 3, Panel A, two of the three partial effects implied by this contrast are significant, but precision
does not differ between the two low authority conditions, i.e., active, as opposed to non-active status motives, given
low authority (p = 0.233). We do not believe that this insignificant partial effect threatens the validity of our
inferences, as there are no differences between the three conditions with -1 a priori contrast weights (see Table 3)
and we are aware of no compelling alternative theory that is both consistent with our findings but contrary to the
logic of our theory. Reasoning ex post, it is possible that high decision authority in itself is sufficient to decrease
precision and that support for our prediction reflects active status motives amplifying this main effect, at least in our
decision context.

25
recommendations under certain conditions. On balance, these results suggest granting a high level

of decision authority to a consultant may actually hinder the effectiveness of consultation as a

quality control.

4.4 EVIDENCE ON IMPROVING THE AUDIT OF ESTIMATES

We now examine the potential effects of consultant recommendations on ultimate audit

effectiveness. Because managements estimates are likely to be consistent with their incentives

and to influence an advice-seekers opinions (Hackenbrack and Nelson [1996]; Kadous et al.

[2003]; Dechow et al. [2010]), consultation can influence audit effectiveness by encouraging

estimates that better reflect underlying fundamentals rather than managements incentives. The

degree to which the audit teams conclusion about a client estimate actually improves after

receiving a consultants advice depends on the relationship of that advice to the prior beliefs of

the audit team and the extent to which the consultants advice is utilized. Nonetheless, we believe

our findings have a generalizable and straightforward link to audit quality of interest to regulators,

practitioners, and scholars focused on improving the audit of estimates.

Auditing standards state there may be a range of reasonable values for a given estimate or

valuation (e.g., AICPA AU 312 [2006]). If an estimate falls outside this range, then an auditor

can accept a valuation at the extreme end of the range closest to the estimate. Our experimental

setting strongly implies that managements incentives are to overstate fair value. Thus, the lower

bound of reasonable discount rates will be the most likely estimate because it results in the highest

valuation that is acceptable to the auditor. To recommend an estimate less likely to be biased by

management, a consultant could recommend increasing the lower bound of the acceptable range

of discount rates, which provides less latitude for management to overstate earnings. The upper

26
bound is less likely to influence the ultimate estimate because managements initial position is

likely to be closer to the aggressive end of the range.

In our previous tests, we find that status-motivated, highly knowledgeable consultants did

not take a contrary position but did provide more precise recommendations. In our setting, less

contrary recommendations are, by construction, less conservative and more consistent with

managements incentives. Accordingly, because high knowledge consultants appear unwilling to

be contraryeven when status motives are activeaudit quality may not benefit in situations in

which an auditor would hope to obtain the most benefit from consultation, i.e., with a high

knowledge consultant.

If a consultants unwillingness to be contrary is unlikely to improve audit quality, then

what about their willingness to be more precise? More precise advice from a consultant could help

limit a clients latitude to report estimates opportunistically (Bamber et al. [2010]; Dechow et al.

[2010]). However, increased precision in our experiment is most likely to do so if the increased

precision is due to an increase in the lower bound of the range of reasonable discount rates. If the

increased precision is due to a decrease in the upper bound, then the underlying valuation is less

likely to be influenced by the consultants advice, because negotiation between the client and audit

team is likely to focus on the lower bound.

<<<<< Insert Figure 3 about here >>>>>

In untabulated analyses, an ANOVA model with status motives, specialized knowledge,

and decision authority as independent variables and the lower bound as the dependent measure is

insignificant (p = 0.539), suggesting increased precision does not lead to recommendations that

substantively decrease the latitude for client opportunism.13 This implies that changes in precision

13
We also conduct one-sample t-tests to examine whether consultants lower bounds statistically differ from the
advice-seekers opinion in any of our conditions. There is a significant difference between the lower bound and the

27
due to active status motives are most likely driven by differences in the upper bounds. A visual

inspection of Figure 3, which depicts the mean upper and lower recommended bounds, confirms

this trend. We also run an untabulated ANOVA with status motives, specialized knowledge, and

decision authority as independent variables and the upper bound of the range of discount rates as

the dependent measure. There is a significant status motive by specialized knowledge interaction

(F1,94 = 6.29, p = 0.014, untabulated), which is driven by high knowledge consultants decreasing

the upper bounds (17.43 versus 16.13, F1,94 = 7.73, p = 0.007, untabulated).

Specialized knowledge develops primarily through experience in situations that do not

involve misstatements (Solomon et al. [1999]). As a result, consultants with higher specialized

knowledge have superior ability to explain why specific evidence does not suggest a misstatement.

In our experiment, consultants with higher specialized knowledge satisfy status motives with

increased precision. However, this is unlikely to lead the audit team to conclude that the valuation

is misstated, which could increase audit quality but also create frictions between the audit team

and the client if the audit team felt they needed to act on that advice. This would require consultants

to recommend increasing the lower bounds. Rather, these consultants react to status motives by

reducing the upper bound of the reasonable range. This reaction is unlikely to affect audit quality

but does allow the consultant to satisfy status motives in a manner that does not cause friction with

the client or audit team. If the consultants goal is to signal self-confidence through incrementally

precise recommendations, then tightening the upper bound is the least adversarial way to do so.

