THE ROLE OF TACIT KNOWLEDGE IN

AUDITOR EXPERTISE AND HUMAN CAPITAL DEVELOPMENT

JASMIJN C. BOL
Tulane University
jbol@tulane.edu

FRANK MOERS
Maastricht University
f.moers@maastrichtuniversity.nl

MARK E. PEECHER
University of Illinois at Urbana-Champaign
peecher@illinois.edu


April 2013




















Author Note: We thank Chris Agoglia, Bradley Bennett, William Kinney, Lisa Koonce, David Piercey,
Brian White, Michael Williamson and workshop participants at the Harvard Business School, University of
Illinois, University of MassachusettsAmherst, and the University of Texas at Austin for helpful
comments.

1
1. INTRODUCTION
In this paper, we investigate how tacit knowledge influences auditor expertise and
audit-firm human capital. Tacit knowledge is a form of procedural knowledge that helps
people better manage themselves, others, and their tasks (Wagner and Sternberg [1985];
Horvath and Sternberg [1999]; Hedlund and Sternberg [2000]). Tan and Libby [1997]
introduce tacit knowledge to the accounting literature by predicting and finding it to
significantly discriminate between auditors who receive top versus bottom annual
performance evaluations, but only if they are relatively experienced (i.e., this result holds
for managers, but not for seniors and staff). A prevailing interpretation of this result is
that tacit knowledge is a key determinant of expertise for experienced auditors, but not
for inexperienced auditors (e.g., Abdolmohammadi, Searfoss, and Shanteau [2004]).
By contrast, we predict and find that when inexperienced auditors immediate
supervisors perceive that they have higher tacit knowledge, these supervisors also believe
that their inexperienced auditors currently have long-term leadership potential.
Supervisors perceptions of their subordinates tacit knowledge therefore play a
previously undocumented, forward-looking role in inexperienced auditors human capital
development. In addition, we predict and find empirical support for three direct
organizational consequences of experienced auditors tacit knowledge. Specifically, we
show that experienced audit supervisors tacit knowledge: (1) improves development of
their subordinates tacit knowledge, (2) influences the relative value supervisors place on
subordinates tacit versus technical knowledge when assessing annual performance, and
(3) increases higher tacit knowledge subordinates commitment to the firm.
2
We use a mixed-method research approach to empirically test these four
predictions. We conduct a qualitative field study, collect proprietary archival
performance evaluation data, and survey employees all at a medium-sized audit firm in
Europe. Our survey includes the firms audit subordinates (staff and seniors), immediate
supervisors (managers), and top management (partners). As such, we base our analyses
on real-world pairings of audit supervisors and subordinates (e.g., Kennedy and Peecher
[1997]).
Our study makes several contributions to the accounting literature. It adds to the
audit expertise literature by providing evidence that tacit knowledge differentiates more
from less promising auditors considerably earlier in their careers than is indicated in the
extant accounting literature. We acknowledge that it may be years before tacit knowledge
systematically discriminates top from bottom auditors in terms of annual performance
evaluations, as in Tan and Libby [1997]. However, our findings suggest that this delay in
discrimination is partially attributable to behaviors and evaluations of supervisors who
lack tacit knowledge themselves. We find that supervisors with relatively low tacit
knowledge put more relative weight on actions and behaviors that signal technical
knowledge than actions and behaviors that signal tacit knowledge when evaluating their
subordinates annual performance. Thus, the lack of discriminatory power of annual
performance measures might be explained by the fact that some supervisors put relatively
little value on their subordinates tacit knowledge when evaluating performance.
These findings are important because identification of key capabilities early in
professionals careers facilitates managing the organizations human capital and
3
allocating its mentoring resources (Kram [1985]).
1
In addition, the forward-looking role
of inexperienced auditors current tacit knowledge suggests potential implications for
firms recruiting practices and for educators curriculum planning. Both practitioners and
educators may benefit from more carefully developing, measuring, and providing
feedback about novice auditors tacit knowledge.
Our study also contributes to the management accounting literature on human
capital development by being the first to identify direct organizational consequences of
audit supervisors tacit knowledge. Whereas Tan and Libby [1997] provide evidence that
managers themselves benefit from having higher tacit knowledge, i.e., they tend to
receive more favorable evaluations, they do not examine human-capital consequences at
the organizational level. Theory posits that people acquire tacit knowledge experientially,
by observing and imitating behaviors modeled by influential others (Brown Collins and
Duguid [1989]). Consequently, we predict and find that audit supervisors with higher
tacit knowledge better cultivate their subordinates tacit knowledge. An organizational
implication is that some audit subordinates those whose supervisors have lower tacit
knowledge likely are at a disadvantage in developing their tacit knowledge.
Our theory and findings also contribute to the performance evaluation literature in
accounting regarding supervisors use of discretion. Research has shown that supervisors
use of discretion in incentive contracting is influenced by their monitoring intensity
(Hppe and Moers [2011]), personal incentives (Bol [2011]), and by their personality and

1
Mentoring is a form of support that senior colleagues provide to ensure junior colleagues
personal/psychological growth and career development (e.g., Kraimer, et al. [2011]). Behaviors that fall
under the mentoring umbrella include sponsorship, coaching, managing introductions to key others, and
steering challenging work towards promising junior colleagues (Dreher and Ash [1990]; Lankau and
Scandura [2002]).
4
general attitudes toward compensation contracting (Bol, Hecht and Smith [2012]).
2
Our
study complements this literature stream by highlighting the importance of supervisors
own expertise. Prior studies indicate that supervisors with more favorable performance
evaluations themselves also have superior performance evaluation skills. For example,
higher rated supervisors evaluate with less partiality (Tan and Jamal [2001]) and provide
better feedback (Dreher and Ash [1990]). Our results show that supervisors skill sets not
only influence how supervisors evaluate but also what they evaluate and reward. We
demonstrate that supervisors with relatively high tacit knowledge tend to put relatively
more weight on performance dimensions that signal tacit knowledge versus technical
knowledge, compared to supervisors with relatively low tacit knowledge. While our
theory is silent regarding the optimal value supervisors should give to subordinates tacit
versus technical knowledge, this finding is important because subordinates generally
have at least short-term incentives to improve on those performance dimensions that are
valued by their immediate supervisors. As such, audit firms may want to rethink the
degree, as well as, the type of discretion that they provide based on how tacit knowledge
likely is distributed across supervisors.
Finally, our study contributes to the audit literature on expertise and task
performance by finding that higher tacit knowledge subordinates are more strongly
committed to the firm when their supervisors also have relatively high tacit knowledge.
While prior work suggests that professionals with greater firm commitment exert greater
effort on firm tasks and want to stay longer with their firm (Ferris and Larcker [1983];

2
Bol et al. [2012] show that supervisors willingness to make discretionary adjustments is affected by their
general attitudes toward a purely objective system, the controllability principle, and the attainability of
evaluation accuracy.
5
Patten [1995]; Brundage and Koziel [2010]), our study reveals how supervising auditors
tacit knowledge strengthens firm commitment of promising subordinates.
We organize the rest of this paper as follows. Section 2 reviews prior research and
develops the four hypotheses. We discuss the research setting and empirical design in
Section 3 and present the results in Section 4. Finally, Section 5 concludes and discusses
directions for future research.

2. LITERATURE REVIEW AND HYPOTHESIS DEVELOPMENT
As Nelson and Tan [2005, 49] observe, audit expertise research is important
because we want to know why and how superior performance is attained, and the nature
of the transition from a novice to an expert (also see, e.g., Davis and Solomon [1989];
Libby and Luft [1989]). Similarly, to sustain their competitive advantage, audit firms
want to identify and develop talented experienced auditors as well as retain expert
auditors (e.g.,Grant [1996]; Hatch and Dyer [2004]; Brundage and Koziel [2010]).
Although prior research indicates tacit knowledge in experienced auditors has value for
audit firms (Tan and Libby [1997]), little empirical evidence suggests tacit knowledge in
inexperienced auditors has value to audit firms (e.g., Shankar and Tan [2006]). In
addition, no study of which we are aware tests predictions about the organizational
consequences of higher versus lower tacit knowledge levels among experienced and
inexperienced employees. Our study helps fill this void by empirically testing hypotheses
about the role of tacit knowledge in the identification of talent and in the development,
evaluation and retention of inexperienced auditors.

6
2.1 Tacit Knowledge
Tacit knowledge is a form of procedural knowledge. Procedural knowledge is a
portfolio of skills and strategies that enables people to know how to perform or to apply
their declarative knowledge, i.e., knowledge of facts about the world (Anderson [1982]).
People typically acquire tacit knowledge from experience rather than from instruction
and training (Polanyi [1966]; Henlund and Sternberg [2000]). In business settings, tacit
managerial knowledge helps people manage themselves, others, and diverse tasks
encountered in ones career (Wagner and Sternberg [1985]; Horvath and Sternberg
[1999]; Hedlund and Sternberg [2000]).
Tacit knowledge about managing self pertains to managing ones self-motivation,
self-organization, and emotions, both on a daily basis and during high stress times, so as
to prioritize and attain ones important goals, such as productivity. Tacit knowledge about
managing others refers to knowledge about how to navigate social structures and
complex interpersonal relationships, including how to interpret and handle nonverbal
hints and emotional information. Tacit knowledge about managing tasks refers to
knowledge about how to plan, alter, and evaluate the success of projects as well as about
how to convince superiors and others of the worth of one's ideas or products (Wagner and
Sternberg 1985; Henlund and Sternberg [2000]).
3


3
The relation between tacit knowledge and other kinds of knowledge is the subject of considerable
research. Tacit knowledge differs from technical knowledge and problem-solving ability (Tan and Libby
[1997]) as well as from big five personality variables (Hedlund and Sternberg [2000]). In addition, tacit
knowledge tends to be uncorrelated or slightly negatively correlated with academic intelligence measures,
such as raw verbal reasoning or analytical reasoning ability (Hedlund and Sternberg [2000]; Grigerenko
and Sternberg [2001]; Sternberg et al. [2001]). The relation between tacit knowledge and emotional
intelligence, however, is a matter of debate. This debate arises in part because emotional intelligence has
been defined both very broadly to include all facets of intelligence other than IQ (see Goleman [1995]), and
more narrowly to be ones ability to recognize, understand, and regulate emotions in others and in oneself
(Mayer et al. [2000]). The latter definition conceptually seems to overlap with tacit knowledge, but we are

