I:\28947 Ind Law Rev 47-1\47masthead.wpd SEPARATION ANXIETY: A CAUTIOUS ENDORSEMENT OF THE INDEPENDENT BOARD CHAIR LISA M. FAIRFAX* INTRODUCTION In 2013, for the second consecutive year, major shareholders at JP Morgan Chase & Co. (“JPMorgan”) strenuously urged JPMorgan to appoint an independent director as chair of its board of directors, and thus to separate the roles of CEO and board chair.1 While JPMorgan indicated that separating the roles of CEO and board chair could cause “uncertainty, confusion, and inefficiency in board and management function and relations,”2 shareholder advocates insisted that combining such roles creates a conflict of interest that weakens the board’s ability to engage in effective oversight, undermining strong corporate governance and corporate performance.3 In 2012, JPMorgan shareholders relied on federal and state investigations aimed at JPMorgan in the wake of the financial crisis to support their call for an independent board chair.4 In 2013, JPMorgan shareholders renewed their calls for independent board leadership, pinpointing the “‘London Whale’ trading fiasco, in which [JPMorgan] recorded $5.8 billion of principal transactions losses.”5 In both years, JPMorgan shareholders argued that separating the roles of CEO and board chair not only would improve directors’ ability to perform their oversight responsibilities, but also could provide independent board leadership necessary to curtail the kind of risky and inappropriate managerial behavior that had contributed to the company’s financial woes.6 * Leroy Sorenson Merrifield Research Professor of Law, George Washington University Law School. Special thanks to Steven Davidoff and all of the students in the Fall 2013 Law and Capital Markets Seminar at the Ohio State University Moritz College of Law for their helpful comments and insights. All errors, of course, are mine. 1. See 2013 Notice of Annual Meeting of Shareholders and Proxy Statement, JPMORGAN CHASE & CO., Apr. 20, 2013, at 44, available at http://files.shareholder.com/downloads/ONE/ 2493146808x0x652544/b2a9705c-e6d5-4060-aaf2-418933ed0001/JPMC_2013_Definitive_ Proxy_Statement_r65_web_post_.pdf archived at http://perma.cc/9552-EE5Q [hereinafter JP Morgan 2013 Proxy Statement]; 2012 Notice of Annual Meeting of Shareholders and Proxy Statement, JPMORGAN CHASE & CO., Apr. 4, 2012, at 38, available at http://files.shareholder.com/ downloads/ONE/2493146808x0xS19617-12-185/19617/filing.pdf, archived at http://perma.cc/ GW97-FBNP [hereinafter JPMorgan 2012 Proxy Statement]. This Article uses the term “independent chair” or “independent board chair” to refer to a board chair who does not concurrently serve as the CEO. 2. JPMorgan 2012 Proxy Statement, supra note 1, at 40. 3. Id. at 38-39. 4. Id. at 39. 5. JPMorgan 2013 Proxy Statement, supra note 1, at 44. 6. Id. (noting that an independent board chair would be “particularly constructive” because the London Whale scandal had tainted the CEO’s reputation as a risk manager and raised questions 238 INDIANA LAW REVIEW [Vol. 47:237 Although the effort to separate the CEO and board chair positions failed at JP Morgan in both years,7 it reflects part of a growing movement by shareholders and others in support of independent board chairs; a movement that has grown in intensity since the financial crisis.8 Indeed, as the authors of one recent study note, when, as a result of a financial crisis, public corporations and their boards come under fire for a lack of accountability and appropriate oversight, the issue of separating the CEO and board chair roles “is often front and center.”9 Activist shareholders, institutional investors, and regulators alike, believe that separating such roles increases the board’s independence from management, thereby enhancing the board’s monitoring and oversight functions while simultaneously reducing the potential for managerial misconduct.10 Propelled by these sentiments, the percentage of companies that have separated the CEO and board chair roles has steadily climbed since the financial crisis.11 In 2007, 65% of board chairs at S&P 500 companies also held the office about the board’s oversight); JPMorgan 2012 Proxy Statement, supra note 1, at 39 (noting that an independent board chair would be “particularly constructive” in light of federal and state investigations aimed at JPMorgan). 7. The vote only received 32.2% shareholder approval in 2013. See Barry B. Burr, Shareholders Fall Short on J.P. Morgan Chase Independent Chair Vote, PENSIONS & INVESTMENTS (May 21, 2013), http://www.pionline.com/article/20130521/DAILYREG/130529981, archived at http://perma.cc/4BK9-E4QR. The 2012 vote received 40% shareholder approval. See Jessica Silver-Greenberg & Susanne Craig, Stockholder Power Faces Test at JPMorgan, N.Y. TIMES, May 19, 2013, http://mobile.nytimes.com/blogs/dealbook/2013/05/19/jpmorgan-chase-vote-tests- stockholders-power/ archived at http://perma.cc/4WBX-DKAW. 8. DELOITTE DEVELOPMENT LLC, HOT TOPICS: 2012 PROXY SEASON: LOOKING AHEAD TO 2013, at 2 (2012) [hereinafter HOT TOPICS], available at http://www.corpgov.deloitte.com/binary/ com.epicentric.contentmanagement.servlet.ContentDeliveryServlet/USEng/Documents/Deloitte %20Periodicals/Hot%20Topics/December%202012%20Hot%20Topics_Deloitte_2012%20Prox y%20Season_Looking%20Ahead%20to%202013_Final.pdf, archived at http://perma.cc/V93V- AQN6; William Kelly & Mutya Harsch, Director Notes: Directors’ Duties Under the New SEC Rules Disclosure Enhancement, THE CONFERENCE BOARD, Feb. 2010, at 7, available at http:// w w w . d a v i s p o l k . c o m / f i l e s / P u b l i c a t i o n / 7 d 3 ff4 1 3 - 0 d 1 c - 4 1 1 f- b 4 9 9 - 0 1 2 2 3 e 8 7 0 d 4 c / Presentation/PublicationAttachment/4807ffe7-2ecf-477f-8072-09333153775a/DN-005-10.pdf, archived at http://perma.cc/U2DM-4FWP (noting that the number of proposals in this area has increased in recent proxy seasons). 9. Richard Leblanc & Katharina Pick, Director Notes: Separation of Chair and CEO Roles, THE CONFERENCE BOARD, Aug. 2011, at 1, available at http://www.yorku.ca/rleblanc/publish/ Aug2011_Leblanc_TCB.pdf, archived at http://perma.cc/533A-DUCQ; see HOT TOPICS, supra note 8, at 3 (noting heightened focus on board leadership structure and accountability by shareholders). 10. Leblanc & Pick, supra note 9 at 1112. 11. SPENCER STUART, 2012 SPENCER STUART BOARD INDEX 12 (2012) [hereinafter 2012 SPENCER STUART BOARD INDEX], available at http://content.spencerstuart.com/sswebsite/pdf/lib/ Spencer-Stuart-US-Board-Index-2012_06Nov2012.pdf, archived at http://perma.cc/5WLU-AE34. 2014] SEPARATION ANXIETY 239 of CEO.12 By 2012, that number had declined to 57%.13 Thus, 43% of S&P 500 boards currently have separated the roles of CEO and board chair.14 Moreover, in the five years from 2007-2012, the number of companies with truly independent board chairs (i.e., board chairs who are not current or former executives of the companies at which they currently serve as chair) had nearly doubled, going from 13% to 23%.15 Like the shareholders at JPMorgan, advocates insist that separating the roles of CEO and board chair will improve board oversight, leading to better corporate governance and improved corporate performance.16 In their view, such separation negates the concentration of power and conflicts of interests inherent in a board leadership structure that combines the two roles.17 Thus, the separation facilitates the checks and balances necessary for appropriate managerial accountability that enables the board to effectively carry out its responsibilities.18 Of course, boards and other commentators disagree about the benefits associated with splitting the CEO and board chair roles. Opponents of such a split insist that the separation not only ignores the benefits of CEO duality (a board structure that combines the two roles), but also ignores the costs associated with a board leadership structure that relies on an independent board chair.19 These concerns appear to be reflected in the kind of support shareholder proposals seeking a split of the CEO and board chair roles have received from the broader shareholder class and proxy advisory firms. Unlike other corporate governance proposals that have witnessed average shareholder supports of 50% or more in recent years,20 shareholder proposals calling for a split of the CEO and board roles have received more modest levels of shareholder support that, on average, fall short of a majority.21 Then too, proxy advisory firms have been equivocal in their support for the independent board chair, suggesting that other board leadership models may be as appropriate depending on the company and company specific needs.22 12. Id. 13. Id. 14. Id. 15. Id. 16. Leblanc & Pick, supra note 9, at 2; Thuy-Hga T. Vo, To Be or Not to Be Both CEO and Board Chair, 76 BROOK. L. REV. 65, 84 (2010). 17. Leblanc & Pick, supra note 9, at 2; Vo, supra note 16, at 84. 18. Paul Hodgson & Greg Ruel, The Costs of a Combined Chair/CEO, GMIRATINGS, June 2012, at 1, available at http://info.gmiratings.com/Portals/30022/docs/gmiratings_ceochaircomp_ 062012.pdf, archived at http://perma.cc/5YDC-76JA; Oded Palmon & John Wald, Are Two Heads Better than One? The Impact of Changes in Management Structure on Performance by Firm Size, 8 J. CORP. FIN. 213, 224-25 (2002). 19. James A. Brickley et al., Leadership Structure: Separating the CEO and Chairman of the Board, 3 J. CORP. FIN. 189, 192-96 (1997); Vo, supra note 16, at 78. 20. See infra note 41. 21. See infra notes 44-45. 22. See infra notes 126-27 and accompanying text. 240 INDIANA LAW REVIEW [Vol. 47:237 This Article critically examines the competing arguments related to splitting the roles of CEO and board chair. Although the campaign for independent board chairs has received increased attention from shareholders and regulators,23 there has been very little academic analysis of such campaign.24 This Article seeks to fill this void not only by examining the campaign, but also by assessing its implications in light of the available empirical evidence and normative claims. Based on this assessment, this Article offers two conclusions. First, while there appear to be costs associated with splitting the roles of CEO and board chair, those costs likely have been overstated. Second, there are clear benefits associated with having an independent board chair. However, whether a corporation can take advantage of those benefits may depend upon various factors and circumstances, some of which may be difficult to achieve. Whether corporations can realize the benefits of separating the board and CEO roles may depend on whether corporations have truly independent board chairs, and many corporations do not. It also may depend on corporate size as well as the extent to which corporations have in place structures and processes ensuring that their outside board chairs have access to appropriate and diverse information sources so that they need not rely solely on their inside CEOs and thus can be effective monitors and leaders. Hence, this Article offers conditional support for splitting the roles of CEO and board chair. As a result, this Article argues that efforts to mandate such a split at all public companies could be counterproductive because such efforts may not appropriately consider the costs of such a split; and those efforts may not appropriately consider that while there are clear benefits to such a split, whether those benefits can be realized may depend on several variables that may not be present at every company. In this regard, when considering whether to split the roles of CEO and board chair, caution is warranted. Part I of this Article demonstrates the manner in which the corporate governance landscape has shifted toward a board- leadership structure that embraces the independent board chair. Part II discusses the empirical evidence associated with the costs and benefits of that embrace as it relates to financial performance, and will then draw important conclusions based on that evidence. Part II also pinpoints the limitations associated with the admittedly large body of empirical evidence in this area. Part III examines the normative case related to the independent board chair. Part IV offers a conclusion. I. A GRADUAL SHIFT TOWARDS THE INDEPENDENT CHAIR Although the issue regarding whether to separate the CEO and board chair 23. See infra notes 28, 29. 24. To be sure, there is a significant body of empirical literature aimed at assessing the financial impact of splitting the roles of CEO and board chair. See infra Part II. However, my research unearthed only two comprehensive scholarly discussions on the issue. See Constance Bagley & Richard Koppes, Leader of the Pack: A Proposal for Disclosure of Board Leadership Structure, 34 SAN DIEGO L. REV. 149, 152 (1997) (discussing the merits of independent board leadership in the context of a proposal for same); Vo, supra note 16, at 118. 