661 3 p

E X C H A N G E

Pay without Performance: Overview of the issues Lucian A. Bebchuk and Jesse M. Fried*

Executive Overview In a recent book, Pay without Performance: The Unfulfilled Promise of Executive Compensation, Bebchuk and Fried critique existing executive pay arrangements and the corporate governance processes that produce them. They also put forward proposals for improving both executive pay and corporate governance. This paper provides an overview of the main elements of their critique and proposals. The authors show that, under current legal arrangements, boards cannot be expected to contract at arm’s length with the executives whose pay they set. They discuss how managers’ influence can explain many features of the executive compensation landscape, including ones that researchers subscribing to the arm’s-length contracting view have long considered as puzzling. The authors also explain how managerial influence can lead to inefficient arrangements that generate weak or even perverse incentives, as well as to arrangements that make the amount and performance- insensitivity of pay less transparent. Finally, they outline proposals for improving the transparency of executive pay, the connection between pay and performance, and the accountability of corporate boards.

In judging whether Corporate America is serious about reforming itself, CEO pay remains the acid test. To date, the results aren’t encouraging.

—Warren Buffett, letter to shareholders of Berkshire Hathaway, Inc., February 2004

I n Pay without Performance and several prior and accompanying papers,1 we seek to provide a full account of how managerial power and influence

have shaped the executive compensation land- scape. The dominant paradigm for financial econ- omists’ study of executive compensation has as-

sumed that pay arrangements are the product of arm’s-length contracting— contracting between ex- ecutives attempting to get the best possible deal for themselves and boards seeking to get the best possible deal for shareholders. This assumption also has been the basis for the corporate law rules governing the subject. We aim to show, however, that the pay-setting process in publicly traded com- panies has strayed far from the arm’s-length model.

Our analysis indicates that managerial power has played a key role in shaping managers’ pay arrangements. The pervasive role of managerial power can explain much of the contemporary landscape of executive compensation. Indeed, it can explain practices and patterns that have long puzzled financial economists studying executive compensation. Furthermore, managerial influence

This paper draws on their earlier work on executive compensation, especially the authors’ recent book, Pay without Performance: The Unfulfilled Promise of Executive Compensation (Harvard University Press, 2004). The paper is a revision of a paper prepared for a Journal of Corporation Law symposium on this book. For financial support, they would like to thank the Guggenheim, Lens, and Nathan Cummins Foundations and the John M. Olin Center for Law, Economics, and Business (Bebchuk); and the Boalt Hall Fund and the U.C. Berkeley Committee on Research (Fried).

* Lucian A. Bebchuk is William J. Friedman and Alicia Townsend Friedman Professor of Law, Economics Finance and Director of the Program on Corporate Governance, Harvard Law School. Contact: [email protected]. Jesse M. Fried is Professor of Law and Co-Director of the Berkeley Center for Law, Business and the Economy at the School of Law, University of California, Berkeley. Contact: [email protected].

2006 5Bebchuk and Fried

over the design of pay arrangements has produced considerable distortions and costs to investors and the economy. It has distorted pay arrangements, diluted managers’ incentives to enhance firm value, and even provided perverse incentives to take actions that reduce long-term firm value.

Executive compensation has long been a topic of heated debate. The rise in executive pay has been the subject of much public criticism, which further intensified following the corporate governance scan- dals that began erupting in late 2001. This wave of corporate scandals shook confidence in the perfor- mance of public company boards and drew attention to potential flaws in their executive compensation practices. As a result, there is now recognition that many boards have employed compensation arrangements that do not serve shareholders’ in- terests. But there is still substantial disagreement about the scope and source of such problems and, not surprisingly, about how to address them.

Many take the view that concerns about exec- utive compensation have been exaggerated. There are some who maintain that flawed compensation arrangements have been limited to a relatively small number of firms, and that most boards have carried out effectively their role of setting execu- tive pay. Others concede that flaws in compensa- tion arrangements have been widespread, but maintain that these flaws have resulted from hon- est mistakes and misperceptions on the part of boards seeking to serve shareholders. According to this view, now that the problems have been rec- ognized, corporate boards can be expected to fix them on their own. Still others argue that, even though regulatory intervention was necessary, re- cent reforms that strengthen director indepen- dence will fully address past problems. Accord- ingly, at least going forward, one can expect boards to set pay policies in shareholders’ interest.

Our work seeks to persuade readers that such complacency is hardly warranted. To begin, flawed compensation arrangements have not been limited to a small number of “bad apples:” they have been widespread, persistent, and systemic. Furthermore, the problems have not resulted from temporary mistakes or lapses of judgment that boards can be expected to correct on their own;

rather, they have stemmed from structural defects in the underlying governance structures that en- able executives to exert considerable influence over their boards. The absence of effective arm’s- length dealing under today’s system of corporate governance has been the primary source of prob- lematic compensation arrangements. Finally, while recent reforms that seek to increase board independence will likely improve matters, they will not be sufficient to make boards adequately accountable; much more needs to be done.

Another, broader aim of our work has been to contribute to a better understanding of some of the basic problems afflicting the corporate gover- nance system. The study of executive compensa- tion opens a window through which we can ex- amine our current reliance on boards to act as guardians of shareholders’ interests. Our corporate governance system gives boards substantial power and counts on them to monitor and supervise the company’s managers. As long as corporate direc- tors are believed to carry out their tasks for the benefit of shareholders, current governance ar- rangements—which insulate boards from inter- vention by shareholders—appear acceptable. Our analysis of the executive pay landscape casts doubt on the validity of this belief and the wisdom of insulating boards from shareholders.

A full understanding of the flaws in current compensation arrangements, and in the gover- nance processes generating them, is necessary for addressing these problems. After providing a full account of the existing problems, our work also puts forward a set of proposals for improving both executive pay and corporate governance. We pro- vide detailed suggestions for making pay, and its relationship to performance, more transparent. Such transparency will provide a better check on managers’ power to influence their own pay. It will also eliminate existing incentives to choose compensation arrangements that are less efficient but more effective in camouflaging the amount of pay or its insensitivity to performance.

Furthermore, our analysis of the myriad ways in which pay is decoupled from performance and weakens or distorts incentives provides a basis for recommending how firms could better tie pay to performance and provide incentives more cost-

6 FebruaryAcademy of Management Perspectives

effectively. Finally, we put forward reforms that make directors not only more independent of in- siders but also more dependent on shareholders, thus improving board accountability to sharehold- ers. Such reforms may well offer the most prom- ising route for improving executive compensation and corporate governance more generally.

In this paper, we outline some of the main elements of our critique of contemporary execu- tive compensation and corporate governance ar- rangements, as well as of our proposals and sug- gested reforms. We start by describing the limitations of the official arm’s-length model of executive compensation. We then turn to the man- agerial power perspective, and discuss how manage- rial influence can explain many features of the com- pensation landscape, as well as the flaws and problems with existing pay arrangements, including their weak relationship to managers’ own perfor- mance and their inadequate disclosure. We conclude with a discussion of our proposals for making pay more transparent, improving the design of pay ar- rangements, and increasing board accountability.

Before proceeding, we wish to emphasize that our strong critique of existing pay arrangements and pay-setting processes should not be understood as a claim that directors and executives are less ethical or have acted with less decency than one would expect from others if they were placed in the same circumstances. Our problem is with the system of arrangements and incentives within which direc- tors and executives operate, not with the moral virtue or caliber of directors and executives.

As currently structured, the system unavoid- ably creates incentives and psychological and so- cial forces that distort pay choices. They can be expected to lead anybody (who is not a saint) to support, at least as long as they remain within prevailing practices and conventions, that favor themselves, their colleagues, or people who can in turn favor them. If we were to maintain the basic structure of our corporate governance system and merely replace directors and executives with an entirely different group of people, their replace- ments would be exposed to the very same incen- tives and forces and, by and large, we would not expect them to act differently. To address the

problems, we need to change the basic arrange- ments that produce these distortions.

The Stakes

W hat is at stake in the debate over executive pay? Some might question whether execu- tive compensation has a significant eco-

nomic impact on shareholders and the economy. The problems with executive compensation, it might be argued, do not much affect shareholders’ bottom line and are mainly symbolic.