7
Tacit knowledge is firm and industry specific and ultimately is determined by top
management (Matthew and Sternberg, [2009]). Although there likely is considerable
overlap between firms especially in related industries, leaders of each specific firms will
have their own view on the best ways to manage oneself, others and task performance.
This, however, does not mean that newly hired auditors must have impoverished tacit
knowledge. To the contrary, because of variation in learning opportunities while still a
student, e.g., internships, case competitions, office visits, leadership roles in student
organizations, new auditors likely already vary considerably in terms of their tacit
knowledge for particular organizations.
Tan and Libby [1997] are the first to examine tacit knowledge in auditing. They
show that experienced managers with top performance evaluations are distinguished by
their tacit knowledge, while tacit knowledge is unrelated to whether inexperienced
auditors receive top or bottom performance evaluations. These results, however, still
allow for the tacit knowledge levels possessed by many relatively inexperienced auditors
to be substantial. Along these lines, Exhibit 1 recasts basic findings in Tan and Libby
[1997] so as to better emphasize both mean differences and variation in auditor tacit
knowledge by rank. As shown, both bottom-ranked and top-ranked staff auditors fairly
often possess tacit knowledge levels rivaling that of top-ranked managers. This provides
motivation for examining whether audit firms value these relatively high levels of tacit
knowledge possessed by this subset of inexperienced auditors.


unaware of robust empirical evidence on their relation. Hedlund and Sternberg [2000, 158] stress that the
decision to call certain kinds of intelligence as social intelligence, practical intelligence, emotional
intelligence or tacit knowledge ultimately should be borne out by empirically testing them and by
conceptually integrating them into a unified model to avoid a proliferation of intelligences.
8
2.2 Identification of Potential Future Leaders
Prior research on auditor expertise indicates that different kinds of knowledge and
skills enable performance on the portfolios of tasks performed by inexperienced versus
experienced auditors (Tan [1999]; Abdolmohammadi and Shanteau [1992];
Abdolmohammadi, Searfoss, and Shanteau [2004]). Related, extant analyses of diverse
audit tasks focus on the cognitive abilities and technical knowledge requirements for
successful performance (e.g., Bonner and Pennington 1991; Bonner and Walker 1994;
Bonner 1999), but they do not focus on important social exchanges typically occurring
before, during, or shortly after performance of such tasks.
4

Because social exchanges pervade in auditing contexts, we believe that superior
tacit knowledge improves task performance and makes favorable impressions for
experienced and inexperienced auditors alike. To elaborate, even though it may take
years before auditors annual performance evaluations begin to qualitatively differ
systematically across auditors who have relatively high or low tacit knowledge for their
given rank (Tan and Libby [1997]), we predict that even lower-ranked auditors can still
distinguish themselves as potential leaders if they have relatively high tacit knowledge
for their rank.
The theoretical appeal of this prediction is strengthened by focusing on the
aforementioned, easily overlooked findings in Tan and Libby [1997]. Because
meaningful differences in tacit knowledge exist not only across hierarchical ranks, but

4
Some common contexts featuring rich social exchanges are informal and formal consultations (e.g.,
Kadous, Leiby and Peecher [2012], the audit review process (e.g., Brazel, Agoglia, and Hatfield [2004]),
and audit team brainstorming (or similar) sessions (e.g., Lynch, Murthy, and Engle [2009]). We leave the
task of overlaying typical, particular social dynamics onto various audit tasks to future research. Here, we
simply argue that rich social interactions occur within and across many audit tasks, and that this matters for
auditors of all ranks.
9
also within particular ranks (see also e.g., Sternberg and Wagner [1985]; [1987]), with
some auditors at low ranks nearing expert levels of tacit knowledge, some auditors likely
begin to emerge as potential leaders early in their careers.
Focusing on variation in inexperienced auditors tacit knowledge also recognizes
that individuals are thought to acquire tacit knowledge with different speeds. And that,
the early, proactive acquisition of tacit knowledge is likely a means to signal ones
potential as a future leader of an organization. As such, we predict that supervisors use
subordinates tacit knowledge as a predictor of the subordinates potential for being
future leaders of their firms. This leads to the following hypothesis:
H1: Subordinates who are perceived to have relatively high tacit knowledge are
more likely to be identified as future leaders of their organization.


The results in Tan and Libby [1997] also show significant variation in
supervisors (managers) tacit knowledge. As depicted in Exhibit 1, many managers
possess tacit knowledge levels that are no higher or even lower than that of inexperienced
auditors. This is potentially problematic for the organization for two reasons. If H1 holds,
it would support the idea that supervisors identify talent based on their idiosyncratic
perceptions of their subordinates tacit knowledge. Supervisors perceptions of their
subordinates promise and tacit knowledge more likely align with views of top
management when supervisors themselves have relatively high tacit knowledge. In
contrast, when supervisors have relatively low tacit knowledge, their views on talent are
less likely to align. As a result, these supervisors can jeopardize the firms competitive
advantage by mistakenly judging their subordinates tacit knowledge as high, or perhaps
10
more importantly, not recognizing the talent and high levels of tacit knowledge of some
of their subordinates. These mistakes can result in misallocation of mentoring resources.
Second, since tacit knowledge includes the ability to manage others (Polanyi
[1966]; Wagner and Sternberg [1985]; Horvath and Sternberg [1999]), the variation in
supervisors tacit knowledge will impact the effectiveness of their interactions with their
subordinates. Specifically, we argue that supervisors tacit knowledge will influence how
supervisors evaluate performance and affect the development of subordinates tacit
knowledge, as well as, their commitment to the firm. Hence, supervisors with relatively
low tacit knowledge can negatively affect human capital development. We examine this
issue in more detail below.
2.3 The Effect of Supervisors Tacit Knowledge on Subordinates Development
As Bonner [2008, 70] observes, it is important to understand factors that affect
auditors development of different kinds of knowledge relevant to task performance
(Ferris and Larcker [1983]; Davis and Solomon [1989]; Libby [1989]; Libby and Luft
[1993]). Shankar and Tan [2006] show that higher tacit knowledge enables staff and
senior auditors to develop stronger justifications for their conclusions on an ill-structured
task, which is a key performance benchmark on such tasks (e.g., Kennedy, Kleinmuntz,
and Peecher 1997). Thus, it is important to improve our understanding of inexperienced
auditors tacit knowledge development.
Since the development of tacit knowledge occurs through an interactive
socialization process of observation, imitation, and practice, we argue that the levels of
tacit knowledge of the experts with whom subordinates regularly interact influence
subordinates development of tacit knowledge. Specifically, because supervisors interact
11
with their subordinates on a daily basis and they are trusted, powerful and well-connected
individuals, subordinates are likely to mimic their behavior and model themselves after
them (Brown Collins and Duguid [1989]; Swap et al. [2001]). Immediate supervisors
therefore play a key role in subordinates tacit knowledge development.
5
When the
immediate supervisor has relatively high levels of tacit knowledge, the subordinate will
likely develop higher levels of tacit knowledge. If the immediate supervisor has relatively
low levels of tacit knowledge, however, the subordinate likely will acquire lower levels
of tacit knowledge by developing similar bad habits.
6

Moreover, since interactions with supervisors moderate the transmission of
knowledge to and personal learning of subordinates (Lankau and Scandura [2002]; Vera-
Munoz, et al. [2006]), supervisors who are better at managing others, i.e., those with
higher tacit knowledge, likely better coach and develop their subordinates. For example,
they are likely to better provide both informal and formal feedback to subordinates
regarding matters such as their preliminary plans, interacting with client management,
and forming judgment calls (Dreher and Ash [1990]). Consequently, we argue that high
tacit knowledge supervisors superior ability to manage others will improve subordinates

5
Public accounting firms often have formal mentors assigned to inexperienced auditors. However, although
active mentoring is an effective way to transfer tacit knowledge (e.g., Mullen [1994]; Wilson and Elman
[1990]), having an official mentor frequently differs from having a day-to-day supervisor as these mentors
often do not regularly interact with their mentee auditors. In addition, it is likely that greater variation in
tacit knowledge exists across day-to-day supervisors than across mentors, as mentors are specifically
chosen to be a role model and often rank two or more hierarchical levels above the persons they mentor.
6
One might object by asking why would a subordinate model themselves after supervisors with relatively
low tacit knowledge? The answer is because many subordinates will not be equipped to differentiate their
supervisors tacit knowledge and even supervisors with relatively low tacit knowledge still have authority
and power over their subordinates evaluations and compensation. Consequently, unless their own tacit
knowledge is already quite high, subordinates will tend to model themselves after their superior,
independent of whether the superior has relatively high or low tacit knowledge. Moreover, the knowledge
transfer through social interactions often happens even in the absence of any reinforcements and whether or
not the individual intends to learn, subordinates might therefore not even be conscious of their mimicking
behavior (Bandura [1977]; Anderson [1995]).
12
development of tacit knowledge. In sum, because acquiring tacit knowledge is an
interactive socialization process, the tacit knowledge possessed by the subordinates
immediate supervisor positively influences the subordinates' tacit knowledge.
7

H2: The tacit knowledge of a subordinates immediate supervisor positively affects
the subordinates tacit knowledge.


