The Ties that Bind: Knowledge-Sharing Networks and Auditor Performance**
Monika Causholli
Von Allmen School of Accountancy
University of Kentucky
Lexington, KY 40506 USA
Theresa Floyd
Department of Management and Marketing
University of Montana
Missoula, MT 59812 USA
Nicole Thorne Jenkins
Von Allmen School of Accountancy
University of Kentucky
Lexington, KY 40506 USA
Scott M. Soltis
Department of Management
University of Kentucky
Lexington, KY 40506 USA
November2016
** Please do not quote without authors’ permission. We appreciate comments received from Pietro Bianchi, Brian Goodson, Sandra Vera-Muñoz, and participants at The 2016 University of Kentucky Department of Management Brown Bag. Monika Causholli and Nicole Jenkins would like to express appreciation to Steve Borgatti. Nicole Jenkins acknowledges research support from the EY Research Fellowship at the University of Kentucky. Monika Causholli gratefully acknowledges financial support from the Von Allmen School of Accountancy.
The Ties that Bind: Knowledge-Sharing Networks and Auditor Performance
Abstract
The dissemination of knowledge in audit firms is a critical process that has gone relatively unexamined by researchers. By surveying a Big 4 audit firm and applying social network analysis, we document the factors that determine the knowledge seeking activities of auditors. We find that auditors tend to seek knowledge from fellowauditors whom they consider their friends, those they are on the same audit team with, those of similar rank, and auditors with longer tenures. We also examine the association betweenthe types and patterns of knowledge seeking ties and individual auditor performance. We find that auditor performance is associated with the number of knowledge ties, the strength of a tie,and the extent of connectednessin the knowledge-seeking network. Taken together, we demonstrate that social ties between auditors improve knowledge acquisition and the latter in turn has implications for auditor performance.
JEL Classification:
Keywords: Social Networks, Knowledge-Sharing, Auditor Performance
- Introduction
Audit firms are under increasing pressure to perform high quality audits efficiently. Regulation emphasizes auditor knowledge and expertise as means to achieving high quality audits (IAASB 2014;PCAOB 2015).[1]However, audits are performed by audit teams and knowledge and expertise are not evenly distributed among the team members (Chow, Ho, and Vera-Muñoz,2008). In order to deliver high quality audits in a cost efficient manner,audit team members must collaborate and share knowledge. Knowledge-sharing allows firms to leverage the expertise possessed by individual auditors (Vera-Muñoz, Ho, and Chow,2006). Audit quality therefore depends not only on individual auditor knowledgebut also on the extent to which knowledge is shared withteam members (Danos, Eichenseher, and Holt, 1989).
While the majority of existing audit research has examined how auditor knowledge affects auditor judgments and audit quality, there has been little research on the determinants of knowledge-sharing and its effects on audit outcomes (Vera-Muñoz et al., 2006; Chow et al., 2008). The extant research has focused on interactions during the audit review process, after audit procedures have been completed by staff auditors (e. g. Gibbins and Newton, 1994; Ramsay, 1994). In contrast, we approach the question of how knowledge is disseminated among auditors during the audit and whether it affects auditor performance from a social network paradigm.
The social network paradigmrecognizes that individuals are embedded in webs of social relationships which have important implications for knowledge transfer and thus performance (Borgatti, Mehra, Brass, and Labianca, 2009). Specifically, in our first research question we are interested in better understanding how social ties affect auditor knowledge seeking activities. Individual auditors interact with other auditors in order to complete their tasks. Although some interactions are determined by task structure and hierarchy, individuals also interact informally through other ties such as friendship ties. Consider an auditor faced with a challenging auditing task, such as determining thereasonableness of the fair value of non-exchange traded investments, for which there is no simple solution available in the audit firm’s knowledge database. There are several plausible options for the auditor at this point, including but not limited to, working on the problem by themselves until they find a solution, seeking help from a fellow auditor whom they consider to be a friend, discussing with a fellow auditor on their current audit team or with whom they have worked with in the past, contacting a more experienced auditor, or perhaps conferring with colleagues who are at the same level as them. We utilize social network analysis to discern between these options tounderstandthe underlying mechanisms ofknowledge seeking among auditors.
