Bebchuk fried pay without performance; the unfulfilled promise of executive compensation (2004)

Preface AWAVE OF corporate scandals that began in late 2001 shook confidence in the performance of public company boards and drew attention topotential flaws in their executive compensatio

Pay without Performance

Harvard University Press

Cambridge, Massachusetts, and London, England 2004

Copyright 䉷 2004 by Lucian Bebchuk and Jesse Fried

All rights reserved

Printed in the United States of America

Library of Congress Cataloging-in-Publication Data

Bebchuk, Lucian

Pay without performance: the unfulfilled promise of executive compensation / Lucian Bebchuk and Jesse Fried.

p cm.

Includes bibliographical references and index.

ISBN 0–674–01665–3 (alk paper)

1 Executives—Salaries, etc 2 Corporate governance I Fried, Jesse II Title HD4965.2.B43 2004

331.2'816584—dc22 2004052253

For Alma, Alon, and Yonatan

—L.B.

For Naomi, Joshua, Avital, and Ayelet

—J.F.

Contents

Preface ix

Part I The Official View and Its Shortcomings

1 The Official Story 15

2 Have Boards Been Bargaining at Arm’s Length? 23

3 Shareholders’ Limited Power to Intervene 45

4 The Limits of Market Forces 53

Part II Power and Pay

5 The Managerial Power Perspective 61

6 The Relationship between Power and Pay 80

7 Managerial Influence on the Way Out 87

8 Retirement Benefits 95

9 Executive Loans 112

Part III Decoupling Pay from Performance

11 Windfalls in Conventional Options 137

12 Excuses for Conventional Options 147

13 More on Windfalls in Equity-Based Compensation 159

14 Freedom to Unwind Equity Incentives 174

viii Contents

Part IV Going Forward

15 Improving Executive Compensation 189

16 Improving Corporate Governance 201

Preface

AWAVE OF corporate scandals that began in late 2001 shook confidence

in the performance of public company boards and drew attention topotential flaws in their executive compensation practices There is nowrecognition that many boards have employed compensation arrangementsthat do not serve shareholders’ interests But there is still substantial dis-agreement about the scope and source of such problems and, not sur-prisingly, about how to address them

Many take the view that concerns about executive compensation havebeen exaggerated Some of these observers believe that flawed compen-sation arrangements have been limited to a small number of firms andthat most boards have carried out effectively their role of setting executivepay Others concede that flaws in compensation arrangements have beenwidespread, but maintain that these flaws have resulted from honest mis-takes and misperceptions on the part of boards seeking to serve share-holders In this view, now that the problems have been recognized, cor-porate boards can be expected to fix them on their own

Our aim in this book is to persuade readers that such complacency ishardly warranted Flawed compensation arrangements have been wide-spread, persistent, and systemic, and they have stemmed from defects inthe underlying governance structure that enable executives to exert con-siderable influence over their boards Given executives’ power, directorscould not have been expected to engage in arm’s-length bargaining withexecutives over their compensation The absence of effective arm’s-lengthdealing under today’s system of corporate governance—not temporarymistakes or lapses of judgment—has been the primary source of prob-lematic compensation arrangements

Another, broader aim of this book is to contribute to a better standing of some basic problems of the corporate governance system Our

under-x Preface

study of executive compensation opens a window through which we canexamine our current reliance on boards to act as guardians of share-holders’ interests Our corporate governance system gives boards sub-stantial power and counts on them to monitor and supervise the com-pany’s managers As long as it is believed that corporate directors carryout their tasks for the benefit of shareholders, current governance ar-rangements—which insulate boards from intervention by shareholders—appear acceptable Our work casts doubt on the validity of this belief and

on the wisdom of insulating boards from shareholders

By providing a full account of how and why boards have failed to servetheir critical role in the executive compensation area, we hope to con-tribute to efforts to improve compensation practices and corporate gov-ernance more generally Understanding the source of existing problems

is essential for assessing reforms Some observers now concede thatboards have not been sufficiently attentive to shareholders’ interests, butargue that an increase in the role of independent directors—facilitated

by recently adopted stock exchange requirements—will make such lems a matter of the past But this is not the case Our analysis indicatesthat, to address the identified problems, directors must be made not onlymore independent of insiders but also more dependent on shareholders

prob-We therefore put forward reforms that would reduce boards’ insulationfrom shareholders Such reforms may well offer the most promising routefor improving executive compensation and corporate governance

This book builds on our 2001 study with David Walker, “Executive pensation in America: Managerial Power or Extraction of Rents?” which

Com-was published in 2002 in the University of Chicago Law Review under the

title “Managerial Power and Rent Extraction in the Design of ExecutiveCompensation.” It also draws on another piece by the two of us, “Ex-ecutive Compensation as an Agency Problem,” which was published in

the Journal of Economic Perspectives in the summer of 2003 As we indicate

at various points throughout the book, we also build on other work that

we have done on corporate governance

In the course of writing this book, we have incurred considerable debtsthat we wish to acknowledge Special thanks go to David Walker, ourcoauthor on the 2001 study David worked on the first stage of this studyand made an important contribution to its development In the fall of

2000, he entered legal practice for two years, and the demands of this

Preface xi

practice prevented him from continuing to work with us His sound ment and insights contributed much to our 2001 study, and we verymuch missed them after he left

judg-We are also grateful to many individuals for their valuable discussionsand comments on drafts of this book or the earlier pieces on which itdraws These individuals include Marc Abramowitz, Yitz Applbaum, AdiAyal, Franklin Ballotti, Oren Bar-Gill, Lisa Bernstein, Margaret Blair,Victor Brudney, Brian Cheffins, Steve Choi, Bob Cooter, Brad DeLong,Mel Eisenberg, Charles Elson, Allen Ferrell, Merritt Fox, Jeff Gordon,Mitu Gulati, Dan Halperin, Assaf Hamdani, Sharon Hannes, HenryHansmann, Paul Hodgson, Marcel Kahan, Louis Kaplow, Ira Kay, ReinierKraakman, Stuart Gillan, Michael Levine, Saul Levmore, Patrick McGurn,Bob Monks, Kevin Murphy, Richard Painter, Adam Pritchard, Mark Roe,David Schizer, Steve Shavell, Andrei Shleifer, Eric Talley, Timothy Taylor,Randall Thomas, DetlevVagts, Steve Vogel, Michael Wachter, MichaelWaldman, Ivo Welch, and David Yermack

