The Economics of Corporate Executive Pay
Prepared for Members and Committees of Congress
In the past ten years, the pay of chief executive officers (CEOs) has more than doubled, and the
ratio of median CEO to worker pay has risen to 179 to 1. High and rising executive pay could be
an issue of public concern on two different grounds. First, it is contributing to widening income
inequality that may be of concern from an equity perspective. Second, it could be the result of
economically inefficient labor markets. It is difficult to determine whether executive pay is
excessive across the board since executives’ marginal product cannot be directly observed. An
upward trend in pay over time is not sufficient proof that the market is not efficient since factors
determining supply and demand, such as the skills required of the position, can change over time.
To show that pay is excessive from an economic perspective, one must first demonstrate that
there is a market failure that is preventing the market from functioning efficiently. The market
failure could originate in the division in large modern firms between management and ownership,
which is typically dispersed among millions of shareholders. Shareholders’ interests are
represented by a board of directors. Critics of executive pay have argued that boards have all too
often been “captured” by the executive and are no longer negotiating pay packages that are in the
shareholders’ best interests. They point to a number of common practices that they call “stealth
compensation” which are inconsistent with arm’s length contracting. These include “golden
parachutes,” generous severance packages, company-provided perks, and bonuses that are
unrelated to firm performance.
Stock options have been the fastest growing portion of executive pay since the 1990s, and critics
believe this pattern can also be explained through the prism of stealth compensation. Rewarding
executives with employee stock options was often justified in terms of the “pay for performance”
mantra, but options are usually designed to reward absolute, not relative, performance. This
means that in the bull market of the 1990s, when virtually all stock prices were rising, a company
could fall behind its competitors and its executives could still receive handsome options payouts.
Indeed, a sizeable portion of the increase in executive pay in the 1990s was likely due to options
that turned out to be much more valuable than expected because of the unprecedented price
increases of the bull market.
Many of the recent corporate scandals appear consistent with stealth compensation as well. Stock
options backdating, earnings manipulation, and accounting fraud might have been motivated by
attempts to covertly increase executive pay. If short-term fluctuations in the stock price are not
good proxies of firm performance, then tying compensation to the stock price can create
incentives for executives to engage in activities that are detrimental to shareholders. Policy
proposals mostly focus on improving transparency, increasing board independence, and
strengthening shareholder control rather than attempting to curb pay directly. S. 1181 (Obama)
and H.R. 1257 (Frank), which the House approved on April 19, 2007, would give shareholders a
non-binding vote on executive pay. Another proposal would modify the limit on deductibility of
executive pay from corporate taxation. More broadly, income inequality could be reduced by
increasing the progressivity of the tax system. For current developments and legislation, see CRS
Report RS22604, Excessive CEO Pay: Background and Policy Approaches.
Introduc tion ..................................................................................................................................... 1
Data on CEO Pay............................................................................................................................3
The Economics of Executive Pay....................................................................................................5
How Executive Compensation Is Set........................................................................................6
Executive Pay in the Neo-Classical World................................................................................6
The Principal-Agent Problem and the Free-Rider Problem......................................................7
The “Managerial Power” Critique of the Neo-Classical Model................................................8
Stealth Compensation and Outrage Costs.........................................................................10
Performance Based Pay: Part of the Solution or Part of the Problem?..............................11
Criticisms .......................................................................................................................... 16
Executive Pay in the 1990s: Being in the Right Place at the Right Time?..............................18
Can “Excessive Pay” Be Empirically Measured?...................................................................19
A Brief History of the Regulation of Executive Pay.....................................................................21
The Omnibus Budget Reconciliation Act (OBRA) of 1993’s Cap on the Deductibility
of Executive Compensation.................................................................................................22
The Sarbanes-Oxley Act.........................................................................................................22
Requirement that Shareholders Approve Equity-Based Compensation Plans for
NYSE and Nasdaq Listed Firms..........................................................................................23
Requirement that NYSE-Listed Companies Have Compensation Committees Solely
Composed of Outside Directors...........................................................................................23
FASB’s Options Accounting Rule...........................................................................................23
SEC Rules on Disclosure of Executive Pay............................................................................24
Maintain the Status Quo..........................................................................................................25
Eliminating or Significantly Restricting OBRA.....................................................................26
A Cap on the Deductibility of Executive Pay When It Exceeds a Certain Ratio of
Disclosure of Corporate Services Provided by Firms Affiliated with Compensation
Consultants or a Ban on Such Services................................................................................27
Increase Shareholder Roles in the Election of Board Members..............................................28
Give Shareholders a Non-binding Vote on Executive Pay......................................................29
Increase the Progressivity of the Tax Code.............................................................................30
Figure 1. Median CEO Pay at the 350 Largest Publicly Held Companies......................................4
Figure 2. S&P 500: Actual and Projected Based on Historical Average, 1995-1999....................19
Appendix. A Primer on Stock Options..........................................................................................32
Author Contact Information..........................................................................................................33
For years, many observers have characterized the annual pay packages awarded to the executives
and chief executive officers (CEOs) of the nation’s largest firms as excessive. According to one
estimate, the total median CEO pay at the nation’s 350 largest publicly owned firms grew from 1
$2.7 million in 1995 to $6.8 million in 2005. The overall increase in CEO pay has outstripped
inflation and the growth in non-managerial pay over the same period, fueling criticism of
executive compensation packages.
Among the quite vociferous critics of executive pay are public officials, academics, shareholders,
and some Members of Congress. For example, William J. McDonough, chairman of the Public
Company Accounting Oversight Board, has said that there is “... nothing in economic theory to 2
justify the levels of executive compensation that are widely prevalent today.” Warren Buffett has
observed that “...it’s difficult to overpay the truly extraordinary CEO of a giant enterprise. But 3
this species is rare.” Former Securities and Exchange Commission (SEC) Chairman Arthur
Levitt, Jr. has written that “these huge paydays, I believe, undermine corporate governance and 4
send a signal that boards are willing to spend shareholders’ money lavishly...” In addition, a 2006
survey found that 90% of institutional investors, who are the largest owners of outstanding 5
domestic corporate shares, said that executives at most companies are overpaid.
There are also a number of current examples of congressional concerns over executive pay. For
example, the version of H.R. 2 (the minimum wage bill) passed by the Senate on February 1,
2007, included several tax provisions, one of which applies to executive pay. Current tax rules
permit individuals to defer taxes on income that is held in non-qualified deferred compensation
plans. Under the Senate version of H.R. 2, an individual could defer no more than $1,000,000 6
annually from taxable income by contributing to such a plan. The version passed by the House
had no similar provision. Arguing for the cap, Chairman Max Baucus of the Senate Finance
Committee, which unanimously passed the tax provisions that became part of the Senate version
of H.R. 2, observed:
Rank-and-file workers generally have to pay taxes on their compensation when they earn it.
The exception is deferred compensation provided through qualified retirement plans with
statutory limits on contributions and benefits. A 401(k) is the best example. Management, on
1 See “Testimony of Frederic W. Cook & Co. before the House Financial Services Committee on Executive
Compensation,” May 25, 2006. Cook & Co.’s data were provided by Mercer Human Resources Consulting. Total pay
includes salary, bonuses, deferred compensation, and incentive pay such as stock grants and stock options. The median
measurement is the middle observation in the sample, if an average had been used instead, both figures would have
been higher. There are no official estimates for executive pay; however, private estimates have been made based on
data that public corporations are required to file in their annual statements.
2 “Remarks by President William J. McDonough delivering the William Taylor Memorial Lecture,” Public Company
Accounting Oversight Board, September 29, 2002.
3 “To the Shareholders of Berkshire Hathaway Inc.,” Berkshire Hathaway 2005 Annual Report, p. 16.