While a better understanding of what underpins the patterns of knowledge-seeking among auditors speaks to several gaps in the auditing literature, a related and important question is whetherknowledge-seeking activity actually influences auditor performance. Therefore, our second research question relates to the extent to which knowledge seeking ties (or auditor social capital) have implications for auditor performance. Consider our previous example of an auditor faced with a dilemma for which they likely need to access the knowledge of others. Not all potential sources of knowledge are equal, neither are all knowledge-seeking relationships the same. Auditors who seek knowledge from a wide variety of sources can suffer from information overload, contradictory advice or fears ofbeingperceivedas incompetent (all of which may hinder performance)[2]. Auditors who have more stable relationships with a trusted set of advisors may be in a better position to actually leverage the knowledge of fellow auditors in their network (Hansen, 1999). The relationships that exist around an auditor may matter as well. For instance, if an auditor is seeking knowledge from two fellow auditors who counsel one another, it is more likely that he/she will receive redundant information (Burt, 1992). The extent to which information-sharing through social tiesimproves auditor performance is a crucial and understudied question in the auditing literature (Nelson and Tan, 2005).
The extent of knowledge sharing is not immediately apparent in the audit setting. There are several factors that may impede knowledge sharing among auditors (Chow et al. 2008). First, due to time and budget pressures auditors are less willing to share knowledge. Second, because audit teams are highly fluid and auditors change audit teams frequently and also because of the high turnoverrate in the profession, auditors maynot be motivated to share knowledge with others who they perceive to be short timers in the profession. Third, auditors who view knowledge as a competitive advantage from which they personally benefit are not likely to share it with others. Finally, auditors maybe less likely to seek out knowledge if doing so will lead others to think of them as incompetent (Lee 1997).
To answer our research questions, we collaborated with one of the Big 4 accounting firms. Specifically, the human resources department of one of the firm’s regional offices provided the complete roster of all the employees in two offices, the main regional office, and a satellite location. We also obtained employee data including age, race, gender, tenure with the firm, rank, highest degree obtained, university, and certifications. We used the roster to create a social networks questionnaire that asks everyone to select from the roster the individuals they know, those they consider friends, those they work with, etc. We also ask the employees to select individuals they seek for audit related knowledge and advice. Finally, we ask those that are ranked at the level of supervisor or above to provide their evaluation of the performance of individual auditors. Our survey method, which is considered standard in the social networks research, was thus supplemented with audit-related questions. We sought to capture the social network of auditors and compute network related characteristics and relate them to individual auditor performance. In doing so, we are first to utilize social network analysis (see Borgatti and Halgin 2011for a discussion) to address the issue of knowledge-sharing and its effects on auditor outcomes.
Our findings can be summarized as follows. For the question of whom auditors seek out for knowledge we find that auditors are more likely to seek knowledge from their friends, their immediate coworkers, those with whom they have worked on an audit team in the past year, auditors who have been with the company longer, and those who are of a similar rank. These results suggest the importance of diversesocial ties in knowledge acquisition in the audit setting. For the question of whether knowledge-sharing ties affect performance, we find that the number of people that auditors seek knowledge from is negatively related to performance. Conversely, auditor performance increases in strength of tie and when there is a densely connected (cohesive)network around the auditor.[3]
The results presented in this paper have important implications for knowledge management in audit firms and ultimately the delivery of high quality audit service. Knowledge-sharing is increasingly important as business transactions and audit tasks have become more complex, and auditors face a multifaceted regulatory environment. Acquiring and storing expertise in these circumstances is difficult and requires significant firm investment in training, retention, and other resources. Therefore, it is more likely that expertise is concentrated in a few individuals who must share their knowledge with others to ensure the effective and efficient completion of audit engagements. Thus, the creation of social capital via knowledge-sharing in the presence of higher task complexity and task interdependence has become a more critical factor in ensuring high quality audits. Further, the increased complexity and ambiguity of knowledge required to complete audit tasks, makes it harder to capture knowledge in a systematic fashion. Our results suggest that knowledge-sharing occurs through informal channels in audit firms and is likely an important component for individual auditor performance. While audit firms capture and automate some of the audit knowledge into various types of decision aids (check lists, auditing/financial reporting standards written in plain language, pre-populated audit programs, etc.), some knowledge (e.g. which client employee to ask certain questions of, what to do with a discovered exception, etc.) resides with individuals and unless it is shared, will be lost. Moreover, the prevalence of knowledge-seeking ties in our data suggests that individuals may actually obtain knowledge from other auditors in addition to utilizing knowledge contained in firm databases.[4] Finally, by documentingwhere auditors seek knowledge and the positive and negative effects of the knowledge seeking activity, audit firms can carefully consider the formal interpersonal channels they attempt to cultivate via audit team structuring, mentoring or other team-building programs.