We also received very useful suggestions from participants in shops at various universities and conferences Thanks go to workshopparticipants at Boalt Hall, Harvard Law School, the Kennedy School ofGovernment at Harvard, Harvard Business School, the American Asso-ciation of Law Schools Business Law Panel on Executive Compensation,the Berkeley Business Law Journal symposium on “The Role of Law inthe Creation of Long-term Shareholder Value,” and the University ofChicago Law Review symposium on “Management and Control of theModern Business Corporation.”

work-We are also indebted to many research assistants who have greatlyhelped us in this project: Alex Aizin, Chris Busselle, Erin Carroll, MisunIsabelle Chung, LubovGetmansky, Miranda Gong, Nicholas Hecker, Mat-thew Heyn, Ryan Kantor, Jason Knott, Karen Marciano, MatthewMcDermott, Selena Medlen, Jay Metz, Aaron Monick, Jeff Wagner, andDan Wolk For their valuable help and for their dedication we are grateful

to the assistants we had while working on the book, especially MadelineBurgess, Julie Johnson, and Anita Sarrett Our presentation was substan-tially improved by the careful copyediting done by Alexandra McCor-mack, Jean Martin, Emily Ogden, and Moshe Spinowitz

For financial support, we wish to thank the John M Olin Center forLaw, Economics, and Business at Harvard Law School; the Harvard LawSchool; the Berkeley Committee on Research; and the Boalt Hall Fund.Lucian Bebchuk wishes to thank Robert Clark and Elena Kagan, his

xii Preface

former and current deans at Harvard Law School, and Professor SteveShavell, the director of the Olin Center, for their encouragement andsupport Jesse Fried is similarly indebted to Herma Hill Kay, John Dwyer,and Robert Berring, his former and current deans at Boalt Hall

We owe a large debt to our wives, Alma Cohen and Naomi Fried, whohave provided a critical and constant source of support, advice, and un-derstanding Our children—Alon and Yonatan in the case of Lucian, andJoshua, Avital, and Ayelet in the case of Jesse—have been somewhat lessunderstanding, but their sweetness has added much to our lives duringthe period in which we worked on this book

June 2004

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

re-Warren Buffett, letter to shareholders ofBerkshire Hathaway, Inc., February 2004

Introduction

Is it a problem of bad apples, or is it the barrel?

Kim Clark, Dean of the Harvard

Business School, 2003

DURING THE EXTENDED bull market of the 1990s, executive pensation at public companies—companies whose shares are traded onstock exchanges—soared to unprecedented levels Between 1992 and

com-2000, the average real (inflation-adjusted) pay of chief executive officers(CEOs) of S&P 500 firms more than quadrupled, climbing from $3.5 mil-lion to $14.7 million.1Increases in option-based compensation accountedfor the lion’s share of the gains, with the value of stock options granted toCEOs jumping ninefold during this period.2 The growth of executivecompensation far outstripped that of compensation for other employees

In 1991, the average large-company CEO received approximately 140times the pay of an average worker; in 2003, the ratio was about 500:1.3

Executive pay has long attracted much attention from investors, cial economists, regulators, the media, and the public at large The higherCEO compensation has climbed, the keener that interest has become.Indeed, one economist has calculated that the dramatic growth in exec-utive pay during the 1990s was outpaced by the increase in the volume

finan-of research papers on the subject.4

Executive compensation has also long been a topic of heated debate.The rise in pay has been the subject of much public criticism, whichfurther intensified following the corporate governance scandals that beganerupting in late 2001 But the evolution of executive compensation duringthe past two decades has also had powerful defenders In their view,despite some lapses, imperfections, and cases of abuse, executive arrange-ments have largely been shaped by market forces and boards loyal toshareholders

2 Introduction

Our main goal in this book is to provide a full account of how agerial power and influence have shaped the executive compensationlandscape The dominant paradigm for financial economists’ study ofexecutive compensation has assumed that pay arrangements are theproduct of arm’s-length bargaining—bargaining between executives at-tempting to get the possible deal for themselves and boards seeking toget the best possible deal for shareholders This assumption has also beenthe basis for the corporate law rules governing the subject We aim toshow, however, that the pay-setting process in publicly traded companieshas strayed far from the arm’s-length model

man-Our analysis indicates that managerial power has played a key role inshaping managers’ pay arrangements The pervasive role of managerialpower can explain much of the contemporary landscape of executivecompensation Indeed, it can explain practices and patterns that have longpuzzled financial economists studying executive compensation

We seek to contribute to a better understanding of the flaws in currentcompensation arrangements and in the corporate governance processesgenerating them Such an understanding is necessary for addressing theseproblems We show that recent corporate governance reforms, which seek

to increase board independence, would likely improve matters but thatmuch more needs to be done And we put forward reforms that, bymaking directors more accountable to shareholders, would reduce theforces that have in the past distorted compensation arrangements

The Official View and Its Shortcomings

Part I discusses the shortcomings of the “official” view of executive pensation According to this view, which underlies existing corporate gov-ernance arrangements, corporate boards operate at arm’s length from theexecutives whose pay arrangements they decide Seeking to serve share-holders, directors design cost-effective compensation arrangements thatprovide executives with incentives to increase shareholder value

com-Recognition of managers’ influence over their own pay has been at theheart of the criticism of executive compensation in the media and byshareholder activists.5 However, the premise that boards negotiate payarrangements at arm’s length from executives has long been and remains

a central tenet in the corporate world and in most research on executivecompensation Holders of the official view believe it provides a goodapproximation of reality When faced with practices that are hard to rec-oncile with arm’s-length contracting, they seek to explain these “deviant”

Introduction 3

examples as “rotten apples” that do not represent the entire barrel or asthe result of temporary lapses, mistakes, or misperceptions that, onceidentified, will promptly be corrected by boards