4 Arthur Levitt, “Corporate Culture and the Problem of Executive Compensation,” December 2005, p. 2, available at
5 “Corporate Directors Give Executive Pay Model Mixed Reviews,” Watson Wyatt Press Release, June 20, 2006.
6 Joint Committee on Taxation, Description of the Chairman’s Modification of the Provisions of the “Small Business
and Work Opportunity Act of 2007” (JCX-5-07), January 17, 2007, p. 25.
the other hand, has no limit on the amount that can be deferred to nonqualified 7
With respect to CEO pay, Chairman Barney Frank of the House Financial Services Committee
I do not think the boards of directors work as effective independent checks. They are not the
fox guarding the hen house. They are the hens guarding the rooster. And I think the time has 8
come to say we need the shareholders to do this....
In the 110th Congress, Representative Frank has introduced the Shareholder Vote on Executive
Compensation Act (H.R. 1257), which was approved by the House on April 19, 2007. Soon
afterwards, Senator Obama introduced an identically named companion bill, S. 1181.
The bills would require publicly held companies to hold annual non-binding shareholder votes on
their executive compensation plans and any new “golden parachute” compensation offered to
executives during mergers and acquisitions.
Critics argue that excessive executive pay can be significantly traced to the fact that the members
of corporate boards are not sufficiently independent of managerial influence. In this view, boards
formulate executive pay packages that often allow executives to extract hefty compensation deals
that bear little relationship to their contribution to the firms. In addition, if corporate executives
are overpaid, other potentially significant concerns include the following:
• Issues over pay inequality and worker productivity. According to one
estimate, between 1994 and 2005, the ratio of annual median CEO pay to median 9
production worker pay nearly doubled, growing from 90 to 1 to 179 to 1.
Indeed, for many workers, the perceived excessiveness of executive pay has
become the most visible embodiment of growing pay inequality, contributing to a
feeling that workers have not shared in the gains from economic growth. For
example, adjusted for inflation, average worker pay rose 8% from 1995 to 10
2005; median CEO pay at the 350 largest firms rose about 150% over the same
period. And while it is very difficult to quantify the impact of executive pay
packages on worker morale and productivity, there are concerns that both could
be affected detrimentally. For example, there have been several high profile
stories of executives who have laid off employees at the same time their own
compensation was rising.
• A potentially negative direct impact on shareholder returns. If executive
compensation has been receiving a share of corporate resources that far exceeds
the executives’ contribution to the firm’s value, shareholder returns may be
compromised. Although not directly addressing the issue of whether increases in
executive pay have been accompanied by commensurate increases in their value
7 Congressional Record, January 24, 2007, pp. S1011-S1048.
8 “Congressman Barney Frank Outlines His Priorities,” PBS’ Nightly Business Report, November 30, 2006, available at
9 “The Testimony of Frederic W. Cook & Co. Before the House Financial Services Committee on Executive
Compensation,” May 25, 2006.
10 As measured by the Bureau of Labor Statistics data series on average earnings of production and non-supervisory
to the firms, one study compared the total pay given to the top five executives
relative to corporate earnings at S&P 1500 firms between 1993-1995 and 2001-
doubled between the two periods.
• A potential proxy for sub-par board performance in monitoring executives.
When a corporate board is unable to get a handle on excessive CEO pay, it may
be a sign of the board’s failings in its central role of corporate governance.
• A purported link between excessive executive pay and the corporate
scandals of recent years. Critics argue that a misalignment of incentives from
executive pay helps explain the occurrence of scandals involving stock options
backdating, accounting fraud, and earnings manipulation.
But this “managerial power” perspective of how flawed corporate governance serves to inflate
CEO pay is at odds with an alternative notion that the increase has largely been a function of
natural market-driven changes in the demand and supply for CEOs. For example, during a 2006
interview with The Wall Street Journal, then Treasury Secretary John Snow observed that “in an
aggregate sense... [CEO pay] reflects the marginal productivity of CEOs.... Until we can find a 12
better way to compensate CEOs, I’m going to trust the marketplace.” Like Secretary Snow,
others argue that market forces are more likely than government to solve any problems that do
exist with excessive pay. In his state of the economy address on January 31, 2007, President Bush
Government should not decide the compensation for America’s corporate executives, but the
salaries and bonuses of CEOs should be based on their success at improving their companies
and bringing value to their shareholders. America’s corporate boardrooms must step up to
their responsibilities. You need to pay attention to the executive compensation packages that
you approve. You need to show the world that American businesses are a model of
transparency and good corporate governance.
Whether or not policy intervention is merited to tackle the issue of executive pay will depend on
whether the managerial power or the market-based view of executive pay is the accurate one.
Given congressional concerns with the potentially serious implications of excessive executive
pay, this report examines critical and supportive evidence surrounding the premise that the
executives are generally overpaid from an economic perspective. The report also provides a brief
history of executive pay regulation and analyzes policy options. But first, the report looks more
closely at the available data.
Median CEO pay began its rapid rise in 1993. As seen in Figure 1, which goes to 2005, it rose
35% in 2001 alone, which would later prove to be the high point. The current economic
expansion has not been as favorable to CEOs. Contrary to popular belief, executive pay does not
always rise—following the large decline in stock prices, median pay fell 13% from 2001 to 2002.
Since then it has made up lost ground, and median pay in 2005 was nearly equal to its 2001 peak.
11 Lucian Bebchuk and Yaniv Grinstein, “The Growth of Executive Pay,” Oxford Review of Economic Policy, vol. 21,
2005, pp. 283-303.
12 Greg Ip, “Snow Defends President’s Handling of Economy,” Wall Street Journal, March 20, 2006.
The 1990s were not the first time that CEO pay had risen rapidly; it also rose by about one half 13
from 1982 to 1987.
As important as the trend in the level of CEO pay is the trend in the composition of CEO pay.
Since 1995, median annual cash salary has risen relatively slowly, from $0.7 million in 1995 to
about $1 million in 2006. But over the same period, median performance-based pay, which
includes stock and stock option grants, grew 13% a year, from $1.3 million to $4.1 million. As
will be discussed later, the role of stock and stock options in executive pay helps explain why it
rose so quickly in the 1990s, and why it fell in 2002.
Figure 1. Median CEO Pay at the 350 Largest Publicly Held Companies
Source: Mercer Human Resources Consulting.
The disparity between the pay given to U.S. executives relative to the compensation awarded to
CEOs in other nations is often mentioned as evidence that American executives are overpaid. For
example, controlling for various firm characteristics such as size, one study examined pay for
CEOs in the United States and the United Kingdom (UK) in 2003. It found that CEO pay in the
United States to be about 1.3 times that of CEOs in the UK and argued that a significant portion
of the greater U.S. CEO pay could be attributed to the higher proportion of equity-based pay, such 14
as stock options, in the United States, and the higher risk premiums associated with such pay.
Research on Japanese executives found that after controlling for firm size, their pay constituted 15
about a third of the pay of their U.S. counterparts in 2004. However, in Japan, the UK, and
13 Kevin Murphy, “Executive Compensation,” in Orley Ashenfelter and David Card, eds., Handbook of Labor
Economics, vol. 3B, Elsevier Publishers (Amsterdam: 1999), p. 2485.
14 Martin Conyon, John Core, and Wayne Guay, “How High Is U.S. CEO Pay? A Comparison with UK CEO Pay,”
working paper, Social Science Research Network, June 2006.
15 Minoru Nakazato, Mark Ramseyer, and Eric Bennett, “Executive Compensation in Japan: Estimating Levels and
Determinants from Tax Records,” Harvard Law and Economics Discussion Paper No. 567, December 2006.
many other nations of the developed world, there are reports that CEO pay levels abroad have 16
been slowly converging with the United States over time.