However, the question of whether and to what extent pay arrangements are flawed is an impor- tant one for shareholders and policymakers— even if symbolism were unimportant. The existing flaws in compensation arrangements impose sub- stantial costs on shareholders. To begin, there is the excess pay that managers receive as a result of their power: that is, the difference between what managers’ influence enables them to obtain and what they would get under arm’s-length contract- ing. As a recent study by Yaniv Grinstein and one of us documents in detail,2 the amounts involved are hardly pocket change for shareholders.

The study finds that, during the period of 1993- 2003, the aggregate compensation paid by public firms to their top-five executives totaled about $350 billion (in 2002 dollars). This aggregate top- five compensation accounted for 6.6 percent of the aggregate earnings (net income) of these firms during the period under consideration. The aggre- gate compensation paid by public firms to their top-five executives was 9.8 percent of the aggre- gate earnings of these firms during 2001-2003, up from 5 percent during 1993-1995. Note that this study relies on a standard executive compensation dataset that (like other such datasets) does not include various forms of compensation not re- ported in publicly filed summary compensation tables, such as the retirement benefits and pack- ages that comprise a significant component of executives’ total career compensation.

Thus, if compensation could be cut without weakening managerial incentives, the gain to in- vestors would not be merely symbolic. Rather, it would have real practical significance. Further- more, and perhaps even more importantly, man- agers’ influence over compensation arrangements

2006 7Bebchuk and Fried

dilutes and distorts managerial incentives. In our view, the reduction in shareholder value caused by these inefficiencies—rather than that caused by excessive managerial pay— could well be the big- gest cost arising from managerial influence over compensation.

Existing pay arrangements have been produc- ing two types of incentive problems. First, com- pensation arrangements have been providing weaker incentives to reduce managerial slack and to increase shareholder value than would be the case under arm’s-length contracting. Both the non-equity and equity components of managerial compensation have been more severely decoupled from managers’ contribution to company perfor- mance than superficial appearances might suggest. Making pay more sensitive to performance may well benefit shareholders substantially.

Second, prevailing practices not only fail to provide cost-effective incentives to reduce slack but also create perverse incentives. For example, managers’ broad freedom to unload company op- tions and stock can lead managers to act in ways that reduce shareholder value. Executives who expect to unload shares have incentives to misre- port results, suppress bad news, and choose projects and strategies that are less transparent to the market. The efficiency costs of such distor- tions might exceed, possibly by a large margin, whatever liquidity or risk-bearing benefits execu- tives obtain from being able to unload their op- tions and shares at will. Similarly, because existing pay practices often reward managers for increasing firm size, they provide executives with incentives to pursue expansion via acquisitions or otherwise, even when that strategy is not value-maximizing.

The Arm’s-Length Contracting View

A ccording to the “official” view of executive compensation, corporate boards setting pay arrangements are guided solely by shareholder

interests and operate at arm’s length from the executives whose pay they set. The premise that boards contract at arm’s length with executives has long been and remains a central tenet in the corporate world and in most research on executive compensation by financial economists.

In the corporate world, the official view serves

as the practical basis for legal rules and public policy. It is used to justify directors’ compensation decisions to shareholders, policymakers, and courts. These decisions are portrayed as being made largely with shareholders’ interests at heart and therefore deserving of deference.

The premise of arm’s-length contracting has also been shared by most of the research on exec- utive compensation. Managers’ influence over di- rectors has been recognized by those writing on the subject from legal, organizational, and socio- logical perspectives, as well as by media coverage of executive pay. But most of the research on executive pay (especially empirical research) has been done by financial economists, and the premise of arm’s-length contracting has guided most of their work. Some financial economists, whose studies we discuss in our book in detail, have reported findings they viewed as inconsistent with the arm’s-length model.3 However, the majority of work in the field has assumed arm’s-length contract- ing between boards and executives.

In the paradigm that has dominated financial economics, boards, operating at arm’s length from executives, seek to serve shareholder interests by adopting compensation schemes designed to pro- vide managers with efficient incentives to maxi- mize shareholder value. In this paradigm, manag- ers’ pay arrangements are viewed as a (partial) remedy to the agency problem, reducing potential costs from self-serving decisions by managers. Like other rational and informed parties who contract at arm’s length, boards and managers are assumed to have powerful incentives to avoid inefficient provisions that shrink the pie produced by their contractual arrangements. The arm’s-length con- tracting view has thus led researchers to assume that executive compensation arrangements will tend to increase value, which is why we have used the terms “efficient contracting” or “optimal con- tracting” to label this approach in some of our earlier work.4

Financial economists, both theorists and em- piricists, have largely worked within the arm’s- length model in attempting to explain common compensation arrangements as well as variation in compensation practices among firms.5 In fact, upon discovering practices that appear inconsis-

8 FebruaryAcademy of Management Perspectives

tent with the cost-effective provision of incen- tives, financial economists have often labored to come up with clever explanations for how such practices might be consistent with arm’s-length contracting after all. Practices for which no expla- nation has been found have been considered “anomalies” or “puzzles” that will ultimately either be explained within the paradigm or disappear.

In our book, we identified many compensation practices that are difficult to understand under the arm’s-length contracting view but can be readily explained by managerial influence over the pay- setting process. Some of our critics suggested rea- sons why some of these practices could still be consistent with arm’s-length contracting and ar- gued that we have therefore not succeeded in ruling out completely the possibility of arm’s- length dealing. For example, in response to our showing that pay is significantly decoupled from performance, critics argued that it might be desir- able to provide managers with large amounts of non-performance pay.6 This type of response re- flects an implicit presumption in favor of arm’s- length contracting. The burden of proof rests on those skeptical of arm’s-length contracting, and arm’s-length contracting should be assumed true until the skeptics prove otherwise.

The presumption of arm’s-length contracting, however, does not seem warranted. As we discuss below, an examination of the pay-setting process suggests that managerial influence plays a key role. Thus, given the a priori plausibility of managerial influence, one might place the burden of proof on those arguing that the executive pay arrangements produced by existing processes are not signifi- cantly shaped by such influence. In any event, that sophisticated financial economists continue to implicitly or explicitly use arm’s-length con- tracting as their baseline presumption indicates the dominance and power of this long-held view.

Limits of the Arm’s-Length View

T he official arm’s-length story is neat, tractable, and reassuring. However, this model fails to account for the realities of executive compen-

sation. The arm’s-length contracting view recognizes

that managers are subject to an agency problem

and do not automatically seek to maximize share- holder value. The potential divergence between managers’ and shareholders’ interests makes it im- portant to provide managers with adequate incen- tives. Under the arm’s-length contracting view, the board, working in shareholders’ interest, at- tempts to cost-effectively provide such incentives through managers’ compensation packages. How- ever, just as there is no reason to presume that managers automatically seek to maximize share- holder value, there is no reason to expect a priori that directors will either. Indeed, an analysis of directors’ incentives and circumstances suggests that directors’ behavior is also subject to an agency problem.

Directors have had and continue to have var- ious economic incentives to support, or at least go along with, arrangements favorable to the compa- ny’s top executives. Social and psychological fac- tors— collegiality, team spirit, a natural desire to avoid conflict within the board, friendship and loyalty, and cognitive dissonance— exert addi- tional pressure in that direction. Although many directors own shares in their firms, their financial incentives to avoid arrangements favorable to ex- ecutives have been too weak to induce them to take the personally costly, or at the very least un- pleasant, route of resisting compensation arrange- ments sought by executives. In addition, limitations on time and resources have made it difficult for even well-intentioned directors to do their pay-setting job properly. Finally, the market constraints within which directors operate are far from tight and do not prevent deviations from arm’s-length contracting outcomes in favor of executives. Below we briefly discuss each of these factors.

Incentives to Be Re-elected

Most directors might wish to be re-appointed to the board. Besides an attractive salary, a director- ship provides prestige and valuable business and social connections. The financial and nonfinan- cial benefits of holding a board seat give directors an interest in keeping their positions.

In a world where shareholders selected individ- ual directors, board members seeking re-appoint- ment might have an incentive to develop reputa- tions as shareholder-serving. Typically, however,

2006 9Bebchuk and Fried

the director slate proposed by management is the only one offered. The key to a board position is thus being placed on the company’s slate. And because the CEO has had significant influence over the nomination process, displeasing the CEO has been likely to hurt one’s chances of being put on the company slate. Directors thus have had an incentive to “go along” with the CEO’s pay ar- rangement, a matter dear to the CEO’s heart, at least as long as the compensation package remains within the range of what can be plausibly de- fended and justified. In addition, developing a reputation as a director who blocks compensation arrangements sought by executives could hurt rather than help a director’s chances of being invited to join other companies’ boards.