In the corporate world, the official view serves as the practical basis forlegal rules and public policy It is used to justify directors’ compensationdecisions to shareholders, policymakers, and courts These decisions areportrayed as being made largely with shareholders’ interests at heart andtherefore deserving of deference

The official view’s premise of arm’s-length bargaining has also beenshared by most of the research on executive compensation Managers’influence over directors has been recognized by those writing on thesubject from a legal, organizational, or sociological perspective.6But most

of the research on executive pay (especially empirical research) has beendone by financial economists, and the premise of arm’s-length bargaininghas guided most of their work Some financial economists, whose studies

we discuss later, have reported findings they viewed as inconsistent witharm’s-length contracting.7However, the majority of work in the field hasassumed arm’s-length bargaining between boards and executives

In the paradigm that has dominated financial economics, which welabel the “arm’s-length bargaining” approach, the board of directors isviewed as operating at arm’s length from executives and seeking to max-imize shareholder value Rational parties transacting at arm’s length havepowerful incentives to avoid inefficient provisions that shrink the pieproduced by their contractual arrangements The arm’s-length con-tracting approach has thus led researchers to believe that executive com-pensation arrangements will tend to increase value, which is why we haveused the terms “efficient contracting” or “optimal contracting” to labelthis approach in some of our earlier work.8

Financial economists, both theorists and empiricists, have largelyworked within the arm’s-length model in attempting to explain commoncompensation arrangements as well as variation in compensation prac-tices among firms.9In fact, upon discovering practices that appear incon-sistent with the cost-effective provision of incentives, financial economistshave often labored to come up with clever explanations for how suchpractices might be consistent with arm’s-length contracting after all Prac-tices for which no explanation has been found have been considered

“anomalies” or “puzzles” that will ultimately either be explained withinthe paradigm or disappear

The official arm’s-length story is neat, tractable, and reassuring ever, as we explain in part I, this model has failed to account for the

fac-in their firms, their financial fac-incentives to avoid arrangements favorable

to executives have been too weak to induce them to take the personallycostly, or at the very least unpleasant, route of haggling with their CEOs.Finally, limitations on time and resources have made it difficult for evenwell-intentioned directors to do their pay-setting job properly

Some writers have argued that even if directors are subject to erable influence from corporate executives, market forces can be relied

consid-on to force boards and executives to adopt the compensaticonsid-on ments that arm’s-length bargaining would produce Our analysis, how-ever, finds that market forces are neither sufficiently finely tuned norsufficiently powerful to compel such outcomes The markets for capital,

arrange-corporate control, and managerial labor do impose some constraints on

executive compensation These constraints are hardly stringent, however,and they permit substantial deviations from arm’s-length contracting

A realistic picture of the incentives and circumstances of board bers, then, reveals myriad incentives and tendencies that lead directors tobehave very differently than expected under the arm’s-length model Re-cent reforms, such as the new stock exchange listing requirements thatseek to limit CEOs’ ability to financially reward independent directors,may weaken some of these factors but will not eliminate them Withoutadditional reforms, the pay-setting process will continue to deviate sub-stantially from the arm’s-length model

mem-Power and Pay

After analyzing the shortcomings of the arm’s-length contracting view, weturn in part II to the managerial power perspective on executive com-pensation The same factors that limit the usefulness of the arm’s-lengthmodel suggest that executives have had substantial influence over theirown pay Compensation arrangements have often deviated from arm’s-length contracting because directors have been influenced by manage-ment, sympathetic to executives, insufficiently motivated to bargain overcompensation, or simply ineffectual in overseeing compensation Execu-

Introduction 5

tives’ influence over directors has enabled them to obtain efits greater than those obtainable under true arm’s-length bargaining.Although top executives generally have some degree of influence overtheir board, the extent of their influence depends on various features ofthe firm’s governance structure The managerial power approach predictsthat executives who have more power vis-a`-vis their boards should receivehigher pay—or pay that is less sensitive to performance—than their lesspowerful counterparts A substantial body of evidence does indeed indi-cate that pay has been higher, and less sensitive to performance, whenexecutives have more power

“rents”—ben-There are, of course, limits to the arrangements that directors willapprove and executives will seek Markets, such as the market for cor-porate control, might penalize boards that allow pay arrangements thatappear egregious Directors and executives might in such a case also bearsocial costs The constraints imposed by markets and by social forces arefar from tight, however, and they permit substantial deviations fromarm’s-length outcomes The adoption of arrangements favoring executives

is unlikely to impose substantial economic or social costs if the ments are not patently abusive or indefensible

arrange-One important building block of the managerial power approach isthat of “outrage” costs When a board approves a compensation arrange-ment favorable to managers, the extent to which directors and executivesbear economic and social costs will depend on how the arrangement isperceived by outsiders whose views matter to the directors and executives

An arrangement that is perceived as outrageous might reduce holders’ willingness to support incumbents in proxy contests or takeoverbids Outrage might also lead to shareholder pressure on managers anddirectors, as well as possibly embarrass directors and managers or harmtheir reputations The more outrage a compensation arrangement is ex-pected to generate, the more reluctant directors will be to approve it andthe more hesitant managers will be to propose it in the firstplace

share-The critical role of outsiders’ perception of executives’ compensationand the significance of outrage costs explain the importance of yet an-other component of the managerial power approach: “camouflage.” Thedesire to minimize outrage gives designers of compensation arrangements

a strong incentive to try to obscure and legitimize—or, more generally,

to camouflage—both the level and performance-insensitivity of executivecompensation Camouflage thus allows executives to reap benefits at the

6 Introduction

expense of shareholders More importantly, attempts to camouflage canlead to the adoption of inefficient compensation structures that harmmanagers’ incentives and, in turn, company performance, imposing evengreater costs on shareholders