Our study makes several contributions to accounting research. Nelson and Tan (2005) suggest that the topic of auditor interaction is the most under-researched area in auditing. Danos et al. (1989) and Vera-Muñoz et al. (2006) indicate that both formal and informal interactions can facilitate knowledge-sharing and Johnstone, Li, and Luo (2014), suggest that knowledge sharing among auditors is essential for audit quality. Our approach marks the first step in employing social networks analysisto empirically examine knowledge dissemination and its effects on auditor performance. Our methodology allows us to directly capture the direction of the knowledge flows (i.e., who learns from who), thus allowing us to better understand how knowledge is acquired, utilized, and ultimately how it affects performance. Whereas most traditional auditing research focuses on the individual as the level of analysis, social network analysis is built upon dyads—pairs of auditors—inthe workplace. By examining the set of ties surrounding an individual, we provide a comprehensive view of the total knowledge that is transferred and thus extend current knowledge-sharing research that has examined information exchanges between a supervisor and staff in the review process (Ramsay 1994). While important, communication that occurs in the review stage is regulated and thus less likely to develop organically from existing social ties. Our study explores the characteristics of knowledge-sharing across all auditors in the office and all phases of the audit.Second, while the majority of existing audit research has examined how auditor knowledge affects auditor judgments and audit quality (reference), there has been very little research on knowledge-sharing and its effects on auditor performance outcomes.Taken together, our findings increase our understanding of the determinants of knowledge sharing within audit firms and the association between knowledge and auditor performance beyond traditional factors at the individual auditor level.
While not an explicit goal, our study also contributes to the field of social networks research. While many studies in social networks have examined outcomes of social ties, our setting further enriches these findings. First, we demonstrate that the type of social relation in the workplace matters and not just the volume of ties. The fact that we find a negative relationship between the sheer number of knowledge-seeking ties and auditor performance adds to the burgeoning literature that suggests some seemingly positive ties can be detrimental to an organization (e.g. Oldroyd and Morris, 2012; Soltis, Sasovova, Agneessens, and Labianca, 2013). Second, we answer calls for a more nuanced view of advice (knowledge-seeking) networks by exploring domain-specific rather than general advice, advice seeking and receipt, and a variety of interrelationships between advice-seeking and other social ties (Sykes, Venkatesh, and Johnson, 2014; Zhang and Venkatesh, 2013). Finally, since auditing is a highly regulated profession where auditors follow regulations in conducting the audit, we are able to examine the antecedents to and effects of social network in the presence of very strong organizational and institutional norms. That we have counterintuitive findings furthers calls for greater focus on the context in which networks research is taking place to better understand the unanticipated effects of social networks (Emirbayer and Goodwin, 1994; Kilduff and Brass, 2010).
Lastly, our findings can be informative for audit regulators worldwide. While focusing on auditor expertise and knowledge is one way to improve audit quality, it is also important to consider how knowledge possessed by individual auditors is deployed and shared among auditors within audit firms to influence auditor performance. Cumulative expertise can be more effective at improving audit quality when information flows to and is used effectively by less experienced and less knowledgeable auditors.
