Death by Committee? An Analysis of Corporate Board (Sub-) Committees

There is a long history of governance reforms that target corporate boards by mandating specific committee structures and director-type requirement. The Securities and Exchange Commission and the New York Stock Exchange first advocated for separate audit committees following the McKesson & Robbins scandal of 1938 (Birkett, 1986). In the 1970s, in response to widespread bribery by U.S. corporations in foreign countries, the SEC recommended that firms maintain an independent director majority on audit and nominating committees, and the NYSE updated its listing standards to require that firms maintain audit committees (Dundas and George, 1980). Several high-profile corporate failures in the early 1980s brought about the Treadway Commission, whose report was a factor when the major national exchanges recommended audit committee independence (Reeb and Upadhyay, 2010). In the early 2000s, the boards of Enron, Worldcom, Tyco and Parmalat, among others, were blamed for corporate malfeasance. Intending to prevent similar governance failures in the future, the Sarbanes-Oxley Act of 2002 (SOX) and the revised NYSE and NASDAQ listing standards were put in place.

Clearly, regulators believe that board structure is an important determinant of corporate governance. This has not been lost on economic and legal researchers who, for example, have found that independent audit committees can improve governance outcomes. However, corporate scandals have demonstrated a remarkable resilience to governance reforms. While reforms may appear to simply codify board characteristics, they do so by altering the nature of authority within the board.

To better understand the role of authority on corporate boards, we turned to theoretical models of groups. Aghion and Tirole (1997) distinguish between formal authority (“the right to decide”) and real authority (“the effective control over decisions”) within organizations, demonstrating that the allocation of formal authority affects information flow and communication. Li, Rosen, and Suen (2001) suggest that sub-groups can also impair communication. According to these, and other, theories, a sub-group member, serving as a principal with both formal and real authority, will be inclined to disclose information. Yet, if that same person is invited to participate as an unofficial sub-group member (an agent with only real authority), their incentives to share information will change. Strategically withholding or manipulating information becomes a method by which the unofficial member asserts real authority and leads the sub-group to their preferred decision.

Our recent publication in the Journal of Financial Economics, Death by Committee? An Analysis of Corporate Board (Sub-) Committees, studies how the allocation of formal authority from a board (i.e., group) to relatively homogeneous committees (sub-groups) affects the overall group. Our innovation lies in two measures that serve as proxy variables for the allocation of formal authority from the board to its committees. These outsider-only fractions measure the average fraction of total operations that directors carry out in committees that are formally composed entirely of outside directors. The first measure uses a grammatical parsing natural language processing algorithms to extract information on board and committee meetings. The second measure uses data from topic modelling algorithms to identify whether information gathering and decision-making tasks are undertaken by the board or a committee. Thus, our two measures capture the scale (meetings-based measure) and the scope (measure of stated responsibilities) of the information gathering and the decision-making authority allocated from the board to outside director committees.

We focus on outsider-only committees, committees composed entirely of outside directors with exclusive formal authority, when constructing these measures as this maps nicely from theory to our data. While insiders may assert real authority in outsider-only committees when they participate by invitation, Aghion and Tirole’s (1997) model suggests that their contribution would be different if they were official members with formal authority. Our outsider-only fractions accounts for hierarchical structures by distinguishing between boards and their committees, and it accounts for group composition and the role of formal authority by differentiating between outsider-only and other committees.

We note a distinct change in the allocation of authority within boards over time. Disclosed outsider-only stated information gathering and decision-making responsibilities increased from about 25% of board operations in 1996 to around 35% in 2010. There is also a substantial increase of meetings of outsider-only committees, which, by 2010, account for 45% of board operations. The increased workload of committees disproportionately affected outside directors who spend approximately 40% of annual operations (stated responsibilities and meetings) in committee at the start of our sample and about 60% at the end. An opposite result holds for inside directors, whose percent of operations in committees decreased from around 35% to about 23% over the sample period.

Consistent with the idea that the information corporate directors have is related to the board’s authority structure, we find that an increase in the meetings-based activity measure has a negative impact on the short-term market reaction to outsider director stock purchase disclosures. That is, the market views these directors as less informed. Moreover, outsider director purchases also perform worse over the medium-term. Thus, the allocation of formal authority from the board to outside director-only committees appears to impair effective communication among board members

Finally, we expect to find a relationship between the board’s authority structure and corporate outcomes. As inside directors may withhold information to increase their real authority as their formal authority declines, outside directors may need to defer, in part, to the judgement of insiders. This suggests that an increase in formal authority represented in our measures may lead to increased agency costs and poorer decision-making. We find this to be the case for acquisitions, a key corporate decision, and for firm value, which represents the cumulative effect of board decisions. Economically, a one standard deviation increase in our meeting-based measure is associated with a 0.72% decrease in the market reaction to acquirers’ acquisition announcements. Firm value, as measured by Tobin’s q falls by 2.49% (0.45%) using the meeting-based (stated responsibility) measure increase around Sarbanes-Oxley as a benchmark.

As financial economists, we must be cautious about interpreting our results for policy recommendations. However, it seems clear to us that reform-induced codification of responsibilities emphasizes regulatory compliance and performance reporting, not the role of formal and real authority on the board. Our results suggest that regulators may need to trade-off reductions in corporate failures with potential reductions in firm values when operating on formal, observable board characteristics.

The complete paper is available to download here.