We present evidence that compensation arrangements have often beendesigned with an eye to camouflaging rent and minimizing outrage Firmshave systematically taken steps that make less transparent both the totalamount of compensation and the extent to which it is decoupled frommanagers’ own performance Managers’ interest in reduced transparencyhas been served by the design of numerous compensation practices, such

as postretirement perks and consulting arrangements, deferred sation, pension plans, and executive loans Overall, the camouflage motiveturns out to be quite useful in explaining many otherwise puzzling fea-tures of the executive compensation landscape

compen-Performance Pay and the Unfulfilled Promise of Executive Pay

Those applauding the rise in executive compensation have stressed thebenefits to shareholders from strengthening managers’ incentives to in-crease shareholder value Indeed, in the beginning of the 1990s, promi-nent financial economists such as Michael Jensen and Kevin Murphyurged shareholders to be more accepting of large pay packages that wouldprovide high-powered incentives.10 Shareholders, it was argued, shouldcare much more about providing managers with sufficiently strong in-centives than about the amounts spent on executive pay

Indeed, throughout the past decade, shareholders have often acceptedthe increase in executive pay as the price of improving managers’ incen-tives Higher compensation has been presented as essential for improvingmanagers’ incentives and therefore worth the additional cost Unfortu-nately, however, much of the additional value provided to executives hasnot actually been tied to their own performance Shareholders have notreceived as much bang for their buck as possible

As we describe in part III, managers have used their influence to obtainhigher compensation through arrangements that have substantially de-coupled pay from performance Firms could have generated the sameincrease in incentives at a much lower cost, or they could have used theamount spent to obtain more powerful incentives Executive pay is muchless sensitive to performance than has commonly been recognized.Although equity-based compensation has recently drawn the most

Introduction 7

attention, much executive pay comes in forms other than equity, such assalary and bonus The evidence indicates that cash compensation—in-cluding bonuses—has been at best weakly correlated with firms’ industry-adjusted performance Such compensation has been generously awardedeven to managers whose performance was mediocre relative to other ex-ecutives in their industry Furthermore, financial economists have paidlittle attention to the other forms of non-equity compensation that man-agers frequently receive, such as favorable loans, pensions and deferredcompensation, and various perks These less-noticed forms of compen-sation, which can be substantial, have tended to be insensitive to mana-gerial performance

In light of the historically weak link between non-equity compensationand managerial performance, shareholders and regulators wishing tomake pay more sensitive to performance have increasingly looked to, andencouraged, equity-based compensation—that is, compensation based onthe value of the company’s stock Most equity-based compensation hastaken the form of stock options—options to buy a certain number ofcompany shares for a specified price (the “exercise” or “strike” price) Westrongly support equity-based compensation, which in principle can pro-vide managers with desirable incentives Unfortunately, however, man-agers have been able to use their influence to obtain option arrangementsthat have deviated substantially from arm’s-length contracting in waysthat favor the managers Our analysis indicates that equity-based planshave enabled executives to reap substantial rewards even when their per-formance was merely passable or even poor

For instance, firms have failed to filter out stock price rises that aredue largely to industry and general market trends and thus are unrelated

to managers’ own contribution to shareholder value Although there is awhole range of ways in which such windfalls could be filtered out, a largemajority of firms have continued to cling to conventional option plansunder which most of the equity-based compensation paid to managers isnot tied to their own performance In addition, firms have given execu-tives broad freedom to unload options and shares, a practice that hasbeen beneficial to executives but costly to shareholders Unfortunately,most of the boards now changing their equity-based compensation plans

in response to outside pressure are still choosing to avoid plans thatwould effectively eliminate such windfalls Rather, they are moving toplans, such as those based on restricted stock, that fail to eliminate, andsometimes even increase, these windfalls

8 Introduction

Alternative Critiques of Executive Compensation

Criticism of executive compensation practices can come from a variety

of methodological and ideological perspectives It is important to makeclear at the outset how our take on the subject differs from other types

of criticism Indeed, in some respects, our positions are closer to those ofsupporters of current pay arrangements than to those of other critics ofthese arrangements

To begin, there is the “moral,” “fairness-based”—and, some might say,

“populist”—opposition to large amounts of pay In this view, puttingaside practical consequences, paying executives hundreds of times whatother employees get is inherently unfair and unacceptable

Our own criticism does not come from this perspective Our approach

is completely pragmatic and consequentialist, focusing on shareholdervalue and the performance of corporations (and, in turn, the economy

as a whole) We would accept compensation at current or even higherlevels as long as such compensation, through its incentive effects, actuallyserves shareholders We are concerned, however, that the compensationarrangements that have been in place do not meet this standard

It is also important to distinguish our position from the view thatfinancial incentives are not very important in motivating top executivesand that enhancing shareholder value therefore does not call for largepay packages At least since the first half of the past century, some in-dustrial psychologists have maintained that corporate executives, who areall materially well off anyway, are primarily moved by such factors as theneed for esteem, self-actualization, and so forth.11In this view, “The realdriving force which motivates the typical executive is not money, butthe deep inner satisfaction that he is doing a tough job well.”12 Accord-ingly, increasing the pay of already well-paid managers, it is argued, doesnot affect performance and is simply a waste of shareholder money

In contrast to this view, we share the assumption of defenders of rent pay arrangements that executives are influenced by financial incen-tives We agree that paying generously to provide desirable incentives can

cur-be a good compensation strategy for shareholders Indeed, the fact thatexecutives (as well as directors) are influenced by financial incentives andhave an interest in increasing their own pay actually plays an importantrole in our analysis Our concern is simply that executives have partlytaken over the compensation machine, leading to arrangements that fail

to provide managers with desirable incentives

them-of pay, and the showering them-of gratuitous benefits on departing executives.