Thus, increased precision does not necessarily enhance audit effectiveness in the setting where the

consultants knowledge could be of the most potential benefit.

advice-seekers opinion in only one conditionspecifically, lower knowledge consultants with active status motives
(13.96 versus 13.20, t(23) = 2.22, p = 0.037). This is consistent with the rationale of H1a that lower knowledge
consultants with active status motives are most likely to be contrary.

28
5. Conclusions and Future Research Directions

Consultation is a critical component of an audit firms system of quality control. In areas

such as accounting estimates in which consultation is pervasive, the receipt of useful advice is

often critical for improving auditor judgments. Contrariness and precision are both attributes of

consultant recommendations that have the potential to improve the audit of estimates. Yet, the

literature provides little guidance about the motivations or judgments of consultants, let alone the

conditions that influence the nature of their recommendations. In an experiment, we examine the

effects of status motives on the contrariness and precision of consultant recommendations, and

whether the effects a consultants degree of specialized knowledge and decision authority are

conditional on status motives.

Our findings suggest that audit consultations may not be as helpful or useful as generally

expected. Evolutionary psychology predicts that contrariness is a natural response to active status

motives, but we find that higher specialized knowledge constrains this effect. We also find some

evidence that high decision authority further constrains the willingness of high knowledge

consultants to be contrary. We find that contrariness is most likely among lower knowledge

consultants, which raises the possibility that contrariness in audit consultations may manifest as

devils advocate positions, which may or may not be taken seriously. While status motives

would generally be expected to increase both contrariness and precision in consultants, we find

that this effect is constrained by features of an audit setting, i.e., when the consultant is highly

knowledgeable and has been ceded extensive decision authority.

It is noteworthy that high decision authority has significant main and interactive effects

that constrain precision. Low decision authority for consultants is likely warranted if the goal is to

elicit the most useful advice from consultants. However, more complex tradeoffs are possible, and

29
firms, regulators, and scholars must consider the goals of a given consultation in order to best

structure decision authority. Gold et al. [2012] find that auditors are more willing to consult when

they expect consultants to have high decision authority, as auditors try to reduce their own

responsibility for consequential decisions. Thus, higher decision authority has the benefit of

encouraging more frequent consultation, as the mere act of consultation makes complex decisions

more defensible to others even when advice is not followed (Kennedy et al. [1997]). However, we

find conditions in which consultants with high decision authority provide less useful advice,

suggesting that responsibility may be shared for conclusions that arent necessarily better. This

raises the possibility that, under some circumstances in the audit of estimates, higher consultant

decision authority could contribute to an illusion of audit quality, i.e., a tenuous balance of more

frequent consultations producing judgments that are not substantively higher quality, but

nonetheless appear better because auditors are more confident in them and can more easily defend

them to others. Granted, consultation may actually improve judgments, but our findings with

experienced, knowledgeable professionals suggest this is not a guarantee. The possibility of a

tradeoff between greater frequency and greater quality of consultation, as well as boundary

conditions on this potential tradeoff, deserves attention from scholars, firms, and regulators.

We believe our theory generalizes to a wide range of settings, and the multi-dimensional

nature of the consultation construct represents many opportunities for future research. For

example, advice can take the form not only of recommendations for a course of action, but also

recommendations against a course of action, suggestions about a decision-making process, or the

provision of additional information about a problem (Dalal and Bonaccio [2010]). It is worthwhile

to examine drivers of the types of advice that auditors seek on estimates and whether or not the

type of advice interacts with our independent variables. For example, it is possible that status-

30
motivated consultants are more willing to be contrary when asked to bring unconsidered

information to the auditors attention, as opposed to when they are asked to make an explicit

recommendation. Similarly, a consultants knowledge can manifest as technical or tacit

knowledge, and the latter type of knowledge truly differentiates top performing managers (e.g.,

Tan and Libby 1997). It is therefore possible that consultants are more willing to show off and

be contrary on matters involving tacit knowledge, such as personnel or client management.

Our study is subject to limitations that may open avenues of future research. First, we note

that regulators and the auditing profession frequently invoke consultation but do not offer a formal,

consistent definition. We draw on insights from auditing and psychology to offer a scalable

definition of consultation that encompasses a range of formal and informal settings, as we believe

our findings generalize broadly. It is possible that the generality of our results depends on how one

defines consultation, and our theory is least likely to generalize to definitions or settings that

involve extreme degrees of decision authority (high or low). For example, one conception of

consultation is one in which executive-level partners substantively impose their will on the

engagement team. Our theory is silent on this narrow, albeit important set of highly formal

consultations.

Similarly, we believe that our high knowledge participants are reasonable proxies for

specialists, and that our inferences thus shed light on the use of specialists. However, we concede

that the degree to which our findings extend to this area may depend on how one defines a

specialist. For instance, is a specialist someone whose expertise lies outside auditing or

accounting, as regulators propose (PCAOB 2015e), or any person who is highly experienced and

knowledgeable within a domain? We do not know if it is consequential whether or not valuation

specialists are CPAs or have audit experience. Definitional issues are consequential only insofar

31
as they give rise to factors that would interact with our independent variables. If experience and

formal training in valuation (determinants of knowledge) do not substantively differ between non-

auditor valuation specialists and auditors who are highly knowledgeable about valuation, then our

results for specialized knowledge likely generalize broadly. Even if differences exist and auditors

would ultimately consult with non-auditor specialists on this type of issue, prior literature (e.g.,

Gibbins and Emby [1985]) suggests that auditors would seek formal or informal input from

knowledgeable other auditors prior to undertaking a costly additional procedure, such as engaging

a non-auditor specialist. As a result, our findings would still be relevant to auditors beliefs about

an issue when deciding to consult a specialist. This could affect the nature of the consultation or

even whether or not the consultation actually occurs.