High and rising executive pay could be a cause for policy concern on efficiency grounds or on
equity grounds. Current pay levels would be economically inefficient if they resulted from a
market failure that prevented pay levels from reflecting an equilibrium between supply and
demand in the labor market for executives. If so, a policy response might be justified on the
grounds that resources could be allocated in the economy more efficiently. However, even if pay
levels were the result of perfect competition, policymakers could still be concerned that high
executive pay resulted in an inequitable distribution of wealth that had negative social
ramifications. The grounds for concern about executive pay matters greatly in determining the
appropriate policy response. If current pay levels are economically efficient, then policy
responses that disrupted the market outcome would lead to a misallocation of resources and
deadweight loss by distorting people’s career choices. Rather than targeting executive pay
specifically, a more economically efficient response to address equity concerns would be through
the normal channels of government redistribution, such as the progressivity of the tax system and
social welfare spending (although it should be noted that these channels create efficiency
tradeoffs of their own). Equity concerns are, at heart, matters of competing societal values for
which economic theory cannot offer definitive answers. The non-economic gains of policies to
improve equity would need to be balanced against their economic costs.
Some economists argue that the trend in executive pay is not just an equity issue, it is the result of
market imperfections that allow executives to manipulate the market outcome. If this is the case,
then executive pay is causing resources to be misallocated to an extent that is, arguably, non-
negligible. For example, one study estimated that the compensation of the top five executives 17
averaged 6.6% of total corporate earnings from 1993 to 2003. Other economists argue that,
despite its flaws, the market for executive pay delivers results that are close enough to efficiency 18
that potential policy changes run the risk of doing more harm than good. This section first
describes how the market would determine executive pay under perfect competition (the so-called
“neo-classical approach”), and then considers real-world deviations from that model. Readers
who would first like a brief description of the mechanics of setting executive pay should read
“How Executive Compensation Is Set,” in the box below.
16 For example, see Eric Wahlgren, “Spreading the Yankee Way of Pay,” Business Week online, April 18, 2001.
17 Lucian Bebchuk and Jesse Fried, “Pay Without Performance: Overview of the Issues,” Journal of Corporation Law,
vol. 30, no. 4, 2005, p. 647. Of course, only a fraction at most of this compensation is “excessive.” (The definition of
excessive pay is discussed in a later section.) Whether or not “excessive pay” is a negligible share of total corporate
earnings depends on what fraction of the 6.6% is “excessive.”
18 See, for example, R. Glenn Hubbard, “Pay Without Performance: A Market Equilibrium Critique,” Journal of
Corporation Law, vol. 30, no. 4, 2005, p. 717.
In publicly owned firms, the shareholders’ interests are represented by the board of directors. The board of directors
is responsible for setting and approving the compensation for the CEO and the top executives of publicly owned
firms. In a majority of publicly traded U.S. firms, the CEO also serves as chairman of the board.
Since shareholders’ interests are represented by the board, shareholders have little direct say in setting executive
compensation. Generally, the broad details of equity-based compensation plans must be approved by shareholders
before they are implemented. Otherwise, the shareholders must approve the directors when they are nominated and
re-nominated, and the board’s decisions on compensation are one of a number of factors that shareholders take into
account when determining whether or not to approve a nomination.
Boards typically cede the setting of executive pay to the compensation committee, a board committee generally
composed of outside directors who are not employed by the firm. The compensation committee recommends
executive pay packages to the board for approval. To help guide the committee in the setting of executive pay,
corporations often hire outside pay experts known as compensation consultants. Central to the advice that they
provide to the firms is a survey of compensation levels and practices at so-called peer firms.
In a “neo-classical” model of perfect competition, the compensation of an executive (or any 19
worker) is determined by the executive’s marginal product. In other words, the executive will be
paid just as much as the revenue he contributes to the firm. If the firm tries to pay him less than
his marginal product, he will seek employment elsewhere. If the firm tries to pay him more, it
will be undercut by competitors (who are paying their executives at a competitive rate) and
become unprofitable. A number of economists and business experts believe this world view is a
reasonable facsimile of reality. For example, Roy Smith, a professor of finance at New York
University, has written that
The best chief executives have proven track records demonstrating management ability in
large, complex corporate situations. Such people are always in demand. But most get only
one shot at being a CEO, and they want to make the most of it. They also know that the rate
of CEO turnover at large corporations has increased significantly in recent years, and that 50
percent of departures are the result of mergers or performance issues. If things go wrong,
their contracts may be all they have to hang on to, so they negotiate the best ones they can
going in.... Today’s CEOs are paid well when they deserve to be (and sometimes when they
don’t), but most corporate directors will tell you that over the past 20 years, public
companies have become better managed, with increased profit margins, productivity and 20
returns on investment...
In this model, high pay is not a sign of an executive being overpaid, it is a sign that the executive
is highly productive. It therefore follows that the burden of proof should be on critics to show that
the high level of executive pay is not simply a reflection of the executive’s strong skills, hard
work, and successful business strategies. Nor, from this perspective, is the relative increase in
executive pay compared to worker pay necessarily a sign of excessive pay. The superior
performance of U.S. firms on average over the past decade, relative to foreign firms and U.S.
firms in the past, could be taken as evidence in favor of the neo-classical model. It could also be
that the increase in pay is driven by a relative increase in the demand for executives. If running a
19 In economic theory, perfect competition exists in a market with many buyers and sellers, no barriers to entry, and
perfect information, and when the costs and benefits of the transaction are completely borne by the buyer and seller.
20 Roy C. Smith, “Worth Every Last Million,” Washington Post, January 21, 2007, p. B-1.
company requires more skill now than in the past—because, for example, markets are now more
competitive or firms are more complex—it would be expected in a neo-classical world that firms
would be willing to pay more now in order to attract these skills. For example, economists have
described a “superstar effect,” where the pay of entertainment celebrities rose once their market 21
draw increased through the development of mass media. Because so much of the market is
captured by a few individuals, small differences in talent or public preferences lead to large
differences in pay. Some economists have suggested a similar effect may be at work in the market
for executives: as firms have gotten larger and begun to operate in a global rather than domestic
market, the value that executives can add to a firm has increased, and their compensation has
followed suit. Using this logic, one study concluded that “the six-fold increase of CEO pay 22
between 1980 and 2003 can be fully attributed to the six-fold increase in market capitalization.”
In the neo-classical model, executive pay is held in check not just by competition in the labor
market for executives, but also by competitive forces in capital markets and the firm’s product
markets. The key point made by this model is that for an executive to be overpaid, the firm must
be generating excess profits that the executive is able to skim off, and in a world of perfect
competition, no excess profits (profits beyond a normal, risk-adjusted rate of return) exist for the 23
executive to capture. (It could be argued that the amount the executives are skimming off is too
negligible to affect the firm’s profitability; but, if so, excessive compensation is presumably of
rather minor importance.) There is ample evidence that capital markets and most product markets
are efficient, so for executives to be overpaid without placing their firms at a competitive
disadvantage, there must be something in the compensation arrangement that is placing most
firms at the same disadvantage. Otherwise, the company would be unable to raise sufficient 24
capital or become vulnerable to a hostile takeover. In a world of perfect competition, there is no
need to try to measure whether executives are overpaid, because excessive pay is an
unsustainable outcome that would be driven out by market forces. For executives to be overpaid,
something must be impeding the efficient market allocation of resources. Economists who have
argued that executives are overpaid have pointed to the principal-agent problem as the market
failure that could lead to executives being widely overpaid.