The new stock exchange listing requirements, which attempt to give independent directors a greater role in director nominations, weaken but do not eliminate executives’ influence over direc- tor nominations. The CEO’s wishes can be ex- pected to continue to influence the decisions of the nominating committee; after all, the directors appointed to the board are expected to work closely with the CEO. As a practical matter, di- rector candidates opposed by the CEO are not expected to be offered board nomination and would likely turn it down even if they were to receive such an offer.7

Even if the CEO had no influence over nom- inations, fighting with the CEO over the amount or performance sensitivity of her compensation might be viewed unfavorably by independent di- rectors on the nominating committee. These di- rectors might prefer to keep off the board an individual whose poor relationship with the CEO undermines board collegiality. They might also wish to avoid the friction and unpleasantness likely to accompany disputes over the CEO’s pay. Finally, the independent directors also might side with the CEO for other reasons to be discussed below.

CEOs’ Power to Benefit Directors

There are a variety of ways in which CEOs can benefit individual directors or board members as a group. For example, CEOs have influence over director compensation, in which directors have a

natural interest. As the company leader, usually as a board member, and often as board chairman, the CEO can choose to either discourage or encourage director pay increases. Independent directors who are generous toward the CEO might reasonably expect the CEO to use her bully pulpit to support higher director compensation. At a minimum, generous treatment of the CEO contributes to an atmosphere that is conducive to generous treat- ment of directors. Indeed, a study finds that com- panies with higher CEO compensation have higher director compensation as well—and that this relationship is caused by cooperation between directors and the CEO rather than by company performance.8

In the past, CEOs have often used their power over corporate resources to reward cooperative directors. The new stock exchange listing stan- dards now place some limits on CEOs’ ability to reward independent directors, but they do leave CEOs with substantial power in this area. For example, these requirements do not prohibit ad- ditional compensation to an independent direc- tor. Rather, they only limit such compensation to $100,000 annually, and do not restrict payments to immediate family members who are non-exec- utive employees.

Similarly, the requirements limit but do not prohibit business dealings between a company and an independent director’s firm, and they place absolutely no limit on the firm’s dealing with the director’s firm before or after the director qualifies for independent director status. And the standards permit unlimited contributions to charitable or- ganizations that independent directors run, are affiliated with, or simply favor. In sum, executives’ control over corporate resources continues to enable them to provide many directors with rewards ex- ceeding the small direct personal cost to most direc- tors of approving pay arrangements that deviate from those expected under arm’s-length contracting.

Friendship and Loyalty

Many independent directors have some prior so- cial connection to, or are even friends with, a company’s CEO or other senior executives. Even directors who did not know the CEO before their appointment may well have begun their service

10 FebruaryAcademy of Management Perspectives

with a sense of obligation and loyalty to the CEO. The CEO often will have been involved in bring- ing the director onto the board— even if only by not blocking the director’s nomination. With such a background, directors often start serving with a reservoir of good will toward the CEO, which will contribute to a tendency to favor the CEO in setting her pay. This kind of reciprocity is expected and observed in many social and profes- sional contexts. Not surprisingly, studies find that compensation committees whose chairs have been appointed after the CEO takes office have tended to award higher CEO compensation.9

Collegiality and Authority

In addition to friendship and loyalty consider- ations, there are other social and psychological forces that make it difficult for directors to resist executive-serving compensation arrangements. The CEO is the directors’ colleague, and directors are expected in most circumstances to treat their fellow directors collegially. The CEO is also the firm’s leader, the person whose decisions and vi- sions have the most influence on the firm’s future direction. In most circumstances, directors treat the CEO with respect and substantial deference. Switching hats to contract at arm’s length with one’s colleague and leader is naturally difficult.

Cognitive Dissonance and Solidarity

Many members of compensation committees are current and former executives of other companies. Individuals are known to develop views consistent with their self-interest. Executives and former ex- ecutives are likely to have formed beliefs that support the type of pay arrangements from which they have benefited. An executive who has ben- efited from a conventional option plan, for exam- ple, is more likely to resist the view that such plans provide executives with excessive windfalls.

Further reinforcing such cognitive dissonance, an executive who serves as a director in another firm might identify and feel some solidarity or sympathy with that firm’s executives; she natu- rally would be inclined to treat these executives the same way she would like to be treated by her own board of directors. Not surprisingly, there is evidence that CEO pay is correlated with the pay

levels of the outside directors serving on the com- pensation committee.10

The Small Costs of Favoring Executives

Directors typically own only a small fraction of the firm’s shares. As a result, the direct personal cost to board members of approving compensation ar- rangements that are too favorable to executives— the reduction in the value of their sharehold- ings—is small. This cost is therefore unlikely to outweigh the economic incentives and social and psychological factors that induce directors to go along with pay schemes that favor executives.

Ratcheting

It is now widely recognized that the rise in exec- utive compensation has in part been driven by many boards seeking to pay their CEO more than the industry average; this has led to an ever- increasing average and a continuous escalation of executive pay.11 A review of reports of compen- sation committees in large companies indicates that a large majority of them used peer groups in determining pay and set compensation at or above the fiftieth percentile of the peer group.12 Such ratcheting is consistent with a picture of boards that do not seek to get the best deal for their shareholders but rather are happy to go along with whatever can be justified as consistent with pre- vailing practices.

Limits of Market Forces

Some writers have argued that even if directors are subject to considerable influence from corporate executives, market forces can be relied on to force boards and executives to adopt the compensation arrangements that arm’s-length contracting would produce. Our analysis, however, finds that market forces are neither sufficiently finely tuned nor sufficiently powerful to compel such outcomes. The markets for capital, corporate control, and managerial labor do impose some constraints on executive compensation. These constraints are hardly stringent, however, and they permit sub- stantial deviations from arm’s-length contracting.

Consider, for example, the market for corpo- rate control—the threat of a takeover. Firms fre- quently have substantial defenses against take-

2006 11Bebchuk and Fried

overs. For example, a majority of companies have a staggered board, which prevents a hostile ac- quirer from gaining control before two annual elections are held, and often enables incumbent managers to block hostile bids that are attractive to shareholders. To overcome incumbent opposi- tion, a hostile bidder must be prepared to pay a substantial premium.13 The disciplinary force of the market for corporate control is further weak- ened by the prevalence of golden parachute pro- visions, as well as by payoffs made by acquirers to target managers to facilitate the acquisition. The market for corporate control thus exerts little dis- ciplining force on managers and boards, leaving them considerable slack and ability to negotiate manager-favoring pay arrangements.

New CEOs

Some critics of our work assumed that our analysis of departures from arm’s-length contracting did not apply to cases in which boards negotiate pay with a CEO candidate from outside the firm.14 However, while such negotiations might be closer to the arm’s-length model than negotiations with an incumbent CEO, they still fall quite short of this benchmark.

Among other things, directors negotiating with an outside CEO candidate know that, after the candidate becomes CEO, she will have influence over their re-nomination to the board and over their compensation and perks. The directors will also wish to have good personal and working re- lationships with the individual who is expected to become the firm’s leader and a fellow board mem- ber. And while agreeing to a pay package that favors the outside CEO hire imposes little finan- cial cost on directors, a breakdown in the negoti- ations, which might embarrass the directors and force them to re-open the CEO selection process, would be personally costly to them. Finally, direc- tors’ limited time forces them to rely on informa- tion shaped and presented by the company’s hu- man resources staff and compensation consultants, all of whom have incentives to please the incom- ing CEO.

Firing of Executives

Some critics of our work have suggested that the increased willingness of directors to fire CEOs over the past decade, especially in recent years, provides evidence that boards do in fact deal with CEOs at arm’s length.15 Although the incidence of firing has gone up over time, firings are still limited to unusual situations in which the CEO is accused of legal or ethical violations (e.g., Fannie Mae, AIG, Boeing, Marsh) or is viewed by revolt- ing shareholders as having a terrible record of performance (Morgan Stanley, HP). Without strong outside pressure to fire the CEO, mere mediocrity is far from enough to get a CEO pushed out. Furthermore, in the rare cases in which boards fire executives, boards often provide the departing executives with benefits beyond those required by the contract to sweeten the CEO’s departure and alleviate the directors’ guilt and discomfort. All in all, boards’ record of dealing with failed executives does not support the view that boards treat CEOs at arm’s length.