Also, our task focuses on intra-firm consultation, and auditors sometimes hire specialists

from other firms. It is possible that differences in judgment may arise due to policy or cultural

differences between the auditors and specialists firms. These issues warrant future examination.

Further, our study focuses on consultation related to accounting estimates, but this is only

one area in which consultation is common and valuable. Other likely areas include revenue

recognition, impairments, and assessing the risk of fraud. While we can think of no a priori reason

why our inferences would not generalize to these settings, future research is warranted to examine

the potential for differences across audit issues. For instance, estimates involve a continuum of

possible outcomes, whereas issues like fraud assessments and revenue recognition often involve

discrete outcomes. When outcomes are discrete, it is possible that consultants motivations to be

contrary or precise differ or that auditors willingness to seek or use advice differs.

Also, in our study, the advice-seeker and the consultant have access to the same evidence

set. However, it is possible that the advice-seeker may stylize the information that they provide to

32
the consultant, especially if the advice-seeker is committed to their opinion and simply seeks

corroboration from the consultant. Future research could examine the conditions under which

advice-seekers motivations for consulting differ, the conditions under which stylization is more

or less likely, and the commons means of stylization.

In sum, this is the first study in auditing to examine status motives and to apply an

evolutionary psychology framework. It is likely that other evolutionary motives besides status

routinely activate in audit settings, e.g., coalition formation or self-protection motives might

influence auditor judgment. We believe that it is potentially fruitful to develop a more complete

understanding of the full range of auditor motivations, as such as an understanding facilitates the

development of mechanisms to improve auditor judgment. Other professional fields could also

benefit from the evolutionary psychology framework. We find limited evidence that status motives

have the same effects in professional settings as they do in general or consumer settings, which we

attribute to the consensus pressures in auditing and the specialized nature of professional

judgments. These pressures exist in other professions and even differ within branches of a

profession (e.g., general medicine has consensus norms whereas surgeons have individualistic

norms, see Lee [1997]). Research could examine whether active status motives reliably increase

contrariness in professional settings with individualistic orientations. It is further worthwhile

examining whether consultants facing strong consensus pressures use means in addition to

precision to satisfy active status motives.

33
REFERENCES

AMERICAN INSTITUTE OF CERTIFIED PUBLIC ACCOUNTANTS (AICPA). Statement of


the Accounting Principles Board No. 4. (1970) New York, NY: AICPA.
AMERICAN INSTITUTE OF CERTIFIED PUBLIC ACCOUNTANTS (AICPA). SAS 99:
Consideration of Fraud in a Financial Statement Audit (2002).
AMERICAN INSTITUTE OF CERTIFIED PUBLIC ACCOUNTANTS (AICPA). AU Section
312: Audit Risk and Materiality in Conducting an Audit. (2006).
AMERICAN INSTITUTE OF CERTIFIED PUBLIC ACCOUNTANTS (AICPA). A Firms
System of Quality Control (2012).
ANDERSON, C., S. SRIVASTAVA, J. BEER, S. SPATARO, AND J. CHATMAN. Knowing
your place: Self-perceptions of status in face-to-face groups. Journal of Personality and
Social Psychology 91(2006): 1094-1110.
BAMBER, L., K. HUI, AND E. YEUNG. Managers eps forecasts: Nickeling and diming the
market? The Accounting Review 85 (2010): 63 95.
BARROWS, S., V. MAHIDHAR, AND A. KAMBIL. Judgment sustained. Deloitte Review
(2010).
BEDARD, J., AND S. BIGGS. The effect of domain-specific experience on evaluation of
management representation in analytical procedures. Auditing: A Journal of Practice
and Theory 10 (1991): 77 95.
BELL, T., M.E. PEECHER, and I. SOLOMON. The 21st Century Public Company Audit:
Conceptual Elements of KPMGs Global Audit Methodology. (2005) Montvale, NJ:
KPMG International.
BELL, T.B., AND J.B. GRIFFIN. Commentary on auditing high-uncertainty fair value
estimates. Auditing: A Journal of Practice and Theory 31(2012): 147 155.
BIGGS, S., M. SELFRIDGE, AND G. KRUPKA. A computational model of auditor
knowledge and reasoning processes in the going-concern judgment. Auditing: A Journal
of Practice of Theory 12 (1993): 82 97.
BONACCIO, S., and R. DALAL. Advice taking and decision making: An integrative literature
review, and implications for the organizational sciences. Organizational Behavior and
Human Decision Processes 101 (2006): 127-151.
BORITZ, J.E., N. KOCHETOVA-KOZLOSKI, AND L. ROBINSON. Are fraud specialists
relatively more effective than auditors at modifying audit programs in the presence of
fraud risk? The Accounting Review 90(2015): 881 915.
BUCKLESS, F., AND S. RAVENSCROFT. Contrast coding: A refinement of ANOVA in
behavioral analysis. The Accounting Review 65(1990): 933 945.
COSMIDES, L., AND J. TOOBY. Better than Rational: Evolutionary Psychology and the
Invisible Hand. The American Economic Review 84(1994): 327 332.
CROWE, E., AND E.T. HIGGINS. Regulatory focus and strategic inclinations: Promotion and
prevention in decision making. Organizational Behavior and Human Decision
Processes 69(1997): 1171-32.