The Stakes

How important is the subject of executive pay? Why should one read awhole book on the subject? Some might wonder whether executive com-pensation has a significant economic impact on the corporate sector Theproblems existing in the area of executive compensation, it might beargued, do not substantially affect shareholders’ bottom line and are thusmainly symbolic

Even if symbolism were unimportant, however, the subject of executivecompensation is of substantial practical importance for shareholders andpolicymakers Flaws in compensation arrangements impose substantialcosts on shareholders To begin with, there is the excess pay that managersreceive as a result of their power, that is, the difference between whatmanagers’ influence enables them to obtain and what they would getunder arm’s-length contracting As a current study by Yaniv Grinsteinand one of us seeks to document in detail,13 the amounts involved arehardly pocket change for shareholders

During the five-year period 1998–2002, the compensation paid to thetop five executives at each company in the widely used ExecuComp data-base, aggregated over the 1500 companies in the database, totaled about

$100 billion (in 2002 dollars) And the capitalized present value of gregate top-5 compensation in publicly traded U.S companies is rathersubstantial During the past ten years, the growth rate of aggregate ex-ecutive compensation has kept pace with that of total stock market cap-italization Assuming that aggregate executive compensation continues togrow in tandem with market capitalization or that managers’ share of

ag-10 Introduction

aggregate corporate profits remains at current levels, the capitalizedpresent value of aggregate top-5 compensation in publicly held U.S firmscould be on the order of half a trillion dollars Thus, if compensationcould be cut without weakening managerial incentives, the gain to inves-tors would not be merely symbolic It would have real practical signifi-cance

Furthermore, and perhaps even more important, managers’ influence

on compensation arrangements dilutes and distorts the incentives thatthey produce In our view, the reduction in shareholder value caused bythese inefficiencies—rather than that caused by excessive managerialpay—could well be the biggest cost arising from managerial influenceover compensation

We discuss two incentive problems that current pay arrangements havebeen producing First, compensation arrangements have been providingweaker incentives to reduce managerial slack and to increase share-holder value than would be the case under arm’s-length contracting.Both the non-equity and equity components of managerial compensationhave been more severely decoupled from managers’ contribution to com-pany performance than superficial appearances might suggest Makingpay more sensitive to performance may well benefit shareholders sub-stantially

Second, prevailing compensation practices have also created perverseincentives For example, managers’ ability to unload options and shareshas provided them with incentives to misreport results, suppress badnews, and choose projects and strategies that are less transparent to themarket Improving compensation schemes could thus considerably ben-efit shareholders by reducing the costs resulting from such distorted be-havior

Going Forward

In part IV, we turn to some of the implications of our analysis, both forexecutive compensation and for corporate governance more generally.Given the difficulty of achieving an arm’s-length relationship betweenboards and managers, it is important that shareholders scrutinize exec-utive compensation for inefficient pay structures Our analysis highlightsthe types of schemes that institutional investors should support Suchinvestors should, for example, pressure firms to use equity-basedschemes that filter out windfalls, tie pay tightly to management’s own

man-be in the public domain and fully accessible and understood by a limitednumber of market professionals Given the importance of outsiders’ per-ceptions and scrutiny, transparency is critical for reducing the extent towhich managers’ influence distorts compensation arrangements.

Getting compensation arrangements to be more consistent with length contracting and getting boards to be more effective monitors ofmanagers generally are likely to be more difficult than many expect Re-cent reforms require companies listed on the major stock exchanges (theNew York Stock Exchange, NASDAQ, and the American Stock Exchange)

arm’s-to have a majority of independent direcarm’s-tors—direcarm’s-tors who are not erwise employed by the firm or in a business relationship with it Thesecompanies must also staff compensation and nominating committeeswith such directors and submit equity compensation plans to a share-holder vote Our analysis indicates that even though these reforms arelikely to be beneficial, they cannot be relied on to produce the kind ofarm’s-length relationship between directors and executives on which ourcorporate governance system relies

oth-What else should be done? To induce boards to perform their criticalrole well, we should focus not only on insulating directors from the in-fluence of executives, but also on reducing their current insulation fromshareholders While we should lessen directors’ dependence on executives,

we should also seek to increase directors’ dependence on shareholders.Even in the wake of poor performance and shareholder dissatisfaction,directors now face very little risk of being ousted in a proxy contest or ahostile takeover This state of affairs should not continue To improve theperformance of corporate boards, arrangements that insulate directorsfrom removal by either a proxy fight or a takeover should be eliminated

or reduced Additionally, boards should not have veto power—whichcurrent corporate law grants them—over changing the governance ar-rangements in the company’s charter

Although we do sketch out some of the implications of our analysis,our aim in this book is not to offer a fully detailed blueprint for changes

in pay arrangements and corporate governance Rather, we focus on someprior and crucial steps: demonstrating that, contrary to the official story

12 Introduction

of executive compensation, boards have not been bargaining at arm’slength with managers over their pay; explaining how managerial influenceand rent seeking have played important roles in shaping executive com-pensation; and providing a full account of the range and adverse conse-quences of the resulting deviations from arm’s-length bargaining.This is an area in which the very recognition of problems may help toalleviate them Managers’ ability to influence pay structures depends onthe extent to which the resulting distortions are not too apparent tomarket participants—especially institutional investors Thus, recognition

of how managerial influence can produce substantial deviations from ficient contracts may serve as a useful check simply by reducing managers’ability to camouflage rents

ef-To address the problems that we examine in corporate governance,additional structural reforms in the allocation of power between boardsand shareholders are necessary To make such reforms possible, share-holders and public officials must have a fuller understanding of howpervasive, systemic, and costly the current flaws in governance have been.Helping to bring about such an understanding is a main aim of this book

Part I

The Official View and Its Shortcomings

1

The Official Story

The board of directors is the oversight mechanism charged

with monitoring management and providing shareholders

with accountability

Ira Millstein, “The Professional Board,” 1995

WE SHOULD START by describing briefly the “official” view of utive compensation—that boards, bargaining at arm’s length with CEOs,negotiate pay arrangements designed to serve shareholders’ interests Thisview underlies corporate law’s approach to executive compensation, helpslegitimize compensation arrangements, and informs much of the largebody of research on executive compensation carried out by financialeconomists

exec-The Agency Problem

Our focus in this book is on publicly traded American companies without

a controlling shareholder This diffuse ownership structure is the norm

in the United States, though not in other countries.1 The dispersion ofshareholder interests was first documented in 1932 by Adolph Berle and

Gardiner Means in their classic study The Modern Corporation and Private

Property,2and it remains the dominant form of ownership among publiclytraded companies in the United States

The dispersed owners of a typical publicly traded company cannotmonitor or direct managers’ actions, so the executives who exert day-to-day control in such companies often have considerable discretion In such

a situation, ownership and control are separated Shareholders own thecompany, but the managers exercise a substantial amount of control overhow it is run

The separation of ownership and control creates what financial

econ-16 The Official View and Its Shortcomings

omists call an “agency relationship”: a company’s managers act as agents

of its shareholders The principals (the shareholders) cannot directly sure that the agents (the managers) will always act in the principals’ bestinterests As a result, the manager-agents, whose interests do not fullyoverlap those of the shareholder-principals, may deviate from the bestcourse of action for shareholders This is called the “agency problem.”3

en-Managers’ departures from shareholder-regarding strategies in turn mayinvolve “inefficient” behavior—behavior that reduces the size of the cor-porate pie The reductions in aggregate company value caused by suchdeviations are called “agency costs.”