In sum, a realistic picture of the incentives and circumstances of board members reveals myriad incentives and tendencies that lead directors to behave very differently than boards contracting at arm’s-length with their executives over pay. Re- cent reforms, such as the new stock exchange listing requirements, may weaken some of these factors but will not eliminate them. Without ad- ditional reforms, the pay-setting process will con- tinue to deviate substantially from arm’s-length contracting.

Power and Pay

T he same factors that limit the usefulness of the arm’s-length model in explaining executive com- pensation suggest that executives have had sub-

stantial influence over their own pay. Compensation arrangements have often deviated from arm’s-length contracting because directors have been influenced by management, sympathetic to executives, insuffi- ciently motivated to insist on shareholder-serving compensation, or simply ineffectual. Executives’ influence over directors has enabled them to ob- tain “rents”— benefits greater than those obtain- able under true arm’s-length contracting.

12 FebruaryAcademy of Management Perspectives

In our work, we find that the role of managerial power can explain many practices and aspects of the executive compensation landscape. It is worth emphasizing that our conclusion is not based on the amount of compensation received by execu- tives. In our view, high absolute levels of pay do not by themselves imply that compensation ar- rangements deviate from arm’s-length contract- ing. Our finding that such deviations have been common is based primarily on an analysis of the processes by which pay is set, as well as on an examination of the inefficient, distorted, and non- transparent structure of pay arrangements. For us, the “smoking gun” of managerial influence over pay is not high levels of pay, but rather such things as the correlation between power and pay, the systematic use of compensation practices that ob- scure the amount and performance insensitivity of pay, and the showering of gratuitous benefits on departing executives.

Power-Pay Relationships

Although top executives generally have some de- gree of influence over their boards, the extent of their influence depends on various features of the firm’s governance structure. The managerial power approach predicts that executives who have more power vis-à-vis their boards should receive higher pay— or pay that is less sensitive to perfor- mance—than their less powerful counterparts. A substantial body of evidence does indeed indicate that pay has been higher, and less sensitive to performance, when executives have more power.

To begin, there is evidence that executive compensation is higher when the board is relatively weak or ineffectual vis-à-vis the CEO. In particular, CEO compensation is higher when the board is large, which makes it more difficult for directors to organize in opposition to the CEO; when more of the outside directors have been appointed by the CEO, which could cause them to feel gratitude or obligation to the CEO; and when outside directors serve on three or more boards, and thus are more likely to be distracted.16 Also, CEO pay is 20 to 40 percent higher if the CEO is the chairman of the board, and it is negatively correlated with the stock ownership of compensation committee members.17

Second, studies find a connection between ex-

ecutive pay and the presence of a large outside shareholder. Such presence is likely to result in closer monitoring and thus can be expected to reduce managers’ influence over their compensa- tion. One study finds a negative correlation be- tween the equity ownership of the largest share- holder and the amount of CEO compensation; doubling the percentage ownership of the outside shareholder reduces non-salary compensation by 12 to 14 percent.18 Another study finds that CEOs in firms that lack a 5 percent (or larger) external shareholder tend to receive more “luck-based” pay—that is, pay associated with profit increases that are entirely generated by external factors (e.g., changes in oil prices and exchange rates) rather than by managers’ own efforts.19 This study also finds that, in firms lacking large external shareholders, the cash compensation of CEOs is reduced less when their option-based compensa- tion is increased.

Third, there is evidence linking executive pay to the concentration of institutional shareholders, which are more likely to engage in monitoring and scrutiny of the CEO and the board. One study finds that more concentrated institutional owner- ship leads to lower executive compensation as well as to more performance-sensitive compensa- tion.20 Another study finds that the effect of in- stitutional shareholders on CEO pay depends on the types of relationships they have with the firm.21 CEO pay is negatively correlated with the presence of institutions that have other business relationship with the firm and thus concerned only with the firm’s share value (“pressure-resis- tant” institutions); however, CEO pay is posi- tively correlated with the presence of firms with business relationships with the firm (e.g., manag- ing a pension fund) and thus vulnerable to man- agement pressure (“pressure-sensitive” institutions).

Finally, studies find a connection between pay and anti-takeover provisions that make CEOs and their boards less vulnerable to a hostile takeover. One study finds that CEOs of firms adopting anti- takeover provisions enjoy above-market compen- sation before adoption of the anti-takeover provi- sions and that adoption of these provisions increases their excess compensation significant- ly.22 This pattern is not readily explainable by

2006 13Bebchuk and Fried

arm’s-length contracting; indeed, if managers’ jobs are more secure, shareholders should be able to pay risk-averse managers less. Another study finds that CEOs of firms that became protected by state anti-takeover legislation enacted during the pe- riod of 1984-1991 reduced their holdings of shares by an average of 15 percent, apparently because the shares were not as necessary for maintaining con- trol.23 Arm’s-length contracting might predict that a CEO protected by anti-takeover legislation would be required to buy more shares to restore her incentive to increase shareholder value.

Limits to Managerial Influence

There are, of course, limits to the arrangements that directors will approve and executives will seek. Although market forces are not sufficiently powerful to compel arm’s-length outcomes, they do impose some constraints on executive compensation. If a board were to approve a pay arrangement viewed as egregious, for example, shareholders would be less willing to support incumbents in a hostile takeover or proxy fight. In addition, directors and executives adopting such an arrangement might bear social costs. The constraints imposed by markets and by social forces are far from tight, however, and they permit substantial deviations from arm’s-length outcomes. The adoption of arrangements favoring executives is unlikely to impose substantial eco- nomic or social costs if the arrangements are not patently abusive or indefensible.

One important building block of the manage- rial power approach is that of “outrage” costs. When a board approves a compensation arrange- ment favorable to managers, the extent to which directors and executives bear economic and social costs will depend on how the arrangement is per- ceived by outsiders whose views matter to the directors and executives. Outrage might also lead to shareholder pressure on managers and directors, as well as possibly embarrass directors and manag- ers or harm their reputations. The more outrage a compensation arrangement is expected to gener- ate, the more reluctant directors will be to ap- prove it and the more hesitant managers will be to propose it in the first place.

There is evidence that the design of compen- sation arrangements is indeed influenced by how

outsiders perceive them. One study finds that, during the 1990s, CEOs who were the target of shareholder resolutions criticizing executive pay had their annual (industry-adjusted) compensa- tion reduced over the following two years.24

Camouflage and Stealth Compensation

The critical role of outsiders’ perception of exec- utives’ compensation and the significance of out- rage costs explain the importance of yet another component of the managerial power approach: “camouflage.” The desire to minimize outrage gives designers of compensation arrangements a strong incentive to try to legitimize, justify, or obscure— or, more generally, to camouflage—the amount and performance-insensitivity of execu- tive compensation.

After the board’s compensation committee ap- proves the compensation package, firms use com- pensation consultants and their reports to justify the compensation to shareholders. Wade, Porac, and Pollack find that companies that pay their CEOs larger base salaries, and firms with more concentrated and active outside ownership, are more likely to cite the use of surveys and consult- ants in justifying executive pay in their proxy reports to shareholders.25 This study also finds that, when accounting returns are high, firms em- phasize these accounting returns and downplay market returns.

For our purposes, attempts to justify compensa- tion arrangements are less important than how managers’ interest in camouflage affects the choice of arrangements in the first place. The latter is quite important because the desire to camouflage might lead to the adoption of com- pensation structures that are less efficient for in- centive generation (and thus hurt managerial in- centives and firm performance) but offer camouflage benefits. In our work we present evi- dence that compensation arrangements have of- ten been chosen and designed with an eye to camouflaging the amount of pay and the extent to which it is decoupled from performance. Overall, the camouflage motive turns out to be quite useful in explaining many otherwise puzzling features of the executive compensation landscape.

Among the arrangements that camouflage the

14 FebruaryAcademy of Management Perspectives

amount and the performance-insensitivity of com- pensation are executive pension plans, deferred compensation arrangements, and post-retirement perks. Most of the pension and deferred compen- sation benefits given to executives are not eligible for the large tax subsidy granted to the standard retirement arrangements provided to other em- ployees. In the case of executives, such arrange- ments merely shift tax liability from the executive to the firm. The efficiency grounds for providing compensation through in-kind retirement perks are also far from clear. All of these arrangements, however, make pay less salient.