34
DEELSTRA, J., M. PEETERS, W. SCHAUFELI, W. STROEBE, F. ZJILSTRA, AND L. VAN
DOORNEN. Receiving instrumental support at work: When help is not welcome.
Journal of Applied Psychology 88 (2003): 324-331.
DEZOORT, T., AND J. STANLEY. Fair presentation in the Sarbanes-Oxley era: An
assessment framework and opportunities for forensic accountants. Journal of Forensic
Accounting 7 (2006): 279 294.
GIBBINS, M., AND C. EMBY. Evidence on the nature of professional judgment in public
accounting. In Auditing Research Symposium, A. Abdel-khalik, I. Solomon (Eds.),
University of Illinois, Champaign, IL (1985).
GIBBINS, M., AND J. NEWTON. An empirical exploration of complex accountability in
public accounting. Journal of Accounting Research 32(1994): 165 186.
GINTIS, H., E. SMITH, AND S. BOWLES. Costly Signaling and Cooperation. Journal of
Theoretical Biology 213 (2001): 103-119.
GLOVER, S., AND D. PRAWITT. Enhancing auditor professional skepticism: The
professional skepticism continuum. Current Issues in Auditing 8(2014): 1 10.
GOLD, A., R. KNECHEL, AND P. WALLAGE. 2012. The effect of the strictness of
consultation requirements on fraud consultation. The Accounting Review 87 (2012):
925-949.
GRIFFIN, J.B. The effects of uncertainty and disclosure on auditors fair value materiality
decisions. Journal of Accounting Research 52(2014): 1165 1193.
GRIFFITH, E. How do auditors use valuation specialists when auditing fair values? Working
paper, University of Wisconsin Madison (2015) Available at:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2460970&rec=1&srcabs=2424718&a
lg=1&pos=1
GRIFFITH, E., J. HAMMERSLEY, AND K. KADOUS. Audits of complex estimates as
verification of management numbers: How institutional pressures shape practice.
Contemporary Accounting Research 32(2015): 833 863.
GRIFFITH, E., K. KADOUS, AND D. YOUNG. How insights from the new JDM research
can improve auditor judgment: Fundamental research questions and methodological
advice. Auditing: A Journal of Practice and Theory 35(2016): 1 22.
GRISKEVICIUS, V., N. GOLDSTEIN, C. MORTENSEN, J. SUNDIE, R. CIALDINI, AND D.
KENRICK. "Fear and loving in las vegas: Evolution, emotion, and persuasion." Journal
of Marketing Research 46(2009a): 384 - 395.
GRISKEVICIUS, V., J. TYBUR, S. GANGESTAD, E. PEREA, J. SHAPIRO, AND D.
KENRICK. "Aggress to Impress: Hostility as an Evolved Context-Dependent Strategy."
Journal of Personality and Social Psychology 96(2009b): 980-994.
GRISKEVICIUS, V., J. TYBUR, AND B. VAN DEN BERGH. "Going Green to Be Seen:
Status, Reputation, and Conspicuous Conservation." Journal of Personality and Social
Psychology 98(2010): 392-404.
GRISKEVICIUS, V., AND D. KENRICK. "Fundamental motives: How evolutionary needs
influence consumer behavior." Journal of Consumer Psychology 23(2013): 372 - 386.
HACKENBRACK, K., AND M.W. NELSON. "Auditors' incentives and their application of
financial accounting standards." The Accounting Review (1996): 43 - 59.

35
HAMMERSLEY, J. Pattern identification and industry specialist auditors. The Accounting
Review 81 (2006): 309-336.
HERZOG, S., AND R. HERTWIG. "The wisdom of many in one mind: Improving individual
judgments with dialectical bootstrapping." Psychological Science 20(2009): 231 - 237.
HIGGINS, E.T. Beyond pleasure and pain. American Psychologist 52(1997):1280 1300.
INTERNATIONAL AUDITING AND ASSURANCE STANDARDS BOARD (IAASB). ISA
220, Quality Control for an Audit of Financial Statements. (2005)
JASKIEWICZ, P., and A. LUCHAK. Explaining performance differences between family firms
with family and nonfamily CEOs: Its the nature of the tie to the family that counts!
Entrepreneurship Theory and Practice 37 (2013): 1361 1367.
JONAS, E., AND D. FREY. Information search and presentation in advisor-client interactions.
Organizational Behavior and Human Decision Processes 91 (2003): 154-168.
JONAS, E., S. SCHULZ-HARDT, AND D. FREY. Giving advice or making decisions in
someone elses place: The influence of impression, defense, and accuracy motivation on
the search for new information. Personality and Social Psychology Bulletin 31(2005):
977 990.
KADOUS, K., J. KENNEDY, AND M.E. PEECHER. The effect of quality assessment and
directional goal commitment on auditors' acceptance of client-preferred accounting
methods. The Accounting Review 78 (2003): 759-778.
KADOUS, K., J. LEIBY, AND M.E. PEECHER. How do auditors weight informal contrary
advice? The joint influence of advisor social bond and advice justifiability. The
Accounting Review 88(2013): 2061 2087.
KANODIA, C., R. BUSHMAN, AND J. DICKHAUT. Escalation errors and the sunk cost effect:
An explanation based on reputation and information asymmetries. Journal of Accounting
Research 27 (1989): 59-77.
KAPLAN, S., AND P. RECKERS. An examination of information search during initial audit
planning. Accounting, Organizations and Society 14 (1989): 337 345.
KAPLAN, S., C. MOECKEL, AND J. WILLIAMS. Auditors hypothesis plausibility
assessments in an analytical review setting. Accounting, Organizations and Society 11
(1992): 50 65.
KENNEDY, J., D. KLEINMUNTZ, AND M.E. PEECHER. Determinants of justifiability of
performance in ill-structured audit tasks. Journal of Accounting Research 35 (1997):
105-123.
KENRICK, D., N. LI, AND J. BUTNER. "Dynamical evolutionary psychology: Individual
decision rules and emergent social norms." Psychological Review 110(2003): 3-28.
KENRICK, D., V. GRISKEVICIUS, J. SUNDIE, N. LI, Y.J. LI, AND S. NEUBERG. "Deep
rationality: The evolutionary economics of decision making." Social Cognition 27(2009):
764-785.
KRAY, L. Contingent weighting in self-other decision making. Organizational Behavior and
Human Decision Processes 83(2000): 82-106.
KUNDA, Z. The case for motivated reasoning. Psychological Bulletin 108 (1990): 480-498.
LARRICK, R., AND J. SOLL. "Intuitions about combining opinions: Misappreciation of the
averaging principle." Management Science 51(2006): 111 - 127.

