Corporate fraud and external social connectedness of independent directors

Abstract

We examine the effects of independent directors' external social connectedness on corporate fraud commission and detection. The results show that well-connected independent directors do not affect the likelihood of fraud commission but significantly reduce the likelihood of fraud detection given occurrence of a fraud. In particular, with a one-standard-deviation increase in independent directors' connectedness, the likelihood of fraud detection reduces by 22.5%. We also find that the consequences of fraud commission faced by firms with well-connected independent directors are less severe as fraud remains undetected for a longer period of time and fewer people are charged with fraud when independent directors are well connected. We further show that independent directors' connections to fraud firms significantly increase a firm's propensity to fraud commission and the likelihood of fraud detection is also higher. Overall, our results suggest that directors' personal networks have a “dark side”. Regulators should be aware of unintended consequences associated with directors' external social connections when considering how to prevent and detect corporate fraud.

Introduction

Corporate fraud has devastating consequences: corporate empires collapse, market confidence erodes, the image of the accounting profession is tarnished, and management and directors are fired, prosecuted, and incarcerated (Association of Certified Fraud Examiners, 2014, Free and Murphy, 2015, PwC, 2014). Studies have advocated that having strong corporate governance and promoting independent director supervision will effectively deter occurrence of the next corporate fraud (Agrawal and Chadha, 2005, Beasley et al., 2000). However, an improved understanding of how independent directors function to deter and detect fraud is essential (Davis and Pesch, 2013, Trompeter et al., 2012). The literature suggests that corporate fraud often interplays with social connections (Free and Murphy, 2015). In this study we focus on external social connectedness (i.e. networks outside the focal firm) of independent directors. In particular, we examine the influence of independent directors' social connectedness on fraud commission and fraud detection, as well as on the consequences of fraud.

Independent directors play a prominent role in implementing corporate fiduciary duty (Avci et al., 2017, Coles et al., 2014). The literature argues that independent directors' external connectedness is indicative of their social influence and capability (Beasley, 1996, Ferris et al., 2003). But it remains unclear whether independent directors' external connectedness implies better monitoring. Independent directors may utilize their external social connections to improve accessibility to information and bargaining power over management (He and Huang, 2011, Mizruchi, 1996, Mol, 2001). On the other hand, social networking requires time, effort, and attention, which can potentially be detrimental to director monitoring efficacy. For example, external connections impose greater demands on multitasking, thereby distracting directors and diminishing their capacity to function effectively (Ahn et al., 2010). As well-connected directors may be over-committed externally, they devote less time and effort to monitoring—and even overlook the signs of managerial opportunism, which provide CEOs opportunities to engage in self-interested activities (Cashman et al., 2012, Ferris et al., 2003, Fich and Shivdasani, 2006, Shivdasani and Yermack, 1999). Thus the overall effect of directors' external connectedness toward fraud commission is yet unclear.

With respect to fraud detection, despite their important role in providing oversight and uncovering corporate misdeeds, independent directors have incentives to conceal corporate wrongdoings. This is because independent directors have to bear huge losses in human capital and future wealth when fraud is detected (Fich and Shivdasani, 2007, Karpoff et al., 2008). Even when independent directors are not actively involved in wrongdoings they still suffer from the adverse consequences of fraud revelation because the exposure of fraud will tarnish their reputation as an effective monitor and ruin their reputation in the director labor markets (Cowen and Marcel, 2011). Reputation has significant economic value in the socially connected corporate world (Fombrun, 1996, Grey and Balmer, 1998, Shane and Cable, 2002). Compared to less connected directors, directors who are externally better connected have greater career concerns and a larger proportion of their wealth depends on their reputation, which is perceived by the markets (Kang, 2008, Weigelt and Camerer, 1988). The potentially significant reputational loss concerns well-connected directors, which explains their incentives to utilize social connections and reduce the likelihood of fraud detection. Prior work indeed shows that social connections are often employed as a means of influencing the SEC enforcement actions and the legal justice system (Correia, 2014, Dorminey et al., 2012, Sutherland, 1944). We thus expect that, conditional upon fraud commission, well-connected independent directors will take advantage of their social influence to minimize the likelihood of fraud detection. We further expect that there will be less severe consequences following fraud detection in firms with well-connected directors.

We adopt a bivariate probit modelling approach and separately model fraud commission and detection (Wang et al., 2010, Wang, 2013). Using a sample consisting of 17,688 observations from fiscal years 1999 to 2013, our results are consistent with our expectations. We show that independent directors' external social connectedness has no significant association with a firm's propensity of committing fraud, but well-connected independent directors are associated with a lower likelihood of fraud detection given the occurrence of fraud. We perform further analyses to examine the effects of director connectedness on consequences of fraud commission. We find that in firms with well-connected independent directors, fraud remains undetected for a longer period of time and fewer people are eventually charged with fraud, suggesting that independent directors' external connectedness is instrumental in minimizing the costs of fraud commission. In terms of economic significance, a one-standard deviation increase in independent directors' connectedness decreases the likelihood of fraud detection by 22.5%.