The principal-agent problem is a well-known market failure that exists when a principal’s
(owner’s) interests are represented by an agent (in this case, the CEO) whose goals or incentives
diverge from the principal, and the principal is unable to closely enough monitor the agent to
keep their interests aligned. The agent could be pursuing any number of private interests,
including minimizing his own effort, maximizing his own reputation (by unprofitably expanding
21 Sherwin Rosen, “The Economics of Superstars,” American Economic Review, vol. 71, no. 5, December 1981, p. 845.
22 In the study, Gabaix and Landier present a theoretical model showing that with scarce CEO talent, CEOs’ pay would
increase when firms get larger. They believe this explains why pay has risen over time and why it is higher on average
in America than abroad. Xavier Gabaix and Augustin Landier, “Why Has CEO Pay Increased So Much?,” National
Bureau of Economic Research working paper 12365, July 2006.
23 Most economists believe that markets are typically better described by a model of imperfect competition, where
firms have some market power, than perfect competition, but the results for both are the same. In a model of imperfect
competition, there are no barriers to entry, so firms are still not generating excess profits that executives could
potentially skim off. Some economists argue that even in markets where one firm wields significant market power, the
threat of potential entry is enough to keep the firm from collecting significant excess profits.
24 Admittedly, there are barriers to corporate takeovers, such as “poison pills,” that make the takeover market less than
perfectly efficient as well.
the firm, for example), covering up his own mistakes, minimizing risk, and so on. The key point
is that the agent is not always pursuing the principal’s goal: profit maximization. Profit
maximization is in the agent’s self-interest only insofar as it maximizes his own compensation or
well-being. Yet there is a tradeoff, since an increase in the agent’s pay decreases the company’s
profits, all else equal.
For a publicly listed corporation, the principal-agent problem is particularly hard to avoid because
of another well-known market failure: the free-rider problem. For a publicly listed corporation,
the (collective) principal is the firm’s thousands of shareholders. Assuming for a moment that
barriers to shareholder action (which will be discussed below) did not exist, it is still unlikely that
shareholders could avoid the principal-agent problem because it would be highly costly for any
individual shareholder or group of shareholders to monitor the agent. Any benefits that resulted
from monitoring the executives would not be captured solely by the shareholders undertaking the
monitoring; they would flow to all shareholders. Hence, it is in any individual shareholder’s self-
interest to leave monitoring the executives to other shareholders since that will lead to the same
benefit with none of the cost of personal monitoring. As a result, every shareholder chooses not to
monitor, and this is the essence of the free-rider problem.
In theory, the firm’s board of directors is meant to safeguard the shareholders’ interests and
monitor the firm’s executives: the free-rider problem has been overcome by the shareholders
banding together and appointing a board to represent them. But does the board really represent
the shareholders’ interests? For the board faces a principal-agent problem of its own: since each
board member’s financial stake in the firm typically makes up a negligible share of the firm’s
total value, the board members may also have an incentive to pursue private interests other than
profit maximization. This incentive opens the door to a number of ways in which the executives
could “capture” the board so that the board members were not truly “independent,” as will be
In this context, it is easy to see how an endemic problem of overpaying executives could 25
potentially arise. Since neither the executives nor the board are acting solely in the shareholders’
interests, they may agree to pay the executive more than his marginal product. Since executives at
competing firms are also overpaid, the firm’s performance would not fall behind its competitors.
If the burden of proof is on critics to prove that the real world outcome does not match the neo-
classical ideal, then the “managerial power” critique, discussed in the following section, attempts
to make a comprehensive case that the principal-agent problem has led to excessive executive
Does the neo-classical or the principal-agent model more accurately describe executive pay
setting in the United States today? In support of the principal-agent model, law professors
Bebchuk and Fried (hereafter, BF) have set out a “managerial power” critique, which identifies
the numerous ways in which the structure of executive pay and the relationship between the 26
executive and the board differs from the neo-classical ideal. Although the managerial power
25 For an in-depth exposition of this case, see Lucian Bebchuk and Jesse Fried, “Executive Compensation as an Agency
Problem,” Journal of Economic Perspectives, vol. 17, no. 3, Summer 2003, p. 71.
26 Lucian Bebchuk and Jesse Fried, Pay Without Performance, Harvard University Press, (Cambridge, MA: 2004). For
summaries of their views, see Lucian Bebchuk and Jesse Fried, “Executive Compensation as an Agency Problem,”
critique cannot numerically estimate how much executives are overpaid, BF argue it is prima
facie evidence that executives are paid excessively and pursue the goal of maximizing pay rather
than shareholder value. The primary elements of the managerial power critique are laid out below.
The neo-classical model assumes that executive pay is determined through “arm’s length
contracting”: in negotiating the executive’s pay, the board of directors (typically through its
compensation committee) is charged with driving the best bargain it can obtain on behalf of
shareholders. BF argue that the typical board structure has little in common with this ideal. In
many corporations, the CEO is also the chairman of the board and other company “insiders,” who
may be loyal to the CEO, serve on the board. BF argue these arrangements are susceptible to the
boards being “captured” by the executives.
Some fears of board capture were allayed in 2003 when the NYSE required that boards of its
listed companies have a majority of outside directors, and the boards’ nominating, compensation,
and audit committees consist solely of outside directors. (The NASDAQ has similar rules, except
for those pertaining to the compensation and nominating committees.) But even “outsiders” who
serve on the board may owe their positions—and hence, BF argue, their allegiance—to the CEO.
For example, the only candidates who appear in the official proxy given to shareholders for
election are individuals selected by the board’s nominating committee. It is commonly perceived
that the CEO strongly influences the nominating committee’s selections. Shareholders can
withhold support from a director but cannot vote against a director, which means that a director
technically needs only one favorable vote to be elected. There are few ways that the shareholders
can directly select members of the board. Alternative directors proposed by shareholders are
costly, difficult, and rare.
Board members may feel that they must keep on the CEO’s “good side” in order to be re-
nominated when their terms expire. Furthermore, board members are frequently executives at
other corporations, which may make them predisposed to view executive compensation requests
favorably and may make them hesitant to be critical board members since their criticisms could
be used against them at their own companies. Boards are also dependent on executives to furnish
them with information and advice in order to fulfill their duties, and must maintain the
executives’ good will in order to receive this information. In fact, the CEO compensation
packages that boards approve typically originate from the company’s human resources
department, which, of course, is subordinate to the CEO.
If boards are captured by the executives, one would expect that arm’s length bargaining would be
compromised. There is evidence that less independent boards are correlated with higher executive 27
pay, all else equal. Some studies have shown that CEO pay is 20%-40% higher when the CEO is
Journal of Economic Perspectives, vol. 17, no. 3, Summer 2003, p. 71; Lucian Bebchuk and Jesse Fried, “Pay Without
Performance,” Journal of Corporation Law, Summer 2005, vol. 30, no. 4, p. 647. Unless otherwise noted, references to
BF throughout the report refer to these publications.
27 The literature is surveyed in Kevin Murphy, “Executive Compensation,” in Orley Ashenfelter and David Card, eds.,
Handbook of Labor Economics, vol. 3B, Elsevier Publishers (Amsterdam: 1999), p. 2485. Although most studies found
that greater board independence leads to lower executive pay, one study found that a higher proportion of independent
directors (who are less likely to be captured by the executives) led to higher pay for the CEO and top five executives.
Wan also found that firms with more independent directors perform no better than firms without. Kam-Ming Wan,
also the chairman of the board.28 One study showed that executive compensation is higher when 29
the CEO picked the board’s outside directors. There are studies that found that executive pay is
lower in situations where shareholders have more influence over the board—for example, when
there is a single large shareholder or institutional shareholders with a significant ownership 30
Companies often hire outside compensation consultants to help them determine the level and
characteristics of the executive compensation package. But BF argue that these consultants can
also be “captured” by management because the consultants are eager for more business from the
firm (including business unrelated to executive compensation); as a result, they tell the executives
what they want to hear.