The remainder of the paper is organized as follows. Section 2 provides a review of auditing and related social network research and states our two research questions. Section 3 explains the data and research method. Section 4 describes the main results as well as some additional analyses. Section 5 concludes the paper.
- Literature Review and Research Questions
Auditing is a knowledge-based profession and audit firms must manage knowledge that resides within the firm efficiently in order to remain competitive and earn profits (Gibbins and Wright, 1999; Chow et al., 2008). To perform an effective audit, auditors must possess knowledge about accounting and auditing standards, knowledge about industry trends, client specific knowledge such as the accounting information system, personnel, internal controls etc.Audit knowledge can be categorized as either explicit or tacit. Explicit knowledge can be more easily captured and disseminated through technology such as knowledge databases (checklists, pre-populated audit programs, plain language summaries of standards, decision making aids. etc.) that auditors can consult when necessary. However, there are challenges associated with capturing all knowledge that individual auditors possess, and with its retrieval including how to get auditors to recalltheir experiences (Gibbins and Wright, 1999). Tacit knowledge, defined as “habitual practices and mental models of individuals” (Vera-Muñoz et al., 2006, p. 135), is more difficult to capture in a systematic way as it frequently manifests itself in the form of intuition or insights (e.g. who at the client to go to for specific information, how to efficiently execute complex tasks, etc.). Given the large amounts of tacit knowledge that resides within individuals at audit firms, it becomes critical to identify pathways for its dissemination (Knechel, 2000; Vera-Muñoz et al., 2006).
2.1 Knowledge-sharing in Audits
Auditors perform many tasks during the planning and execution of the audit, and prior research has examined the effects of auditor knowledge, ability and motivation on the quality of these tasks (e. g.,Bonner and Lewis, 1990; Libby and Luft, 1993; Solomon, Shields, and Whittington, 1999;Owhoso, Messier, and Lynch, 2002). However, because audit team members possess varying levels of knowledge, an effective audit depends not only on the amount of knowledge possessed by individual auditors, but also on whether that knowledge is shared with others on the team (Vera-Muñoz et al., 2006). Only a handful of studies have examined the determinants of information sharing and how communication between auditors affects task performance.[5]Danos et al. (1989) examine the patterns of communication among audit team members and between team members and other specialized personnel within the firm. They elicit information regarding communication patterns and find evidence consistent with the hierarchical nature of consultation within the audit team. That is, auditors turn to their superiors for audit knowledge. They also find that auditor communication is limited to the local office rather than across offices and that knowledge is contained within the team, even industry knowledge. Danos et al. (1989) argue that communication patterns might be influenced by social relations but they do not explicitly explore this possibility.
Dirsmith and Covaleski (1985) also examine informal communication patterns and their effect on performance using data collected from interviews. They find that staff and senior auditors believe that social interactions are important for successful completion of audit tasks. They also find that mentoring relationships arose when individuals were comfortable with each other, suggesting the importance of social relations. Finally, they find that mentoring and informal communications are related to performance of audit tasks. Similarly, Du, Lehman, and Wilson (2014) find that auditors are more likely to share knowledge with individuals whom they have positive prior experiences. Kadous, Leiby and Peecher (2013) suggest that social ties affect how auditors perceive information they receive from their advisors. Specifically, non-expert auditors tend to put more weight on the information they receive from advisors with whom they share a social bond, suggesting negative effects of social bonds.Bedard, Biggs, and Maroney(1998) find that group dynamics determine whether knowledge is shared. Group interaction can lead to improved performance because members can pool their knowledge and correct each other’s errors while ineffective group work occurs when one individual dominates the conversation. A more recent study by Bobek, Daugherty, and Radtke (2012), uses an experiential questionnaire and finds that communication with audit team members is the most important factor in successfully resolving audit issues whilecommunication with others outside the audit team is not important.Nelson and Tan (2005) call for additional research in the area of auditor interactions. We seek to contribute to this stream of literature using a different approach than that utilized in prior studies. Specifically, we utilize social network analysis to examine the role of auditor social ties on knowledge-sharing and how this in turn impacts auditor performance.