2014] SEPARATION ANXIETY 241 roles has been debated for years,25 the financial crisis and other governance failures have thrust it into the spotlight as advocates insist that such a separation can help improve board oversight and thus prevent corporate wrongdoing.26 Governance experts insist that separation of the CEO and board chair roles represents the most optimal board structure.27 Shareholder advocates echo this sentiment. In its 2010 policy survey, the proxy advisory firm Institutional Shareholder Services (“ISS”) found that a substantial majority of investors believed that the CEO should not concurrently hold the role of board chair.28 On the heels of the financial crisis, regulators took up the calls for an independent board chair. In response to that crisis, legislators in both the Senate and House introduced bills that would have required public companies to have an independent board chair.29 In December 2009, the Securities and Exchange Commission (the “SEC”) approved new rules requiring a company to disclose (a) whether and why the company has chosen to combine or separate the principal executive officer and board chair positions, and (b) why the company believed that its leadership structure is the most appropriate.30 The SEC stated that the new rules were “not intended to influence a company’s decisions regarding its board 25. Leblanc & Pick, supra note 9, at 1 (noting that since the early 1980s the “combination (or separation) of the chair and CEO roles” has been among “the most hotly debated structural feature”). 26. Id. at 2; see also HOT TOPICS, supra note 8, at 3. 27. Leblanc & Pick, supra note 9, at 1. 28. See INSTITUTIONAL SHAREHOLDER SERVICES, INC., 2010-2011, POLICY SURVEY SUMMARY OF RESULTS 9 (2010), available at http://www.issgovernance.com/files/ISS2010- 2011_PolicySurveyResults.pdf, archived at http://perma.cc/SRM3-ULEY (finding that 76% of investors and 41% of issuers believe that the chair and CEO roles should be separate); see also INSTITUTIONAL SHAREHOLDER SERVICES, INC., 2011-2012, POLICY SURVEY SUMMARY OF RESULTS 16 (2011), available at http://www.issgovernance.com/files/PolicySurveyResults2011.pdf, archived at http://perma.cc/T5N3-9P65 (revealing that 70% of investors believe that a company should commit itself to adopting an independent chair model after the current combined CEO/chair leaves; only 11% of issuers held such a belief). 29. Shareholder Bill of Rights Act of 2009, S. 1074, 111th Cong. 8-9 (2009), available at http://www.gpo.gov/fdsys/pkg/BILLS-111s1074is/pdf/BILLS-111s1074is.pdf; Shareholder Empowerment Act of 2009, H.R. 2861, 111th Cong. 5-7 (2009), available at http://www.gpo.gov/ fdsys/pkg/BILLS-111hr2861ih/pdf/BILLS-111hr2861ih.pdf (requiring that the chairman of the board of directors of a public company be an independent director who has not previously served as an executive officer). 30. Proxy Disclosure Enhancements, SEC Release No. 33-9089, at 43 [hereinafter SEC Release No. 33-9089], available at http://www.sec.gov/rules/final/2009/33-9089.pdf, archived at http://perma.cc/4YYJ-SJDB (for companies that combine the roles of CEO and chair, and rely on a lead independent director, the new rules require disclosure regarding why the company has a lead independent director and the specific role the lead independent director plays in the leadership of the company). See Bagley & Koppes, supra note 24, at 152 (As early as 1997, two professors recommended that listing agencies adopt a similar disclosure policy regarding board leadership.). 242 INDIANA LAW REVIEW [Vol. 47:237 leadership.”31 However, the new rules were part of a host of rules aimed at responding to the financial crisis by enhancing corporate governance and board accountability.32 Given that context, such rules may be viewed as playing a role in facilitating, if not supporting the general push for the separation of the CEO and board chair roles. Importantly, the pressure for additional regulation remains. For example, in 2010, the United States Congress introduced three proposals calling for the separation of the CEO and board chair functions.33 In addition to agitating for regulatory reform, shareholder advocates have focused their sights on altering the board structures at major companies, particularly those embroiled in corporate scandal. The fight to separate the CEO and chair position at JPMorgan received significant attention given the prominence of the company and the significance of its financial woes. And the fight was part of a larger effort in this area. The 2012 proxy season witnessed a record number of proposals calling for the separation of the CEO and board chair roles.34 Shareholders submitted thirty eight such proposals in 2012, as compared to twenty-five in 2011.35 In 2012 and 2013, proposals to split the CEO and board roles were the second most prevalent shareholder proposal type—second only to proposals related to political spending.36 As of May 2013, eighteen Fortune 250 companies had faced or were being faced with proposals to split the CEO and board chair roles.37 Shareholder support for such proposals can best be described as strong but cautious. As noted in the introduction, both proposals failed at JPMorgan.38 Moreover, a relatively small number of proposals have garnered majority support. Only three such proposals passed in 2011,39 while only two passed in 2012.40 31. SEC Release No. 33-9089, supra note 30, at 43. 32. Press Release, Sec. Exch. Comm., SEC Approves Enhanced Disclosure About Risk, Compensation and Corporate Governance (Dec. 16, 2009), available at http://www.sec.gov/news/ press/2009/2009-268.htm, archived at http://perma.cc/NL95-LGYW. 33. Tina Yang & Shan Zhao, CEO Duality, Competition, and Firm Performance, at 3-4 (2012) available at http://www.lehigh.edu/~jms408/yang_2012.pdf, archived at http://perma.cc/ 3SGX-FR3W. 34. HOT TOPICS, supra note 8 at 3; SULLIVAN & CROMWELL LLP, 2012 PROXY SEASON REVIEW 1 (2012) [hereinafter SULLIVAN REVIEW], available at http://www.sullcrom.com/files/ Publication/fdd28332-7b79-4d37-9ada-89da9bc111a9/Presentation/PublicationAttachment/ d546858a-ee4a-43af-b379-170d4995e41c/2012_Proxy_Season_Review-7-20-2012.pdf, archived at http://perma.cc/SQY9-86RD. 35. SULLIVAN REVIEW, supra note 34, at 1. 36. Press Release, Proxy Monitor, 2013 Season Under Way, available at http://proxymonitor. org/Forms/2013Finding2.aspx, archived at http://perma.cc/Q9V6-DCSE [hereinafter Proxy Monitor 2013 Season] 37. Id. 38. Burr, supra note 7 (the failure of the proposition in 2013); Silver-Greenberg & Craig, supra note 7 (the failure of the proposition in 2012). 39. The companies were Aetna, Moody’s, and Vornado Realty. Ted Allen et al., 2011 U.S. Postseason Report, INSTITUTIONAL SHAREHOLDERS SERVICES INC., Sep. 29, 2011, at 24, available 2014] SEPARATION ANXIETY 243 Then too, the average level of shareholder support falls short of the level of shareholder support for other governance proposals in the wake of the financial crisis and corporate governance scandals, such as majority voting or board declassification41 where the average support has been 50% or higher for several years.42 Nonetheless, shareholder support can still be considered relatively strong. In 2012, such proposals averaged 35% of the votes cast,43 with a similar level of support for 2011.44 As of May 2013, although the average support for the separation of CEO and chair roles among Fortune 250 companies was down to 27%,45 shareholder proposals at three companies in the Fortune 250 received over 40% shareholder support.46 at http://www.issgovernance.com/files/private/2011_US_PostSeason_Report_0929.pdf, archived at http://perma.cc/HS5U-EB64. A proposal at Cedar Fair, a non-Russell 3000 company, also won majority support. Id. As a result of the votes, Aetna expanded the duties of its presiding director, Cedar Fair appointed an independent chair, and Moody’s agreed to appoint an independent chair in 2012. See Shirley Westcott, 2012 Proxy Season Review: Shareholder Proposals, ALLIANCE ADVISORS, Sep. 2012, at 5 n.12, available at http://allianceadvisorsllc.com/dimages/file_49.pdf, archived at http://perma.cc/Q4CQ-ZQ4P. 40. HOT TOPICS, supra note 8, at 3 (citation omitted). 41. See Lisa M. Fairfax, Making the Corporation Safe for Shareholder Democracy, 69 OHIO STATE L.J. 53, 66-67, 70-71 (2008) [hereinafter Fairfax, Making the Corporation Safe]. 42. Id. See RISKMETRICS GROUP, RISKMETRICS GROUP POSTSEASON REPORT: A NEW VOICE IN GOVERNANCE: GLOBAL POLICYMAKERS SHAPE THE ROAD TO REFORM 5 (2009) (The average shareholder support for majority voting proposals was 56% in 2009 and 50% in 2008 and 2007. In 2009, the average shareholder support for board declassification was 63%, compared to 67% in 2008 and 6363636463% in 2007. In 2011, the average shareholder support for majority voting was 56.6%, while the average shareholder support for board declassification was 70%.), available at https://www.governanceexchange.com/repository/KnowledgeGateway/pubs/2009_PSR_Public_ final.PDF, archived at http://perma.cc/B9Z3-UBRW; see also INSTITUTIONAL SHAREHOLDER SERVICES, 2011 U.S. PROXY SEASON SCORECARD (2011), available at http://www.issgovernance. com/files/private/2011ProxySeasonScorecard_20110606.pdf, archived at http://perma.cc/XH5V- MDEP. 43. See SULLIVAN REVIEW, supra note 34, at 1 (such proposals received an average of 34% support at Fortune 200 companies in 2012); James R. Copland et al., Proxy Monitor 2012: A Report on Corporate Governance and Shareholder Activism, PROXY MONITOR, Fall 2012, at 18, available at http://www.proxymonitor.org/pdf/pmr_04.pdf, archived at http://perma.cc/D4LJ-MF2H; ERNST & YOUNG LLP, FOUR KEY TRENDS OF THE 2012 PROXY SEASON: ENGAGEMENT DRIVES CHANGE 5 (2012) (average support at Russell 3000 companies was as high as 37%), available at http://www.ey.com/Publication/vwLUAssets/Four_key_trends_of_the_2012_proxy_season/$FI LE/1207-1372854_ProxyGovernance_CF0035_071612.pdf, archived at http://perma.cc/W3CB- 6QU4. 44. See SULLIVAN REVIEW, supra note 34, at 1 (revealing average support of 34%); Allen et al., supra note 39, at 24 (revealing that proposals earned 32.8% average support at Russell 3000 companies). 45. Proxy Monitor 2013 Season, supra note 36. 46. Id. (those companies included Honeywell, Boeing and IBM). 244 INDIANA LAW REVIEW [Vol. 47:237 Irrespective of whether these proposals have received widespread shareholder support, the push to separate the roles of CEO and board chair has prompted many corporations to voluntarily change their board leadership structure. In 2002, only 25% of S&P 500 companies maintained boards with separate roles for CEO and board chair.47 By 2007, that number had increased to 35%.48 By 2012, 43% of S&P 500 companies had adopted a leadership structure comprising separate roles for the board chair and CEO.49 Many prominent companies across industries have separated their CEO and board chair positions, including Proctor & Gamble, Visa, Starbucks, and FedEx.50 Importantly, the percentage of corporations with truly independent board chairs has nearly doubled in five years. A truly independent board chair is someone who has no significant relationship with the corporation outside of being a board member and is neither a current executive nor a former CEO or executive of the company for which she is currently serving as board chair.51 In 2007, 13% of board chairs were truly independent.52 That number had increased to 23% in 2012.53 Thus, the percentage of both independent board chairs and truly independent chairs has increased since the financial crisis. To be sure, the shift in board leadership structure is not necessarily permanent. A 2012 survey revealed that “[o]nly eighteen companies . . . report[ed] having a formal policy requiring the separation of the CEO and chair roles.”54 The lack of such a policy means that corporations are free to alter their board leadership structure whenever they choose. Corporations have taken advantage of this freedom. For example, Disney separated the roles of CEO and board chair in 2005 and then restored them in 2012.55 Similarly, the CEO of Dell relinquished his chair position in 2004, only to step back into that role three years later.56 Consistent with this anecdotal evidence, in 2012, eight companies that had separated the CEO and board chair roles had returned to combining them.57 Nonetheless, the overall empirical evidence reflects a growing trend towards independent board chairs. 47. See 2012 SPENCER STUART BOARD INDEX, supra note 11, at 10. 48. See id. 49. Id. 50. Elizabeth Olson, Why the CEO-Chair Split Matters, CNN MONEY (Mar. 12, 2013), http://management.fortune.cnn.com/2013/03/12/ceo-chair-split/, archived at http://perma.cc/UAK6- CGWC. 51. See Hodgson & Ruel, supra note 18, at 2 (describing characteristics of a non-independent board chair). 52. 2012 SPENCER STUART BOARD INDEX, supra note 11, at 10. 53. Id. 54. Id. at 23. 55. See Olson, supra note 50. 