Among other things, under existing disclosure rules, firms do not have to place a dollar value on—and include in the firm’s publicly filed sum- mary compensation tables—amounts provided to executives after they retire. Although the exis- tence of executives’ retirement arrangements must be noted in certain places in the firm’s public filings, this disclosure is less salient because out- siders focus on the dollar amounts reported in the compensation tables. Indeed, the standard com- pensation datasets generally used by media report- ers and researchers do not include information about executives’ retirement benefits.

In a recent empirical study, Robert Jackson and one of us use the information provided in proxy statements to estimate the value of the executive pension plans of S&P 500 CEOs.26 About two- thirds of CEOs have such plans, and the study provides estimates of the value of these plans for all the CEOs who recently left their firms or are close to retirement age. For the median CEO in the study’s sample, the actuarial value of the CEO’s pension was $15 million, which comprised about one-third of the total compensation (both equity-based and non-equity) they had received during their service as CEOs. When pension value is included in calculating executive pay, compen- sation is much less linked to performance than commonly perceived. Such inclusion increases the fraction that is salary-like (basic salary during the CEO’s service and pension afterwards) from 16 percent to 39 percent. The study documents that the current omission of retirement benefits from standard compensation datasets has distorted in- vestors’ picture of pay arrangements. In particular,

this omission has led to: (i) significant under- estimations of the total magnitude of pay; (ii) considerable distortions in comparisons among executive pay packages; and (iii) substantial over- estimations of the extent to which executive pay is linked to performance.

While firms do not make the value of executive pensions transparent, they do disclose the infor- mation that enables one to estimate the value of these pensions. In contrast, the information pro- vided about deferred compensation arrangements does not allow even the most diligent outsider to estimate with any precision the value conferred on executives through these arrangements. Thus, this form of compensation is especially effective in camouflaging potentially large amounts of non- performance pay. How large these amounts are for any given executive is not something that we can currently estimate.

Gratuitous Goodbye Payments

In many cases, boards give departing CEOs pay- ments and benefits that are gratuitous—that are not required under the terms of a CEO’s compen- sation contract. Such gratuitous “goodbye pay- ments” are common even when CEOs perform so poorly that their boards feel compelled to replace them. For example, when Mattel CEO Jill Barad resigned under fire, the board forgave a $4.2 mil- lion loan, gave her an additional $3.3 million in cash to cover the taxes for forgiveness of another loan, and allowed her unvested options to vest prematurely. These gratuitous benefits were of- fered in addition to the considerable benefits that she received under her employment agreement, which included a termination payment of $26.4 million and a stream of retirement benefits ex- ceeding $700,000 per year.

It is not easy to reconcile such gratuitous pay- ments with the arm’s-length contracting model. The board has the authority to fire the CEO and pay the CEO her contractual severance benefits. Thus, there is no need to “bribe” a poorly perform- ing CEO to step down. In addition, the signal sent by the golden goodbye payment will, if anything, only weaken the incentive of the next CEO to perform.

The making of such gratuitous payments, how-

2006 15Bebchuk and Fried

ever, is quite consistent with the existence of managerial influence over the board. Because of their relationship with the CEO, some directors might be unwilling to replace the existing CEO unless she is very generously treated. Other direc- tors might be willing to replace the CEO but prefer to accompany the move with a goodbye payment, either to reduce the personal discomfort they feel in forcing out the CEO, or to make the difficult separation process more pleasant and less contentious. In all of these cases, directors’ will- ingness to make gratuitous payments to the (poorly performing) CEO results from the CEO’s relationship with the directors.

It is important to note that, taking managerial power as given, providing gratuitous payments to fired CEOs might be beneficial to shareholders in some instances. If many directors are loyal to the CEO, such payments might be necessary to assem- ble a board majority in favor of replacing him. In such a case, the practice would help shareholders when the CEO’s departure is more beneficial to shareholders than the cost of the goodbye pay- ment. For our purposes, however, what is impor- tant is that these gratuitous payments—whether they are beneficial to shareholders or not—reflect the existence and significance of managerial in- fluence.

The Decoupling of Pay from Performance

T hose applauding the rise in executive compen- sation have emphasized the benefits of strengthening managers’ incentives to increase

shareholder value. Indeed, in the beginning of the 1990s, prominent financial economists such as Michael Jensen and Kevin Murphy urged share- holders to be more accepting of large pay packages that would provide high-powered incentives.27 Shareholders, it was argued, should care much more about providing managers with sufficiently strong incentives than about the amounts spent on executive pay.

Indeed, throughout the past 15 years, investors have often accepted increases in executive pay as the price for improving managers’ incentives. Higher compensation has been presented as essen- tial for improving managers’ incentives and there- fore worth the additional cost. Pay, however, is

hardly as tied to managers’ own performance as investors commonly assume. Shareholders have not received as much bang for their buck as pos- sible. Firms could have generated the same in- crease in incentives at a much lower cost, or they could have used the amount spent to obtain more powerful incentives. Executive pay is much less sensitive to performance than has commonly been recognized.

Non-Equity Compensation

Although the equity-based fraction of managers’ compensation has increased considerably during the past decade and has therefore received the most attention, non-equity compensation contin- ues to be substantial. In 2003, non-equity com- pensation comprised on average about half the total compensation (as reported in the standard ExecuComp dataset) of CEOs, as well as of other top-five executives, in S&P 1500 companies not classified as new economy firms.28

Although significant non-equity compensation comes in the form of base salary and sign-up “golden hello” payments that do not purport to be performance-related, much non-equity compensa- tion comes in the form of bonus compensation which purports to be performance-based. None- theless, empirical studies have failed to find any significant correlation between non-equity com- pensation and managers’ own performance during the 1990s.29

A close examination of firms’ practices suggests why non-equity compensation is not tightly con- nected to managers’ own performance. To begin, many firms use subjective criteria for at least some of their bonus payments. While subjective criteria could play a useful role in the hands of boards guided solely by shareholder interests, boards fa- voring managers can use discretionary criteria to ensure that managers are well paid even when, because of poor performance, bonuses based on objective criteria are low.

Furthermore, when firms do use objective cri- teria, these criteria and their implementation do not seem to be designed to reward managers for their own performance. Firms commonly do not base bonuses on how the firm’s operating perfor- mance or earnings increased relative to peer firms.

16 FebruaryAcademy of Management Perspectives

Instead, some firms base bonuses on how earnings or other financial variables compared to prior year figures. However, bonuses that are paid whenever there is improvement over prior year outcomes will often reward managers whose results fluctuate from year to year around a level reflecting poor performance.

Other firms base bonuses on how financial per- formance fared relative to a threshold specified by the board. In such cases, how well an executive fares depends not only on how well the executive performs but also on how low the goal is set. By setting goals low enough, directors can ensure executives receive rich bonuses. And when the firms fail to meet the established targets, they can reset the target (as happened at Coca-Cola in 2001 and AT&T Wireless in 2002) or compen- sate the executives by setting especially low figures going forward. Importantly, boards rarely attempt to filter out improvements in financial performance reflecting industry-wide changes that have nothing to do with the managers’ own performance.

Many boards award bonuses to managers for buying other firms. In about 40 percent of large acquisitions, the CEO of the acquiring firm re- ceives a multi-million dollar bonus for completing the deal.30 But making acquisitions appears hardly something for which managers should receive a special reward beyond whatever positive effects the acquisition might have on the value of the managers’ options and shares and earning-based bonuses. Executives do not lack incentives to make acquisitions. If anything, investors’ concern is that executives may engage in empire-building and make too many acquisitions. Thus, although the making of a large acquisition might provide a convenient excuse for a large bonus, acquisition bonuses are not called for by incentive consider- ations.

Windfalls in Equity-Based Compensation

In light of the historically weak link between non-equity compensation and managerial perfor- mance, shareholders and regulators wishing to make pay more sensitive to performance have increasingly looked to, and encouraged, equity- based compensation—that is, compensation based on the value of the company’s stock. Most equity-

based compensation has taken the form of stock options— options to buy a certain number of com- pany shares for a specified price (the “exercise” or “strike” price). We strongly support equity-based compensation, which in principle can provide man- agers with desirable incentives. Unfortunately, how- ever, the conventional design of options enables executives to reap substantial rewards even when their performance was merely passable or even poor.

Rewards for Market-Wide and Industry-Wide Movements: Conventional stock options enable executives to gain from any increase in the nom- inal stock price above the grant-date market price. Thus, executives can gain even when their per- formance is unimpressive or mediocre relative to their peers, as long as the firm’s stock price rises largely due to market-wide and industry-wide movements. In fact, much of the variation in executives’ payoffs from options comes from such fluctuations rather than from firm-specific move- ments that might be due to the manager’s own performance.