Next, we examine the effects of interlocking directorships in fraud firms on the fraud commission and detection processes. We find that directorships to firms that are currently involved in fraud will significantly increase a firm's propensity to commit fraud, which is in line with prior work showing that firm practices, including those potentially detrimental to firm value, become “contagious” via interlocking directorates (Collins et al., 2009, Ertimur et al., 2012, Shropshire, 2010). Further, the detection rate of fraud is higher in firms sharing interlocking directors with fraud firms, suggesting that regulators are likely aware of possible fraud contagion and accordingly increase their scrutiny toward firms that share directorships with a fraudulent firm. More importantly, after controlling for interlocking directorships to fraudulent firms, we still find that firms with socially influential independent directors are associated with a lower likelihood of fraud detection.

One general concern in corporate governance studies relates to the potential endogeneity problem (Faleye et al., 2014, Wintoki et al., 2012). In our study, the concern arises due to the existence of confounding factors other than director connectedness that may explain fraud commission and detection. We perform a series of analyses to mitigate the endogeneity concern, including employing a two-stage least squares (2SLS) estimation method and a propensity-score-matching approach, exploring potential channels through which directors' social influence plays a prominent role in the fraudulent processes, conducting analysis in a sample where selection issue is less of a concern, investigating how well-connected directors would be affected once fraud is detected, and examining the robustness of the findings toward an array of alternative explanations. Our conclusions are drawn on the basis of consistent findings.

This study makes several important contributions to the literature. First, it contributes to corporate fraud research by examining the effects of independent directors' social connectedness in facilitating the development of fraud. In a related study, Khanna et al. (2015) investigate how internal relationships between CEOs and directors within a firm relate to corporate fraud. However, the effects of external social networks on corporate control system and fraud development are under-explored in the literature. Networking outside the focal firm plays a vital role in all social and economic transactions; internal and external social networks are subject to distinct economic costs and benefits (Kilduff and Tsai, 2003, Tian et al., 2011). We extend Khanna et al. (2015) by considering independent directors' external connectedness while controlling for CEO-director internal connections. Our work thus provides a more complete picture of the development of corporate fraud.

We also extend the literature on corporate governance. To the best of our knowledge, we are the first to examine the effects of independent directors' social connectedness on corporate fraud. Our evidence demonstrates a potential cost a firm has to consider when appointing directors who are externally well connected. We show that rather than publicly revealing the wrongdoings these directors possess a greater incentive to deter the detection of fraud. Furthermore, fraud perpetrated in firms with well-connected directors are subject to less severe consequence, indicating that independent directors' self-interests adversely affect the best interest of shareholders and even society (Adams and Ferreira, 2007, Brochet and Srinivasan, 2014, Duchin et al., 2010).

Our study contributes to a stream of burgeoning literature that explores how personal social networks influence corporate behavior and firm decisions (see Cohen et al., 2008, Engelberg et al., 2012, Engelberg et al., 2013, Uzzi, 1997; among others). Our findings suggest that it is the connections to influential firms, such as those in a major market index, which matters the most in explaining a firm's fraudulent behavior. In contrast, social connections to small firms do not exhibit any significant power in explaining the occurrence of corporate fraud. In additional analysis, we show that both work-related and friendship ties of independent directors have significant effects on fraud detection and thus provide further evidence that business is embedded in various types of social networks. Our study thus represents a timely response to a recent call to examine the substantial influence of social ties of independent directors (Tian et al., 2011).

We further provide evidence on the “burden of fame” that well-connected directors have to bear (Wade et al., 2006). When a director becomes more externally and socially connected they are subject to greater potential reputational costs and career damage once fraud is publicly exposed. Our results indeed show that such directors are faced with a dimming likelihood of receiving board seats in another firm after fraud detection in the focal firm, supporting our argument that the reputational costs are especially substantial for well-connected board directors.

Our study further adds to the interlocking directorate literature (Collins et al., 2009, Ertimur et al., 2012, Shropshire, 2010). We show that interlocking directorships facilitate the diffusion of corporate fraudulent practices. Our findings provide practical implications to regulators. Recent years have witnessed regulators and institutional shareholders advocating a fully independent board of directors as the “cure-all” for corporate failures (Avci et al., 2017, Solomon, 2013). We show that externally well-connected independent directors want to avoid reputational losses and will deter the uncovering of corporate fraud at its occurrence, thereby suggesting that merely increasing the representativeness of independent directors on board, in particular those that are externally well connected, does not necessarily provide an optimal solution to curbing corporate scandals.

The rest of the paper is organized as follows. Section 2 reviews related literature and develops our hypotheses. Section 3 presents our models, describes the construction of our sample, and provides variable definitions. In Section 4 we discuss our main findings. Section 5 reports how we handle potential endogeneity issues and discuss the robustness of our results to various checks. Section 6 concludes the paper.