56. Nell Minow, Independent Chairmen Are Smart Investments, BLOOMBERG (Jul. 17, 2012), http://www.bloomberg.com/news/2012-07-17/independent-chairmen-are-smart-investments-nell- minow.html, archived at http://perma.cc/HX3S-XXP3. 57. 2012 SPENCER STUART BOARD INDEX, supra note 11, at 23. 2014] SEPARATION ANXIETY 245 This trend is consistent with the broader shift towards greater independence on the board as a whole. Indeed, as a result of federal regulations and a growing consensus related to best practices,58 “[t]he percentage of independent directors on S&P 500 boards has increased from 79% in 2002 to 84% in 2012.”59 Today, almost every corporation has fewer than two non-independent directors on their board.60 As a result, the CEO increasingly represents the only non-independent director on the board. In 2002, the CEO was the only non-independent director on 31% of S&P 500 boards.61 By 2012, that number had risen to 59%.62 As these statistics reveal, this number has nearly doubled in the past decade. This data suggests that if shareholder advocates are successful in their efforts to create more independent board chairs, a sizeable majority of boards will be composed solely of independent directors. The remainder of this Article weighs the costs and benefits of this phenomenon. II. ASSESSING THE FINANCIAL PERFORMANCE DATA Grappling with the empirical evidence on the impact of independent board chairs on financial performance represents a daunting task.63 There is a significant body of empirical literature focused on this issue, comprising more than 30 empirical studies and meta-analyses.64 Several commentators have 58. See generally Lisa Fairfax, The Uneasy Case for the Inside Director, 96 IOWA L. REV. 127, 136-37 (2010) [hereinafter Fairfax, The Uneasy Case]. 59. 2012 SPENCER STUART BOARD INDEX, supra note 11, at 6. 60. See Fairfax, The Uneasy Case, supra note 58, at 136 (noting that 91% of companies had two or fewer inside directors in 2004). 61. 2012 SPENCER STUART BOARD INDEX, supra note 11, at 15. 62. Id. (In 2007, CEOs comprised the only non-independent directors at 43% of S&P 500 boards.). 63. This Article does not seek to separately evaluate the quality of the empirical studies themselves, but rather evaluates them based on their own stated conclusions and limitations. 64. See Protecting Shareholders and Enhancing Public Confidence by Improving Corporate Governance: Hearing Before the Subcomm. on Sec., Ins., and Inv. of the S. Comm. on Banking, Hous., & Urban Affairs, 111th Cong. 7 (2009) [hereinafter Coates Testimony] (statement of Prof. John C. Coates IV), available at http://www.banking.senate.gov/public/index.cfm?FuseAction= Hearings.Testimony&Hearing_ID=c754606c-0b95-4139-a38a-63e63b4b3fa9&Witness_ ID=49f23bdb-ae69-42a8-a6d5-82d7fb82502a, archived at http://perma.cc/XF7D-ZJDX. The analysis in this section is limited to those studies that focus on financial performance. There is also empirical evidence on the impact of separating the roles of CEO and board chair in the context of other issues (e.g., there is some empirical support for the proposition that independent board chairs can help curb corporate misconduct). See Hodgson & Ruel, supra note 18, at 4. The Hodgson and Ruel study suggest that combining the roles of CEO and board chair creates a greater potential for governance and management failures. Companies in the study that combined the roles faired far worse in ratings that tested for fraud and financial restatements. The study did include some companies that had split the roles, but they also fared poorly on such ratings. However, on average, 246 INDIANA LAW REVIEW [Vol. 47:237 characterized the empirical evidence regarding the impact of separating the CEO and board chair roles as mixed or weak.65 By contrast, at least one commentator has described the overall set of empirical evidence related to independent board chairs as providing strong support in favor of such separation.66 This Article argues that an overall assessment of the available data supports at least three conclusions. First, there may be costs associated with separating the two roles, but many of those costs have been exaggerated and outweighed by the benefits of such separation, particularly for long-term shareholders and at larger corporations. Second, there are clear benefits associated with separating the two roles, though the strengths of those benefits and whether they can be realized vary. Third, truly independent board chairs impact financial performance, but that impact also varies. A. Evidence of Impact on Financial Performance Available research only revealed one study that appeared to unequivocally support the view that CEO duality positively correlates with firm performance. The 1991 study found that CEO duality was associated with higher levels of average return on equity.67 However, the authors were careful about drawing any broad conclusions, and qualified their results by suggesting that combining the roles of CEO and board chair does not produce adverse consequences for corporate performance, and actually could produce benefits.68 While a few other studies identified a positive connection between CEO duality and firm performance, all of those studies included important limitations or countervailing evidence. One study examined a period between 1979 and 1998, and found that companies with CEO duality outperformed those companies with separate roles in environments where there is increased competition.69 This was primarily because firms with CEO duality can access information more quickly, which is an important benefit when competition intensifies.70 However, the positive effects of CEO duality disappeared once the study controlled for other governance mechanisms.71 Moreover, the study did not distinguish between non-CEO board chairs and those chairs who also serve as a former or present companies with an independent board chair scored better on metrics aimed at identifying items that might signal potential fraudulent accounting statements 65. See Leblanc & Pick, supra note 9, at 2 (describing the evidence as “not definitive”); Coates Testimony, supra note 64, at 7 (describing the evidence as “more mixed” than research on other board proposals). 66. See Vo, supra note 16, at 118. 67. Lex Donaldson & James H. Davis, Stewardship Theory or Agency Theory: CEO Governance and Shareholder Returns, 16 AUSTL. J. MGMT. 49, 56 (1991). 68. Id. at 61. 69. Yang & Zhao, supra note 33, at 1-2, 22, 31. 70. Id. 71. Id. at 31. 2014] SEPARATION ANXIETY 247 employee of the firm.72 These limitations could undermine the saliency of the study’s findings because the latter chairs may not be viewed as truly independent. Along similar lines, after assessing over 600 firms, a 1997 study concluded that firms that split the CEO and board chair roles “do not necessarily have lower accounting returns” than firms that combined the roles.73 The study’s authors warned that their results “should be interpreted with caution,” because their tests “do not control for other potential determinants of firm performance.”74 Three other studies revealed both positive and negative connections between firm performance and CEO duality. A 1995 study found support for both CEO duality and separating the CEO and board chair functions.75 Similarly, a 2002 study of 157 announcements of board changes from 1986 to 1999 found that small firms experienced negative abnormal returns when changing from a combined board leadership model to a split model, while large firms experienced positive abnormal returns.76 Along these same lines, a 2012 study found that “[i]n the short term, companies with combined chair and CEO roles fared much better than those companies with a separate CEO and chair” roles.77 However, over a longer period, companies with separate CEO and board chair roles had shareholder returns nearly 28% higher than those with the combined roles.78 At least one broader assessment of available empirical data suggests that there is no connection between board leadership structure and company performance. Thus, a 2007 meta-analysis of the empirical literature found no evidence to support the connection between corporate performance and leadership structure.79 By comparison, several studies have reflected a weak connection between board leadership structure and performance. A 1998 meta-analysis of thirty-one studies concluded that there was relatively little evidence of a systemic relationship between financial performance and board leadership structure.80 Similarly, a 1996 study found only weak evidence that combining the roles of CEO and board chair negatively impacts long-term performance (performance over a five year period from 1986-1991), after controlling for other factors that 72. Id. at 32. 73. Brickley et al., supra note 19, at 211 (finding mixed evidence on the impact of CEO duality on firm performance). 74. Id. 75. Brian K. Boyd, CEO Duality and Firm Performance: A Contingency Model, 16 STRATEGIC MGMT. J. 301, 309 (1995). 76. Palmon & Wald, supra note 18, at 216, 222. 77. See Hodgson & Ruel, supra note 18, at 4. 78. Id. (study was over five years). 79. Dan Dalton et al., The Fundamental Agency Problem and Its Mitigation: Independence, Equity, and the Market for Corporate Control, THE ACADEMY OF MANAGEMENT ANNALS, 1, 13 (Royston Greenwood ed., 2007) (“[t]here is no evidence of substantive, systematic relationships between corporate financial performance and board leadership structure”). 80. Dan Dalton et al., Meta-Analytical Reviews of Board Composition, Leadership Structure, and Financial Performance, 19 STRATEGIC MGMT. J. 269, 278 (1998). 248 INDIANA LAW REVIEW [Vol. 47:237 might impact performance.81 In contrast, several studies reveal a strong link between financial performance and an independent board chair. A 1978 study, one of the earliest, found that Fortune 200 companies that combined the CEO and board chair roles had significantly lower stock price appreciation and return on equity than companies that separated such roles.82 A 1991 study found that during 1978 and 1983, companies that separated the board chair and CEO positions had significantly higher returns on investment, average returns on equity, and average profit margins than companies with combined positions.83 A 1993 study found that, on average, between 1988 and 1990, companies in the banking industry that had separated the roles consistently outperformed those with CEO duality.84 A 2001 meta-analysis of twenty-two samples across 5,751 companies concluded that independent board leadership has a significant influence on performance, though the correlation varies by context.85 In support of their argument that an independent board chair has been found to improve financial performance, sponsors of the JPMorgan shareholder proposal pinpointed a 2007 Booz Allen Hamilton & Co. study which found that, “[i]n 2006, all of the underperforming North American CEOs with long tenure had either held the additional title of company chairman or served under a chairman who was the former CEO.”86 The Booz Allen Hamilton study indicated that investors enjoy higher returns over the long run when the chair is independent of the CEO.87 A 2010 meta-analysis of over fifteen studies, including several other meta-analyses, concluded that the empirical evidence provides “a convincing case that separating the CEO and Chair positions has a positive impact on corporate performance from both financial and nonfinancial perspectives.”88 81. B. Ram Baliga et al., CEO Duality and Firm Performance: What’s the Fuss?, 17 STRATEGIC MGMT. J. 41, 41 (1996). 82. Sanford V. Berg & Stanley K. Smith, CEO and Board Chairman: A Quantitative Study of Dual vs. Unitary Board Leadership, 3 DIRECTORS AND BOARDS 34, 35 (1978). 83. Paula L. Rechner & Dan R. Dalton, CEO Duality and Organizational Performance: A Longitudinal Analysis, 12 STRATEGIC MGMT. J. 155, 158-59 (1991). 84. Lynn Pi & Stephen Timme, Corporate Control and Bank Efficiency, 17 J. BANKING & FIN. 515, 529 (1993). 85. Dawna L. Rhoades et al., A Meta-analysis of Board Leadership Structure and Financial Performance: Are “Two Heads Better than One”?, 9 CORP. GOVERNANCE: AN INT’L REV. 311, 311 (2001). 86. See Chuck Lucier et al., The Era of the Inclusive Leader, 47 Strategy+Business, BOOZ & CO. 2, 4 (2007), available at http://www.strategy-business.com/media/file/sb47_07205.pdf, archived at http://perma.cc/D9VY-7VA7; JPMorgan 2013 Proxy Statement, supra note 1, at 44 (citing study); JPMorgan 2012 Proxy Statement, supra note 1, at 39 (same). 87. See Lucier et al., supra note 86, at 6. 88. Vo, supra note 16, at 118. 