2.4 The Effect of Supervisors Tacit Knowledge on Performance Evaluation
As discussed above, Shankar and Tan [2006] examine the joint influence of tacit
knowledge and technical knowledge on the task performance of staff and senior auditors
and find that tacit knowledge improves third party experts assessments of staff and
senior auditors justifications of their conclusions.
8
Thus, even though evidence in Tan
and Libby [1997] provides no reason to conclude that tacit knowledge systematically
affects how favorably inexperienced auditors overall annual performance is evaluated,
theory-consistent evidence in Shankar and Tan [2006] indicates the value of tacit
knowledge for performance at more junior levels.
9
The theory discussed and tested in this
study helps reconcile this apparent disconnect. In particular, we propose that the absence

7
Our theory does not specify a single direction of causality. In some instances subordinates will possess
enough tacit knowledge to understand the importance of proactively seeking out learning opportunities or
supervisors with relatively high tacit knowledge will seek out talented subordinates, i.e., those with
relatively high tacit knowledge. In other instances, relatively low tacit knowledge subordinates happen to
be at the right place at the right time and fortuitously grow their tacit knowledge.
8
In their experimental task subordinates first assess whether or not to require an additional allowance for
doubtful accounts and then learn of their supervising managers preliminary conclusion. Subordinates are
held accountable by being told their managers will review their judgments and supporting justifications.
The supervising managers conclusion is manipulated between-subjects to either agree or disagree with the
subordinates. When supervisors agree with them, higher tacit knowledge subordinates more frequently
emphasize consistent and downplay inconsistent evidence. When supervisors disagree, higher tacit
knowledge subordinates more frequently list pros and cons of alternative conclusions and cover a greater
breadth of issues, but only if they also possess relatively high levels of technical knowledge. These
strategic modifications to their justifications improved how favorably third party experts evaluated high
tacit knowledge subordinates task performance.
9
Shankar and Tan [2006] is the only accounting study of which we are aware that examines the influence
of tacit knowledge on relatively inexperienced professionals task performance.
13
of a significant association between staff and senior auditors tacit knowledge and their
annual performance evaluation is largely non-diagnostic as to whether or not tacit
knowledge can be viewed as valuable and important for inexperienced auditors.
We argue that because of the importance of social interactions in the audit
profession at all levels, tacit knowledge is expected to improve performance at all levels.
However, since some low-level supervising auditors have relatively low tacit knowledge
(see exhibit one taken from Tan and Libby [1997]), these supervisors probably have little
understanding of the importance of tacit knowledge, especially for auditors at lower
organizational ranks. Lacking tacit knowledge themselves, they are unaware of their own
incompetence when it comes to appropriately valuing subordinates tacit knowledge (cf.
Argyris [1990]). By contrast, supervisors who possess relatively high levels of tacit
knowledge will value tacit knowledge in their subordinates as they realize the importance
of tacit knowledge in the audit profession at all organizational levels.
Supervisors own understanding of what are the relevant skills for subordinates to
develop and demonstrate will influence performance evaluation ratings if supervisors are
provided with discretion in the evaluation and rewarding process.
10
When applying
discretion supervisors will attribute more value to knowledge, actions and/or behaviors
that better align with their own understanding of what constitutes a contribution to firm
value. We posit that for supervisors who have relatively high tacit knowledge this means
that they will put relatively more value on the subordinates tacit knowledge, while
supervisors with relatively low tacit knowledge will put relatively more value on

10
Prior literature, e.g. Woods [2012], confirms that supervisor discretion is very common in the
performance evaluation of auditors.
14
subordinates technical abilities.
11
Hence, conceivably it is because some supervisors give
little value to tacit knowledge in their immediate subordinates that empirical evidence
does not necessarily show an association between subordinates current performance
evaluation rating and their tacit knowledge levels.
12

In sum, we predict that supervisors with relatively high tacit knowledge
understand the importance of tacit knowledge at all ranks and will therefore relatively
strongly value subordinates tacit knowledge compared to the subordinates technical
ability and this will be reflected in the subordinates performance ratings through the
supervisors discretion. Supervisors with relatively low tacit knowledge, on the other
hand, will not understand the value of tacit knowledge for current performance at lower
ranks and will consequently put relatively little value on tacit knowledge, while putting
relatively more value on technical skills when applying discretion in the evaluation
process.
H3: Supervisors with relatively high tacit knowledge put greater relative value on
tacit versus technical knowledge in evaluating subordinates annual
performance compared to supervisors with relatively low tacit knowledge.


2.5 The Effect of Supervisors Tacit Knowledge on Subordinates Firm Commitment
A well-known problem for audit firms is that they invest enormous resources in
training and grooming their professionals only to have many (and sometimes their most

11
In our particular setting the compensation system uses subjective weights. Hence, the supervisors value
assessments are reflected in the subjective weightings. Our theory, however, generalizes to all forms of
discretion as they can all be used to highlight tacit versus technical knowledge.
12
One might wonder why tacit knowledge at higher levels of the organization is positively associated with
performance ratings if some supervisors put relatively little weight on it. First, at higher organizational
levels there are more output measures available and the organization is therefore not as depended on
subjective assessment from supervisors. Moreover, at higher levels of the organization supervisors with
lower tacit knowledge have likely been weeded out and hence the limited subjective assessments will
likely be more calibrated.
15
promising) leave the firm (Patten [1995]; Vera-Munoz, Ho and Chow 2006]). Turnover
rates in larger audit firms historically have been high, with recent estimates ranging from
20% to 28% per year (Satava [2003]; Brundage and Koziel [2010]). Such turnover is
costly and drains firms client-specific technical and tacit knowledge (Francis and Yu
[2009]; Brundage and Koziel [2010]).
13

We argue that subordinates immediate supervisors tacit knowledge likely plays
a key role in managing the organizations investment in human capital by moderating
promising subordinates commitment to the audit firm. Inexperienced auditors who
possess relatively high tacit knowledge and who are mentored, cultivated and evaluated
by supervisors with relatively high tacit knowledge likely more strongly embed their
professional identity with the firm (Vera-Munoz et al. [2006]), which will make them
more committed to the firm. That is, inexperienced auditors with relatively high tacit
knowledge are more likely to appreciate and benefit from a high quality role models and
better feedback, which are more likely supplied by supervisors with relatively high tacit
knowledge (see H2). The resulting organizational commitment will enhance their task
performance and make them more likely to stay with the firm (Ferris and Larcker [1983];
Meyer, Stanley, and Vandenberg [2012]).
14

On the other hand, inexperienced, promising auditors who are mentored,
cultivated and evaluated by supervisors with relatively low tacit knowledge will probably

13
Archival data indicate that over half of the auditors who leave audit firms do so despite the firms desire
to keep them for a longer period of time (Rhode, Sorensen, and Lawler [1977]).
14
Ferris and Larcker [1983] provide a theoretical model and empirical evidence that greater organizational
commitment improves auditor task performance. Conceptually, the Ferris and Larcker [1983] model
includes a broader set of determinants of auditor performance (i.e., motivation, task related knowledge and
ability, and organizational commitment) than more recent auditor expertise models, which focus on
auditors knowledge and abilities (Libby and Luft [1989]). Libby and Luft [1989] and related empirical
studies (e.g., Bonner and Lewis [1990]; Tan and Libby [1997]) refine the kinds of knowledge and abilities
that are important to auditors of different experience levels and areas of specialization.
16
be directed by their supervisor to focus on the technical and more mundane aspects of the
profession, which will hamper their personal growth and motivate them to consider
looking elsewhere for such growth. By contrast, firm commitment among inexperienced
auditors with relatively low tacit knowledge likely is less sensitive to their immediate
supervisors tacit knowledge, and supervisors with high tacit knowledge likely devote
relatively little of their scarce mentoring resources on lower tacit knowledge
subordinates. In sum, we predict that relatively high tacit knowledge subordinates will be
more committed to the firm and more inclined to stay when working under relatively high
versus low tacit knowledge supervisors. Formally:
H4: Firm commitment is stronger for subordinates with relatively high tacit
knowledge when their immediate supervisors also possess relatively high tacit
knowledge.


3. RESEARCH DESIGN
3.1 The Company and Research Methods
Our research setting is a mid-sized regional European audit firm with 4 offices
and over 275 employees. The company is a strong player in their regional market and
services small, mid-size and larger national and international organizations.
In order to get an in-depth understanding of the human capital development and
evaluation processes at the firm, and allow for quantitative analysis, we used a mix-
method research approach. We gathered information for our research by: (1) conducting a
qualitative field study, (2) collecting proprietary archival data and (3) conducting an
extensive survey, details of which are provided below.
We gathered the qualitative field study data in two main ways. First, we carried
out eight in-depth semi-structured interviews. Specifically, we interviewed three partners,
17
one manager, three subordinates at different job-levels, and the head of the HR
department. The semi-structured interviews lasted about an hour and were aimed at
getting a general idea of their human capital development and evaluation processes.
Second, we studied internal documents, including the descriptions of the evaluation
systems, which allowed us to better understand the structure of the organization and the
internal procedures. The qualitative field study, did not only help us get a deep
understanding of our setting, it was also instrumental in the design of our questionnaire
and the choice of our proxies. For example, we felt comfortable asking supervisors to
assess the potential of their subordinates to become a manager and partner as we learned
from the interviews that the firm keeps track of their most talented employees.
To gather quantitative archival data, we hand-collected the personnel records of
all accounting related subordinates in the firm.
15
This process resulted in 92 performance
documents. The performance documents provide information on the subordinates
position, department, and the subordinates immediate supervisor. The performance
documents also indicate the performance measures, performance ratings, and the
subordinates yearly compensation. Moreover, the HR department provided us with
demographic information such as job tenure, firm tenure, age and contact information.
Based on these documents, we were able to establish 101 subordinate-supervisor/partner
dyads.
In order to capture the subordinates and immediate supervisors tacit knowledge,
as well as, get more detailed information on subordinates firm commitment, we collected
quantitative survey data. We approached all accounting related employees of the firm