Although there is a whole range of ways in which such windfalls could be filtered out, a large majority of firms have failed to adopt equity-based plans that filter out such windfalls. Unfortunately, most of the boards now changing their equity- based compensation plans in response to outside pressure are still choosing to avoid plans that would effectively eliminate such windfalls. Rather, they are moving to plans such as those based on re- stricted stock that fail to eliminate, and sometimes even increase, these windfalls.

Rewards for Short-Term Spikes: Option plans have been designed, and largely continue to be designed, in ways that enable executives to make considerable gains from temporary spikes in the firm’s stock price, even when long-term stock per- formance is poor. Firms have given executives broad freedom to unwind equity incentives, a practice that has been beneficial to executives but costly to share- holders. In addition to giving executives freedom to exercise their options as soon as they vest and sell the underlying stock, firms have given executives substantial control over the timing of sales, enabling executives to benefit from their inside information. Indeed, many firms have not only failed to limit the unloading of options but have also adopted reload

2006 17Bebchuk and Fried

plans that encourage executives to lock in short- term spikes in stock prices.

The features of option plans that reward man- agers for short-term spikes not only provide man- agers rewards that might not reflect their long- term performance but also provide perverse incentives to manipulate earnings. There is, in fact, significant evidence linking executives’ free- dom to unload options with earnings manipula- tion and financial misreporting.31

Compensation at and after Departure

As already noted, a substantial portion of execu- tives’ compensation is not reported with a dollar figure in firms’ public disclosures and conse- quently not included in standard executive com- pensation datasets. This “stealth compensation” includes executive pensions, deferred compensa- tion arrangements, and post-retirement consulting contracts and perks. These less-noticed forms of compensation have tended to be insensitive to managerial performance, thus further contributing to the decoupling of pay from performance.

Take, for example, Franklin Raines, who was forced to retire as Fannie Mae’s CEO in late 2004. Upon departure, Fannie owed him (and his sur- viving spouse after his death) an annual pension of approximately $1.4 million, an amount speci- fied without any connection to the firm’s perfor- mance under Raines. In a case study of his com- pensation, we estimated the value of this non- performance element of Raines’s pay at about $25 million.32

Further decoupling pay from performance are severance payments given to departing executives. Executives who are pushed out by their boards due to extremely poor performance are typically paid a severance equal to their compensation over a multi-year period, often two or three years’ worth. These payments are not reduced even when the firm’s performance has been objectively dismal. Furthermore, standard severance provisions do not reduce the severance payment even if the executive quickly finds other employment.

Interestingly, although non-executive employ- ees are generally more likely to be terminated, they rarely receive such generous financial protec- tion. If anything, executives’ wealth and generous

retirement benefits are likely to make them more capable of bearing the risk of termination. More importantly, if executives’ large compensation is justified by the importance of providing them with incentives, one should expect executives’ compensation to be more sensitive to performance and provide less protection in the event of dismal failure. The existing severance practices that firms use for their executives not only fail to contribute to the link between pay and performance but also affirmatively operate to weaken it. They weaken the payoff difference between good and poor per- formance, a difference that shareholders spend much to create.

Improving Transparency

W e now turn to the implications of our anal- ysis—to our proposals for improving pay ar- rangements and the governance processes

generating them. We start with reforms that we view as no-brainers, ones for which we see no reasonable basis for opposition. Specifically, firms should be required to make the amount and struc- ture of pay more transparent.

Financial economists have paid insufficient at- tention to transparency because they often focus on whether information is disclosed and, there- fore, whether the information can become incor- porated into market pricing. It is widely believed that information can be reflected in stock prices as long as it is known and fully understood by even a limited number of market professionals.

In the case of executive compensation, there is already significant disclosure. As we have dis- cussed, SEC regulations require detailed disclosure of the compensation of a company’s CEO and of the four other most highly compensated execu- tives. In our view, however, it is important to recognize the difference between disclosure and transparency, and it is transparency that should receive the most attention.

The primary goal of requiring the disclosure of executive compensation is not to enable accurate pricing of the firm’s securities. Rather, this disclo- sure is primarily intended to provide some check on arrangements that are too favorable to execu- tives. This goal is not well served by disseminating information in a way that makes the information

18 FebruaryAcademy of Management Perspectives

understandable to a small number of market pro- fessionals but opaque to others.

Public officials, governance reformers, and in- vestors should work to ensure that compensation arrangements are and remain transparent. Trans- parency would provide shareholders with a more accurate picture of total pay and its relationship with performance and thereby provide some check on departures from arrangements that serve shareholder interests. Furthermore, transparency would eliminate the distortions that currently arise when pay designers choose particular forms of compensation for their camouflage value rather than for their efficiency. Finally, transparency would impose little cost on firms because it would simply require them to disclose clearly informa- tion they have or can obtain at negligible cost.

Although we support improved mandatory dis- closure requirements, nothing prevents companies in the meantime from voluntarily making pay more transparent. Investors should demand more openness, and companies should not continue to follow a “lawyerly” approach of not disclosing more than required. The following measures could substantially increase the transparency of pay ar- rangements.

(As this paper went to print, the SEC began a formal consideration of expanded disclosure re- quirements. The proposals put forward by the SEC staff include the first measure discussed below, and we hope that the other measures below will also be included during the process of the SEC’s consid- eration of the subject.)

1. Placing a Monetary Value on All Forms of Compensation

Companies should be required to place a dollar value on all forms of compensation and to include these amounts in the summary compensation ta- bles contained in company SEC filings. Firms have been able to provide executives with sub- stantial “stealth compensation” by using pensions, deferred compensation, and post-retirement perks and consulting contracts. Although some details of these arrangements have appeared elsewhere in companies’ SEC filings, firms have not been re- quired to place a dollar value on these benefits and to include this value in the summary tables, which

receive the most attention from investors and the media. These benefits have not even been in- cluded in the standard database used by financial economists to study executive compensation.

In our view, companies should be required to place a monetary value on each benefit provided or promised to an executive, and to include this value in the summary compensation table the year in which the executive becomes entitled to it. Thus, for example, the compensation tables should include the amount by which the expected value of an executive’s promised pension pay- ments increases during the year. In addition, it might be desirable to require companies to place a monetary value on any tax benefit that accrues to the executive at the company’s expense (for ex- ample, under deferred compensation arrange- ments)—and to report this value.

2. Disclosing All Non-Deductible Compensation

Efficient arrangements should take into account their effect on the combined tax bill of the com- pany and the executive. The tax code permits companies to deduct certain payments to execu- tives but not others. Companies routinely include in their disclosure boilerplate language putting shareholders on notice that some of the arrange- ments may result in the firm being unable to take a deduction for the compensation paid to execu- tives. But firms now do not provide details about what particular amounts end up not being deduct- ible. Firms should provide full details about the components of pay that are not deductible, place a monetary value on the costs of this non-deduct- ibility to the firm, and disclose this dollar cost to investors.

3. Disclosing the Relationship Between Pay and Performance

Companies should make transparent to their shareholders how much of managers’ profits from equity and non-equity compensation is due to general market and industry movements. This could be done by requiring firms to calculate and report the gains made by managers from the ex- ercise of options (or the vesting of restricted shares, in the case of restricted share grants) and to report what fraction, if any, was due to the

2006 19Bebchuk and Fried

company’s superior performance over its industry peers. Such disclosure would make much more transparent the extent to which the company’s equity-based plans reward the managers’ own per- formance.

4. Disclosure of Option and Share Unloading

Companies should be required to make transpar- ent to shareholders on a regular basis the extent to which their top five executives have unloaded any equity instruments received as part of their com- pensation. Although a diligent and dedicated re- searcher can obtain this information by sifting through stacks of executive trading reports filed with the SEC, requiring the firm to compile and report such information would highlight for all investors the extent to which managers have used their freedom to unwind incentives.

Improving Pay Arrangements

W ell-designed executive compensation can provide executives with cost-effective in- centives to generate shareholder value. We

have argued, however, that the promise of such arrangements has not yet been realized. Below we note various changes that firms should consider, and investors should urge them to adopt, in order to strengthen the link between pay and perfor- mance and improve executives’ incentives.