36
LEE, F. "When the going gets tough, do the tough ask for help? Help seeking and power
motivation in organizations." Organizational Behavior and Human Decision Processes
72(1997): 336 - 363.
LEVY, G. Anti-herding and strategic consultation. European Economic Review 48 (2004): 503-
525.
LOW, K. The effects of industry specialization on audit risk assessments and audit-planning
decisions. The Accounting Review 79 (2004): 201-219.
MESERVY, R., M. ROMNEY, AND M. ZIMBELMAN. Certified fraud examiners: A survey
of their training, experience, and curriculum requirements. Journal of Forensic
Accounting 7 (2006): 163 184.
MISSAL, M. United States Bankruptcy Court for the District Delaware in re: New Century TRS
Holdings, Inc. Final Report of Michael J. Missal, Bankruptcy Court Examiner. (2008)
Available at:
http://graphics8.nytimes.com/packages/pdf/business/Final_Report_New_Century.pdf
Accessed 10 February 2016.
MOORE, D.A., AND P. HEALY. The trouble with overconfidence. Psychological Review
115(2008): 502 517.
NELSON, M.W. A model and literature review of professional skepticism in auditing.
Auditing: A Journal of Practice and Theory 28(2009): 1 34.
NG, T., AND P. SHANKAR. Effects of technical departments advice, quality assessment
standards and client justifications on auditors propensity to accept client-preferred
accounting methods. The Accounting Review 85 (2010): 1743-1761.
PEECHER, M.E. The influence of auditors justification processes on their decisions: A
cognitive model and experimental evidence. Journal of Accounting Research 34(1996):
125 140.
PEECHER, M. E., M.D. PIERCEY, J. S. RICH, AND R. M. TUBBS. The effects of
supervisors active intervention in subordinates judgments, directional goals, and
perceived technical knowledge advantage on audit team judgments. The Accounting
Review 85 (2010): 1763-1786.
PEECHER, M.E., I. SOLOMON, AND K.T. TROTMAN. An accountability framework for
financial statement auditors and related research questions. Accounting, Organizations
and Society 38(2013): 596-620.
POLMAN, E. Self-other decision making and loss aversion. Organizational Behavior and
Human Decision Processes 119 (2012a): 141-150
POLMAN, E. Effects of self-other decision making on regulatory focus and choice overload.
Journal of Personality and Social Psychology 102(2012b): 980-993.
PRENDERGAST, C. A theory of yes men. The American Economic Review 83(1993): 757
770.
PUBLIC COMPANY ACCOUNTING OVERSIGHT BOARD (PCAOB). AU Section 230: Due
Professional Care in the Performance of Work. (2006)
PUBLIC COMPANY ACCOUNTING OVERSIGHT BOARD (PCAOB). QC Section 20:
System of Quality Control for a CPA Firms Accounting and Auditing Practice. (2012)