If shareholders thought that the board was allowing executives to be overpaid, institutional
barriers would limit their influence over the board. The general terms of the company’s overall
performance-based compensation packages, such as stock options plans, may be subject to an up
or down vote by the shareholders, but shareholders cannot modify the details of the package or
offer alternatives. Only fifteen of 2000 options plans were voted down by shareholders from July 31
attract a majority of voters, the resolutions may not be binding.
BF argue that one of the main constraints on executive compensation levels are “outrage costs”—
a fear on the part of executives that if their compensation is too excessive they will face hostility 33
from shareholders and negative publicity from the media. BF argue that outrage costs have two
effects. First, they reduce executive compensation levels below what they would otherwise be.
(This factor works to the extent that outrage exists—BF argue one reason why executive pay rose
in the late 1990s is because CEOs were exalted during the stock market boom, causing outrage
costs to diminish.)
“Independent Directors, Executive Pay, and Firm Performance,” EFMA 2003 Helsinki Meetings, working paper, 2003.
28 See Richard Cyert, et al., “Corporate Governance, Takeovers, and Top-Management Compensation,” Management
Science, vol. 48, no. 4, 2002, p. 453; John Core, et al., “Corporate Governance, Chief Executive Compensation, and
Firm Performance,” Journal of Financial Economics, vol. 51, no. 3, March 1999, p. 371.
29 John Core, et al., “Corporate Governance, Chief Executive Compensation, and Firm Performance,” Journal of
Financial Economics, vol. 51, no. 3, p. 371. See also David Larcker, et al., “Director Networks, Executive
Compensation, and Firm Performance,” working paper, 2006.
30 Jay Hartzell and Laura Starks, “Institutional Investors and Executive Compensation,” Journal of Finance, vol. 2351,
2003; Richard Cyert, et al., “Corporate Governance, Takeovers, and Top-Management Compensation,” Management
Science, vol. 48, no. 4, p. 453.
31 Matthew Malone, “Are Compensatory Stock Options Worth Reforming?,” Gonzaga Law Review, vol. 38, no. 535,
32 Lucian Bebchuk and Jesse Fried, Pay Without Performance, Harvard University Press, (Cambridge, MA: 2004), p.
33 For a study on the media’s influence on executive compensation, see John Core, et al., “The Power of the Pen and
Executive Compensation,” Social Science Research Network, working paper, October 2005, available at
http://ssrn.com/abstract=838347. The authors find that the media focus more on the absolute level of executive pay than
the relationship between pay and performance.
Second, they encourage executives to seek “stealth compensation”—compensation delivered in a
form that is difficult for shareholders and the public to understand—in order to avoid outrage.
“Golden parachutes,” generous severance or retirement packages (details of which do not have to 34
be disclosed beforehand), “golden hellos” (additional incentives to join a company), life
insurance, deferred compensation, personal loans (before the Sarbanes-Oxley Act), bonuses that
are not linked to company performance, and company-provided perks in the form of vehicles,
aircraft, and club memberships are all examples of stealth compensation.
BF believe that the rapid growth in executive stock options is explained by the desire for stealth
compensation. From a neo-classical perspective, it is difficult to understand why stock options are
awarded rather than stock—when the options are too far “underwater” (the stock price is below 35
the option’s strike price), there is no incentive effect, and when it is close to the strike price,
executives may be extremely risk averse for fear that the option will fall below the strike price.
After they are exercised, options cease to have any influence on executives’ behavior. And
options, unlike stock, are not necessarily aligned with the shareholder’s interest since they
discourage the payment of dividends (i.e., the money can be used to buy back the stock to boost
its price instead). But BF argue that from the managerial power perspective, the prevalence of
stock options is logical because accurately pricing options at the time they are granted is complex
and controversial. Until recently, options were not expensed on a firm’s balance sheet, so some
argued this treatment made their effect on shareholder value less transparent. Stock options may
also be a way to increase compensation without generating as much shareholder outrage since
options are worth more when the stock price is high and shareholders are doing well, without
exposing the executives to any risk when the share price falls below the strike price.
BF also attribute the “ratcheting up” of executive pay to outrage costs. Executive pay is often set
relative to some (self-selected) peer group, typically identified by the compensation consultant.
To justify compensation levels, the board sets CEO pay somewhat higher than the industry
average, reasoning that its executive is an above average performer (which critics have compared 36
to the mythical “Lake Wobegon,” where all of the children were above average). Of course, it is
mathematically impossible for everyone’s performance to be above average, but as long as
compensation is close to the industry norm, shareholders are unlikely to complain and the
company is unlikely to attract unwanted publicity. But if each executive is having their pay set
above the average, then the average will rise over time—inadvertently, the practice will cause pay
to be ratcheted up to levels that may have little to do with marginal product. If pay levels are set
based on peer groups, they will also be insensitive to performance, the subject of the next section.
In the 1980s and early 1990s, many academics complained that executives were well
compensated regardless of whether their firm performed well or poorly. In a much cited article,
Jensen and Murphy argued
34 Julie Creswell, “Pay Packages Allow Executives to Jump Ship With Less Risk,” New York Times, December 29,
35 The strike price is the predetermined stock price at which the option allows the stock to later be bought. For a basic
primer on stock options, see the Appendix.
36 John Bizjak, et al., Has the Use of Peer Groups Contributed to Higher Levels of Executive Compensation?, Portland
State University working paper, June 2004.
The relentless focus on how much CEOs are paid diverts public attention from the real
problem—how CEOs are paid. In most publicly held companies, the compensation of top
executives is virtually independent of performance.... Is it any wonder then that so many
CEOs act like bureaucrats rather than the value-maximizing entrepreneurs that companies 37
According to their estimates, a $100 increase in firm value led to a 26 cents median increase in 38
CEO wealth, of which cash compensation rose only 4 cents. They argued that the way to solve
the principal-agent problem was to align executive incentives with profit maximization by
making compensation more sensitive to the firm’s performance. (They estimated executive
compensation was no more sensitive to firm performance than workers’ wages.) One way to
accomplish this goal was through performance-based bonuses. Another way was by compensating
executives with firm stock or stock options, so that when executives made decisions that made the
firm more profitable, the firm’s stock price would rise, causing the value of the executive’s
compensation to rise. In theory, executives would no longer wish to pursue private goals that
undermine firm profitability because it would reduce their compensation.
The authors argued that if pay were more closely linked to performance, average pay would
probably rise because bad executives would be forced out, more talented individuals would be
drawn to managing, executives would be motivated to work harder, and firm performance would
improve. In addition, pay would need to be higher if it were riskier (since incentive-based pay
would fall if the executive missed the performance benchmarks) because individuals must be
compensated to take on risk. They argued that performance-based pay could exacerbate inequality
but society as a whole would be better off overall.
While most studies confirmed that cash pay and bonuses were fairly unresponsive to firm
performance, the literature is divided about the relation of total pay to performance, once
executive stock holdings and stock options are included. For example, Hall and Liebman argue
that pay for performance rose throughout the 1980s and early 1990s because of the rising share of 39
compensation that consisted of stock and stock options. They estimate that CEO wealth rose
about sixty cents for every $100 increase in firm value in 1994, significantly higher than Jensen
and Murphy’s earlier estimate. Furthermore, they argue that executive pay is much more sensitive
to performance than Jensen and Murphy’s measure implies because the firms involved are so
large that even if the ratio of the change in wealth and the change in firm value is small, the dollar
change in wealth is very large. For example, they estimated that a CEO who increased the
performance of his company from below average to above average would increase the median
value of his stock holdings by $4 million, all else equal. Similarly, Core et al. estimated that the
median portfolio value of an S&P 500 CEO’s own firm holdings was $30.1 million in 2003, and a
37 See, for example, Michael Jensen and Kevin Murphy, “It’s Not How Much You Pay, But How,” Harvard Business
Review, no. 3, May 1990, p. 138.