2014] SEPARATION ANXIETY 249 B. Concluding Assessments and Limitations of the Data First, while there is some support for the proposition that splitting the roles of the CEO and board chair can be costly, that support is relatively weak and tentative, suggesting that the costs of such a split—at least in terms of financial performance—may have been exaggerated.89 Indeed, in light of the wealth of empirical evidence on this issue, it is worth noting that there is only one study that appears to unequivocally find a positive association between CEO duality and firm performance.90 Importantly, the authors of that study seem less than confident in their conclusions, suggesting that the case for a positive correlation is tentative at best.91 Moreover, the remaining studies that demonstrate positive connections between combining the CEO and board chair roles and corporations’ financial performance do so only in the context of pinpointing countervailing negative results. Taken together, this data suggests that while there may be costs associated with splitting the roles of CEO and board chair, it is not clear how significant those costs are, and such costs must be weighed against the benefits, particularly benefits that flow to larger shareholders and those that flow over the long run when the CEO and board chair positions are split.92 Second, the empirical evidence provides ample support for arguments in favor of splitting the CEO and board chair roles. While the bulk of the empirical evidence admittedly falls along a spectrum, on balance such evidence does appear to tilt in favor of supporting the proposition that independent board chairs can enhance financial performance.93 There are certainly several studies as well as meta-analyses indicating little to no connection between board leadership structure and financial performance.94 Such findings suggest little reason to prefer one board structure over another, at least with respect to the potential for such a structure to enhance corporate performance.95 Moreover, they suggest that it is appropriate for boards to determine their leadership structure on a case-by- case basis. On the other hand, many more studies reflect a positive connection between firm performance and splitting the CEO and board functions.96 Thus, the case for splitting the roles of the CEO and board chair is much stronger than the case against such a split, supporting the proposition that there are important benefits to be gained from such a split. Third, and consistent with the first two conclusions, the empirical evidence suggests that whether splitting the roles of CEO and board chair produces positive benefits may depend on the context and circumstances. At least two studies 89. See supra notes 67-74 and accompanying text. 90. See Donaldson & Davis, supra note 67, at 56. 91. Id. 92. See Hodgson & Ruel, supra note 18, at 4; Palmon & Wald, supra note 18, at 216. 93. See infra Part II.A. 94. See supra notes 77-79 and accompanying text. 95. See Leblanc & Pick, supra note 9, at 3 (noting that “[n]o structural attribute of boards has ever been linked consistently to company financial performance”). 96. See supra notes 82-88 and accompanying text. 250 INDIANA LAW REVIEW [Vol. 47:237 revealed that the benefits of such a split are more pronounced in the long run.97 Those studies suggest that splitting the roles of CEOs and board chairs in larger companies may be more optimal, because larger more complex companies benefit from a structure that facilitates more effective checks and balances.98 The authors of a 2012 study, finding that companies with separate CEO and board chair roles fared significantly better in the long term, concluded: “Strong shareholder returns and sustainability extend beyond separating the role of CEO and chair, however, the decision to have the roles separate is likely to set off a chain reaction of decisions at the company that are made with the proper checks and balances in place.”99 Confronted with this kind of evidence, even an opponent to mandating the separation of the CEO and board chair roles conceded that the empirical evidence indicated that the split might be a good idea for larger companies.100 Fourth, the evidence suggests that truly independent board chairs impact performance, but that impact may vary. The one study (a 2013 study) that focuses specifically on truly independent board chairs found a much stronger connection between such board chairs and corporate performance.101 Importantly, that study found that there was virtually no impact on firm performance when the CEO and board chair roles were split, but the board chair had some connection to the corporation, thereby making him non-independent.102 However, a significant connection emerged for board chairs who were truly independent.103 The authors of the study concluded that separating the CEO and board chair positions can be beneficial when there is true independence of the board chair, but they also emphasized that even with a truly independent board chair, the benefits could be realized only under the right circumstances.104 With respect to this data on truly independent board chairs, it is important to point out that, outside of this study, none of the empirical evidence seeks to distinguish between the impact of non-independent board chairs and truly independent chairs. Indeed, the number of truly independent board chairs has been relatively low throughout many of the periods focused on in the available empirical research. In 1988, almost no major firm in the United States had an independent outsider as board chair.105 Instead, in almost all cases, the chair was the former CEO or a person with ties to the firm.106 As recently as 2007, only 97. See Hodgson & Ruel, supra note 18, at 4; see also Palmon & Wald, supra note 18, at 216. 98. Palmon & Wald, supra note 18, at 223. 99. Hodgson & Ruel, supra note 18, at 4. 100. Coates Testimony, supra note 64, at 77. 101. Ryan Krause & Matthew Semadeni, Director Notes: CEO-Board Chair Separation—If It Ain’t Broke, Don’t Fix It, THE CONFERENCE BOARD, Jun. 2013, at 2-3, available at http://www. conference-board.org/retrievefile.cfm?filename=TCB-DN-V5N11-13.pdf&type=subsite. 102. Id. 103. Id. 104. Id. 105. Brickley et al., supra note 19, at 218. 106. Id. 2014] SEPARATION ANXIETY 251 13% of directors were truly independent.107 In 2005, only 9% of directors were truly independent.108 In light of those small numbers, one study acknowledged the difficulties of pinpointing truly independent board chairs, but also acknowledged that the failure to tease out the impact of truly independent board chairs on firm performance could limit the saliency of the study.109 Other studies do not appear to recognize that this distinction that could impact their data. The fact that there may be considerable differences in corporate performance associated with truly independent board chairs as compared to board chairs that have some connection to the corporation, and that almost no study accounted for these differences, raises questions about the strength of the available empirical evidence as a whole. To summarize, the literature points to the conclusion that separating the CEO and board chair roles can yield positive financial results. However, (1) those results may depend on the context, (2) those results must be weighed against the admittedly tentative evidence on the costs of such separation, and (3) those results likely depend significantly on whether the board chair can be classified as truly independent. III. THE NORMATIVE CASE FOR THE INDEPENDENT BOARD CHAIR While the governance community insists that splitting the roles of CEO and board chair represents the most optimal board leadership structure, its opponents contend that such a split can be costly. This section reveals that while there may be costs associated with splitting the roles, such costs may have been overstated, and such costs may be outweighed by the benefits that flow from an independent board chair. In this regard, this assessment is consistent with the conclusions drawn from the empirical data on financial performance. A. The Perils of the CEO/Board Split 1. The Merits of Unity.—Those who oppose efforts to mandate independent board chairs contend that there are important benefits associated with combining the roles of CEO and board chair. As an initial matter, some wonder whether the appointment of an independent board chair raises additional agency costs because there may be no one to monitor such a chair.110 More importantly, separating the two roles creates possible confusion both within and outside of the corporation.111 In contrast, CEO duality creates clear and unambiguous lines of authority and accountability, which is essential to effective management and leadership.112 At least one study of boards in the United States and the United Kingdom revealed 107. 2012 SPENCER STUART BOARD INDEX, supra note 11, at 10. 108. SPENCER STUART, 2010 SPENCER STUART BOARD INDEX 8 (2010), available at http://content.spencerstuart.com/sswebsite/pdf/lib/ssbi2010.pdf. 109. Yang & Zhao, supra note 33, at 32. 110. Brickley et al., supra note 19, at 194. 111. Id. at 195. 112. Leblanc & Pick, supra note 9, at 2. 252 INDIANA LAW REVIEW [Vol. 47:237 that separating the roles of CEO and board chair paves the way for confusion that leads to struggles over power, territory, and accountability.113 Consistent with this study, some point to the public disagreement between GM’s chair and CEO as an example of the pitfalls associated with separating such roles.114 Of course, this anecdotal and empirical evidence does not necessarily condemn separation, but rather suggests that when companies separate such roles, it is also imperative to clearly define the roles and lines of authority associated with each position. Indeed, the same study that pinpointed concerns about confusion stemming from splitting the roles of CEO and board chair not only revealed that too often there was no clear and defined job description for those who served as an independent chair, but also that there was a wide range of roles and activities that fell under the purview of board chairs at different companies.115 This suggests that the confusion associated with separation is not endemic to the separation itself, but instead may stem from the failure to clearly pinpoint the roles and responsibilities of the two positions. Then too, separating the roles also can lead to better clarity for those who occupy the positions. The board chair is the leader of the board and its team of monitors. By contrast, the CEO is head of the corporate managerial team. Combining the roles serves to blur the distinction between these two functions.116 Recent years have ushered in an increased expectation that the board would take its monitoring role seriously, as shareholders and others increasingly look to the board to engage in active monitoring over corporate affairs. As the pressure increases on directors to take their monitoring obligations more seriously, pressure mounts to more clearly define directors’ monitoring obligations and distinguish them from the managerial role.117 Creating clear lines of power and authority may be especially important for CEOs because it may encourage them to recognize and pay heed to the boundaries of their authority.118 Additionally, studies revealing the benefits associated with a combined board leadership structure must be weighed against the drawbacks, and the drawbacks appear to be more acute at larger companies and over the long term. Thus, studies reveal that there are distinct benefits to having an independent board 113. Id.; Jay Lorsch & Andy Zelleke, Should the CEO be the Chairman?, 2 MIT SLOAN MGMT. REV. 46, 70-74 (2005). 114. Amy Goodman et al., Considerations for Public Company Directors in the 2012 Proxy Season, GIBSON, DUNN & CRUTCHER LLP, Jan. 3, 2012, available at http://www.gibsondunn. com/publications/pages/ConsiderationsforPublicCompanyDirectors-2012ProxySeason.aspx, archived at http://perma.cc/T7NX-DNYM. 115. Leblanc & Pick, supra note 9, at 2 (citing Lorsch & Zelleke, supra note 113, at 70-74). 116. See Olson, supra note 50 (noting that “[c]ombining the two roles often seems a default for companies despite studies showing that the arrangement muddies clear lines of authority”). 117. Id. 118. See Bagley & Koppes, supra note 24, at 158 (noting that “the existence of a separate chair serves as a reminder to the CEO that he or she reports to, and serves at the pleasure of, the company’s board of directors”). 2014] SEPARATION ANXIETY 253 chair, including their ability to monitor more effectively.119 Moreover, studies suggest that because there is a greater potential for abuse in larger firms, these monitoring benefits outweigh the costs associated with separating the roles of CEO and board chair for larger firms.120 In light of these benefits, even those who oppose legislation that would mandate splitting the two positions concede the appropriateness of such legislation for larger companies.121 2. Independent Leadership by Any Other Name?