37
PUBLIC COMPANY ACCOUNTING OVERSIGHT BOARD (PCAOB). Report on 2014
Inspection of Deloitte & Touche LLP (2015a).
PUBLIC COMPANY ACCOUNTING OVERSIGHT BOARD (PCAOB). Report on 2014
Inspection of PricewaterhouseCoopers LLP (2015b).
PUBLIC COMPANY ACCOUNTING OVERSIGHT BOARD (PCAOB). Report on 2014
Inspection of KPMG LLP (2015c).
PUBLIC COMPANY ACCOUNTING OVERSIGHT BOARD (PCAOB). Report on 2014
Inspection of Ernst & Young LLP (2015d).
PUBLIC COMPANY ACCOUNTING OVERSIGHT BOARD (PCAOB). Staff Consultation
Paper No. 2015-01: The Auditors Use of the Work of Specialists (2015e).
RADZEVICK, J., AND D. MOORE. Competing to be certain (but wrong): Market dynamics
and excessive confidence in judgment. Management Science 57(2011): 93-106.
RANZILLA, S., R. E. CHEVALIER, G. HERRMANN, S. M. GLOVER, AND D. F.
PRAWITT. Elevating Professional Judgment in Auditing: The KPMG Professional
Judgment Framework. KPMG LLP (2011).
ROCKE, D., H. BEUMER, S. THOMAS, and W. LEE. Effect of physician perspective on
allocation of medicare resources for patients with advanced cancer. Otolaryngology
Head and Neck Surgery 150 (2014): 792 800.
SALTERIO, S., AND R. DENHAM. Accounting consultation units: An organizational memory
analysis. Contemporary Accounting Research 14 (1997): 669-691.
SOLOMON, I., M. SHIELDS, AND R. WHITTINGTON. What do industry specialist auditors
know? Journal of Accounting Research 37 (1999): 191-208.
TAYLOR, M. The effects of industry specialization on auditors' inherent risk assessments and
confidence judgements. Contemporary Accounting Research 17 (2000): 693-712.
TETLOCK, P. Expert Political Judgment: How Good Is It? How Can We Know? (2005)
Princeton, NJ: Princeton University Press.
UBEL, P., A. ANGOTT, and B. ZIKMUND-FISCHER. Physicians recommend different
treatments for patients than they would choose for themselves. Archives of Internal
Medicine 171 (2011): 630 634.
VAUGHAN, D. The Challenger Launch Decision: Risky Technology, Culture, and Deviance at
NASA (1997) Chicago, IL: University of Chicago Press.
VUL, E., AND H. PASHLER. Measuring the crowd within: Probabilistic representations within
individuals. Psychological Science 19(2008): 645 647.
WRIGHT, S., AND A. WRIGHT. The effect of industry experience on hypothesis generation
and audit planning decisions. Behavioral Research in Accounting 9 (1997): 273 294.
YANIV, I. Receiving other peoples advice: Influence and benefit. Organizational Behavior
and Human Decision Processes 93 (2004): 1-13.
ZIKMUND-FISCHER, B., B. SARR, A. FAGERLIN, AND P. UBEL. A matter of perspective:
Choosing for others differs from choosing for yourself in making treatment decision.
Journal of General Internal Medicine 21(2006): pp. 618-622.

38
Appendix
Variable Descriptions

Dependent Variables:

Contrariness is the signed difference between participants recommended rate and the rate of the
advice-seeker. Higher values indicate more contrariness.

Precision is the absolute difference between the size of each participants range of reasonable
discount rates and the largest range size in our sample. That is, each participant assessed an
upper bound and a lower bound for the reasonable rates. We calculate the range size for each
participant as the difference between these two values. We then we subtract the largest range size
(7.00 points) in our sample from each observation and use the absolute value as our measure of
precision. Thus, the largest range size has precision of 0. Higher values indicate more precision.

Independent Variables:

Status Motives are manipulated with a 600 word prime prior to the task that either describes
looking for and finding lost concert tickets (Not Active) or starting a prestigious new job and
competing with co-workers (Active).

Specialized Knowledge is an indicator that equals one if the participants self-reported


securitization knowledge is above the sample median (Higher knowledge), and zero if it is below
the median (Lower Knowledge).

Decision Authority is manipulated by telling participants that the advice-seeker has been directed
to either document and take actions on the basis of the auditors recommendation (High decision
authority) or to simply seek the auditors opinion and make their own decision (Low decision
authority).

39
Table 1
Descriptive Statistics

Table 1 reports descriptive statistics for the dependent variables contrariness and precision.
Contrariness is the signed difference between participants recommended rate and the rate of the
advice-seeker. Higher values indicate more contrariness. Precision is the absolute difference
between the size of each participants range of reasonable discount rates and the largest range
size in our sample. Higher values indicate more precision.

Panel A reports contrariness descriptive statistics partitioned by the independent variables status
motives and specialized knowledge. The Status Motives variable is manipulated with a 600 word
prime prior to the task that either describes looking for and finding lost concert tickets (Not
Active) or starting a prestigious new job and competing with co-workers (Active). The
Specialized Knowledge variable is an indicator that equals one if the participants self-reported
securitization knowledge is above the sample median (Higher knowledge), and zero if it is below
the median (Lower Knowledge).

Panel B reports precision descriptive statistics partitioned by the independent variables status
motives and decision authority. Status Motives are described above. The Decision Authority
variable is manipulated by telling participants that the advice-seeker has been directed to either
document and take actions on the basis of the auditors recommendation (High decision
authority) or to simply seek the auditors opinion and make their own decision (Low decision
authority).

40
Table 1 (continued)
Descriptive Statistics

Panel A: Status Motives by Specialized Knowledge- Means, (Standard Deviations), Cell Counts
Contrariness Precision
Not Active Active Not Active Active
Status Motives Status Motives Status Motives Status Motives
Lower 1.08 2.19 3.91 3.51
knowledge (0.88) (1.78) (1.36) (1.07)
N = 27 N = 24 N = 27 N = 24

Higher 1.53 1.14 3.34 4.50


Knowledge (1.43) (0.84) (1.75) (1.41)
N = 26 N = 25 N = 26 N = 25

Panel B: Status Motives by Decision Authority-Means, (Standard Deviations), Cell Counts


Contrariness Precision
Not Active Active Not Active Active
Status Motives Status Motives Status Motives Status Motives
Low 1.14 1.90 3.87 4.36
Decision (1.10) (1.16) (1.38) (1.06)
Authority N = 23 N = 23 N = 23 N = 30

High 1.43 1.44 3.37 3.65


Decision (1.26) (1.69) (1.71) (1.49)
Authority N = 30 N = 26 N = 23 N = 26

41
Table 2
ANOVA Results

Table 2 reports the results of 2 (Motive: status active, status not active) X 2 (Specialized
knowledge: higher, lower) X 2 (Decision authority: high, low) between-participants ANOVA
tests, and the significance of related simple effects tests. The dependent measure in Panel A is
contrariness, which is the signed difference between participants recommended rate and the rate
of the advice-seeker. The dependent measure in Panel B is precision, which is the absolute
difference between the size of each participants range of reasonable discount rates and the
largest range size in our sample. All p values are two tailed.

