The evidence on the firm-to-firm spread of financial reporting behavior via board networks contributes to a little-studied area in accounting that should be important. The authors contribute to the corporate governance literature by offering evidence that contagion effects vary with board positions. They show that board supervision of management is important for ensuring high-quality financial reporting and that board linkages affect the success of this supervision. Regulators concerned about improving financial reporting quality should consider the board connectivity of companies.

Chiu, P. C., S. H. Teoh, and F. Tian. 2013. Board Interlocks and Earnings Management Contagion. Accounting Review 88 (3): 915-944.

In the corporate world, behavior may spread through board of director networks. A board link exists between two firms whenever a director sits on both firms’ boards. A typical board in the sample has nine directors, and the median number of interlocks with other boards is approximately five. In this way, firms are widely connected by their board networks, which potentially serve as conduits for spreading behaviors from firm to firm.

In this study, the authors investigate whether financial reporting behavior spreads through interlocking corporate boards. The test design emphasizes contagion of “bad” financial reporting choices, specifically, earnings management that results in a subsequent earnings restatement, although it also allows for inferences about “good” reporting contagion. The authors use restatements to identify firms that have managed earnings and the period when the manipulation occurred. They refer to a firm that later restates earnings as contagious. The authors define the contagious period as starting in the first year for which earnings are restated and ending two years after. Any firm that shares an interlocked director with the contagious firm during the contagious period is therefore exposed to an earnings management infection via the board network. They consider a multiyear contagious period to allow the earnings management infection to incubate, which is analogous to an epidemiological setting for viral infections. The key test investigates whether an exposed firm is more likely to manage earnings during the contagious period as compared to an unexposed firm.

The authors use the U.S. Government Accountability Office’s (GAO) first release of restatements between January 1, 1997 to June 30, 2002 to identify contagious firms and their contagious periods. They keep only the earliest restatement within the sample period when a firm has multiple restatements. The authors obtain director names from Risk Metrics. In the 1997–2001 sample period the authors identify a sample of 118 observations.

• The authors find strong evidence that a firm is more likely to manage earnings when exposed within a three-year period to earnings management from a common director with an earnings manipulator. The contagion effect is economically substantial.
• The regression odds ratio suggests that a board link to a manipulator doubles the likelihood that the firm will manage earnings.
• The authors also find evidence for good financial reporting contagion. A board link to a non-manipulator significantly decreases the likelihood of the firm being a manipulator.
• Both bad and good accounting behaviors are contagious across board networks.
• Earnings management contagion is stronger when the shared director has a leadership position as board chair or audit committee chair, or an accounting-relevant position as an audit committee member, in the exposed firm.
• The contagion is also stronger when the linked director is the board chair or CEO of the contagious firm, but not when the linked director is the CEO of the exposed firm.
• Earnings management contagion is exacerbated when the exposed firm is located within 100 miles of the contagious firm and shares a common auditor with the contagious firm.
• The evidence supports the proposition that earnings manipulation spreads through board networks.

http://www.auditingresearchsummaries.org