15
Administrative personnel like secretaries and HR employees were not included. Moreover, the firm also
has a significant tax and corporate finance department that was not included
18
with the request to participate in our survey. We conducted two separate surveys, one for
the supervisors (managers and partners), in which they were asked to fill out the tacit
knowledge scale and assess the potential of their subordinates, and one for the
subordinates, which also included the tacit knowledge scale and asked questions about
their involvement and intentions to leave the organization. We designed the surveys using
the online tool Qualtrics, and sent all subordinates, supervisors, and partners an email
with a link to the survey. In the introduction to the survey, as well as in the email, we
informed participants that each participant will receive a movie voucher for 12,50 and
that we will randomly select one participant to receive an Apple iPad. The HR manager
of the firm also emailed all potential participants to indicate the companys support for
the research study and to encourage participation. After two weeks we sent out a follow
up email to thank those who had participated and to encourage those who had not. Out of
the 101 subordinates that we contacted based on our subordinate-supervisor/partner
dyads, 94 (93%) participated, while 18 of the 20 (90%) contacted partners and
supervisors filled out the survey, which allow a link to 99 subordinates. An overview of
the observations in each separate dataset, as well as a description of the different samples
used in this paper is provided in Table 1.
3.2 Performance Evaluation and Rewarding System
The organizations compensation system consists of five different dimensions:
Goal Achievement, Technical Performance, Competencies, Training, and Leadership.
Whether or not all these dimensions are used in the annual evaluation depends on the job-
level of the subordinate. For example, until a subordinate reaches a supervisory level the
leadership dimension is not part of the evaluation. However, the technical performance
19
and competencies dimensions apply to all employees. Each dimension consists of several
performance measures that are specifically chosen for each subordinate and that are
agreed upon at the beginning of the contracting period. At the end of the contracting
period, the supervisor evaluates the subordinates performance on each relevant
dimension on a scale from 1 to 4 with the extreme anchors being unsatisfactory (1) and
excellent (4). The system uses subjective weighting, that is, the supervisors are provided
with the discretion to compose the final score based on the scores of the relevant
dimensions, instead of the dimension scores being combined using a pre-specified
formula. We collected information on the subordinates rating on each individual
performance measure, the different dimensions, and the final total score.
3.3 Dependent, Independent and Control Variables
In our test of H1, we examine whether tacit knowledge is a good indicator of
future leadership potential. We capture leadership potential by asking each supervisor to
indicate whether each of their subordinates a) has the potential to become a manager
(Potential Manager) and b) has the potential to become a partner (Potential Partner).
The main independent variable in our first analysis is the supervisors perception
of the subordinates tacit knowledge. We capture subordinates tacit knowledge by using
Tan and Libbys [1997] modified tacit knowledge scale.
16
This scale presents ten work-
related scenarios and nine to eleven options associated with each scenario. The

16
This scale was originally developed by Wagner and Sternberg [1985] but modified by Tan and Libby
[1994] to fit the instrument to the public accounting context. The ten work-related scenarios related to (1)
activities which lead to promotion from manager to partner, (2) the importance of different tasks to audit
firm management, (3) activities which lead to success in public accounting, (4) proper responses to a
personnel problem, (5) motivations for pursuing a public accounting career, (6) personal characteristics of
successful public accountants, (7) personal work strategies of successful managers, (8) components of a
good reputation as a manager, (9) experiences relevant to becoming a good manager, and (10) the
importance of candidate attributes in employment selection for entry-level positions in public accounting.
For more information see Tan and Libby [1994].
20
subordinates are asked to rate each option on a seven-point scale, with the extreme
anchors being extremely unimportant and extremely important to success (see Appendix
1). The subordinates immediate supervisors also filled out the same scale and, following
Tan and Libby [1997], we measure subordinates tacit knowledge by calculating the
absolute difference between a subordinates importance ratings and the ratings of the
subordinates immediate supervisor. Higher consistency in the importance ratings
(smaller absolute difference) indicates higher levels of perceived tacit knowledge, using
the immediate supervisor as a benchmark. Finally, we take the highest observed score
minus the score of the subordinate, so that higher values indicate higher subordinate tacit
knowledge (Tacit Knowledge Subordinate Supervisor Benchmark).
Since we want to examine the effect of subordinates tacit knowledge considering
their experience level, we need to control for experience. We do this by including the
subordinates firm tenure, job level and age (Firm Tenure, Job Level, and Age). Further,
in order to examine whether tacit knowledge provides additional information on
subordinates potential to become a future leader beyond information already captured by
the performance evaluation process, we also include the subordinates final evaluation
score (Overall Evaluation Rating).
To test H2, we measure subordinates and supervisors tacit knowledge using
their partner as the benchmark. That is, we measure subordinates (supervisors) tacit
knowledge by calculating the absolute difference between a subordinates (supervisors)
importance ratings and the ratings of the subordinates (supervisors) immediate partner
on the tacit knowledge scale. We again rescale this measure such that higher values
21
indicate more subordinate (supervisor) tacit knowledge (Tacit Knowledge
Subordinate/Supervisor Partner Benchmark).
To test H3, we capture the value that supervisors ascribed to different knowledge
categories by analyzing the weights that supervisors give to the different behaviors and
actions of their subordinates in determining their annual performance evaluation scores.
Recall that supervisors evaluate subordinates on at least two and at most five different
dimensions, depending on their rank. For each dimension, supervisors must provide a
score from 1 to 4, where 4 is the best. In addition, the supervisor subjectively combines
these dimensions to reach an overall assessment of the subordinate (Overall Evaluation
Rating). We use the subordinates score on the technical performance dimension to proxy
for the subordinates technical skills (Technical Performance Rating) and the
performance rating on the competencies dimensions to capture the subordinates level of
tacit knowledge (Competencies Rating). The competencies dimension is broader than
tacit knowledge, nevertheless, we believe the performance rating is a good a priori proxy
for the subordinates tacit knowledge as it includes sub-dimensions such as customer
orientation, self-confidence, independence and willingness to learn from mistakes. These
sub-dimensions seem to be related to the ability to manage oneself, others and diverse
tasks encountered in ones career.
Finally, to examine H4, we capture the subordinates commitment to the firm by
using the following four items: (1) I see myself working at [the firm] in the long-run,
(2) Unless [the firm] actively undertakes steps, I consider leaving the company
(reverse-coded), (3) Even if I could do my job at another company while being paid
more or equal, I would stay at [the firm] because [the firm] offers me other advantages,
22
and (4) What is the chance that you will still work at [the firm] in 12 months?. We use
a fully-anchored 7-point scale specific to each item. Principal component analysis reveals
one factor with an Eigenvalue greater than 1, which explains 68% of the total variance.
The (absolute) factor loadings of the four items range from 0.76 to 0.90. We compute the
construct Firm Commitment by averaging the scores of the four items.
17


4. RESULTS
4.1 Descriptive Statistics
Tables 2 and 3 provide descriptive statistics and simple Pearson and Spearman
correlations for our variables. Several results are notable. One is the significant but not
perfect positive correlation (r = +0.56) between subordinates tacit knowledge, as
benchmarked by their immediate supervisors responses to the tacit knowledge
questionnaire and subordinates tacit knowledge as benchmarked by the questionnaire
responses of the partners who supervise these immediate supervisors. This indicates that
supervisors themselves, on average, do not have a perfect understanding of how the
firms top management thinks one should manage oneself, others and task performance.
However, the high correlation coefficient also suggests that immediate supervisors and
partners views on how one ought to respond to complex social situations are reasonably
aligned. In line with H2, the average subordinates tacit knowledge is significantly higher
when the immediate supervisor is used as a benchmark than when the partner is used as a
benchmark (i.e., 79.81 versus 73.04, p
(two-tailed)
= 0.001). Desirable or not, subordinates
ideas of how one ought to respond are more closely aligned with their immediate

17
Our results are robust to using the factor score as a measure of Firm Commitment.
23
supervisors views than with the views of their partners. As suggested before, the
difference between tacit knowledge as perceived by the supervisor and tacit knowledge
according to top management might be problematic. We leave it to future research to
examine whether this misperception has real organizational consequences. Finally, we
note that the overall performance evaluation ratings heavily gravitate to 3 (1
st
quartile,
median, and 3
rd
quartile all equal 3), consistent with research showing that performance
evaluations exhibit both a centrality bias and a leniency bias (Prendergrast [1999]; Moers
[2007]; Bol [2011]).
4.2 Test of Hypothesis 1: Identification of Talent
To test H1, which predicts that subordinates who have relatively high perceived
tacit knowledge are more likely to be identified as future leaders of their organization, we
estimate the following logit model:
( )

where Potential refers to either Potential manager or Potential partner and Tacit
knowledge subordinate is measured using the subordinates immediate supervisor as the
benchmark, as this supervisor is implicitly charged to identify whether or not specific
subordinates have long run potential. We run this model both with and without Overall
Evaluation Rating.
The results, presented in Table 4, show a consistent pattern regarding the
relationship between tacit knowledge of the subordinate and the probability of being
identified as a potential manager/partner. More specifically, consistent with H1 perceived
tacit knowledge is positively associated with the supervisors assessment of the
subordinates potential to become a manager/partner.
24
To examine whether the audit firms annual performance evaluations already fully
impound the forward-looking value of subordinates tacit knowledge, we include Overall
Evaluation Rating as a control. We find that the results for potential to become an audit
manager lose statistical significance when Overall Evaluation Rating is included. The
results for Potential Partner, however, are more robust. That is, tacit knowledge is
significantly positively associated with the potential to become a partner, irrespective of
whether Overall Evaluation Rating is included in the analysis. Consistent with H1, these
results imply that, as the tacit knowledge of the subordinate increases (using the
immediate supervisor as a benchmark), the supervisor is more likely to identify the
subordinate as a potential future leader. They also imply that annual performance
evaluations do not fully reflect the forward looking value of subordinates tacit
knowledge for their potential to become an audit partner.
Regarding the control variables, we find that both Job Level and Overall
Evaluation Rating are significantly positively associated with both Potential Manager
and Potential Partner. This indicates that subordinates higher up the hierarchy and/or
with a better performance evaluation are more likely to be identified as future potential
leaders. Further, Age has a significant negative effect on the probability to become a
manager, but not on the probability to become partner, while Firm Tenure is insignificant
in all models.
Although we find results consistent with H1, it could be that tacit knowledge is
seen as a general predictor of success, not solely as a forward-looking measure. If this is
true, the tacit knowledge of a subordinate should be positively associated with the
potential to make any promotion, not only to manager and/or partner. We therefore run an
25
additional analysis where we replace Potential Manager / Potential Partner by an
indicator variable that indicates one if the supervisor thinks the subordinate had potential
to be promoted beyond their current level and zero otherwise. Untabulated results show
that tacit knowledge does not in general predict promotions, which indicates that
supervisors recognize the importance for tacit knowledge for leadership positions but do
not perceive it as an indicator of success at the next hierarchical level in general.
18

