Empirical Findings

Although social ties can help facilitate information exchange between CEOs and outside directors, they can cast doubt on the independence of outside directors and the effectiveness of their monitoring activities.³⁵Studies have examined the effects that social ties can have on outside directors’ monitoring function. In testing the monitoring effectiveness of outside directors with social ties, studies have used different proxies including CEO-compensation level, earnings management, and financial-reporting quality.

32 Adam and Ferreira, supra note 18.

33 James D. Westphal, Collaboration in the Boardroom: Behavioral and Performance Consequences of CEO-Board Social Ties, 42 ACAD.MGMT J.7, 16-19 (1999). Social ties among corporate leaders could help secure outside resources the firm might need and facilitate inter-firm resource dependence. See James D. Westphal et al., The Strategic Impetus for Social Network Ties: Reconstituting Broken CEO Friendship Ties, 27 STRAT.MGMT.J. 425 (2006).

34It is common in social network research to ask the respondent to identify their perceived level of friendship with others. Westphal, supra note 33, at 13-14; Westphal et al., supra note 33, at 433.

EALE 2013 Warsaw (Sept. 26-28, 2013) Yu-Hsin Lin

Recent theoretical work has presented economic models that probe into the relations between social networks and corporate governance. Theory suggests that social connections between board members tend to impede governance effectiveness because boards, desirous of preserving their social capital, are reluctant to undertake an intense monitoring of CEOs. In addition, social networks can reduce the precision of information collected and used by board members in deciding resource allocation. While precise information may improve resource allocation, it could also raise the probability of detecting CEO’s siphoning of corporate assets.

Therefore, board members with social ties to CEOs tend to reduce the precision level of information they collection order to preserve social capital.³⁶

Subrahmanyam presents empirical evidence on social networks and corporate governance.

The study uses age differences, occupation, ethnicity, gender, and familial relationships between CEOs and board members as proxies for social networks. Subrahmanyam's empirical tests present compelling evidence that when boards consist of both fewer members who are CEOs and greater non-Caucasian representation, the corresponding firms are better governed and executive compensation is lower than would otherwise be the case.³⁷ Given that the majority of board members of U.S. companies are Caucasian, research results suggest that firms’ governance can improve when social networks between board members and CEOs are minimal or absent.³⁸ In sum, Subrahmanyam's work suggests that social ties can reduce board's monitoring capabilities and increase CEO compensation, which lead to lower shareholder value.

Hwang and Kim examined the social ties between board directors and CEOs within individual Fortune 100 firms and the association between these ties and executive

36AvanidharSubrahmanyam, Social Networks and Corporate Governance, 14 EUR.FIN.MGMT. 633, 636-45 (2008).

37 The quality of governance was measured by the governance index obtained from Andrew Metrick's website. Id., at 649.

38Id., at 647-53.

EALE 2013 Warsaw (Sept. 26-28, 2013) Yu-Hsin Lin

compensation.³⁹ The researchers identified social ties by noting board directors and CEOs' shared backgrounds, including mutual alma mater, military service, regional origin, academic major, and industry, and found that the percentage of independent boards dropped from 87% to 62% when screenings indicated that there were shared backgrounds.⁴⁰ Using CEO compensation as a proxy for directors’ monitoring level, the researchers tested the association of social ties on CEO compensation. They found that the CEOs of companies with socially independent boards received significantly lower compensation than CEOs of companies with non-independent boards, suggesting that social ties do impair the impartiality of outside directors and diminish their monitoring function.⁴¹

Following up on their prior research, Hwang and Kim further examined the effects of social ties between CEOs and audit committee members on audit committees’ oversight capabilities and, in particular, on the firms’ financial reporting process. The researchers found that social ties, defined by shared backgrounds, between audit committee members and CEOs are more prevalent than conventional ties, as captured by the law since Sarbanes-Oxley Act of 2002 (“SOX”). The samples of this study include Fortune 100 firms as declared in 1996 and 2005. The researchers built a social index which represents the average number of ties each audit committee member has with the CEO. In 25% of the sample firms, the average social ties of each audit committee member to the CEO is greater than 1.0; in 2.4% of the samples firms, the average number is greater than 2.0, suggesting a strong presence of social ties in Fortune 100 firms. Each audit committee member, on average, has 0.6 social ties and 0.1 conventional ties to the CEO.⁴²Since the samples span the period from 1996 to 2005, it is expected that conventional ties

39H&K 2009, supra note 10.

EALE 2013 Warsaw (Sept. 26-28, 2013) Yu-Hsin Lin

will not be prevalent because all audit committee members are supposed to meet the independence requirements set by SOX and later-published rules of NYSE and NASDAQ.

Hwang and Kim further examined the association between audit committee members' social ties and the firms’ earnings-management practices. The major responsibility of an audit committee is to oversee the integrity of a firm’s financial reporting system. Earnings management refers to attempts by the management to influence or manipulate reported earnings.

Earnings management practices are usually associated with fraud and threaten the integrity of a firm’s financial reporting system. Hwang and Kim used abnormal (i.e. discretionary) accruals as proxies for earnings-management activities and examined the association between social ties of audit committee members and abnormal accruals.⁴³ They found that the association between abnormal accruals and the extent of audit-committee members’ social ties to the CEO is substantially stronger than the association between abnormal accruals and audit-committee members’ conventional ties to the CEO. This finding suggests that mutual qualities foster relationship building and that social ties can impair the oversight ability of audit committees, as evidenced by facilitation of earnings-management practices.⁴⁴

Empirical research shows that social ties between independent directors and CEOs do compromise the monitoring ability of independent directors. However, because social ties could foster information exchange, which is essential for independent directors’ execution of their responsibilities, social ties between independent directors and CEOs could foster board collaboration and improve boards’ advisory function. Finally, whether social ties increase or decrease firm value will depend on each firm’s specific current status, such as a firm’s development stage and a firm’s complexity. The optimal composition of boards may vary among

43Id., at 7-8.

44Id., at10-12.

EALE 2013 Warsaw (Sept. 26-28, 2013) Yu-Hsin Lin

different firms and at different stages of development.