Chairman and CEO: The Controversy over Board Leadership

David Larcker is Professor of Accounting at Stanford Graduate School of Business. This post is based on a paper authored by Professor Larcker and Brian Tayan, Researcher with the Corporate Governance Research Initiative at Stanford Graduate School of Business.

Our paper, Chairman and CEO: The Controversy over Board Leadership, examines the circumstances under which companies decide to combine or separate the chairman and CEO roles and shareholder response to this decision.

In recent years, companies have consistently moved toward separating the chairman and CEO roles. According to Spencer Stuart, just over half of companies in the S&P 500 Index are led by a dual chairman/CEO, down from 77 percent 15 years ago. In theory, an independent chairman improves the ability of the board of directors to oversee management. However, separation of the chairman and CEO roles is not unambiguously positive, and there is little research support for requiring a separation of these roles. Still, shareholder activists and many governance experts remain active in pressuring companies to divide their leadership structure.

Given the controversy over chairman/CEO duality, we examined in detail the leadership structures of publicly traded corporations and the circumstances under which they are changed. Our sample includes the 100 largest and 100 smallest companies in the Fortune 1000 in 2016. The measurement period includes the 20-year period 1996-2015.

We find that board leadership structures are not stable. Only a third (34 percent) of companies made no changes during the entire 20-year measurement period. Slightly under half of these consistently maintained separate chairman and CEO positions (such as Costco, Intel, and Walgreens); slightly more than half of these consistently combined them (such as Amazon, Berkshire Hathaway, and ExxonMobil). Still, these companies are the exception rather than the rule. It is significantly more likely that a company makes at least one change to board leadership structure (combination or separation) over time. On average, companies made 1.7 changes, or approximately 1 change every 12 years. Changes are more frequent among large companies (2.2 changes, on average) than smaller companies (1.3 changes). In both cases, companies are slightly more likely to separate the roles than to combine them.

Most separations occur during the succession process, with the former CEO, founder, or other officer continuing to serve as chair on either a temporary or permanent basis. Of the 171 separations in our sample, 134 (78 percent) are associated with an orderly succession. This is true of both small and large companies. However, large companies are significantly more likely to separate the roles temporarily, whereas smaller companies are more likely to do so permanently.

Approximately a quarter (22 percent) of separations are not part of an orderly succession. Nine percent follow an abrupt resignation of the CEO, 6 percent a governance issue (such as accounting restatement or CEO scandal), 3 percent a merger, 2 percent a shareholder vote, and 2 percent are required of the company as part of a government bailout.

The decision to combine the chairman and CEO roles tends to be more uniform. The vast majority of combinations (91 percent) involve an orderly succession at the top. Only 9 percent are associated with a merger, sudden resignation, or governance-related issue. In 90 percent of combinations, the current CEO is given the additional title of chair; in 10 percent of cases, a new CEO is recruited to become dual chair/CEO.

Most interesting, perhaps, is the frequency with which companies “permanently” separate the leadership roles only to recombine them at a later date. Slightly over one-third (34 percent) of companies in our sample permanently separated the chairman and CEO roles and later recombined them during the 20-year measurement period. Best Buy split the roles for nearly 13 years when founder and chairman Richard Schultze stepped down as CEO in 2002; Schultze eventually resigned from the board and when his successor as chairman retired in 2015, then-CEO Hubert Joly was given the additional title of board chair. The company gave no public explanation of its decision to recombine the roles. Bank of America and Walt Disney both separated the chairman and CEO roles following shareholder votes and subsequently recombined them 5 and 9 years later, respectively, under different management. In both cases, the board justified the decision to recombine as rewarding the successful leadership of the current CEO.

In the cases of Bank of America and Walt Disney, the decisions to recombine the roles were highly controversial. Across the entire sample, however, shareholder response was unexpectedly varied. Only 34 percent of the companies that separated and recombined the chairman and CEO roles were targeted by shareholder-sponsored proxy proposals to require separation. Average support for these proposals was 33 percent, not significantly different from companies that consistently maintain a dual chairman/CEO structure (34 percent support) or that separate the roles temporarily during succession (36 percent support). It was also not significantly different from the average support across the total universe of companies that face shareholder-sponsored proposals requiring separation (32 percent).

Finally, it is interesting to note that pressure to separate the chairman and CEO roles seems to center almost exclusively on large companies. Only 3 of the 95 small companies in our sample were the target of a shareholder proposal to require an independent chairman over the entire 20-year measurement period, even though their board leadership structures are not significantly different from those of larger corporations. By contrast, a majority (56 out of 92) of large companies were targeted at least once. This suggests that the companies that shareholders target to advocate for independent board leadership might not necessarily be those with the most egregious governance problems but instead those that are the most visible public targets.

The full paper is available for download here.

Both comments and trackbacks are currently closed.