In sum, we find support for H1; subordinates who are perceived to have relatively
high tacit knowledge are more likely to be identified as future leaders of their
organization.
4.3 Tests of Hypotheses 2: Development of Tacit Knowledge
To examine the hypothesized positive relation between the tacit knowledge of a
subordinates immediate supervisor and a subordinates tacit knowledge, we first test
whether the tacit knowledge of subordinates differs between two types of supervisors,
i.e., those with above-the-median tacit knowledge and those with below-the-median tacit
knowledge (using the partner as the benchmark). We limit our sample to supervisors and
subordinates who have been in their current job for more than a year, in order to ensure
that supervisors had the opportunity to impact their subordinates. This reduces the sample
to 12 supervisors and 51 subordinates for whom we have tacit knowledge data. Of these
51 subordinates, 23 (28) are associated with supervisors with above-the-median (below-
the-median) tacit knowledge. To assure that we compare individual subordinates at the
same job-level, we match subordinates of supervisors with above-the-median tacit

18
Note that these analyses are based on the supervisors perceptions of promotability. The fact that
supervisors, on average, do not in general perceive tacit knowledge to be an indicator of promotions might
be driven by some supervisors own low levels of tacit knowledge (see H3).
26
knowledge to subordinates of supervisors with below-the-median tacit knowledge based
on the subordinates job level. As a result, the final sample consists of 12 supervisors and
42 subordinates (21 matches).
We find that the mean tacit knowledge of subordinates (using the partner as a
benchmark) of relatively high tacit knowledge supervisors is significantly higher than
that of subordinates of relatively low tacit knowledge supervisors (78.14 vs. 67.33;
p=0.03, two-tailed). We find even stronger results when comparing the medians (76 vs.
63; p<0.01, two-tailed). These results are further corroborated by the Wilcoxon rank-sum
test, which shows that subordinates of relatively high tacit knowledge supervisors tend to
have relatively higher levels of tacit knowledge than those of relatively low tacit
knowledge supervisors (p=0.02, two-tailed). Finally, the Kolmogorov-Smirnov test
confirms that the two samples do not come from the same distribution (p=0.02).
Our second test takes a more continuous perspective and correlates the level of
tacit knowledge of supervisors with that of their subordinates. Given that all supervisors
have multiple subordinates, the number of which also varies across supervisors, we
calculate, for each supervisor separately, the median level of tacit knowledge of all of
his/her subordinates. We then correlate these medians with the tacit knowledge of the
supervisors (n=12). The result of this analysis shows that the correlation coefficient is
fairly high (=0.56) and significant (p=0.06, two-tailed), indicating a positive association.
Collectively, these results suggest that because tacit knowledge is acquired through an
interactive socialization process, immediate supervisors play a key role in subordinates
tacit knowledge development.
27
An alternative explanation for these results is that relatively high tacit knowledge
subordinates and/or supervisors self-select into working with one another. During our
field examination, however, it became apparent that self-selection was unlikely because
there were few opportunities for subordinates and supervisors to switch
supervisor/subordinate given the mid-size of the audit firm. To confirm this field
observation, we explore whether self-selection is evident in our survey data. If self-
selection were a key explanation for these results, we would expect to observe a positive
relationship between the tacit knowledge of subordinates and their immediate
supervisors, even for quite new subordinate-supervisor relationships. We have data for 12
subordinate-supervisor dyads that have worked together for a year or less. Of these 12
dyads, five (seven) have supervisors with below-median (above-median) tacit knowledge.
We observe no significant difference in the tacit knowledge of subordinates with
relatively high versus relatively low tacit knowledge supervisors (means: 74.00 and 76.2;
medians: 77 and 78, respectively). Furthermore, the correlation between the tacit
knowledge of these subordinates and their immediate supervisors is essentially zero (=-
0.02; p=0.95, two-tailed). Overall, we provide strong evidence in support of H2.
4.4 Tests of Hypotheses 3: Performance Evaluation
To test H3 we examine the impact of supervisors tacit knowledge on the weight
placed on different performance dimensions in subordinates performance evaluations.
Following Ittner, Larcker and Meyer (2003), we run a regression of the Overall
Evaluation Rating on the ratings of the two dimensions of performance that are included
28
for all subordinates and use the resulting regression coefficients as our estimates of the
subjective weights.
19

To examine the impact of supervisors tacit knowledge on subjective weights, we
allow the regression coefficients to vary between supervisors with below-the-median tacit
knowledge (HiTKMgr=0) and supervisors with above-the-median tacit knowledge
(HiTKMgr=1). As a result, we run the following regression specification:

Model (I) in Table 5 presents the results of the above specification. In particular,
the main effect of Technical Performance Rating and Competencies Rating, which
represents the subjective weight on these dimensions by below-the-median tacit
knowledge supervisors, is positive and significant, indicating that these supervisors
significantly weight these dimensions in their overall assessment. Our test of H3,
however, requires examination of how the coefficients on these dimensions change with
the supervisors tacit knowledge. The coefficient on the HiTKMgr*Technical
Performance Rating interaction is negative and significant, which indicates that
supervisors with relatively high tacit knowledge put less weight on the subordinates
technical performance than supervisors with relatively low tacit knowledge. The
interaction term HiTKMgr*Competencies Rating is positive, as predicted but not
significant.

19
When we run a regression of the Overall Evaluation Rating on the ratings on all five dimensions and take
the absence or presence of the other three dimensions into account by adding an indicator variable that
equals 1 (0) if the dimension is present (absent) we find similar results (untabulated).
29
Panel B of Table 5 is a more pointed test of H3. It examines whether relatively
high tacit knowledge supervisors and relatively low tacit knowledge supervisors have
significantly different relative weights for Technical Performance Rating and
Competencies Rating. Consistent with H3, relatively high tacit knowledge supervisors
place significantly greater weight on subordinates Competencies Rating than their
Technical Performance Rating (+0.66
tacit
versus -0.02
technical
, p < 0.01). By contrast,
relatively low tacit knowledge supervisors weight their subordinates Technical
Performance Rating as much or more than their Competencies Rating (+0.53
technical

versus +0.33
tacit
, n.s.).
To examine whether our results are driven by supervisors with relatively high
tacit knowledge supervising subordinates who happen to be at a higher job level, we
rerun our model after including Job Level and its interaction. The results, reported as
Model (II) in Table 5, show that our H3 results are not driven by job level. That is, the
added variables are all not significant, while the results for the variables of interest to H3
are in line with Model (I).
20

Overall the results clearly show a different pattern in performance evaluation
behavior between supervisors with below-the-median versus above-the-median tacit
knowledge. In particular, supervisors with below-the-median tacit knowledge put
relatively more weight on the dimension that provide an indication of technical
knowledge (Technical Performance Rating), and put relatively less weight on the

20
We also run our model after replacing HiTKMgr by Job Level. The (untabulated) results show that Job
Level and its interactions with the two performance dimensions are all not significant.
30
dimension that most likely is an indication of tacit knowledge (Competencies Rating).
21

Given that subordinates are expected to respond to and/or anticipate this behavior,
subordinates who are supervised by supervisors with below-the-median tacit knowledge
might misdirect their effort and underinvest in their tacit knowledge development.
4.5 Tests of Hypothesis 4: Firm Commitment
In H4, we predict that firm commitment is stronger for relatively high tacit
knowledge subordinates when their immediate supervisors also possess relatively high
tacit knowledge. To test this prediction, we regress subordinates firm commitment on
their tacit knowledge and allow the regression coefficient to vary between supervisors
with below-the-median tacit knowledge and supervisors with above-the-median tacit
knowledge. We further control for subordinates job level, firm tenure, age, and overall
evaluation rating. As a result, we run the following regression specification:

where Tacit Knowledge Subordinate is measured using the partner as the benchmark.
Model (I) in Panel A of Table 6 presents the results of the above specification. We
find that the main effect of Tacit Knowledge Subordinate is not significantly different
from zero, which implies that more promising auditors are on average as committed to
the firm as less promising ones when they are mentored by supervisors with below-the-

21
An assumption underlying this conclusion is that low tacit knowledge supervisors do not differ from high
tacit knowledge supervisor with respect to their technical knowledge. Although we have not included a
measure of technical knowledge in our survey, we do have, for the majority of the supervisors, archival
data on their Technical Performance Rating. Using this rating as a measure of technical knowledge, we
find that low and high tacit knowledge supervisors have (statistically) indistinguishable technical
knowledge, which corroborates the validity of our assumption and conclusion.
31
median tacit knowledge (see also Panel B of Table 6). Consistent with our prediction, the
coefficient on the HiTKMgr*Tacit Knowledge Subordinate interaction is positive and
significant, which indicates that subordinates with relatively high tacit knowledge are
more committed to the firm when mentored by a supervisor with above-the-median tacit
knowledge as compared to a supervisor with below-the-median tacit knowledge. Panel B
of Table 6 further corroborates this effect and shows that, if subordinates are mentored by
supervisors with above-the-median tacit knowledge, then the subordinates tacit
knowledge is significantly positively associated with firm commitment (p<0.01, two-
tailed). Regarding the control variables, we find that only Job Level has a significant
effect on firm commitment. In particular, the higher the subordinate is on the job ladder
the less likely s/he is to stay with the firm.
Given the significant effect of the subordinates job level on firm commitment,
we examine a sample of subordinates of supervisors with above-the-median tacit
knowledge matched to subordinates of supervisors with below-the-median tacit
knowledge based on the subordinates job level to better control for this effect. We run
the following specification on this sample

The result of this analysis, presented as Model (II) in Panel A of Table 6, leads to
the same inferences as before. We find that the main effect of Tacit Knowledge
Subordinate is not significant, while the effect of the HiTKMgr*Tacit Knowledge
Subordinate interaction is positive and significant. The main difference is that the results
become stronger in magnitude and significance. In sum, our results show that supervisors
32
with relatively high tacit knowledge are more able to commit and retain subordinates
with relatively high tacit knowledge, providing support for H 4.