38 Michael Jensen and Kevin Murphy, “Performance Pay and Top-Management Incentives,” Journal of Political
Economy, vol. 98, no. 2, p. 225. The estimate covers 1975 to 1988. The remaining $0.22 was due to changes in the
value of stock holdings and options, and the likelihood of dismissal. Updated results suggest that the sensitivity of cash
compensation had about doubled for 1990-1996 (see Kevin Murphy, “Executive Compensation,” in Orley Ashenfelter
and David Card, eds., Handbook of Labor Economics, vol. 3B, Elsevier Publishers (Amsterdam: 1999), p. 2485).
39 Brian Hall and Jeffrey Liebman, “Are CEOs Really Paid Like Bureaucrats?,” Quarterly Journal of Economics, vol.
CXIII, Issue 3, August 1998, p. 653.
In the context of this report, this shift to so-called “performance-based” pay turns out to be
important because it was the source of most of the rise in executive pay in the 1990s and 2000s.
The estimated value of options granted to CEOs (at the time of the grant) increased nine-fold
from 1992 to 2000, while other types of compensation rose three-fold over that period.
BF argue that the growing share of non-cash executive compensation should not be taken as a
sign that pay is now tied more closely to performance. The managerial power critique identifies
several standard components of executive pay that may appear to be performance-based pay, but
arguably allow executives to circumvent the link between pay and performance in practice.
For example, severance packages often handsomely reward executives for leaving their jobs, on
good or bad terms. These packages have only negative effects on incentives since they send the
message that regardless of whether an executive succeeds or fails, he will be rewarded after the
fact. Bonuses should be an important part of performance-based pay, but studies have found that 41
instead they are often paid for reasons unrelated to profit maximization.
BF identify several common characteristics of stock options that seem inconsistent with the pay
for performance mantra. Stock options are almost always designed to reward absolute 42
performance rather than relative performance. For example, if a company’s stock price rises by
just as much as the overall market, then the executive’s compensation will have risen because of
trends largely beyond his control. Even if a company is outperformed by its competitors, the
options could still have value in a bull market since “a rising tide lifts all ships.” Furthermore,
stock options are usually granted “at the money” (they have a strike price equal to the market
price at the time the option is granted). Since stocks usually rise over time, this will generally
reward executives for doing nothing, a phenomenon Warren Buffett characterized as “a royalty 43
for the passage of time.” Once exercised (usually after a vesting period), the options no longer
provide executives with any incentives.
There is also a more basic question to ask: in instances where firms offered incentive-based pay,
why did it typically supplement rather than substitute for cash salary? The thrust of the pay for
performance argument was that executives should be paid differently, but in practice they were
40 John Core, et al., “Is U.S. CEO Compensation Inefficient Pay Without Performance?,” Vanderbilt Law and
Economics Research Paper 5-05, January 2004.
41 Lucian Bebchuk and Jesse Fried, “Pay Without Performance,” Journal of Corporation Law, Summer 2005, vol. 30,
no. 4, p. 647; Kevin Murphy, “Executive Compensation,” in Orley Ashenfelter and David Card, eds., Handbook of
Labor Economics, vol. 3B, Elsevier Publishers (Amsterdam: 1999), p. 2485; Yaniv Grinstein and Paul Hribar, “CEO
Compensation and Incentives,” Journal of Financial Economics, vol. 73, 2004, p. 119.
42 Linking options to relative performance would require a more complex derivative than the standard options offered
now. For example, instead of giving the option a fixed strike price, the strike price could be set relative to the prevailing
value of a stock index. A shift to options based on relative performance could cause executive pay to rise because the
holder would demand more options to offset the greater risk, making them more expensive for the company to offer.
43 Quoted in Patrick Bolton, et al., “Pay for Short-Term Performance,” National Bureau of Economic Research,
working paper 12107, March 2006.
also paid more. Instead of rewarding good executives and punishing incompetent executives as
many academic obeservers prescribed, everyone was made better off.
Besides pay, some argue that executives have a degree of job security that is at odds with the pay
for performance mantra. For example, Murphy found that the correlation between firm
performance and CEO turnover was low and fell in the 1990s to a statistically insignificant 44
relationship. CEOs also have more contractual job protection than employees: Schwab and
Thomas found that only 25 of 375 of the CEOs they examined served “at will,” like regular 45
employees, while the remainder could only be dismissed without penalty for “just cause.”
The corporate scandals of recent years have brought greater scrutiny to executive pay practices,
particularly stock-based and option-based compensation. Linking compensation to the stock price
can only mitigate the principal-agent problem if the stock price is an accurate proxy for firm
value. The assumption that it is a good proxy is at the heart of the efficient market hypothesis, and
most economists would consider it true over long periods of time. But it is difficult to argue that
the stock price is a good proxy for firm value over short time periods when stock prices tend to
fluctuate widely and unpredictably, implying that the value of the firm can fluctuate by millions
or even billions of dollars in a matter of days. If it is not a good proxy, a potential problem arises
because the agent’s incentive is linked to a goal that is not perfectly aligned with the principal’s
interests. In other words, the agent can increase his compensation by taking advantage of—or
even causing—short-term changes in the firm’s stock price, which could be detrimental to the
long-term performance of the firm. Unsurprisingly, the agent may decide to take actions that
boost the short-term stock price when the firm’s long-term prospects are bleak. In this situation,
the large stock and stock option holdings of the typical executive arguably offer a powerful
incentive to take actions to boost the stock price long enough to divest one’s stock holdings. In
this context, performance-based pay exacerbates the principal-agent problem.
This problem was the essence of many of the corporate scandals that followed the 2001 recession.
Some of the scandals involved executives pumping up the firm’s stock price by manipulating 46
reported earnings in order to increase their performance-based pay (or prevent it from falling).
Other scandals revolved around executives attempting to hide financial problems from the public
to prevent their stock and option holdings from losing value. In other cases, executives used
inside information to exercise their options before the firm’s share price fell. Likewise, the
ongoing scandal involving the backdating of stock options revolves around executives
retroactively manipulating the timing of the option grant to a day when the firm’s stock price was 47
unusually low in order to increase the option’s value when it is exercised. These scandals
suggest that, on balance, the principal-agent problem seems to have survived the shift to
performance-based pay intact. In fact, the growth of performance based pay and accounting
44 Kevin Murphy, “Executive Compensation,” in Orley Ashenfelter and David Card, eds., Handbook of Labor
Economics, vol. 3B, Elsevier Publishers (Amsterdam: 1999), p. 2485.
45 Stewart Schwab and Randall Thomas, “An Empirical Analysis of CEO Employment Contracts,” Washington and
Lee Law Review, vol. 63, no. 1, winter 2006, p. 231.
46 See CRS Report RS21392, Stock Options: The Accounting Issue and Its Consequences, by Bob Lyke and Gary
47 For more information, see CRS Report RL33926, Stock Options: The Backdating Issue, by James M. Bickley and
problems seems to have gone hand-in-hand—700 firms issued financial restatements from 1997 48
to 2000, compared with 11 from 1992 to 1993, when performance-based pay was less prevalent.
Burns and Kedia find that firms whose executives have a higher share of option-based 49
compensation are more likely to restate their earnings.
While the incentives for executives to manipulate earnings seems straightforward, the rationale
for the board’s failure to prevent it seems less clear: the board would not profit substantially from 50
manipulation and could face high costs if the manipulation is detected, unless “board capture” 51
has occurred or the board is operating under the same short-term time horizon as the executives.
