Corporate Governance: Separating the CEO and the Chairman Roles
A growing chorus of voices is calling for public companies to make the separation of the Chairman and CEO functions the default governance structure. This movement, which may have the support of the new SEC Chair, appears likely to lead to some type of "adapt or explain" approach. Increasing evidence that the companies where the CEOs also act as board Chair are likelier to have "certain troubling governance characteristics" will likely encourage shareholder interest in the initiative as well.
The idea of separating the two roles is hardly new, but it has gained significant support from a wide variety of sources recently. First, on March 30, 2009, the Chairmen’s Forum of the Millstein Center for Corporate Governance at Yale School of Management issued a report entitled "Chairing the Board: The Case for Independent Leadership in Corporate North American" (here) calling on all North American companies to "voluntarily adopt independent chairmanship as the default model of board leadership," and if they chose to take a different course "to explain to their corporate shareholders why doing so represents a superior approach to optimizing long-term shareholder value."
(The Chairman’s Forum is a group of more than 50 current and former board chairs, directors and CEOs convened at the Millstein Center.)
The Millstein Center’s March 30, 2009 press release (here) reports that while in the U.k. only 5% of the FTSE 350 companies combine the chairman and CEO roles, over 60% of the S&P 500 companies have boards that are chaired by their CEOs. The press release quotes one commentator as saying that the independent chair model "has been adopted successfully by many companies in many regions of the globe as a means to further ensure and empower board independence."
The press release also quote the former chairman of Northwest Airlines as saying that combining both roles puts both functions in one person who "is obviously conflicted in the essential duty of providing oversight and monitoring the CEO and management team."
A March 30, 2009 Wall Street Journal article discussing the Chairmen’s Forum’s report can be found here.
Recent remarks from, Mary Schapiro, the SEC’s new Chair, in her April 6, 2009 speech to the Council of Institutional Investors (here), seem to suggest the possibility of an SEC move to a disclosure based approach toward separating the two roles. Among other things, she said that "we’ll also be considering whether boards should disclose to shareholders their reasons for choosing their particular leadership structure—whether that structure includes an independent chair, a non-independent, or a combined CEO/Chair."
As Professor Jay Brown has suggested on his Race to the Bottom blog (here), Schapiro’s remarks may suggest a SEC attempt to influence corporate governance through disclosure. Professor Brown has been a vocal advocate in favor of separating the two roles (as shown here).
The logic of targeting this particular issue as an important corporate governance objective was reinforced by the research recently released by The Corporate Library. As described in their March 25, 2009 press release (here), companies whose CEOs also serve as board Chair are "more likely to have certain troubling corporate governance characteristics than companies where the roles are separated."
The troubling characteristics, which are "associated with board entrenchment or lessened oversight of management," include relatively long CEO tenures; fewer board meetings per year; classified board structure; and "the presence of executive committees, which are typically given the power to act on behalf of the entire board, potentially allowing for a concentration of power."
The Corporate Library’s findings raise the possibility that having a single person as the Chair and CEO could be a risk factor for D&O insurance underwriters to assess. Along those lines, it is worth considering, as noted by the Chairmen’s Forum report, that "the overwhelming majority of financial institutions had combined roles before the current crisis erupted" – including, among others, Bear Stearns, Lehman Brothers, Citigroup, Washington Mutual and Wachovia.
On the other hand, there may be limits to how much can be expected or discerned from this single governance trait. As the Chairmen’s Forum’s report also notes, "splitting the role of chairman and CEO does not guarantee the application of independent oversight," adding that "it is no secret that certain companies, featured in some of the most famous corporate debacles, had separate CEOs and chairmen." Splitting the roles must be accompanied by other steps "in order for the independent chairman to fulfill the important leadership role."
In other words, while the continued combination of the two roles in a single person may (particularly in the current climate) represent something of a risk factor, the separation of the two functions alone is no guarantee of the absence of risk.
In any event, it seems likely that pressure for change will continue for all companies, and that companies that do not change will find themselves increasingly called upon to explain.
Hi Kevin; thanks for the thoughtful post. It's interesting how the report handles the available empirical evidence on whether splitting the Chair / CEO matters much. In particular, most academics who study this stuff tend to agree that the evidence does not systematically support the proposition that CEO/Chair separation systematically predicts better performance in economically / statistically significant respects. There are a few that do, but there are also some that point in the opposite direction (i.e,. sometimes it predicts better performance, sometimes worse, but it usually doesn't predict much at all). Although the report acknowledges this stalemate, describing the state of affairs as being "inconclusive," it nonetheless advocates splitting the roles as a form of chicken-soup-couldn't-hurt palliative. That conclusion strikes me as rash, for two reasons: First, there are -- at the very least -- transition/transaction costs to splitting the role for companies that haven't already done so; to make that cost worth bearing, as a board member I'd want to know that it's a positive net-present-value proposition. Second, and more troublingly, the report seems to disregard how "selection bias" may interrelate to the inconclusive empirical evidence. In particular, the report's chicken-soup prescription effectively presumes that firms' past decisions about whether to split the Chair/CEO is essentially random, and thus we should expect the inconclusive effects in the literature to carry over to the corporations that have yet to split the roles. But that prior choice probably isn't random. As a first approximation, it is plausible that firms that have already split the roles are systematically the ones for which the benefits are the highest and the costs are lowest. Vice versa for the non-splitters as observed today. If that's what's happening (and this is hard to test), we'd expect to see exactly the inconclusive empirical findings that we in fact observe; but it would tend to refute the conclusion that forcing transition upon the heretofore non-splitters is a advisable, even on a "couldn't hurt" basis.
Kevin: Are you aware of any studies of litigation outcomes (shareholder class actions or derivative actions)that would show differences in outcomes for companies with COB/CEO combined positions vs. separate positions?
Bill