5. CONCLUSION
In this study we investigate how tacit knowledge influences auditor expertise and
human capital development. In contrast to relatively common interpretations of findings
in Tan and Libby [1997], we posit that tacit knowledge already begins to differentiate
more from less promising auditors early in their career. Specifically, we predict and find
that inexperienced auditors perceived tacit knowledge levels significantly predict
whether or not they currently have perceived leadership potential. Tacit knowledge
therefore plays a previously undocumented, forward-looking role in inexperienced
auditors expertise development. In addition, we predict and find evidence supporting
three direct organizational consequences of supervisors tacit knowledge: (1) it improves
development of their subordinates tacit knowledge, (2) it increases the relative value
they place on subordinates tacit versus technical knowledge when assessing annual
performance, and (3) it increases the firm commitment of subordinates with relatively
high tacit knowledge. Collectively, these findings contribute to the audit expertise, human
capital development and incentive contracting literatures in accounting.
Our results show the pivotal role of supervisors own tacit knowledge in the
development of their subordinates tacit knowledge. Subordinates who work for
supervisors with relatively high tacit knowledge have, on average, higher levels of tacit
knowledge. Interestingly, these results do not hold when the relationship between the
subordinate and supervisor is less than a year, indicating that these findings are unlikely
33
driven by (self-)selection effects. These findings suggest that the proactive management
of tacit knowledge development by exposing the most talented employees to the
supervisors with relatively high tacit knowledge will likely bear fruit.
Our study also provides valuable insights for compensation system designers and
firms as it suggests there could be value in assessing the tacit knowledge of the
supervisors who have discretion in how they cultivate, evaluate and reward subordinates.
Our results show that the skill sets and knowledge of supervisors not only influence how
supervisors evaluate but also what they evaluate and reward. We provide empirical
evidence indicating that supervisors with relatively high tacit knowledge put significantly
more weight on behaviors and actions that signal tacit versus technical knowledge
compared to supervisors with relatively low tacit knowledge. Because subordinates are
motivated to develop those skills that their supervisors evaluate and reward, incentive
systems that are executed by supervisors with relatively low tacit knowledge provide
fewer incentives to develop tacit knowledge. Hence, our study suggests there could be
costs associated with endowing the same amount of discretion to relatively low and high
tacit knowledge supervisors.
After analyzing the data we returned to the firm to report our findings. Our
findings resonated with top management. They indicated that they had noticed that there
were some supervisors that, although technically strong, did not seem to have good
people skills (we confirmed that they were referring to the tacit knowledge concept).
The development of the subordinates that worked for these supervisors seemed not to be
as rapid as those that worked for stronger supervisors. There was especially one case
that came to mind; one supervisor who lacked people skills supervised two
34
subordinates that were both considered to be high potential. They mentioned that they
were worried that these two subordinates would leave the firm if nothing would change
as they seemed to be frustrated by the lack of learning opportunities. These conversations
with the firm confirmed our empirical findings. Future research could examine whether
and how firms deal with supervisors that have relatively low tacit knowledge.
As with all studies, this one has limitations. One limitation is that the tacit
knowledge scale used in this paper measures whether subordinates recognize actions that
represent higher tacit knowledge, not whether they are able or inclined to actually
perform them. Hence, consistent with earlier studies on tacit knowledge (e.g. Tan and
Libby [1997]; Sternberg and Wagner [1985]; [1987]), we assume that auditors who
recognize actions consistent with high tacit knowledge are more likely than other auditors
to perform these actions. Another limitation is that our data, though rich and gathered
from several research approaches, come from one mid-sized accounting firm in Europe.
This is a similar limitation that applies to all of the extant tacit knowledge research in
auditing of which we are aware. Tan and Libby [1997], for example, examine tacit
knowledge for an audit firm in Singapore. Finally, we do not presume that higher tacit
knowledge generally produces higher quality audits. On the one hand, greater auditor
tacit knowledge likely leads to better-justified decisions and to more skillfully handled
inquiries with client management that reveal helpful audit evidence. On the other hand,
greater tacit knowledge can provide auditors who are so inclined with greater ability to go
the extra mile to please client management and assuage audit committee member or
regulator concerns. Examining the audit quality implications, as well as broader societal
35
implications of auditors with higher levels of tacit knowledge, is a promising topic for
future research.
Another future research question raised by our study is whether relatively low
tacit knowledge supervisors pose incremental human capital risk to audit firms by
unwittingly failing to identify promising subordinates or by mistakenly identifying below
average subordinates as having substantial promise. Our results indicate that supervisors
use subordinates perceived tacit knowledge as a predictor of the subordinates potential
for being future leaders and that there is a difference between tacit knowledge as
perceived by the supervisor and tacit knowledge according to top management. Future
research could examine whether these misperceptions have real organizational
consequences.
This study also motivates additional work along the lines of Shankar and Tan
[2006], which is the only study of which we are aware to show how greater tacit
knowledge enables inexperienced auditors to more skillfully deploy their technical
knowledge to reach better-justified decisions. In particular, do lower tacit knowledge
supervisors fail to pick up on these tacit-knowledge-driven performance differences in
justification quality? Related, do they mistakenly attribute any performance
improvements to better technical knowledge and discount the importance of tacit
knowledge?
Finally, our study motivates developing a refined measure of tacit knowledge that
more convincingly applies to the financial-statement auditing context. Such a scale could,
for example, present auditors with an apparent impasse between client management and
an auditor and provide them with alternative negotiation strategies (e.g., Sanchez,
36
Agoglia and Hatfield [2007]), or with a complex estimate that may benefit from informal
and/or formal consultation with other auditors and provide them with alternative ways to
sequence the consultation process (e.g., Kadous, Leiby, and Peecher [2012]). It may be
that tacit knowledge would be shown to play an even more pivotal role for audit
expertise, regardless of auditor rank, if scholars were to refine the measure of the tacit
knowledge construct, as it pertains to the audit context.
37
REFERENCES
ABDOLMOHAMMADI, M.J.; D.G. SEARFOSS; AND J. SHANTEAU. An
Investigation of the Attributes of Top Industry Audit Specialists. Behavioral
Research in Accounting 16 (2004): 1-17.
ABDOLMOHAMMADI, M.J., AND J. SHANTEAU. Personal Attributes of Expert
Auditors. Organizational Behavior and Human Decision Processes 53 (1992): 158-
172.
ACCOUNTING AND CORPORATE REGULATORY AUTHORITY. Talent Attraction
and Retention in Larger Accounting Firms. August (2012).
http://www2.accaglobal.com/documents/talent_attraction.pdf
ARGYRIS, C. Overcoming organizational defenses: Facilitating organizational
learning. Boston, MA: Allyn and Bacon, 1990.
ANDERSON, J. R. Cognitive Psychology and Its Implications. 4th ed. New York: W.H.
Freeman and Company, 1995.
ANDERSON, J. R. Acquisition of Cognitive Skill Psychological Review 98, 4 (1982):
369-406.
BAILEY, W.J., G. HECHT, and K.L. TOWRY. 2011. Dividing the pie: The influence of
managerial discretion extent on bonus pool allocation. Forthcoming, Contemporary
Accounting Research.
BANDURA, A. Social Learning Theory. Englewood Cliffs, NJ: Prentice Hall, 1977.
BENABOU, C., and R. BENABOU. 1999. Establishing a Formal Mentoring Program of
Organization Success. National Productivity Review 18 (1999): 7-14.
BHAMORNSIRI, S., and R.E. GUINN. The Road to Partnership in the Big Six Firms:
Implications for Accounting Education. Issues in Accounting Education 6 (1991): 9-
24.
BOL, J. C. 2011. The Determinants and Performance Effects of Managers Performance
Evaluation Biases The Accounting Review 86(5): 1549-1575.
BONNER, S. A Model of the Effects of Audit Task Complexity. Accounting,
Organizations and Society 19 (1994): 213-234.