1. Reducing Windfalls in Equity-Based Compensation

Investors should encourage firms to adopt equity compensation plans that filter out at least some of the gains in the stock price that are due to general market or industry movements. With such filter- ing, the same amount of incentives can be pro- vided at a lower cost, or stronger incentives can be provided at the same cost. This can be done not only through indexing of the exercise price but also in other ways. For example, by linking the exercise price of options to the stock price of the worst-performing firms in the industry, market- wide movement can be filtered out without im- posing excessive risk on executives. It is important to note that moving to restricted stock is not a good way to address the windfalls problem; in fact,

restricted-stock grants provide an even larger windfall than conventional options do.

2. Reducing Windfalls in Bonus Plans

For similar reasons, companies should design bo- nus plans that filter out improvements in financial performance due to economy- or industry-wide movements. Even assuming that it is desirable to focus on accounting performance rather than stock price performance, as bonus plans seek to do, rewarding executives for improvements shared by all firms in the industry is not a cost-effective way to provide incentives. Thus, bonus plans should not be based on absolute increases in earnings, sales, revenues, and so forth, but rather on such increases relative to peer companies.

3. Limiting the Unwinding of Equity Incentives

Investors also should seek to limit executives’ broad freedom to unwind the equity-based incen- tives provided by their compensation plans. It may well be desirable to separate the vesting of options and managers’ ability to unwind them. By requir- ing that executives hold vested options (or the shares resulting from the exercise of such options) for a given period after vesting, firms would ensure that options that already belong to the executive will remain in his or her hands for some time, continuing to provide incentives to increase shareholder value. Furthermore, such restrictions would eliminate the significant distortions that can result from rewarding executives for short- term spikes in the stock price even when long- term stock returns are flat. To prevent circumven- tion, such restrictions should be backed by contractual prohibitions on executives’ hedging or using any other scheme that effectively eliminates some of their exposure to declines in the firm’s stock price.

In addition, it might be desirable, as one of us proposed some time ago, to require executives to disclose in advance their intention to sell shares, providing detailed information about the in- tended trade, including the number of shares to be sold.33 Providing executives with opportunities to sell their shares when their inside information indicates the stock price is about to decline can dilute and distort their incentives.

20 FebruaryAcademy of Management Perspectives

4. Tying Bonuses to Long-Term Performance

Even assuming it were desirable to reward manag- ers for improvements in accounting performance, such rewards should not be given for short-term fluctuations but rather only for improvements over a considerable period of time. Rewarding executives for short-term improvements is not an effective way to provide beneficial incentives and indeed might create incentives to manipulate short-term accounting results.

Similarly, compensation contracts should gen- erally include claw-back provisions that require managers to return payments based on accounting figures that are subsequently restated. Such return of payments is warranted, regardless of whether the executive was in any way responsible for the misreporting. When the board believes it is desir- able to tie executive payoffs to a formula involving a metric whose value turns out to have been inflated, correctly applying the formula requires reversing payments that were based on an errone- ous value. The principle should be: “What wasn’t earned must be returned.”

5. Be Wary of Paying for Expansion

Because running a larger firm increases managers’ power, prestige, and perquisites, executives might have an excessive incentive to expand the com- pany. Executive compensation arrangements should seek to counter rather than reinforce this incentive. Thus, the common practice of paying executives bonuses for making acquisitions and otherwise rewarding managers for firm expansion can create perverse incentives. While the in- creased difficulty of running a larger firm might make it necessary to pay executives of bigger firms additional compensation, boards should keep in mind that such practices provide executives with ex ante incentives to expand (say, by making ac- quisitions) even when expansion is not value- maximizing.

6. Dividend-Neutrality

Under current option plans, terms are not updated to reflect the payment of dividends, and as a result executives’ payoffs are reduced when they decide to pay a dividend. Indeed, there is evidence that

executives whose pay has a large option compo- nent tend to issue lower dividends. Instead, they resort to share repurchases, which have a less adverse effect on the value of managers’ options but may not always be the most efficient form of payout.34 To reduce distortions in managers’ pay- out decisions, all equity-based compensation should be designed in such a way that it is divi- dend-neutral; that is, it neither encourages nor discourages the payment of dividends. In particu- lar, in the case of option plans, the exercise price of options should be adjusted in any case where a dividend is paid.

7. Rethinking Executive Pensions

There are reasons to doubt the efficiency of the widespread practice of using Supplemental Executive Retirement Plans (SERPs) to provide executives with a major component of their career compensation. Unlike pension plans used for non- executive employees, SERPs do not enjoy a tax subsidy. And given that firms have been generally moving away from defined benefit plans to de- fined contribution plans for non-executive em- ployees, it is far from clear that providing execu- tives with defined benefit plans is required by risk-bearing considerations. While defined benefit plans shift the risk of investment performance from the employee to the firm, executives do not seem to be less able to bear such risk than other employees.

While the efficiency benefits of SERPs are far from clear, SERPs provide executives with pay that is largely decoupled from performance and thus weakens the overall link between total pay and performance. Firms thus would do well to reconsider their heavy reliance on SERPs.

8. Avoiding Soft-Landing Arrangements

Soft-landing arrangements, which provide manag- ers with a generous exit package when they are pushed out due to failure, dilute executives’ incen- tives. While firms spend large amounts on produc- ing a payoff gap between performing well and performing poorly, the money spent on soft-land- ing arrangements works in the opposite direction, narrowing the payoff gap between good and poor performance.

2006 21Bebchuk and Fried

At present, executives are commonly promised generous severance arrangements in the event of termination, unless the termination is triggered by an extremely narrow set of circumstances (such as criminal indictment or “malfeasance”). Even if firms stick to the existing broad definition of ter- mination without cause, the payoff in such a ter- mination should depend in part on the firm’s performance relative to its peers during the exec- utive’s service. An executive who is terminated against the background of extremely poor stock performance should get less than an executive who is terminated when the company’s perfor- mance is reasonable. Furthermore, firms should consider provisions that make the termination payoff depend on the reasons for the executive’s termination.

Improving Board Accountability

P ast and current flaws in executive pay arrange- ments, we argue, have resulted from underlying problems within the corporate governance sys-

tem: specifically, directors’ lack of sufficient in- centive to focus solely on shareholder interests when setting pay. If directors could be relied on to focus on shareholder interests, the pay-setting pro- cess, and board oversight of executives more gen- erally, would be greatly improved. The most promising route to improving pay arrangements is thus to make boards more accountable to share- holders and more focused on shareholder interests.

Increasing accountability to shareholders would transform the arm’s-length contracting model into a reality and lead to improved pay- setting processes. Accountability would thus lead to better-designed compensation arrangements as well as improved board performance more gener- ally.

Recent reforms require most companies listed on the major stock exchanges (the New York Stock Exchange, NASDAQ, and the American Stock Exchange) to have a majority of indepen- dent directors— directors who are not otherwise employed by the firm or in a business relationship with it. These companies must also staff compen- sation and nominating committees entirely with independent directors. These reforms are likely to reduce managers’ power over the board and im-

prove directors’ incentives somewhat. But they fall far short of what is necessary.

Our analysis shows that the new listing require- ments weaken executives’ influence over directors but do not eliminate it. More importantly, there are limits to what independence can do by itself. Independence does not ensure that directors have incentives to focus on shareholder interests or that directors will be well-selected. In addition to be- coming more independent of insiders, directors also must become more dependent on sharehold- ers. To this end, we should eliminate the arrange- ments that currently entrench directors and insu- late them from shareholders.

To begin, shareholders’ power to replace direc- tors should be turned from myth into reality. Even in the wake of poor performance and shareholder dissatisfaction, directors now face very little risk of being ousted. Shareholders’ ability to replace di- rectors is extremely limited. A recent study by one of us provides evidence that, outside the hostile takeover context, the incidence of electoral chal- lenges to directors has been practically negligible in the past decade.35 This state of affairs should not continue.

To improve the performance of corporate boards, impediments to director removal should be reduced.36 To begin, shareholders should be given the power to place director candidates on the corporate ballot. Secondly, proxy contest challengers that attract sufficient support should receive reimbursement for their expenses.

Furthermore, it would be desirable to eliminate staggered boards, which most public companies now have, and have all directors stand for annual election. Staggered boards provide a powerful pro- tection from removal in either a proxy fight or a hostile takeover. A recent empirical study by Alma Cohen and one of us finds that staggered boards bring about an economically significant reduction in firm value.37

In addition to making shareholder power to remove directors viable, boards should not have veto power—which current corporate law grants them— over changing governance arrangements in the company’s charter. Shareholders should have the power, which they now lack, to initiate and adopt changes in the corporate charter.