The agency problem can affect a wide variety of managerial choices:how much effort to exert, how many perks to consume, and which stra-tegic and business choices to make for the company In each instance,managers’ interests may not match those of shareholders Consider, forexample, decisions about how much effort to exert Because managerswill bear the entire cost of their own effort, but will not fully enjoy itsbenefit to the company, they will tend to exert less effort than is optimal

On the other hand, because managers will fully enjoy the perks theyconsume while not bearing their entire cost, they will have an incentive

to consume too much

Managers’ private interests may also distort business decisions that fect company size CEOs may engage in empire building, which can in-crease their prestige, perks, compensation, and other private benefits.4

af-Because managers derive more private benefits from being at the helm

of a larger company, they may make acquisitions and additional ments that reduce shareholder value They may also fail to reorganize andreduce the scope of operations when downsizing is called for Moreover,

invest-to avoid company contraction and perhaps also invest-to facilitate future empirebuilding, managers may retain too much cash, failing to distribute excessfunds to shareholders even when their firms do not have profitable in-vestment opportunities.5

Agency problems are likely to affect other business decisions as well.Overall, managers may run companies in ways that are personally moresatisfying or convenient even if they come at the expense of shareholders.They may be tempted to pursue pet projects, for example, or they mayfail to take actions that are personally costly, such as firing mediocresubordinates who also happen to be their friends

Finally, managers may continue to run their companies even when theyare no longer the best people for the job.6They may turn down attractiveacquisition offers or block takeover attempts that would increase share-

The Official Story 17

holder value but cost them their positions And they have an incentive

to take entrenching actions (such as adopting antitakeover measures) thatmake it more difficult to replace them

The Board as Guardian of Shareholder Interests

The official theory of executive compensation recognizes that there is anagency problem in publicly traded companies with separation of owner-ship and control This agency problem is supposed to be addressed bythe board’s supervision and monitoring of managers Under the rules ofcorporate law, the power to run the company is not given to the CEOand other officers Rather, this power is vested in the board, under whosedirection the business and affairs of the corporation are supposed to bemanaged.7

Although the board has formal authority and control, directors are notexpected to manage the company themselves The directors of publiclytraded companies have other primary careers and thus do not performtheir board roles full-time In addition, many directors sit on more thanone board Directors are therefore generally expected to delegate ongoingmanagement to the company’s officers and especially to the CEO But theboard’s power to intervene is supposed to keep managers in line Thethreat of board intervention is expected to curb managers’ tendency to-ward self-serving behavior, thereby reducing agency costs.8

After selecting and hiring executives, directors are therefore supposed

to monitor their performance, replacing them as necessary Major porate decisions, such as how to respond to an acquisition offer, areultimately reviewed by the board, which has full power to accept or rejectmanagement’s recommendations

cor-In carrying out its supervisory duties, the board must be guided by theinterests of the corporation and its shareholders The directors have afiduciary duty to the company and its shareholders In addition to thisduty, the directors are assumed to have an incentive to serve shareholderinterests Failure to do so, it is widely believed, may lead shareholders toreplace the board by voting in a different slate of directors or by sellingtheir shares to a hostile acquirer

Arm’s-Length Bargaining over Compensation

Given executives’ natural interest in being paid more and working less,permitting them to set their own pay would obviously generate large

18 The Official View and Its Shortcomings

agency costs Therefore, the board is directly entrusted with these sions Under the official theory of executive compensation, the board isassumed to bargain at arm’s length with executives over their pay, solelywith the interests of the corporation and its shareholders in mind Thatpremise underlies corporate law’s treatment of board compensation de-cisions Following a strong presumption that directors exercise their busi-ness judgment to serve shareholders, courts generally defer to boards’decisions on compensation issues As the Delaware Supreme Court re-cently wrote, “The size and structure of executive compensation are in-herently matters of judgment” that are entitled to “great deference” bythe courts.9

deci-The same premise underlies most of the large volume of research thatfinancial economists have done on executive compensation The domi-nant model in the economics literature assumes that, in negotiating com-pensation, directors take an independent position vis-a`-vis executives Theboard is viewed as bargaining with management with the exclusive goal

of serving shareholder interests In an effort to understand executive pensation practices, financial economists have done substantial workwithin this model of arm’s-length bargaining.10 Even after the wave ofcorporate scandals that began erupting in late 2001, financial economistshave continued to use arm’s-length contracting as the main lens throughwhich to view compensation arrangements

com-Efficient Contracting and Paying for Performance

What would characterize an executive compensation arrangement duced by arm’s-length bargaining between the executive and a boardseeking to maximize shareholder value? To begin with, the contract mustprovide enough value to induce the executive to accept and remain inthe position being offered Thus, the contract must provide benefitswhose value meets or exceeds the value of the other opportunities avail-able to the candidate (the executive’s “reservation value”)

pro-In addition, when rational, self-interested buyers and sellers transact,their contracts tend to avoid inefficient terms, that is, terms that reducethe size of the pie produced by the contractual arrangement and shared

by the transacting parties Thus, for example, employment contracts tend

to take advantage of forms of compensation that are tax subsidized andthus can increase the parties’ combined wealth, and tend to avoid waysthat impose an unnecessary tax cost and therefore reduce the parties’

The Official Story 19

combined wealth For this reason, when studying compensation ments from the perspective of the arm’s-length model, financial econo-mists have viewed terms that seem inefficient as puzzling and have sought

arrange-to show that they might be value enhancing after all

Economists have long believed that efficient compensation contractsshould link pay with performance to provide executives with desirableincentives Indeed, according to the standard view, the compensation ar-rangement is an important mechanism for reducing agency costs Andthe significance of the agency problem makes it crucial to use this in-strument effectively