38
BONNER, S. E., and N. PENNINGTON. Cognitive Processes and Knowledge as
Determinants of Auditor Expertise. Journal of Accounting Literature 10 (1991): 1-
50.
BONNER, S. E., and P. L. WALKER. The Effects of Instruction and Experience on the
Acquisition of Auditing Knowledge. The Accounting Review 69 (1994): 157-178.
BONNER, S. E. Judgment and Decision-Making Research in Accounting. Accounting
Horizons 13 (1999): 385-398.
BONNER, S.E. Judgment and Decision-Making Research in Accounting. Upper Saddle
River, NJ: Pearson Prentice Hall, 2008.
BRAZEL, J. F., C. AGOGLIA, and R. C. HATFIELD. Electronic versus Face-to-Face
Review: The Effects of Alternative Forms of Review on Auditors Performance. The
Accounting Review 79,4 (2004): 949-966.
BROWN, J.S.; A. COLLINS; and P. DUGUID. Situated Cognition and the Culture of
Learning. Educational Researcher 18 (1989): 32-42.
BRUNDAGE, H. and M. KOZIEL. Retaining Top Talent Still a Requirement for Firms
Journal of Accountancy 209, 5 (2010): 38-44.
CAMPBELL, D. Nonfinancial Performance Measures and Promotion-Based
Incentives. Journal of Accounting Research 46 (2008): 297-332.
DAVENPORT, T. H. and D. E. SMITH. Managing Knowledge in Professional Service
Firms. In The Knowledge Management Yearbook 2000-2001, edited by J.W.
Cortada, and J. A. Woods. Reed Elsevier, 2000.
FERRIS, K. R. Organizational Commitment and Performance in a Professional
Accounting Firm Accounting Organizations and Society 6, 4 (1981): 317-325.
GIBBINS, M. Propositions about the psychology of professional judgment in public
accounting. Journal of Accounting Research 22 (1984): 103-125.
GIBBINS, M., and C. EMBY. Evidence on the nature of professional judgment in public
accounting. In Proceedings of the Auditing Research Symposium 1984, edited by
A.R. Abdel-khalik and I. Solomon. Champaign, Ill: College of Commerce and
Business Administration. 1985.
GIBBINS, M., and K. TROTMAN. Audit review: Managers' interpersonal expectations
and conduct of the review. Contemporary Accounting Research 19 (2002): 411-444.
39
GOLD, A., R. KNECHEL, and P. WALLAGE. The effect of the strictness of
consultation requirements on fraud consultation. The Accounting Review 87 (2012):
925-949.
GRANT, R. M. Toward a Knowledge-Based Theory of the Firm. Strategic
Management Journal 17 (1996): 109122.
GRIGORENKO, E. L. and R. J. STERNBERG. Analytical, creative, and practical
intelligence as predictors of self-reported adaptive functioning: a case study in Russia.
Intelligence 29, 1 (2012): 57-73.
HATCH, N. W. and J. H. DYER. Human Capital and Learning as a Source of
Sustainable Competitive Advantage. Strategic Management Journal 25 (2004):
1155-1178.
HEDLUND, J. and R. J. STERNBERG. 2000. Too many intelligences? Integrating
social, emotional, and practical intelligence. In The Handbook of Emotional
Intelligence: Theory, Development, Assessment, and Application at Home, School,
and in the Workplace, edited by R. Bar-On and J. D. A. Parker. John Wiley & Sons,
Inc.
KADOUS, K., J. LEIBY, and M. E. PEECHER. How Do Auditors Weight Informal
Advice? The Joint Influence of Advisor Social Bond and Advice Justifiability.
2012. Available at SSRN: http://dx.doi.org/10.2139/ssrn.1656070
KRAMER, M. L., S. E. SEIBERT, S. J. WAYNE, R. C. LIDEN, and J. BRAVO.
Antecedents and Outcomes of Organizational Support for Development: The Critical
Role of Career Opportunities. Journal of Applied Psychology 96, 3 (2011): 485-500.
LAM, A. Tacit Knowledge, Organizational Learning and Societal Institutions: An
Integrated Framework. Organization Studies 21 (2000): 487-513.
LANKAU, M. J., AND T. A. SCANDURA. An Investigation of Personal Learning in
Mentoring Relationships: Content, Antecedents, and Consequences. Academy of
Management Journal 45 (2002): 779-790.
LIBBY, R., and J. LUFT. Determinants of Judgment Performance in Accounting
Settings: Ability, Knowledge, Motivation and Environment. Accounting,
Organizations & Society 18 (1993): 425-450.
40
LYNCH, A. L., U. S. MURTHY, T. J. ENGLE. Fraud Brainstorming Using Computer
Mediated Communication: The Effects of Brainstorming Technique and Facilitation.
The Accounting Review 84, 4 (2009): 1209-1232.
MATTHEW, C.T. and R. J. STERNBERG. Developing Experience-Based (Tacit)
Knowledge Through Reflection. Learning and Individual Differences 15 (2009):
530-540.
MEYER, J.P., L. J. STANLEY, AND R. J. VANDENBERG. A Person Centered
Approach to the Study of Commitment. Human Resource Management Review.
(2012), http://dx.doi.org/10.1016/j.hrmr.2012.07.007.
MULLEN, E. J. Framing the Mentoring Relationship as an Information Exchange.
Human Resource Management Review 4 (1994): 257-281.
OSLAND, J. S., D. A. KOLB, I. M. RUBIN, and M. E. TURNER. Organizational
Behavior: An Experiential Approach. Prentice Hall. 2006.
PATEL, V. L., J. F. AROCHA, AND D. R. KAUFMAN. Expertise and tacit knowledge
in medicine. In Tacit Knowledge in Professional Practice: Researcher and
Practitioner Perspectives, edited by R.J. Sternberg and J.A. Hovarth. Mahwah, NJ:
Lawrence Erlbaum Associates, 1999.
PEECHER, M. E., M. D. PIERCEY, J. S. RICH, R. M. TUBBS. The Effects of a
Supervisors Active Intervention in Subordinates Judgments, Directional Goals, and
Perceived Technical Knowledge Advantage on Audit Team Judgments. The
Accounting Review 85, 5 (2010): 1763-1786.
POLANYI, M. 1966. The Tacit Dimension. London: Routledge & Kegan Paul, 1966.
RICCHIUTE, D. N. Effects of an Attorney's Line of Argument on Accountants' Expert
Witness Testimony. The Accounting Review 79 (2004): 221-245.
RHODE, J. G., J. E. SORENSEN, AND E. E. LAWLER. Sources of Professional Staff
Turnover in Public Accounting Firms Revealed by the Exit Interview. Accounting,
Organizations & Society, 2 (2): 153-164.
SOLOMON, I. Multi-Auditor Judgment/Decision Making Research. Journal of
Accounting Literature (1987): 1-25.
41
SWAP, W.; D. LEONARD; M. SHIELDS; and L. ABRAMS. Using Mentoring and
Storytelling to Transfer Knowledge in the Workplace. Journal of Management
Information Systems 18 (2001): 95-114.
STERNBERG, R. J. What Do We Know About Tacit Knowledge? Making the Tacit
Become Explicit. In Tacit Knowledge in Professional Practice: Researcher and
Practitioner Perspectives, edited by R.J. Sternberg and J.A. Hovarth. Mahwah, NJ:
Lawrence Erlbaum Associates, 1999.
STERNBERG, R.J., C. NOKES, P. W. GEISSLER, R. PRINCE, F. OKATCHA, D. A.
BUNDY, E. L. GRIGORENKO. The Relationship Between Academic and Practical
Intelligence. Intelligence. 29, 5 (2001): 401-418.
TAN, H. Organizational Levels and Perceived Importance of Attributes for Superior
Audit Performance. Abacus 35 (1999): 77-90.
TAN, H and R. LIBBY. Tacit Managerial Versus Technical Knowledge as Determinants
of Audit Expertise in the Field. Journal of Accounting Research 35 (1997): 97-113.
TORFF, B. Tacit knowledge in teaching: Folk pedagogy and teacher Education. In Tacit
Knowledge in Professional Practice: Researcher and Practitioner Perspectives,
edited by R.J. Sternberg and J.A. Hovarth. Mahwah, NJ: Lawrence Erlbaum
Associates, 1999.
VERA-MUNOZ, S., J. HO, and C. CHOW. Enhancing Knowledge Sharing in
Accounting Firms Accounting Horizons 20 (2), 133155.
VON KROGH, G., K. ICHIJO and I. NONAKA. Enabling Knowledge Creation: How to
Unlock the Mystery of Tacit Knowledge and Release the Power of Innovation.
Oxford: Oxford University Press, Inc., 2000.
WAGNER, R. K. Tacit Knowledge in Everyday Intelligent Behavior. Journal of
Personality and Social Psychology (1987): 1236-1247.
WAGNER, R.K. and R.J. STERNBERG. Practical Intelligence in Real-World Pursuits:
The Role of Tacit Knowledge. Journal of Personality and Social Psychology 49
(1985): 436-458.
WAGNER, R.K. and R.J. STERNBERG. Tacit Knowledge in Managerial Success.
Journal of Business and Psychology 1 (1987): 301-312.
42
WILSON, J.A. and N.S. ELMAN. Organizational Benefits of Mentoring. Academy of
Management Executive 4 (1990): 88-94.
WOODS, A. Subjective Adjustments to Objective Performance Measures 2009
Available at SSRN: http://ssrn.com/abstract=1521491 or
http://dx.doi.org/10.2139/ssrn.1521491



43
APPENDIX A
Tan and Libbys [1997] modified tacit knowledge scale

1. It is your second year as an audit manager in a department of about 30 professional
staff. The evaluation of your first year on the job has been generally favorable.
Though your staff are generally motivated, there are some who seem to go through
the motions but are of little real help. You believe that although you are well-liked,
there is little that would distinguish you in the eyes of your superiors from the nine
other managers at comparable levels in your firm.
Your goal is rapid promotion. The following is a list of things you are considering
doing in the next 12 months. You obviously cannot do them all. Rate the importance
of each by its priority as a means of reaching your goal:
...
d. make an effort to better match the work that needs to be done with the
strengths and weaknesses of individual employees
e. put pressure on staff to cut their time/cost budgets


2. Rate the following characteristics of a job by their importance in leading to a
successful career in a given company:
a. the job will bring your work to the attention of higher-level management
personnel

f. you can master the job with almost no effort


3. During one of your recruiting interviews at the university, a student asks you about
things one can do to increase one's chances for success in public accounting. Rate
each of the following things one can do by its importance to a successful career in a
public accounting firm:

b. avoid drawing attention to yourself at all costs

g. take advantage of opportunities to take on responsibilities beyond the
scope of your immediate assignment

4. You have just been transferred to another department in the firm. You were asked to
take on this new job because of rather serious personnel-related problems in the new
department. Morale in the new department is low. The department is divided into
those who are sorry the former manager has left and those who are sorry the manager
has not left earlier. Performance of the department has been below expectations. The
problem has been around for some time, and you realize that solving them won't
happen overnight. You also believe that this is a chance to show your superiors what
you can do in a tough situation, and you hope that by doing well, you will improve
your opportunities for advancement. Rate the following actions by their importance in
helping you to succeed in your new position:
44
a. always delegate to the most junior person who can be trusted with the task

j. do not try to do too much too soon

5. Rate the following motivations in terms of their importance as incentives for pursuing
a career in public accounting:

e. I enjoy responsibility and the power that goes with it

h. I like a job that I can "leave at the office" at the end of the day

6. In auditing as in other fields, there are often several people who are acknowledged to
do extraordinary work. Rate the following characteristics by how important you
believe they help in contributing to the success of these individuals:

i. power hungry

k. better able than most to grasp and operate in terms of the "big picture", i.e.,
the mission of the firm

7. Rate the following strategies of working according to how important you believe they
are to the day-to-day work of an audit manager:

c. be in charge of all phases of every task you are involved with

k. carefully consider the optimal strategy before beginning a task

8. A number of factors enter into the establishment of a good reputation in a public
accounting firm as a manager. Consider some of these factors and rate their
importance:
a. critical thinking ability

f. extent of education and the prestige of the school attended


9. Rate the following experiences by their importance in becoming a good audit
manager:

e. a strong background in accounting

i. experience as editor of a professional magazine

10. Your company has sent you to a university to recruit and interview potential audit
trainees. You have been considering characteristics of students that are important to
later success in the public accounting firm. Rate the importance of the following
45
student characteristics by the extent to which they lead to later success in public
accounting:
a. motivation

j. the need to win at everything no matter what the cost



46
Exhibit 1: Tacit Knowledge Distributions Implied by Results in Tan and Libby
[1997]



We developed this exhibit by using sample means and standard deviations as reported in
Table 1 of Tan and Libby [1997] and by making the simplifying assumption that tacit
knowledge levels are normally distributed. The blue (red) lines are estimated normal
distributions of tacit knowledge levels for audit managers (staff). The solid (broken) lines
are estimated normal distributions of tacit knowledge levels for auditors with top
(bottom) annual performance evaluations.