Of course, the board may have been unaware of the executives’ actions, but that may raise the
question of why it did not monitor the executives more closely. There may also be a human
tendency to use less scrutiny when results appear positive.
Although the corporate scandals that have unfolded over the past few years may not be proof of
the managerial power theory, they can be shown to be consistent with it. Boosting pay through
accounting fraud and earnings smoothing is consistent with executives’ desire for stealth pay.
Bergstresser and Philippon show that at firms that appear to have managed earnings, executives
exercise options and sell stock at a higher rate than average and have a greater share of 52
performance-based pay. The failure of the board to prevent such activities is consistent with
board capture. Allowing executives to exercise their options at their discretion (after a vesting
period) is primarily what made accounting fraud and manipulation profitable, yet there is little
purpose in allowing options to be exercised at will if the options are meant to keep the executive’s
interests aligned with the shareholders’ on an on-going basis. It is also difficult to see why firms
would allow backdated options except to mask the true size of executive compensation (relative
to the “performance” required to earn the compensation). Firms could have legally paid
executives just as much by either issuing them more options or options with a lower strike price,
but doing so would have drawn more attention from shareholders and the media. One study found
that favorably dated option grants, potentially caused by backdating, was one-third more likely at 53
firms where independent directors did not make up a majority of the board.
48 Patrick Bolton, et al., “Pay for Short-Term Performance,” National Bureau of Economic Research, working paper
12107, March 2006.
49 Natasha Burns and Simi Kedia, “The Impact of CEO Incentives on Misreporting,” Journal of Financial Economics,
50 The notion that board members would not benefit from manipulation by executives was questioned in a recent study
that found that outside directors also received stock options whose timing was characteristic of back dating in 9% of all
cases. Lucian Bebchuk et al., “Lucky Directors,” Harvard Law School working paper, December 2006.
51 Bolton et al. derive a model where earnings manipulation may be in the shareholders’, and by extension the board’s,
self interest. In their model, it is possible for the CEO to artificially increase the stock price temporarily above its
fundamental value. The CEO can then exercise his options and profit from the artificially high price before the price
falls back to its fundamental value. This arrangement could suit current shareholders since they could sell while the
stock price is inflated and profit at the expense of future shareholders (who buy the stock at the inflated price). This
reasoning could also be applied to the board, since it represents the interests of current shareholders but not future
shareholders. Patrick Bolton, et al, “Pay for Short-Term Performance,” National Bureau of Economic Research,
working paper 12107, March 2006. Bolton et al. argue that, contrary to the “managerial power” thesis, reloading
options and revaluing options that are underwater may help keep managers’ incentives aligned with the interests of
52 Daniel Bergstresser and Thomas Philippon, “CEO Incentives and Earnings Management,” Journal of Financial
Economics, forthcoming 2006. See also Messod Beneish and Mark Vargus, “Insider Trading, Earnings Quality, and
Accrual Mispricing,” Accounting Journal, vol. 77, no. 4, October 2002.
53 Lucian Bebchuk et al., “Lucky CEOs,” Social Science Research Network, working paper, November 2006.
Criticisms of the managerial power theory can be split into two broad categories, disagreements
with BF’s findings and alternative explanations for the phenomena that BF identify. On their
findings, critics have disagreed with the following:
• If pay were determined by managerial power, then executives hired from the
outside should not receive the same generous pay packages as insiders, but they 54
do. BF reply that the board and compensation consultants have the same
incentive to reward someone who will have influence over their welfare in the
future as they do over someone who can influence it now.
• BF’s claim that executive pay is not well linked to performance. As noted above,
some studies have argued that executive pay is highly sensitive to performance
because of their large holdings of their firm’s stocks and options, which gives
them powerful incentives to maximize the firm’s profitability.
• BF’s ability to explain the change in executive pay over time. First, BF’s theory
depends on the CEO’s ability to “capture” the board so that it does not act in the
shareholders’ interests. Yet boards have become more independent and active in 55
recent years. For example, in 2003 the NYSE and Nasdaq required that a
majority of a company’s directors must be independent in order to be listed on
their exchanges. Second, changes to SEC rules required more disclosure of
options granted to executives in 1992—at the very time that options began to
increase in value. This pattern seems at odds with BF’s argument that executives
prefer options because they are a form of stealth compensation. Third, Murphy
argues that outrage costs were much higher in the early 1990s, with the
introduction of legislation and new regulations to curb pay, than the late 1990s, 56
when compensation escalated.
• Although executive pay is higher in the United States than abroad, that could be
evidence in favor of efficient markets since the United States stock market has
consistently outperformed foreign stock markets. This finding casts doubt on 57
BF’s claim that excessive pay is undermining firm profitability.
Alternative explanations have been offered for several of the phenomena that the managerial
power theory describes. Some of them are related to the incentives offered by accounting rules.
For example, the shrinking share of executive compensation being paid in cash wages has been
attributed to the million dollar cap on pay that corporations are able to deduct from taxes, which
is described below. Similarly, certain standard characteristics of options (such as setting the strike
price equal to the actual price at time of issue and using options that reward absolute instead of
54 Murphy claims that total compensation is higher for “outsiders” than “insiders.” Kevin Murphy, “Explaining
Executive Compensation: Managerial Power versus the Perceived Cost of Stock Options,” University of Chicago Law
Review, vol. 69, no. 847, summer 2002.
55 Bengt Holmstrom and Steven Kaplan, “Corporate Governance and Takeovers in the U.S.,” Journal of Economic
Perspectives, vol. 15, no. 2, Spring 2001, p. 121; Brian Hall and Kevin Murphy, “The Trouble with Stock Options,”
Journal of Economic Perspectives, vol. 17, no. 3, summer 2003, p. 49.
56 Kevin Murphy, “Explaining Executive Compensation: Managerial Power versus the Perceived Cost of Stock
Options,” University of Chicago Law Review, vol. 69, no. 847, summer 2002.
57 John Core, et al., “Is U.S. CEO Compensation Inefficient Pay Without Performance?,” Vanderbilt Law and
Economics Research Paper 5-05, January 2004.
relative changes in the stock price) were encouraged by accounting rules that, until recently (see
below), did not require that options with those characteristics be expensed.
Hall and Murphy have argued that the managerial power theory cannot adequately explain the
recent burgeoning of stock option compensation because most stock options (90% in 2002) are
paid to non-executives. They offer a “perceived cost theory” of why stock options have become
so popular: because stock options do not require the company to outlay any cash when granted
and were not expensed in the company’s accounting statements before the rule change in 2005, 58
they argue that options became popular because they appeared costless to issue. On the contrary,
options do have a cost—they dilute the existing shareholders’ ownership position when they are
realized. The shortcoming of this explanation is it begs the question of whom Hall and Murphy
believed were being duped into thinking the options were costless—the board or the
shareholders? In the former case, BF argue that “...if directors had so little financial
sophistication, then the board-monitoring model of corporate governance is in even worse shape 59
than our analysis suggests.” In the latter case, the perceived cost argument is consistent with,
but not proof of, the managerial power view that options are a form of stealth compensation.
Alternatively, the perceived cost view could be rejected on the grounds of market efficiency. For
example, although options were not expensed, information about the options were included in the
footnotes to corporations’ financial statements. If market actors use all of the information
available to them, as market efficiency requires, then the stock price would already reflect the
information presented in the footnotes, and the stock price would decline, all else equal, when
options were granted. Otherwise, market participants who were aware of the real costs associated
with options could systematically profit by short-selling firms that offered employees overly
generous stock options.