The Status Motives variable is manipulated with a 600 word prime prior to the task that either
describes looking for and finding lost concert tickets (Not Active) or starting a prestigious new
job and competing with co-workers (Active). The Specialized Knowledge variable is an indicator
that equals one if the participants self-reported securitization knowledge is above the sample
median (Higher knowledge), and zero if it is below the median (Lower Knowledge). The
Decision Authority variable is manipulated by telling participants that the advice-seeker has been
directed to either document and take actions on the basis of the auditors recommendation (High
decision authority) or to simply seek the auditors opinion and make their own decision (Low
decision authority).

42
Table 2 (continued)
ANOVA Results

Panel A: Analysis of Variance (ANOVA) for Contrariness


Source Sum of Squares df Mean Square F p
Status 3.08 1 3.08 1.88 0.174
Decision Authority < 0.01 1 < 0.01 < 0.01 0.965
Knowledge 2.13 1 2.13 1.30 0.258
Status by Knowledge 12.66 1 12.66 7.72 0.007
Status by Decision Authority 2.74 1 2.74 1.67 0.199
Decision Authority by Knowledge 1.71 1 1.71 1.71 0.310
Status by Decision Authority by Knowledge 1.95 1 1.95 1.19 0.278
Error 154.2 94 1.64

Panel B: Analysis of Variance (ANOVA) for Precision


Source Sum of Squares df Mean Square F p
Status 3.67 1 3.67 1.88 0.174
Decision Authority 8.87 1 8.87 4.53 0.036
Knowledge 1.13 1 1.13 0.58 0.450
Status by Knowledge 15.17 1 15.17 7.75 0.006
Status by Decision Authority 0.29 1 0.29 0.15 0.702
Decision Authority by Knowledge 2.27 1 2.27 1.16 0.284
Status by Decision Authority by Knowledge 2.81 1 2.81 1.43 0.234
Error 183.94 94 1.96

43
Table 3
Hypothesis Tests

This table reports our hypothesis tests and related simple effect. The dependent measure for the
tests in the left column is contrariness, which is the signed difference between participants
recommended rate and the rate of the advice-seeker. The dependent measure for the tests in the
right column is precision, which is the absolute difference between the size of each participants
range of reasonable discount rates and the largest range size in our sample.

The Status Motives variable is manipulated with a 600 word prime prior to the task that either
describes looking for and finding lost concert tickets (Not Active) or starting a prestigious new
job and competing with co-workers (Active). The Specialized Knowledge variable is an indicator
that equals one if the participants self-reported securitization knowledge is above the sample
median (Higher knowledge), and zero if it is below the median (Lower Knowledge). The
Decision Authority variable is manipulated by telling participants that the advice-seeker has been
directed to either document and take actions on the basis of the auditors recommendation (High
decision authority) or to simply seek the auditors opinion and make their own decision (Low
decision authority).

Panel A reports the linear contrasts used to test our hypotheses and the partial effects implicitly
tested by the contrast. Full support for H1a, H1b, H2a, and H2b is indicated by significance for
the indicated linear contrast and for related partial effects. The F statistics for all hypothesis tests
for contrariness and precision are calculated by dividing the indicated mean squares by the error
mean squares from the ANOVA models in Table 2, Panel A and Panel B, respectively. The p-
values for our linear contrasts are two-tailed. The p-values for the partial effect tests are one-
tailed because directional predictions are made.

For completeness, Panel B reports simple effect tests for contrariness and precision that are not
tested by our linear contrasts. All p-values for these simple effects in Panel B are two-tailed.

44
Table 3 (continued)
Hypothesis Tests
Panel A: Hypothesis Tests
Contrariness Precision

Hypothesis 1
Linear Contrasts for Status Motive by Specialized Knowledge Interaction

H1a H1b
+3 for Active Status Motives, Lower Specialized +3 for Active Status Motives, Higher
Knowledge, -1 otherwise Specialized Knowledge, -1 otherwise
MS F 1,94 p MS F 1,94 P
14.92 9.09 0.003 18.98 9.70 0.002

Partial Effect Tests of +3 Weighted Cell Mean Versus:


F1,94 p F1,94 p
Active Status Motives, 7.42 0.004 Active Status motives, 6.07 0.008
Higher Knowledge Lower knowledge
Non-Active Status Motives, 3.17 0.039 Non-Active Status Motives, 8.72 0.002
Lower Knowledge Lower knowledge
Non-Active Status Motives, 8.53 0.002 Non-Active Status Motives, 2.25 0.068
Higher Knowledge Higher knowledge

Hypothesis 2
Linear Contrasts for Status Motive by Decision Authority Interaction

H2a H2b
+3 for Active Status Motives, Low Decision Authority, - +3 for Active Status Motives, Low Decision
1 otherwise Authority, -1 otherwise
MS F 1,94 P MS F 1,94 P
3.52 2.15 0.146 9.26 4.73 0.032