22 FebruaryAcademy of Management Perspectives

Under current rules, shareholders can pass only nonbinding resolutions, and a recent empirical study by one of us documents that boards com- monly elect not to follow resolutions that receive majority support from shareholders, even if such resolutions pass two or three times.38

Allowing shareholders to amend the corporate charter would improve over time the entire range of corporate governance arrangements without outside regulatory intervention. If there is concern that shareholders are influenced by short-term con- siderations, shareholder-initiated changes could re- quire approval by majority vote in two successive annual shareholder meetings. But we should not continue denying shareholders the power to change the corporate charter, no matter how widespread and long-lasting shareholder support for such a change is. Allowing shareholders to set governance arrangements would contribute to making boards more accountable to shareholders.

To fully address the existing problems in exec- utive compensation and corporate governance, structural reforms in the allocation of power be- tween boards and shareholders are necessary. Given political realities and the power of vested interests, such reforms would not be easy to pass. But the corporate governance flaws that we have discussed—which we have seen to be pervasive, systemic, and costly— call for such reforms.

Endnotes 1 Bebchuk, L.A. & Fried, J.M. 2004. Pay without

performance: The unfulfilled promise of executive compen- sation. Cambridge: Harvard University Press. Accompa- nying articles include: Bebchuk, L.A.. & Fried, J.M. 2004. Stealth compensation via retirement benefits. Berkeley Business Law Journal 2: 291–325. Bebchuk, L.A. & Fried, J.M. forthcoming. Executive compensation at Fannie Mae: A case study of perverse incentives, non- performance pay, and camouflage. Journal of Corporation Law; Bebchuk, L.A. & Grinstein, Y. The growth of executive pay. 2005. Oxford Review of Economic Policy 21: 283–303. Earlier work by us on which the book draws includes Bebchuk, L.A., Fried, J.M., & Walker, D.I. 2002. Managerial power and rent extraction in the de- sign of executive compensation. University of Chicago Law Review 69: 751– 846; Bebchuk, L.A. & Fried, J.M. 2003. Executive compensation as an agency problem. Journal of Economic Perspectives 17:71–92.

2 Bebchuk & Grinstein, supra note 1. 3 See, e.g., Blanchard, O.J., Lopez-de-Silanes, F., & Shleifer,

A. 1994. What do firms do with cash windfalls? Journal

of Financial Economics 36: 337–360; Yermack, D. 1997. Good timing: CEO stock option awards and company news announcements. Journal of Finance 52: 449 – 476; Bertrand, M. & Mullainathan, S. 2001. Are CEOs r ewarded for luck? The ones without principals are,” Quarterly Journal of Economics 116: 901–932.

4 See Bebchuk, Fried, & Walker, supra note 1; Bebchuk and Fried, Executive Compensation as an Agency Problem, supra note 1.

5 For surveys from this perspective in the finance and eco- nomics literature, see, for example, Abowd, J.M. & Kaplan, D.S. 1999. Executive compensation: Six ques- tions that need answering. Journal of Economic Perspec- tives 13: 145–168; Core, J.E., Guay, W., & Larcker, D.F. 2003. Executive equity compensation and incentives: A survey. Economic Policy Review 9: 27–50.

6 See, e.g., Core, J.E., Guay, W.R. & Thomas, R.S. 2005. Is U.S. CEO compensation inefficient? Michigan Law Re- view 103: 1142–1185.

7 Nasaw, D. 2003. Opening the board: The fight is on to determine who will guide the selection of directors in the future, Wall Street Journal, October 27, 2003, R8.

8 Brick, I.E., Palmon, O. & Wald, J.K. forthcoming. CEO Compensation, director compensation, and firm performance: Evidence of cronyism. Journal of Corporate Finance.

9 Main, B.G.M., O’Reilly III, C.A., & Wade, J. 1995. The CEO, the board of directors, and executive compensation: Economic and psychological perspec- tives. Industrial and Corporate Change 11: 292–332.

10 Main, O’Reilly III, & Wade, supra note 9. 11 Murphy, K.J. 1999. Executive compensation, in Handbook

of Labor Economics, Ashenfelter, O. & Card, D. (Eds.). New York: Elsevier.

12 Bizjak, J.M., Lemmon, M.L., & Naveen, L. 2003. Has the use of peer groups contributed to higher levels of exec- utive compensation? Working paper.

13 Bebchuk, L.A., John Coates IV, J., & Subramanian, G. 2002. The powerful antitakeover force of staggered boards: Theory, evidence, and policy. Stanford Law Re- view 54: 887–951.

14 Murphy, K.J. 2002. Explaining executive compensation: Managerial power vs. the perceived cost of stock options. University of Chicago Law Review 69: 847– 869.

15 See, e.g., Jenkins, H.W. 2002. Outrageous CEO pay re- visited, Wall Street Journal, October 2, 2002, A17.

16 Core, J., Holthausen, R., & Larcker, D. 1999. Corporate governance, chief executive compensation, and firm per- formance. Journal of Financial Economics 51: 371– 406.

17 Core, Holthausen, & Larcker, supra note 16; Cyert, R., Sok-Hyon Kang, S. & Kumar, P. 2002. Corporate gov- ernance, takeovers, and top-management compensation: Theory and evidence. Management Science 48: 453– 469.

18 Cyert, Kang, & Kumar, supra note 17. 19 Bertrand, M. & Mullainathan, S. 2000. Agents with and

without principals. American Economic Review 90: 203– 208.

20 Hartzell, J.C. & Starks, L.T. 2003. Institutional investors and executive compensation. Journal of Finance 58: 2351–2374.

2006 23Bebchuk and Fried

21 Parthiban, D., Kochar, R., & Levitas, R. 1998. The effect of institutional investors on the level and mix of CEO compensation. Academy of Management Journal 41: 200 – 208.

22 Borokhovich, K.A., Brunarski, K.R., & Parrino, R. 1997. CEO contracting and anti-takeover amendments. Jour- nal of Finance 52: 1503–1513.

23 Cheng, S., Nagar, V., & Rajan, M.V. 2005. Identifying control motives in managerial ownership: Evidence from antitakeover legislation. Review of Financial Studies 8: 637– 672.

24 Thomas, R.S. & Martin, K.J. 1999. The effect of share- holder proposals on executive compensation. University of Cincinnati Law Review 67: 1021–1065.

25 Wade, J.B., Porac, J.F., & Pollock, T.G. 1997. Worth, words, and the justification of executive pay. Journal of Organizational Behavior 18: 641– 664.

26 Bebchuk, L.A. & Jackson, Jr., R. 2005. Executive pen- sions. Forthcoming. Journal of Corporation Law.

27 Jensen, M.C. & Murphy, K.J. 1990. Performance pay and top-management incentives. Journal of Political Economy 98: 225–264; Jensen, M.C. & Murphy, K.J. 1990. CEO incentives: It’s not how much you pay, but how. Harvard Business Review 68: 138 –153.

28 Bebchuk & Grinstein, supra note 1. 29 See Murphy, supra note 11. 30 Grinstein, Y. & Hribar, P. 2004. CEO compensation and

incentives: Evidence from M&A bonuses. Journal of Fi- nancial Economics 71: 119 –143.

31 Bergstresser, D. & Philippon, T. forthcoming. CEO in- centives and earnings management: Evidence from the 1990s. Journal of Financial Economics; Summers, S.L. & Sweeney, J.T. 1998. Fraudulently misstated financial statements and insider trading: An empirical analysis. Accounting Review 73: 131–146.

32 Bebchuk & Fried, supra note 1. 33 See Fried, J.M. 1998. Reducing the profitability of corpo-

rate insider trading through pretrading disclosure. South- ern California Law Review 71: 303–392.

34 See Fried, J.M. forthcoming. Informed trading and false signaling with open market repurchases. California Law Review.

35 Bebchuk, L.A. 2003. The case for shareholder access to the ballot. The Business Lawyer 59: 43– 66.

36 For a fuller analysis of the ways in which shareholder power to remove directors could be made viable, see Bebchuk, L.A. 2005.The myth of the shareholder fran- chise. Working paper.

37 Bebchuk, L.A. & Cohen, A. 2005. The costs of en- trenched boards. Journal of Financial Economics 78:409 – 433.

38 Bebchuk, L.A. 2005. The case for increasing shareholder power. Harvard Law Review 18: 833–914.

24 FebruaryAcademy of Management Perspectives