—Most public companies do not dispute the need for independent board leadership; instead they insist that their current leadership structure provides such independence. There appears to be an emerging consensus about the need for independent board leadership.122 The disagreement lies in whether there is an optimal board structure for obtaining such independent leadership. Shareholder advocates favor separating the board chair and CEO roles; many companies insist that independent leadership can be obtained even when the CEO and board chair roles are combined. In its statement against the shareholder proposal for separating the roles of board and chair, JPMorgan insisted that such separation was unnecessary because its leadership structure “already provid[ed]” independent leadership and oversight.123 JPMorgan pointed out that all but one of its directors was independent as defined by the NYSE listing requirements.124 Most importantly, JPMorgan emphasized that it had appointed a presiding director to further provide independent leadership on the board.125 Among other things, the presiding director (1) was appointed by the independent directors, (2) presided over executive sessions and meetings at which the chair was not present, (3) had the authority to call meetings of independent directors, (4) approved board meeting agendas and schedules for each board meeting, and had the authority to add agenda items, (5) approved board meeting materials for distribution to and consideration by the Board, and (6) facilitated communication between the chair, CEO, and independent directors.126 JPMorgan argued that this structure was sufficient to ensure independent board leadership even without separating the roles of CEO and board chair. Empirical evidence reveals a significant trend in favor of lead or presiding directors. The duties of a lead or presiding director may vary by company, but as a general matter, a lead or presiding director is an independent director who, among other things, presides over executive sessions of the board—sessions comprised solely of independent directors. Thus, the lead director can provide a source of independent leadership even when non-independent directors, such 119. Id. at 152. 120. Palmon & Wald, supra note 18, at 223. 121. Coates Testimony, supra note 64, at 7. 122. Goodman et al., supra note 114. 123. JPMorgan 2012 Proxy Statement, supra note 1, at 39. 124. Id. 125. Id.; JPMorgan 2013 Proxy Statement, supra note 1, at 44. 126. JPMorgan 2012 Proxy Statement, supra note 1, at 39; JPMorgan 2013 Proxy Statement, supra note 1, at 44. 254 INDIANA LAW REVIEW [Vol. 47:237 as the CEO, serve on the board. This Article will refer to such a director as a lead director. The number of companies with lead directors has more than doubled in the last decade.127 In 2012, 92% of S&P 500 boards reported having a lead director.128 This is a dramatic change from 1996 when only 27% of companies that had combined the CEO and board chair roles had a lead director.129 There also is a growing consensus among directors that appointing a lead director is a good corporate governance practice. In 2003, 72% of directors believed it was the right thing to do.130 In 2007, 85% of directors agreed that it was appropriate to appoint a lead director.131 ISS has suggested that a board structure that includes a lead director may be an appropriate substitute for separating the CEO and board chair roles. In their guidance, ISS indicated that one of the most critical issues it would consider when deciding whether to recommend a vote in favor of proposals for calling the separation of the CEO and board chair roles was whether the company targeted by the proposal had a lead director with specified duties.132 According to ISS, the lead director’s role should include, among other things, (1) presiding over executive sessions and meetings at which the chair is not present, (2) serving as a liaison between the chair and independent directors, (3) approving information sent to the board, meeting schedules, and meeting agendas for the board, and (4) having the authority to call meetings of the independent directors.133 These duties appear consistent with those identified by JPMorgan. In this regard, ISS has suggested that a board structure that includes a lead director could provide the independent board leadership necessary to promote effective board oversight and good corporate governance. However, it is clear that a lead director is not the same as an independent chair. A board chair has authority that a lead director does not, including the ability to more proactively influence the meeting agenda, and to chair regular board meetings.134 Commentators insist that agenda control is a critical source of power.135 While the lead director has input in the board agenda, evidence suggests that such a director plays less of a role in developing the agenda or 127. 2012 SPENCER STUART BOARD INDEX, supra note 11, at 25 (noting the changes from 2004). 128. Id. 129. Bagley & Koppes, supra note 24, at 159 n.42 (citing KORN/FERRY INTERNATIONAL, 23RD ANNUAL BOARD OF DIRECTORS STUDY 23 (1996)). 130. KORN/FERRY INSTITUTE, 34TH ANNUAL BOARD OF DIRECTORS STUDY 7 (2008). 131. Id. 132. INSTITUTIONAL SHAREHOLDER SERVICES, 2012 U.S. PROXY VOTING SUMMARY GUIDELINES 19-20 (2011), available at http://www.issgovernance.com/files/2012USSummary Guidelines.pdf, archived at http://perma.cc/4EKE-QPHG. ISS also indicated that a company’s problematic governance and management issues would factor into ISS’s recommendation regarding whether to vote in favor of a proposal for splitting the board chair and CEO roles. See id. 133. Id. at 20. 134. Leblanc & Pick, supra note 9, at 10. 135. Bagley & Koppes, supra note 24, at 157 n.35. 2014] SEPARATION ANXIETY 255 otherwise ensuring that critical issues are included on the agenda.136 In this respect, it seems relatively clear that a board chair has more agenda setting power than a lead director. Moreover, because the board chair presides over all normal board sessions, the board chair, rather than the lead director, continues to be the primary source of leadership on the board. Importantly, empirical evidence suggests that the power and influence associated with the board chair means that any negative effects of combining the board chair and CEO functions cannot be entirely offset by installing a lead director.137 The differences between the roles of board chair and lead director mean that the presence of a lead director is not the only factor ISS will consider when making a recommendation regarding shareholder proposals for splitting the CEO and board chair positions. First, ISS recently has altered its policy, indicating that it would look more closely at the distinction between the lead director’s ability to “consult” or “review” materials as opposed to approve, which suggests concerns that a lead director may not have a sufficiently active role on the board to serve as a suitable replacement for an independent board chair.138 Second, ISS has indicated that, even when a company has a lead director with comprehensive duties, ISS would assess a company’s financial performance as well as the extent to which the company has problematic performance or management issues before recommending against a proposal to split the CEO and board chair roles.139 Indeed, ISS recommended a vote in favor of splitting the CEO and board chair positions at JPMorgan despite the fact that JPMorgan’s lead director had duties consistent with those outlined by ISS, based on the belief that the presence of such a director had been insufficient to provide an optimal level of board oversight.140 These nuances in the ISS policy reflect its concern that the lead director may have shortcomings that make splitting the roles of CEO and board chair the most optimal board structure. In fact, ISS considered changing its policy so that it would only recommend against proposals for splitting the CEO and board chair functions when a company has a compelling company-specific reason against such a split, even if the company has a lead director with appropriately defined duties.141 ISS indicated that the intent of such a change would be to encourage companies to 136. 2012 SPENCER STUART BOARD INDEX, supra note 11, at 27 (revealing that the lead director has less responsibility for such role than the combined CEO/Chair). 137. Vo, supra note 16, at 120. 138. SULLIVAN REVIEW, supra note 34, at 2. 139. INSTITUTIONAL SHAREHOLDER SERVICES, 2013 U.S. PROXY VOTING SUMMARY GUIDELINES 19-20 (2013), available at http://www.issgovernance.com/files/2013ISSUSSummary Guidelines1312013.pdf, archived at http://perma.cc/S2W9-4B8H. 140. Dawn Kopecki, JPMorgan Investors Urged to Split Chairman Role, Oust Directors, BLOOMBERG (May 5, 2013), http://www.bloomberg.com/news/2013-05-04/jpmorgan-should-name- chairman-to-watch-ceo-iss-tells-investors.html, archived at http://perma.cc/CXN9-SKAJ. 141. Independent Chair Shareholder Proposals (US), INSTITUTIONAL SHAREHOLDER SERVICES (2011), http://www.issgovernance.com/policy/2011comment/IndepChair, archived at http://perma. cc/S56C-73JS. 256 INDIANA LAW REVIEW [Vol. 47:237 explain why the role of board chair could not be filled by an independent director.142 Although ISS does not appear to have adopted such a change, its consideration of the change suggests a growing belief that the lead director may be suboptimal, and thus, that an independent board chair represented the most appropriate board leadership structure. Overall, this assessment reveals that the costs connected to splitting the CEO and board chair functions may have been overstated. This is because in many cases, not only can those costs be mitigated with appropriate planning, but also those costs may be offset by benefits, particularly those that flow to larger companies. Moreover, the assessment of lead directors reveals that companies with such directors may not necessarily be able to capture the benefits associated with independent board chairs. B. The Benefits of Separating the CEO and Board Chair Roles 1. The Corrupting Potential of Absolute Power.—Advocates for separation of the roles of CEO and board chair argue that such separation protects against potential abuse associated with over-concentration of power. The two most authoritative positions in the corporation and its boardroom are the CEO and the board chair.143 Combining the two positions concentrates the power of such offices within the hands of one individual, which creates the potential for abuse.144 However, concerns about over-concentration of power may be overstated in light of the fact that the power associated with the CEO and board chair roles has been diminished in recent years. In their paper, Embattled CEOs, Professors Marcel Kahan and Ed Rock argue that power has shifted away from the CEO as a result of two key phenomena: changes in the compensation and characteristics of the board, and increased shareholder power.145 The shift in power away from the CEO not only may represent an important accountability check for CEOs, particularly those who serve concurrently as board chair, but also may mitigate concerns regarding over-concentration of power. Both CEO and board chair power have been reduced in recent years by changes in board composition and authority, including the increased independence of the board and the enhanced authority of board committees.146 As a result of perceived best practices, as well as soft rules and regulations at the state and federal level,147 the audit committee, nominating committee, and 142. Id. 143. Hodgson & Ruel, supra note 18, at 1. 144. Id. 145. See Marcel Kahan & Edward Rock, Embattled CEOs, 88 TEX. L. REV. 987, 989 (2010) (noting that the loss of power “represents a significant move away from the imperial CEO who was surrounded by a hand-picked board and lethargic shareholders”). 146. Id. at 1007-11 (noting the impact of the decline in staggered boards and the rise of majority voting for directors). 147. Fairfax, Making the Corporation Safe, supra note 41, at 136-37, 139-45. 