Directors have neither the time nor the information necessary to itor all managerial actions to ensure that they benefit shareholders Giventhe considerable discretion inherent in a CEO’s position, inducing theCEO to focus on shareholder interests and avoid self-serving choices istherefore important The board can influence the CEO to behave in thismanner by designing a compensation arrangement that provides the CEOwith an incentive to increase shareholder value Thus, it is argued, a well-designed compensation scheme can make up for the fact that directorscannot directly monitor or evaluate many of their top executives’ deci-sions Such a well-designed scheme can substantially reduce agency costs,improve performance, and increase shareholder value.11

mon-Linking compensation to performance may require a company to crease an executive’s level of compensation because pay that is sensitive

in-to performance is less valuable in-to managers than fixed pay with the sameexpected value Because the company’s performance will depend on somefuture factors beyond the executive’s control, tying a manager’s compen-sation to performance makes the level of compensation uncertain Man-agers are generally risk averse—they value a dollar paid with certaintymore than they value variable pay with an expected value of a dollar (e.g.,

a 50 percent chance of receiving two dollars) This risk aversion makesperformance-based compensation worth less to an executive Everythingbeing equal, it would take more performance-based compensation thanfixed compensation to meet the executive’s reservation value As long asmanagers’ incentives are important, however, an efficient contract can beexpected to provide a major part of its compensation in ways that induceand reward performance For this reason, economists have long empha-sized the importance of increasing the sensitivity of compensation to per-formance.12

Indeed, the incentive benefits of performance-based pay might be

sub-20 The Official View and Its Shortcomings

stantial enough to make it desirable for shareholders to pay their topexecutives more than the “reservation wage” needed to hire and retainthem.13 It would be worth granting, say, another $1,000,000 ofperformance-based compensation if the effect of the stronger incentives

on the executive’s behavior is expected to increase the value of the pany by more than $1,000,000 Importantly, however, such an increase

com-in compensation beyond the executive’s reservation wage could serveshareholders only if it were given in a performance-based form

In examining whether the empirical reality has been consistent witharm’s-length contracting, it will be useful to focus on the structure ofexecutive compensation Because a board seeking to maximize share-holder value may set high levels of compensation, such levels do not bythemselves demonstrate a departure from arm’s-length contracting.Under the arm’s-length contracting view, however, there is little reason

to expect the widespread persistence of arrangements that, by distortingincentives or otherwise, compensate managers in an inefficient way Thus,evidence that such arrangements are widespread and persistent will be amore telling sign of departures from the arm’s-length model than absoluteamounts could be

It’s the Market

Defenders of existing compensation practices often try to base their case

on analogies to other markets for talent For example, while testifyingbefore a U.S Senate committee, noted compensation consultant Ira Kayargued that “the CEO labor market meets all of the criteria of anymarket.”14 In this view, compensation arrangements are the outcome ofmarket interactions, a product of the combined forces of the supply ofand demand for managerial talent Executive compensation arrangementsproduced by this process, it is argued, are not more problematic, andshould not be questioned more, than the compensation arrangementsobtained by other highly paid individuals, such as star athletes and famousmovie actors

Are CEOs like Star Athletes?

The analogy to star athletes is one that defenders of existing pay practicesoften seek to invoke.15 After all, star athletes are a rather popular andadmired group, and reports about their high salaries are commonly

The Official Story 21

greeted with awe and approval rather than with outrage Furthermore,the compensation of star athletes climbed dramatically during the 1990s.16

Thus, defenders of current executive pay practices can assert that themeteoric rise in executive pay is simply part of the broader phenomenon

of the rise in the value of special talent that has also manifested itself inother markets The rise in executive pay, so the argument goes, is nomore problematic than the fact that top basketball center ShaquilleO’Neal is paid much more than earlier great centers such as Bill Russelland Wilt Chamberlain

But the process generating the compensation of, say, O’Neal is quitedifferent from that producing the compensation arrangements of, say,Disney CEO Michael Eisner Those invoking the market analogy implic-itly rely on the premise of arm’s-length bargaining, which is valid forathletes but not for executives When an athlete’s compensation arrange-ments are set, there is little doubt that the club’s manager is negotiatingwith the athlete at arm’s length The manager is seeking to serve the club’sinterests, not those of the individual player And when transactions occurbetween independent buyers and sellers, the invisible hand of the markettends to produce efficient arrangements

Indeed, it is worth noting that although star athletes are highly paid,some more than the average S&P 500 CEO, their compensation arrange-ments lack the features of executive pay arrangements that managerialinfluence produces After the compensation packages of star athletes arenegotiated, clubs have little reason to try to camouflage the amount ofpay and to channel pay through arrangements designed to make the payless visible While athletes are paid generously during the period of theircontracts, clubs generally do not provide them with a large amount ofcompensation in the form of postretirement perks and payments.Clubs also generally do not provide athletes with complex deferred-compensation arrangements that serve to obscure total pay And whenclubs get rid of players, they do not generally provide them with largegratuitous payments in addition to the players’ contractually entitled pay-outs As we shall see, however, these are all common practices in the area

of executive compensation Executives are not like star athletes

Is It Possible for Most Executives to Be Overpaid?