Two of the main takeaways emphasized in Tan and Libby [1997] and by papers citing
this study are that tacit knowledge levels are significantly higher for managers who
receive top versus bottom performance evaluations, but not significantly different for
audit staff who receive top versus bottom performance evaluations.

We use this exhibit to accentuate another important, but easily overlooked, implication of
their findings: The tacit knowledge levels of a considerable portion of both top- and
bottom-rated staff auditors rival that of top-rated audit managers. In particular, the mean
tacit knowledge level of top-rated managers, bottom-rated managers, top-rated staff, and
bottom-rated staff are 486.3, 447.94, 414.3, and 406.4 (higher numbers are better). The
related standard deviations 41.0, 62.8, 86.9 and 125.7. An implication is that staff
auditors whose tacit knowledge is just one-standard deviation greater than the mean for
their level would approximate the average tacit knowledge level of top-rated managers.

47
Table 1: Description of Datasets and Samples


Datasets / samples #obs. #obs.

(1) # of subordinate-supervisor/partner dyads based on personnel files 101

(2) # of subordinates with non-missing immediate supervisor survey data 99

(3) # of subordinates with non-missing immediate partner survey data 66

(4) # of subordinates with non-missing subordinate survey data 85

(5) # of subordinates with archival data on performance evaluation 92

Hypothesis 1: Intersection of (1), (2), and (4) 85

Hypothesis 1: Intersection of (1), (2), (4), and (5) 79

Hypotheses 2-4: Intersection of (1)-(5) 61


48
Table 2: Descriptive Statistics


Variable N Mean Median Std Q1 Q3

Tacit Knowledge
Subordinate Supervisor
Benchmark

85

79.81

80.0

21.94

66

98

Tacit Knowledge
Subordinate Partner
Benchmark

85

73.04

76.0

15.83

62

83

Firm Tenure 85 82.13 47.0 83.73 27 111

Age 85 31.95 29.0 9.53 26 35

Job Level 85 2.32 2.0 1.36 1 3

Overall Evaluation Rating 79 2.99 3.0 0.20 3 3

Technical Performance
Rating
79 2.98 2.97 0.29 2.9 3

Competencies Rating 79 2.94 2.94 0.18 2.9 3

Firm Commitment 85 5.12 5.50 1.29 4.5 6


Tacit Knowledge Subordinate Supervisor (Partner) Benchmark: Subordinates tacit knowledge measured
using the Tan and Libby [1994] modified tacit knowledge scale and using the immediate supervisor
(partner) as the benchmark.
Firm Tenure: Subordinates tenure with the firm in months.
Age: Subordinates age in years.
Job Level: Subordinates level in the corporate hierarchy, where higher values reflect a higher level.
Overall Evaluation Rating: Performance rating received by the subordinate, ranging from 1 (poor) to 4
(excellent).
Technical Performance Rating: Performance sub-rating received by the subordinate on the dimension
Technical Performance, ranging from 1 (poor) to 4 (excellent).
Competencies Rating: Performance sub-rating received by the subordinate on the dimension
Competencies, ranging from 1 (poor) to 4 (excellent).
Firm Commitment: The average of four items, scaled from 1-7, capturing the extent to which the
subordinate is committed to the firm (principal component analysis reveals that these items load on a single
factor, explaining 68% of the total variance and factor loadings ranging from 0.76-0.90).

49
Table 3: Correlations (Pearson below diagonal; Spearman above diagonal)


Variable 1 2 3 4 5 6 7 8 9


1. Tacit Knowledge Subordinate Supervisor Benchmark 1 0.55 -0.04 -0.01 -0.11 0.14 0.29 0.36 0.19

2. Tacit Knowledge Subordinate Partner Benchmark 0.56 1 -0.12 -0.02 -0.04 0.09 0.18 0.26 0.20

3. Firm Tenure -0.08 -0.17 1 0.77 0.28 0.05 0.16 -0.09 -0.10

4. Age -0.06 -0.05 0.85 1 0.39 -0.04 0.12 -0.19 -0.02

5. Job Level -0.11 0.01 0.21 0.19 1 0.05 0.16 -0.06 -0.19

6. Overall Evaluation Rating 0.16 0.11 0.02 -0.01 0.09 1 0.44 0.48 -0.01

7. Technical Performance Rating 0.19 -0.03 0.08 0.02 -0.07 0.27 1 0.58 0.01

8. Competencies Rating 0.34 0.27 -0.22 -0.23 -0.08 0.55 0.32 1 0.11

9. Firm Commitment 0.19 0.29 0.02 0.02 -0.15 0.07 -0.05 0.10 1



All correlation coefficients that are significant at the 10% level (two-tailed) or better are in bold.
For variable definitions see Table 2.




50
Table 4: H1 - Impact of Subordinates Perceived Tacit Knowledge on Manager/Partner Potential

Independent variables Predicted sign Potential to be Manager Potential to be Partner


Tacit Knowledge Subordinate
Supervisor Benchmark
+ 0.02*
(0.01)
0.01
(0.02)
0.08**
(0.04)
0.08*
(0.05)

Job Level + 0.90***
(0.29)
0.90***
(0.33)
1.33***
(0.49)
1.20**
(0.52)

Firm Tenure ? 0.01
(0.01)
0.00
(0.01)
0.03
(0.03)
0.02
(0.03)



Age + -0.31***
(0.12)
-0.31**
(0.14)
-0.53
(0.34)
-0.36
(0.35)

Overall Evaluation Rating + 3.86*
(2.22)
4.82*
(2.66)

Sample size 85 79 85 79


***,**,* is significant at the 1%, 5%, and 10% level (two-tailed), respectively. Standard errors are in parentheses.
Potential to be Manager (Partner) is a dummy variable that equals 1 if the subordinates immediate supervisor indicates that the subordinate has the potential to
be a manager (partner), zero otherwise. All other variables are defined in Table 2.




Table 5: H3 - Impact of Supervisors Tacit Knowledge on Relative Weights Given to
Perceived Tacit versus Technical Knowledge in Evaluating Subordinates Annual
Performance


Panel A: Relationship between sub-ratings and overall evaluation rating by supervisor type

Independent variables
Predicted
Sign
Overall Evaluation Rating
(I) (II)


Technical Performance Rating + 0.58***
(0.19)
0.54**
(0.21)

Competencies Rating + 0.33*
(0.17)
0.35*
(0.18)

HiTKMgr

? 0.78
(0.77)
1.02
(0.88)

HiTKMgr *
Technical Performance Rating
- -0.60***
(0.21)
-0.78**
(0.32)

HiTKMgr *
Competencies Rating
+ 0.33
(0.26)
0.44
(0.30)

Job Level ? 0.13
(0.31)

Job Level *
Technical Performance Rating
? -0.10
(0.12)

Job Level *
Competencies Rating
? 0.06
(0.11)

Adjusted R-square 36.07% 33.28%

Sample size 61 61


Panel B: Weights on Different Performance Dimensions by Supervisor Type based on Model (I)

LoTKMgr HiTKMgr Hi vs. Lo TKMgr

Weight on Technical Performance 0.58*** -0.02 -0.60***

Weight on Competencies 0.33** 0.66*** 0.33

Weight on Competencies vs.
Technical Performance

-0.25 0.68*** 0.93**



***,**,* is significant at the 1%, 5%, and 10% level (two-tailed), respectively. Standard errors are in
parentheses.
HiTKMgr is a dummy variable that equals 1 if the subordinates immediate supervisor has a level of tacit
knowledge that is above the median, zero otherwise. All other variables are defined in Table 2. Job Level is
centered at the mean.


52
Table 6: H4 - Impact of Immediate Supervisor Tacit Knowledge on High Tacit
Knowledge Subordinates Firm Commitment


Panel A: Interaction Between Tacit Knowledge of Subordinate and Manager Type

Independent variables
Predicted
Sign
Firm Commitment
(I) (II)


Tacit Knowledge Subordinate -
Partner Benchmark
? 0.01
(0.01)
0.00
(0.28)

HiTKMgr ? 0.11
(0.23)
0.33
(0.39)

Tacit Knowledge Subordinate -
Partner Benchmark * HiTKMgr
+ 0.03*
(0.02)
0.05**
(0.02)

Job Level ? -0.17*
(0.09)


Firm Tenure ? -0.00
(0.00)


Age ? 0.02
(0.02)


Overall Evaluation Rating ? 0.16
(0.56)



Adjusted R-square 14.67% 17.85%

Sample size 61 42


Panel B: Effect of Tacit Knowledge Subordinate on Firm Commitment by Manager Type based on
Model (I)

LoTKMgr HiTKMgr Hi vs. Lo TKMgr

Tacit Knowledge Subordinate
Firm Commitment
0.01 0.03*** 0.03*



***,**,* is significant at the 1%, 5%, and 10% level (two-tailed), respectively. Standard errors are in
parentheses.
HiTKMgr is a dummy variable that equals 1 if the subordinates immediate supervisor has a level of tacit
knowledge that is above the median, zero otherwise. All other variables are defined in Table 2.