2014] SEPARATION ANXIETY 257 compensation committee of the board not only are comprised solely of independent directors,148 but also have increasingly more critical roles. Indeed, such committees not only have primary responsibility for key board functions such as oversight of financial reporting, director nominations, and executive compensation, but these committees also have been empowered with the ability to hire their own counsel and advisors.149 The augmentation of board committee power has shifted power away from the CEO, greatly undermining the CEO’s decision-making power.150 It also has shifted power away from the board as a whole, and thus away from the board chair, particularly if that chair is an inside director such as the CEO. Several key indicators underscore CEO’s diminished power. First, a key indicator of committees’ increased power is the number of times they meet.151 That number has not only grown over the years, but remained steady in recent years.152 Second, Kahan and Rock noted that when boards spend more time monitoring the CEO, the CEO is likely to have less power.153 The average number of hours board members work has increased through the years and Kahan and Rock indicate that these changes reflect that the boards are increasingly engaged in monitoring.154 Third, “the percentage of boards with a formal process for evaluating CEOs increased from the high in the sixties in 1997 and 2001 to around 92% in 2007.”155 This suggest a change in boardroom dynamics pursuant to which boards more actively assess CEOs, and thus supports Kahan and Rock’s thesis that CEOs have less power. Consistent with this concept, studies reveal that boards are becoming less tolerant of under-performing CEOs.156 Instead, boards are far more likely to challenge and terminate CEOs for poor performance.157 A Booz Allen Hamilton study revealed Annual turnover of CEOs across the globe increased by 59%between 1995 and 2006. In those same years, performance-related turnover—cases in which CEOs were fired or pushed out—increased by 148. Kahan & Rock, supra note 145, at 1022-23; see 2012 SPENCER STUART BOARD INDEX, supra note 11, at 11 (citing a growth in committee independence over five and ten year periods). 149. Dodd-Frank Wall Street Reform and Consumer Protection Act, Pub. L. No. 111-203, 124 Stat. 1376 (2010) (empowering compensation committee to hire its own independent counsel and compensation consultant). 150. See Kahan & Rock, supra note 145, at 1039 (stating that much of a CEO’s power has been delegated to the board committees of audit, compensation and nomination). 151. Id. at 1027-28. 152. Id. 153. Id. at 1025. 154. Id. at 1029. 155. Id. 156. Lucier et al., supra note 86, at 6 (finding that between 1995 and 2006, CEOs that delivered above-average returns to investors were more than twice as likely to stay in their position for more than seven years than CEOs who delivered subpar returns). 157. Id. 258 INDIANA LAW REVIEW [Vol. 47:237 318%. In 1995, only one in eight departing CEOs was forced from office. In 2006, nearly one in three left involuntarily.158 Increased CEO turnover, and board’s willingness to effect such turnover, is yet another indicator of the board’s enhanced independent oversight, as well as the diminished power of the CEO position.159 All of these indicators reflect reduced power of the CEO and board chair, and therefore may alleviate concerns that combining the CEO and board chair roles represents an unacceptable concentration of power. In recent years, the corporate governance landscape also has shifted so that shareholders have increasingly gained more power and authority over corporate affairs.160 Professors Kahan and Rock argue that increased shareholder power further reflects an erosion of CEO power. In their view, shareholder activism is a key source of the loss of CEO power.161 As shareholders have played a more active role in the shareholder proposal arena, the CEO has lost his or her agenda setting control to shareholders.162 This loss of control undermines CEO’s decision-making power. Shareholders’ ability to influence critical decisions within the corporation further erodes CEO’s power. Thus, shareholders not only play an increasingly greater role in the director nomination and director election process, but also have increasingly greater influence over compensation decisions.163 This enhanced shareholder power lessens the power held by both the CEO and the board chair, thereby potentially weakening the claim that combining these roles reflect an over-concentration of power. These observations, however, do not mean that over-concentration should not remain a concern. To be sure, one factor contributing to Professors Kahan and Rock’s assessment that CEOs are becoming less powerful is the erosion in the practice of having the CEO serve as board chair.164 Moreover, Professors Kahan and Rock argue that changes in boardroom structure, particularly those that focus on enhanced director independence, serve to undercut CEO power and dominance.165 This suggests that separating those roles may still be relevant. Two authors have theorized that increased independence on the board may prove inefficient unless it is coupled with splitting the roles of the CEO and board chair.166 From this perspective, separating the roles of CEO and board chair may continue to be necessary to avoid over-concentration. Increased shareholder power coupled with the increased power of 158. Id. at 3. 159. Kahan & Rock, supra note 145, at 1031-32. 160. Fairfax, Making the Corporation Safe, supra note 41, at 55. 161. Kahan & Rock, supra note 145, at 1000 (noting that activism by hedge funds has become a “prime irritant for CEOs”). 162. Id. at 1038-39. 163. Id. at 1039-40; see Fairfax, Making the Corporation Safe, supra note 41, at 61-79. 164. Kahan & Rock, supra note 145, at 1030. 165. Id. at 1042. 166. Bagley & Koppes, supra note 24, at 163-66. 2014] SEPARATION ANXIETY 259 independent directors and board committees may provide some check on the concentrated power stemming from combining the CEO and board roles. However, those increases in power do not eradicate the latter concerns, and thus, the combination of such roles may still prove problematic. 2. Conflicts of Interest.—Advocates also insist that separating the CEO and board chair roles helps avoid the inherent conflict of interests associated with combining such roles, leading to better and more effective monitoring and oversight. As one former board chair queried: “Is a company a sandbox for the CEO, or is the CEO an employee? If he’s an employee, he needs a boss and that boss is the board. The Chairman runs the board.”167 Monitoring the CEO and top management is perhaps the most fundamental function of the board. As the above quote suggests, when the CEO concurrently serves as the board chair, that structure creates concern regarding how the board can properly perform that function. Such concurrence suggests that the CEO is essentially charged with monitoring himself or herself.168 In other words, such a structure means that the CEO essentially acts as his or her own boss, creating an obvious conflict of interest that undermines effective oversight of the CEO and his or her decision- making. On the one hand, evidence suggests that combining the CEO and board chair positions does not necessarily undermine the board’s ability to evaluate, and in fact, terminate CEOs. As noted above, boards have an increased ability and willingness to actively monitor CEOs and to terminate those CEOs for poor performance.169 Evidence suggests that this increase may be connected to an increase in overall board independence, without regard to the independence of the board chair. Studies also find that companies with outside dominated directors are significantly more likely to remove CEOs on the basis of performance, as opposed to companies with insider dominated boards.170 These studies, however, do not take into account board leadership structure. All of this evidence supports the proposition that boards can effectively monitor CEOs irrespective of board leadership structure. On the other hand, there is clear and convincing evidence that combining the roles of CEO and board chair has a negative impact on the board’s ability to monitor the CEO. A recent study found that board leadership structure has a significant impact on the effectiveness of board’s oversight. That study revealed that CEOs generally have incentives to direct board attention away from active and robust monitoring.171 That study also revealed that boards’ ability to attend to all of the responsibilities associated with carrying out their duties is necessarily 167. Olson, supra note 50 (quoting the former chair of Intel, Andrew Grove). 168. Hodgson & Ruel, supra note 18, at 1. 169. See supra notes 151, 153. 170. See Michael S. Weisbach, Outside Directors and CEO Turnover, 20 J. FIN. ECON., 431, 432 (1988). 171. Christopher Tuggle et al., Commanding Board of Director Attention: Investigating How Organizational Performance and CEO Duality Affect Board Members’ Attention to Monitoring, 31 STRATEGIC MGMT. J. 946, 951 (2010). 260 INDIANA LAW REVIEW [Vol. 47:237 limited.172 As a result, boards do not consistently engage in monitoring, but instead selectively attend to that function.173 Whether and under what circumstances boards focus on monitoring is impacted by whether a corporation combines the roles of CEO and board chair.174 The study revealed that the combining the roles of CEO and board chair gives the CEO the perfect platform to divert the board from monitoring the CEO and top management.175 Thus, the combination of those roles leads to the board allocating lower attention to monitoring.176 First, the power associated with combining the positions of CEO and chair may enable the person holding the position to “create norms in which it is inappropriate to question management’s effectiveness,” thereby reducing the board’s ability to monitor.177 Second, the CEO-chair’s control of the agenda has an appreciable impact on boards’ attention to monitoring not only by focusing attention away from such monitoring, but also by making it difficult to engage in effective CEO evaluation and succession planning.178 Third, the CEO-chair may institute unacceptably low levels of board attention to monitoring during periods of positive performance. This may be done not only because there is a general tendency to focus board’s attention away from monitoring, but also because during periods of positive performance, there are less likely to be compelling reasons for the board to focus on monitoring.179 This is particularly troublesome when issues involving long-term planning are discussed because boards may not give due attention on those issues. Further, boards may not be able to effectively engage in proactive monitoring, which could have negative long term implications on company performance.180 These problems are exacerbated during times of poor performance or managerial misconduct. As one shareholder representative noted: “It is impossible to imagine how board oversight of the company’s affairs will be strengthened while [the CEO] leads the very board that is charged with overseeing his own shortcomings.”181 Combining the CEO and board chair roles “tilts the balance of power in favor of the CEO such that even as firm performance deteriorates, board monitoring can be impeded.”182 As the authors 172. Id. at 947. 173. Id. 174. Id. 175. Id. at 951. 176. Id. at 951, 960. 177. Id. at 951. 178. Id. at 960. 179. Id. 180. Id. 181. Press Release, Am. Fed’n of State, Cnty. & Mun. Employees, Major Investors Call on JPMorgan Chase to Name Independent Board Chair (Feb. 20, 2103), available at http://www. afscme.org/news/press-room/press-releases/2013/major-investors-call-on-jpmorgan-chase-to-name- independent-board-chair, archived at http://perma.cc/4EJE-Y6QN (citing Connecticut Treasurer Denise L. Nappier). 182. Tuggle et al., supra note 171, at 952. 2014] SEPARATION ANXIETY 261 of a study point out, when performance is poor, CEOs are more likely to use their influence to impact board’s monitoring in ways that protect the CEOs jobs.183 Importantly, while poor performance generally enhances all boards’ attention to their monitoring responsibilities, the study found that CEO-chairs disrupt this phenomenon.184 Thus, their evidence suggests that when faced with the threat of poor performance, CEO-chairs “utilize their power to combat the natural tendency of boards to increase attention to monitoring.”185 Another study found that the existence of a predecessor CEO as board chair dampens the ability of the new CEO to deliver performance that deviates from pre-succession levels or otherwise make strategic changes.186 Importantly, it is not clear that a lead director overcomes this problem. This is because the lead director only acts as chair when the board chair is not available. Hence, in most settings the lead director is not able to take the lead in setting the agenda on these critical issues, or otherwise appropriately focus the board’s attention on monitoring. However, merely separating the CEO and board chair may be insufficient to truly reduce the conflict of interest, unless companies also commit to ensuring that the board chair is truly independent. Financial performance data confirms that truly independent board chairs matter. Krause and Semadeni found that a CEO-chair separation without true independence of the board chair has virtually no impact on firm performance other than what one would expect from any CEO succession.187 The Booz Allen Hamilton study similarly reveals that separation without true independence of the board chair is not likely to have an impact on undermining conflicts of interest, and thus ensuring the kind of independent leadership that will enhance the board’s overall monitoring capabilities.188 The study found that “most chairmen who were CEO protect their protégés, reducing the likelihood that the new CEO will be fired for poor performance.”189 In the alternative, former CEOs who were not ready to relinquish their role seek to use their position as chair to interfere with the new CEO or otherwise find faults with the CEO at the first sign of trouble.190 Other studies also reveal that separation without true independence may not remedy the conflicts of interest that impede effective board oversight. As Professor Fred Tung noted, new studies support the proposition that social ties impact directors’ performance of their duties.191 At 183. Id. 184. Id. at 952. 