Defenders of current pay arrangements also argue that the only way toassess whether an individual is overpaid is by reference to the compen-

22 The Official View and Its Shortcomings

sation paid to others in the same market One cannot judge a sation arrangement, it is argued, outside the context of the particularmarket in which it is set In this view, it is impossible, by definition, formost executives to be paid too much Furthermore, in this view, com-pensation arrangements should not be seen as problematic as long asboards use market surveys to set compensation in line with that paid toexecutives of other publicly traded companies

compen-It is true that, by definition, most CEOs cannot be compensated abovethe median level of their peer group But boards acting at arm’s lengthshould be doing more than ensuring that the compensation they set can

be defended as being in line with market levels Boards acting at arm’slength are supposed to try to get, against the background of market con-ditions, the best outcome for shareholders

Furthermore, in the absence of an arm’s-length bargaining process,concerns about executive compensation arrangements cannot be assuaged

by a determination that all executives get similar packages The absence

of arm’s-length bargaining could still mean that managers are paid toomuch or paid in inefficient ways In such a market, compensation levelscould be higher than those that would prevail if arm’s-length bargainingshaped the market Thus, when the market as a whole is distorted by theabsence of arm’s-length bargaining, general conformity to market termscannot allay concerns about the amount and structure of compensation

In the end, then, the validity of the arguments for deference to marketoutcomes depends on whether those outcomes are largely generated byarm’s-length negotiations between executives and self-interested pur-chases of their services The critical question, which we will consider next,

is whether the arm’s-length model is a sufficiently accurate reflection ofreality

2

Have Boards Been

Bargaining at Arm’s Length?

The directors of [joint stock] companies, however, being

the managers rather of other people’s money than of their

own cannot well be expected [to] watch over it with

the same anxious vigilance [as the owners themselves]

Adam Smith, 1776

THE ARM’S-LENGTH CONTRACTING view recognizes that executives

do not automatically serve shareholders It concedes that executives mightwell act in ways that suit their own interests—a tendency that properincentives and board oversight are supposed to curb But the arm’s-lengthmodel implicitly assumes that, unlike corporate executives, corporate di-rectors can be relied on to serve shareholders

Given that executives do not instinctively seek to maximize shareholdervalue, however, there is no reason to expect a priori that directors willact in this way Directors’ own incentives and preferences do matter.Directors have financial and nonfinancial incentives to favor, or at least

to get along with, executives A variety of psychological and social factorsreinforce these incentives Because directors hold only a tiny fraction ofthe firm’s shares, their holdings are insufficient to outweigh their incen-tives and tendencies to side with executives In any event, directors havethus far had neither the time nor the information necessary to serveshareholder interests when determining executive compensation

The pay-setting process is, of course, better in some firms than inothers However, significant deviations from arm’s-length contractinghave been common in widely held public companies And while recentlyadopted stock exchange requirements will probably somewhat improvepay-setting processes, they are unlikely to eliminate substantial and wide-spread deviations from arm’s-length contracting

24 The Official View and Its Shortcomings

The Pay-Setting Process

We begin our exploration of the limitations of the arm’s-length modelwith a brief description of the pay-setting process in large public corpo-rations The board of directors is responsible for determining the com-pensation of the CEO and other top executives.1Boards of large publiccompanies delegate to compensation committees the task of working outthe critical details of executive compensation arrangements

The compensation committee has typically been composed of three orfour directors.2For some time now, most directors serving on compen-sation committees have been “independent.” Directors are generally con-sidered independent if they are not current or former employees of thefirm and are not affiliated with the firm other than through their direc-torship The Investor Responsibility Research Center reported that 73percent of the S&P 1500 compensation committees surveyed during 2002were fully independent.3

Tax rules and court decisions have contributed to the widespread use

of compensation committees made up exclusively of independent tors Since 1994, the federal tax code has penalized corporations lackingsuch committees; publicly traded corporations have not been permitted

direc-to deduct pay in excess of $1 million annually per executive unless theexcess compensation either consists of options or is based on the achieve-ment of performance goals that have been established by a compensationcommittee composed solely of independent directors.4Courts have gen-erally upheld compensation arrangements recommended to the board by

a compensation committee composed of independent directors Thus, theuse of such a committee has largely insulated board compensation deci-sions from judicial review.5

Although already very common, the presence of independent directors

on boards and on compensation committees in particular is expected togrow because of the listing requirements adopted by the NYSE, NASDAQ,and AMEX and approved by the Securities and Exchange Commission(SEC) in 2003.6These new provisions require the boards of most publiclytraded companies to have a majority of independent directors.7Under theNYSE rules, each company must have a compensation committee com-posed only of independent directors Under the rules of the other two ex-changes, the CEO’s compensation must be determined or recommended

to the board by a majority of independent directors or a compensationcommittee composed solely of such directors The stock exchange re-quirements also establish standards for determining independence

Have Boards Been Bargaining at Arm’s Length? 25

Although these new requirements have attracted a great deal of tion, it is important to keep in mind that they merely make mandatory

atten-a pratten-actice thatten-at most public compatten-anies atten-alreatten-ady hatten-ave been following forsome time Thus it seems unlikely that these new requirements, by them-selves, will greatly change the relationship between executives and theirboards Indeed, there are good reasons to doubt that the mere presence

of independent directors on the board and on the compensation mittee can ensure a pay-setting process that approximates arm’s-lengthbargaining

com-Directors’Desire to Be Reelected to the Board

A director receives a number of benefits from serving on a board First,

a board seat provides direct financial benefits In most cases, these benefitsare likely to be economically significant to the director Like executivepay, director pay rose dramatically with the stock market In 2002, di-rector compensation averaged $152,000 in the largest 200 companies and

$116,000 in the largest 1,000 companies.8There are often additional perksand indirect benefits; for example, directors of UAL Corp (which ownsUnited Airlines) can fly United free of charge, and directors of StarwoodHotels get complimentary nights in company hotels.9Moreover, a boardseat often provides directors with prestige and with valuable business andsocial connections The financial and nonfinancial benefits of holding aboard seat give directors a strong interest in keeping their positions.That directors have a desire to be reelected is clear The question, then,

is what incentives this desire provides According to the official view, thedesire to be reelected by shareholders should make directors attentive toshareholder interests; the better their performance, the argument goes,the more likely they are to win reelection

In reality, however, candidates placed on the company’s slate by theboard have been virtually assured of being reelected Dissident share-holders contemplating putting forward their own director slate have con-fronted substantial obstacles.10As a result, the director slate proposed bythe company has almost always been the only one on the ballot In arecent empirical study of the seven-year period 1996–2002, we documentthat, outside the context of hostile takeovers, the incidence of electoralchallenges to the board’s slate was practically nonexistent—no more thantwo a year among firms with a market capitalization exceeding $200 mil-lion.11

The key to a board position is, therefore, getting one’s name on the