Partial Effect Tests of +3 Weighted Cell Mean Versus:


F1,94 p F1,94 p
Active Status Motives, 0.93 0.338 Active Status motives, 3.05 0.042
Higher Decision Authority Higher Decision Authority
Non-Active Status Motives, 0.91 0.343 Non-Active Status Motives, 6.33 0.014
Higher Decision Authority Higher Decision Authority
Non-Active Status Motives, 3.24 0.038 Non-Active Status Motives, 0.54 0.233
Lower Decision Authority Lower Decision Authority

45
Panel B: Simple Effects Not Tested By Hypotheses

Status Motives by Specialized Knowledge (Collapsed Across Decision Authority)


Corresponding to Cell Means Reported in Table 1, Panel A and Graphed in Figure 1
Contrariness Precision
Source F 1,94 p Source F 1,94 P
Status motives | Higher 1.00 0.320 Status motives | Lower 0.99 0.322
knowledge knowledge
Knowledge | Not active 1.39 0.241 Knowledge | Not active 2.13 0.148
status motives status motives

Status Motives by Decision Authority (Collapsed Across Knowledge)


Corresponding to Cell Means Reported in Table 1, Panel B and Graphed in Figure 2
Contrariness Precision
Source F 1,94 p Source F P
1,94
Status motives | Higher < 0.01 0.954 Status motives | Higher 1.41 0.239
authority authority
Authority | Not active 0.76 0.387 Authority | Not active 1.57 0.213
status motives status motives

46
Figure 1
Contrariness and Precision by Status Motives and Specialization

This figure depicts graphed cell means for the contrariness and precision measures. The
dependent measure in Panel A is contrariness, which is the signed difference between
participants recommended rate and the rate of the advice-seeker. Higher values indicate more
contrariness. The dependent measure in Panel B is precision, which is the absolute difference
between the size of each participants range of reasonable discount rates and the largest range
size in our sample. Higher values indicate more precision. The Status Motives variable is
manipulated with a 600 word prime prior to the task that either describes looking for and finding
lost concert tickets (Not Active) or starting a prestigious new job and competing with co-workers
(Active). The Specialized Knowledge variable is an indicator that equals one if the participants
self-reported securitization knowledge is above the sample median (Higher knowledge), and zero
if it is below the median (Lower Knowledge).

47
Figure 1 (continued)
Contrariness and Precision by Status Motives and Specialization
(Collapsed Across Decision Authority)

Panel A: Contrariness

2.19
Contrariness

1.53 Lower specialized


knowledge
Higher specialized
1.14 knowledge
1.08

Not Active Active


Status Motives

Panel B: Precision

4.50
Precision

3.91 Lower specialized


knowledge
Higher specialized
knowledge
3.51
3.34

Not Active Active


Status Motives

48
Figure 2
Contrariness and Precision by Status Motives and Decision Authority
(Collapsed Across Specialized Knowledge)

This figure depicts graphed cell means for the contrariness and precision measures. The
dependent measure in Panel A is contrariness, which is the signed difference between
participants recommended rate and the rate of the advice-seeker. Higher values indicate more
contrariness. The dependent measure in Panel B is precision, which is the absolute difference
between the size of each participants range of reasonable discount rates and the largest range
size in our sample. Higher values indicate more precision. The Status Motives variable is
manipulated with a 600 word prime prior to the task that either describes looking for and finding
lost concert tickets (Not Active) or starting a prestigious new job and competing with co-workers
(Active). The Decision Authority variable is manipulated by telling participants that the advice-
seeker has been directed to either document and take actions on the basis of the auditors
recommendation (High decision authority) or to simply seek the auditors opinion and make their
own decision (Low decision authority).

49
Figure 2 (continued)
Contrariness and Precision by Status Motives and Decision Authority
(Collapsed Across Specialized Knowledge)

Panel A: Contrariness

1.90
Contrariness

Low decision
1.43 1.44 authority
High decision
1.14 authority

Not Active Active


Status Motives

Panel B: Precision

4.36
Precision

3.87 Low Decision


Authority
High Decision
3.65 Authority

3.37

Not Active Active


Status Motives

50
Figure 3
Range of Reasonable Discount Rates
(Collapsed Across Decision Authority)

This figure depicts the range of reasonable discount rates that participants recommended to
advice-seekers. The mean recommended upper bound and mean recommended lower bound of
reasonable discount rates are consultants recommendations of the highest and lowest reasonable
rates, respectively. Higher rates result in lower valuations, smaller gains, and are thus more
conservative. The Status Motives variable is manipulated with a 600 word prime prior to the task
that either describes looking for and finding lost concert tickets (Not Active) or starting a
prestigious new job and competing with co-workers (Active). The Specialized Knowledge
variable is an indicator that equals one if the participants self-reported securitization knowledge
is above the sample median (Higher specialized knowledge), and zero if it is below the median
(Lower specialized knowledge).

17.50
17.00
16.50 Recommended upper
Discount Rate

16.00 reasonable bound


15.50
Recommended lower
15.00
reasonable bound
14.50
14.00 Advice-seeker's initial
13.50 estimate
13.00 (constant at 13.20)
Status Status Status Status
Motives Motives Motives Motives
Not Active Not Active
Active Active
Lower specialized Higher specialized
knowledge knowledge

51

You might also like