185. Id. at 960. 186. See Timothy Quigley & Donald Hambrick, When the Former CEO Stays on as Board Chair: Effects on Successor Discretion, Strategic Change, and Performance, 33 STRATEGIC MGMT. J. 834, 834 (2012) (arguing that retaining a CEO on a board restricts the CEO’s predecessor’s discretion). 187. See Krause & Semadeni, supra note 101. 188. Lucier et al., supra, note 86, at 7. 189. Id. 190. Id. at 7-8. 191. Frederick Tung, The Puzzle of Independent Directors: New Learning, 91 B.U. L. REV. 262 INDIANA LAW REVIEW [Vol. 47:237 least one study reveals that even soft social ties such as “mutual alma mater, military service, regional origin, academic discipline and industry” between the CEO and outside directors had significant effects on firm performance, CEO compensation and CEO turnover.192 The fact that separating the CEO and board chair roles does not necessarily correlate with true independence undermines the extent to which such separation will generate its promised benefits.193 3. Undue Influence.—Advocates for separation of the CEO and board chair functions also argue that such separation will reduce the ability of the CEO to unduly influence and control the board of directors.194 The power and expertise inherent in the CEO role inevitably prompts increased reliance on the part of the board. This reliance is augmented by recent shifts in the corporate governance landscape related to boards. Since the CEO is often the only insider/employee on the board, the CEO may have greater company-specific knowledge, prompting boards to rely more significantly on the CEO’s perspective. Further, the CEO may have more industry-specific knowledge, which increases the likelihood that board members will unduly rely on the CEO. While such reliance is expected, and some reliance in fact may be appropriate, the danger is that boards will rely too heavily on the CEO, effectively rubber-stamping his or her decisions without any critical examination of those decisions. Such behavior, in turn, renders the boards ineffective as monitors. Separating the roles of CEO and board chair is designed to undermine this instinctive reliance, prompting boards to engage in more objective analysis of managerial programs, and thus improving their monitoring of CEOs and those programs.195 Unfortunately, it is not clear that corporations can counter inappropriate reliance merely by removing CEOs from the board rooms.196 In the compensation context, evidence suggests that the compensation committee continues to unduly rely on the CEO despite the fact that the CEO is not a member of that committee. A similar phenomenon has developed in the director nomination process, where the CEO continues to wield authority despite not being a formal member of the nomination committee. Empirical evidence confirms these trends. The evidence reveals that even when the CEO and board chair roles are separated, the board continues to rely on the CEO not only to keep them apprised of developments between board meetings, but also as the primary source for determining the quality, quantity and 1175, 1179-85 (2011). 192. Id. at 1181. 193. See Leblanc & Pick, supra note 9, at 2-3. 194. See Bagley & Koppes, supra note 24, at 157. 195. Nicola Sharpe, Process Over Structure: An Organizational Behavior Approach to Improving Corporate Boards, 85 S. CAL. L. REV. 261, 290-91 (2012) [hereinafter Sharpe, Process Over Structure]. 196. See Nicola Sharpe, Questioning Authority: The Critical Link Between Board Power and Process, 38 J. CORP. L. 1, 36, 46-47 (2012) [hereinafter Sharpe, Questioning Authority] (describing information asymmetries between managers and independent directors). 2014] SEPARATION ANXIETY 263 timeliness of information received from management.197 Directors rely almost exclusively on the CEO and management for information, and directors’ information channels are limited to information filtered through the CEO and the management team the CEO oversees, which undermines the board’s ability to be effective.198 This reliance undermines the board’s ability to monitor the CEO because it limits the extent directors receive and assess unbiased and objective information.199 Even if a CEO is not actively seeking to distort information, studies suggest that the CEO filters and organizes information in ways that are biased towards the CEO’s perspective and aimed at supporting the CEO’s agenda.200 As a result, even when there is a separation of the roles of CEO and board chair, the fact that directors continue to rely almost exclusively on CEOs and information filtered through CEOs means that such separation may not have an impact on inappropriate reliance. Thus, the separation may not yield the positive results that its advocates have promised.201 4. Salary Distortions.—Another benefit of separating the CEO and board functions is to combat excessive executive compensation, an issue that has risen to prominence since the financial crisis.202 Studies reveal that executives who serve jointly as CEO and board chair cost more than when the CEO and board chair positions are separated.203 A 2012 study assessed 180 North American companies with a market capitalization of $20 billion or more to gain a better understanding of how leadership structure impacted large complex corporations.204 The study found that executives with combined CEO and board chair roles received a median compensation of just over $16 million.205 In contrast, companies with a CEO and a separate board chair each paid a combined $11 million in compensation, while companies with a CEO and a separate independent chair each paid a combined $9.3 million in compensation.206 Other studies confirm that CEO pay tends to be higher when the CEO also chairs the board.207 197. 2012 SPENCER STUART BOARD INDEX, supra note 11, at 27 (revealing that most survey respondents rely on the CEO for these information gaps). 198. Sharpe, Questioning Authority, supra note 196, at 13, 44. 199. Id. 200. Id. at 243; see Sharpe, Process Over Structure, supra note 195, at 308. 201. Sharpe, Process Over Structure, supra note 195, at 308. 202. See Ben Protess, S.E.C. Plans Crackdown on Bonuses, N.Y. TIMES, Mar. 3, 2011, http://query.nytimes.com/gst/fullpage.html?res=9F02EFDE1F3EF930A35750C0A9679D8B63 &ref=executivepay, archived at http://perma.cc/WX35-7BQ3; RICHARD FERLAUTO, THE CONFERENCE BOARD TASK FORCE ON EXECUTIVE COMPENSATION 7 (2009) (noting that the economic crisis “has intensified public anger over executive compensation”). 203. Hodgson & Ruel, supra note 18, at 1. 204. Id. 205. Id. 206. Id. 207. Tuggle et al., supra note 171, at 951. See generally Charles O’Reilly & Brian Main, Economic and Psychological Perspectives on CEO Compensation: A Review and Synthesis, 19 264 INDIANA LAW REVIEW [Vol. 47:237 Advocates for separating the CEO and board chair functions not only insist that these increased salaries result from the excessive influence and conflict of interest inherent in having the CEO serve as board chair, but also that separation can overcome such salary distortions. To be sure, the authors of the 2012 study pointed out that one would expect a premium to exist for executives who hold the positions of CEO and board chair.208 However, they expressed concern regarding the size of that premium in light of the fact that employing two different individuals to serve in the two roles only accounted for 75% of the costs of the combined position.209 The authors also noted that the size of the premium for the combined role was particularly problematic because the factors that typically account for variations in compensation were moot.210 In this regard, they concluded that the increased costs is not based on merit, but rather results from the dangers inherent in combining the two positions.211 Notably, the 2012 study found that non-independent chairs earn far more than their independent colleagues.212 The authors stated that while there may be some expectation that CEOs would receive a premium for serving in dual capacity, “there is little or no reason for non-executive positions to earn more.”213 As a result, the inevitable conclusion is that such non-independent chairs earn more because of conflicts of interests associated with their connections to the CEOs. This conclusion not only provides another reason for separating the CEO and board functions, but also underscores the argument that such separation will achieve the most optimal results when the board chair is truly independent.214 The fact that shareholders increasingly have played a role in the context of executive compensation may mitigate some of the concerns about excessive compensation because shareholders may be able to help check any tendencies to award excessive pay packages.215 Shareholders now have a say on pay—a non- binding vote on the compensation packages of executive officers,216 and the available empirical evidence reveals that shareholders’ say on pay vote can impact a corporation’s compensation practices.217 It is possible, therefore, that INDUS. & CORP. CHANGE 675, 686-87 (2010) (discussing CEO’s level of influence on board and the effect of the influence on CEO compensation). 208. Hodgson & Ruel, supra note 18, at 1-2. 209. Id. at 2. 210. Id. 211. Id. at 4-5. 212. Id. The study found that independent chairs are paid less than non-independent chairs.. The median compensation of independent chairs was $417,910 while the median compensation for non-independent chairs was $630,930. 213. Id. 214. See id. at 1-3. 215. See FERLAUTO, supra note 202, at 27 (discussing shareholder advisory votes on questions of executive compensation). 216. See Dodd-Frank Wall Street Reform and Consumer Protection Act, Pub. L. 111-203, § 951, 124 Stat. 1376, 1899 (2010). 217. See generally Fabrizio Ferri & David Maber, Say on Pay Votes and CEO Compensation: 2014] SEPARATION ANXIETY 265 compensation practices will achieve optimal levels without regard to board leadership structure. To be sure, however, studies about say on pay reveal that while say on pay has had some impact on disciplining pay practices at poorly performing firms, it has had no impact on firms that do well.218 One recent study focused on the impact of say on pay in the United States failed to find any major change in the level or structure of CEO compensation based on say on pay votes.219 At the very least, this suggests that in those cases, there will continue to be pay distortions, which may be more problematic in companies that do not separate the CEO and chair positions. This section suggests that there are real benefits to be gained from separating the CEO and board chair roles, including those related to curtailing excessive compensation, preventing undue reliance on executives, and reducing conflicts of interests. All of these benefits, if realized, enhance boards’ ability to effectively monitor the CEO and the corporation enterprise more broadly. However, this section also reveals that there may be factors that undermine corporations’ ability to realize those benefits, raising questions regarding whether separating the CEO and board chair functions promises rewards that it cannot deliver. The next section seeks to respond to those questions. C. Beyond Separation Both the empirical and analytical evidence suggests that whether boards can obtain the benefits of separating the roles of CEO and board chair may depend on various factors. This section focuses on two of those factors: independence and board processes. 1. The Elusive Quest for Independence.—As discussed above, mere separation of the CEO and board chair functions does not guarantee independent board leadership because it does not ensure that the board chair will be truly independent, and thus, free from conflicts and biases that undermine his or her ability to perform the oversight role effectively. Unfortunately, there does not appear to be a clear path towards overcoming this obstacle for at least three reasons. First, the current governance landscape does not include many truly independent CEOs. Indeed, as of 2012, only 23% of chairs of S&P 500 companies could be classified as truly independent, as opposed to the 20% of independent chairs that are, in large part, former company CEOs or current Evidence from the UK, REV. FIN., Feb. 13, 2013, available at http://papers.ssrn.com/sol3/papers. cfm?abstract_id=1420394l (discussing the view that shareholder pay votes add value to an organization); Vicente Cuñat et al., Say Pays! Shareholder Voice and Firm Performance, Feb. 2012, at 7, available at http://www.oecd.org/els/SayonPaySept72012Complete.pdf, archived at http://perma.cc/9LL8-KE2H (noting that the U.K. say on pay votes appear to curb the “pay for failure” scenario). 218. Cuñat et al., supra note 217, at 7; Andrew C.W. Lund, Say on Pay’s Bundling Problems, 99 KY. L.J. 119, 119 (2010-2011). 219. See generally Cuñat et al., supra note 217, at 3 (noting limited effect of say on pay votes on executive compensation in the United States). 266 INDIANA LAW REVIEW [Vol. 47:237 executives.220 At least one recent study confirms that the most prevalent CEO- Board chair separation involves the former CEO retaining the board leadership position.221 This means that when companies separate the roles of chair and CEO, it is most often that the chair is a former CEO, former executive, or other individual who has a significant business relationship with the company other than board service.222 As one recent study noted, “the separation of the two posts in American firms often signifies that the chair is the former CEO—hardly a dispassionate supervisor, and . . . possibl[y] a major obstacle to change.”223 The current status quo raises questions regarding whether it is possible for corporations to have board chairs that are truly independent. Second, evidence suggests that many companies are reluctant to adopt a separation model disallowing the former CEO to serve as board chair because that model is preferred by CEOs, and hence eliminating or restricting the model could entail costs.224 Commentators explain that a common CEO succession process is “passing the baton,” pursuant to which the CEO relinquishes her title but remains in the role of board chair for some period.225 Efforts to install truly independent board chairs run counter to this process. CEOs often bargain for the right to be promoted to board chair after their employment term ends,226 and removing that right may impact the bargaining process in ways that make corporations reluctant to embrace a model of truly independent board chairs, including increasing the costs of such process or otherwise undermining the corporation’s ability to attract top talent.227 This reduces the possibility that such a model will be the norm, and thus, reduces the extent to which we can expect separation of the roles to achieve the true independence necessary to ensure that companies can take full advantage of that separation.228 Finally, the fact that definitions of independence do not take into account social ties may undermine true independence of board chairs.229 As one expert noted, there is a longstanding tradition “that the chairman and the CEO usually come from the same group of people, the circle of friends and acquaintances where people know each other.”230 While this may create an ease of 220. 2012 SPENCER STUART BOARD INDEX, supra note 11, at 23. 221. Krause & Semadeni, supra note 101, at 2. 222. See Hodgson & Ruel, supra note 18, at 2 (describing characteristics of a non-independent board chair). 223. Quigley & Hambrick, supra note 186, at 836. 224. Brickley et al., supra note 19, at 192. 225. Id. at 194. 226. See id. at 195. 227. Id. at 194-95. 228. See Quigley & Hambrick, supra note 186, at 834-85 (describing the advantages of retaining a former CEO as a board chair). 229. See O’Reilly & Main, supra note 207, at 686 (noting the social influence of the CEO over the board). 230. Olson, supra note 50. 2014] SEPARATION ANXIETY 267 communication,231 it also may undermine the objectivity needed to be deemed truly independent. Directors often share thick social ties with the CEO that may undermine their ability to be objective, and hence impede their ability to effectively monitor CEO behavior.232 As an initial matter, ensuring that directors do not share bias-producing social ties with the CEO may not be possible given the pool from which directors are selected, as well as the social ties that are formed with the management team once directors serve on the board.233 Regulators and judges have essentially refused to include, in any meaningful way, considerations of social or professional ties in the determination of director independence,234 and there is no indication that the courts will reverse course with respect to that refusal.235 This creates additional uncertainty regarding whether separating the roles will translate into true independent leadership, which impacts the extent to which the benefits of separation can be realized.236 2. The Importance of Board Process.—Several commentators have noted that the focus on independence and other structural reforms related to the board is too narrow, and therefore, incomplete.237 Thus, in addition to embracing enhanced independence, boards also must consider the processes necessary to ensure that such independence will translate into effective oversight. Behavioral theories buttress the notion that board process is critical to ensuring that boards can successfully perform their responsibilities.238 Thus, merely seeking independence without regard to the processes that are necessary to ensure that such independence translates into effective monitoring is a mistake.239 One important process relates to information. Ensuring that the board, and particularly, an independent board chair, have access to unbiased information, and multiple information gathering channels that do not rely on management, is important for board independence.240 How directors obtain information and who provides the information is critical to an effective decision-making process.241 This is because information that stems from limited and potentially biased 231. Id. 232. Tung, supra note 191, at 1177, 1179-80. 233. Fairfax, The Uneasy Case, supra note 58, at 153; see generally Tung, supra note 191, at 1185 (describing social ties between boards and managers). 234. See Fairfax, The Uneasy Case, supra note 58, at 146-48. 235. See id. at 151-52. 236. See Tung, supra note 191, at 1185 (noting the difficulties with creating a workable definition of independence that captures all of the social experiences that may undermine monitoring incentives, and noting that those difficulties may cause the acknowledgement that there could be limits on the ability to “operationalize independence”). 237. See Sharpe, Process Over Structure, supra note 195, at 264; Leblanc & Pick, supra note 9, at 3. 238. Sharpe, Process Over Structure, supra note 195, at 264; Leblanc & Pick, supra note 9, at 3. 239. Sharpe, Process Over Structure, supra note 195, at 265. 240. Sharpe, Questioning Authority, supra note 196, at 26-27, 36. 241. Id. at 36. 268 INDIANA LAW REVIEW [Vol. 47:237 channels, reduces the full range of information available to the board, and thus necessarily undermines the board’s ability to assess a broad range of available alternatives.242 Without an appreciation for that range of alternatives, directors are hampered in their ability to adequately oversee the choices made by the CEO and management.243 The very independence of independent board chairs hinders their access to information. Consequently, in these boards, a focus on board process is critical to ensuring that such chairs can be effective monitors. An independent board chair by definition serves on a part-time basis. This creates time constraints that may undermine the chair’s ability to effectively absorb information.244 Further, a truly independent board chair by definition has no employment relationship with the company whose board she chairs.245 This makes it more difficult for such chairs to access information, increasing the potential that they will exclusively utilize one mode of information-gathering: reliance on the CEO and management.246 At least one study suggests that a way to enhance the effectiveness of the board chair’s information gathering process is to ensure that the board chair has appropriate industry experience.247 Industry knowledge furthers at least two goals critical to information gathering and effective decision-making processes. First, it generates respect for the chair and her experience, which leads to more efficient and constructive dialogue in the boardroom.248 The study revealed that such industry knowledge provided board chairs with the cloak of respectability that enabled them to lead discussions effectively.249 The industry experience also provided the chairs with a level of in-depth knowledge and sophistication that allowed the chairs to frame critical issues, and provide important insights.250 By contrast, independent board chairs’ lack of industry knowledge “made it difficult for [the] separate chair to establish legitimacy with the directors and management team in a way that allowed him or her to meaningfully shape the board discussion.”251 Second, a board chair’s industry knowledge and experience not only augments her access to information, but also increases the likelihood that she will have ready access to information gathering channels independent of management. As a result, such background provides an important source of objective information upon which the board chair can rely to better process and understand other information being provided to the board, and to more 242. Id. 243. Id. 244. Sharpe, Process Over Structure, supra note 195, at 290. 245. See id. (describing the difficulty with which part-time and independent board members learn about a company). 246. Sharpe, Questioning Authority, supra note 196, at 26-27. 247. Leblanc & Pick, supra note 9, at 8. 248. Id. 249. Id. 250. Id. 251. Id. 2014] SEPARATION ANXIETY 269 appropriately monitor management.252 This suggests that industry experience may play an important role in buttressing board process, and thus ensuring that separate board leadership can prove beneficial. However, retaining board chairs with industry experience may prove challenging. Indeed, the pressure to populate the board with independent directors undermines the ability to recruit directors with company and industry specific ties because there is an increased likelihood that directors with such ties will not be independent.253 Further, restrictions on corporate directorships have become increasingly more common as corporations recognize that effective board service requires increased time commitment.254 Seventy-four percent of S&P 500 companies limit corporate directorship for their board members, as compared to 55% five years ago.255 While this limitation may be appropriate for effective board governance, it limits the pool of available candidates with the industry experience necessary to serve in the board chair role. Reflective of this limit, corporations have expressed frustration with their inability to find enough board members with industry experience who can serve on their boards.256 CONCLUSION Boards are facing significant pressure to engage in more effective monitoring. The financial crisis raised concerns regarding whether boards had paid sufficient attention to their monitoring function, and the crisis spurred reforms aimed at improving such function. One such reform is the separation of the board and CEO functions. The hope is that such separation will reduce conflicts of interest and other factors that impede the board’s ability to objectively oversee the corporation and its operations. The hope is also that such separation will better equip boards to monitor the actions of the CEO and top management team, including the prevention of excessive compensation packages. This Article contends that splitting the board and CEO roles can positively impact the board’s monitoring function. However, it identifies important caveats to that contention. First, it acknowledges that there may be costs associated with that split, some of which may have been overstated, but nevertheless, are worthy of consideration when determining the appropriate board leadership structure. Second, it recognizes that whether separating the roles of CEO and board chair can prove beneficial may depend on several variables including the size of the firm, the relative independence of the chair, and the types of processes— particularly information processes—that are available to the independent board chair. With respect to the latter two variables, achieving them may prove 252. See id. at 8-9 (noting that an experienced board chair can operate independently from management, which helps the chair monitor and oversee the actions of the company). 253. Fairfax, The Uneasy Case, supra note 58, at 166, 167. 254. 2012 SPENCER STUART BOARD INDEX, supra note 11, at 16. 255. Id. 256. Id 270 INDIANA LAW REVIEW [Vol. 47:237 challenging for corporations. This undermines the extent to which the independent board chair can deliver its promised benefits. As a result, this Article only gives conditional support for the separation of the CEO and board chair roles.