It has also been suggested that there are alternative explanations to the managerial power theory
for why executive compensation could exceed marginal product. In a well-known article, Lazear
and Rosen argued that executive pay could be much higher than the executive’s marginal product 60
and still be economically efficient from the perspective of the overall firm. They argued that this
could be true if the firm’s pay structure is likened to a tournament. If the firm has no easy way to
accurately identify each of its workers’ marginal product, then the most efficient way to induce
maximum effort from its workers could be to attach a large monetary incentive to the “winner” of
each promotion. At the top of this ladder is the CEO, with a correspondingly larger prize each
step of the way to motivate workers to strive for further promotions. As long as the CEO’s high
pay motivates his subordinates to work harder to strive to some day replace him, it will be
profitable for the firm to compensate him at that rate regardless of his marginal product. Anabtawi
argues that the tournament model explains why executive pay is not more closely linked to
performance (because reductions would weaken incentives of subordinates to try to win the 61
58 Brian Hall and Kevin Murphy, “The Trouble with Stock Options,” Journal of Economic Perspectives, vol. 17, no. 3,
summer 2003, p. 49. The company also has tax incentives to favor compensation in the form of options.
59 Lucian Bebchuk and Jesse Fried, “Executive Compensation as an Agency Problem,” Journal of Economic
Perspectives, vol. 17, no. 3, summer 2003, p. 87.
60 Edward Lazear and Sherwin Rosen, “Rank-Order Tournaments as Optimum Labor Contracts,” Journal of Political
Economy, vol. 89, no. 841, 1981.
61 Iman Anabtawi, “Explaining Pay Without Performance: The Tournament Alternative,” Emory Law Journal, vol. 54,
iss. 4, fall 2005, p. 1557.
The decision to base executive stock options in the 1990s on absolute performance rather than
relative performance had important implications during the stock market boom of the 1990s.
When the board chooses how much stock option compensation to award to an executive, it is
basing its decision on the ex ante expected value of the options. But actual executive pay depends
on what happens to the stock price ex post. For example, if an executive were granted one option
with a strike price of $40, and the board expected the stock price to rise to $50 when the
executive exercised it, the executive’s expected pay gain ex ante would be $10. But if the stock
price actually rose to $60, his actual pay increase would be $20—double what the board expected
him to earn.
A strong case can be made that since the sharp appreciation in stock prices in the 1990s was
largely unexpected, much of the pay increase tied to the stock price or other metrics of firm
performance was also unexpected windfalls. For example, in the 50 years preceding December
averaged 2.3%. From December 1995 to December 1999, the average real rate of appreciation
was 22.8%. Figure 2 shows the expected value (based on the historical average) and actual value
for a stock option issued in December 1994 and exercised at any point from December 1995 to
1999. If the company set the strike price at the firm’s current stock price and expected the stock
price to follow the historical average, then the company would expect the nominal stock price to
rise about 26% by 1999. Instead, the average stock over that period rose by 214% in nominal
value. Assuming a strike price at the firm’s current stock price, the actual payment from options 63
over that period would be over eight times higher than expected for the average firm.
62 The nominal 50-year average appreciation rate was 6.8%. This figure does not represent the average rate of return on
stocks, which includes dividend payments. Dividend payments are not included since options are based solely on the
stock price. One drawback to a historical comparison is that dividend payments have become a smaller portion of the
rate of return over time.
63 Murphy explains how a rising stock market led to rising option pay in Kevin Murphy, “Executive Compensation,” in
Orley Ashenfelter and David Card, Handbook of Labor Economics, vol. 3B (Amsterdam: Elsevier Publishers, 1999), p.
Figure 2. S&P 500: Actual and Projected Based on Historical Average, 1995-1999
Source: CRS calculations based on data constructed by Robert Shiller (Yale University).
Figure 2 suggests that much of the increase in pay in the 1990s may have been unrelated to
competitive forces or managerial power, but instead was unanticipated on the part of executives
and board directors. In a rapidly rising market, almost all executives, regardless of talent or effort, 64
arguably profited from being in the right place at the right time. But even if the windfall was
unintentional, that begs the question of why executive pay did not fall further in the 2000s when
the stock market declined by nearly one half. It would be expected that boards would react to the
stock option windfalls received by executives in the 1990s by offering less generous stock options
from that point on or tying future options to relative rather than absolute performance. Indeed,
even if directors had made no changes to ex ante option packages, one might expect sharp
declines in option-based pay from 2000 to 2003 since stock prices were falling. Instead, overall
pay fell by only 11% between 2001 and 2003 (which was still higher than it was in 2000) and
returned to its previous peak in 2004, which seems difficult to reconcile with the neo-classical
model. While windfalls are consistent with arm’s length contracting in a world of uncertainty,
executives would not be expected in a neo-classical world to permanently lock in windfall
payments after the source of the windfall had disappeared. This experience does seem consistent
with the managerial power theory, however. Shareholder satisfaction with the large stock market
returns of the 1990s may have greatly reduced the outrage costs associated with the
corresponding increase in executive stock options. Executives may have been able to maintain
those pay increases in the 2000s despite the fall in the stock market by using past pay levels as a
benchmark for future pay as a way to limit outrage.
Excessive pay might be defined from an equity perspective in terms of material need or in
relation to the pay of others in society. In economic discussions concerning the efficient allocation
of resources, these definitions are not likely to be useful. From an economic perspective, the
64 Since the stock market boom was largely unanticipated, the need to design pay packages that avoid windfalls ex ante
was probably unforeseeable at first, although it certainly became more obvious as the boom persisted.
starting point for determining whether pay is excessive is likely to be whether or not pay exceeds
marginal product. Unfortunately, economists have no way to directly measure an executive’s
marginal product (what monetary value should be placed on, say, decision making?), so there is
no way to directly determine whether executives are being overpaid in absolute terms. (In the
neo-classical model, that is not a concern since nobody is overpaid.) All that can be measured is
what specific executives are being paid relative to others, and how their firms have performed
(based on profitability, rate of return, stock price, and so on) relative to others. Many economists
have attempted to determine whether executive pay is correlated with firm success. In these
studies, executives with above average compensation can be thought to have “earned it” if their 65
firm outperformed its rivals. As useful as this exercise may be, it should be noted that it does not
attempt to answer the underlying question of whether, overall, executives are excessively paid on
average, and whether firms could achieve the same results if they paid their executives less.
Some economists define excessive pay more broadly than the marginal product definition,
defining it in terms of executives who are capturing economic rents. Taken literally, an economic
rent is defined as any level of pay above the worker’s reservation wage, which is the lowest wage 66
the worker would be willing to accept to work. Thus, according to this definition, an executive
could be paid less than the value he adds to the firm (i.e., his marginal product), but still be
overpaid since he would have been willing to perform the same job for less. This broader
definition is somewhat problematic from a measurement perspective—since the reservation wage
cannot be observed, classifying an executive as overpaid becomes purely subjective. In a
perfectly competitive labor market, one would expect the reservation wage to converge with
marginal product since there are a very large number of interchangeable workers for any job, and
jobs for any worker. In the market for executives, where the candidates and openings are limited
and not perfectly substitutable, a wedge between the reservation wage and marginal product could
theoretically exist. But it would be expected that the wedge would ultimately be limited by the
fact that any executive paid less than his marginal product could be profitably snatched away by
another firm willing to pay slightly more. Furthermore, the rent will exist whether it is captured
by the executive or the firm’s shareholders. Therefore, equating the executive’s capture of the rent
with excessive pay (and all the negative connotations that entails) implies that one views capital
(shareholders) as a more deserving recipient of the rent than labor. Defining excessive pay as pay
above marginal product remains less controversial, since it has clearer implications for economic
Arguments about efficiency mostly come down to incentive effects—the fear that attempts to
curb executive compensation would also reduce executive productivity. This would be true if
current compensation levels are needed to entice the best possible candidates to become
executives instead of pursuing other career paths that would not maximize these individuals
skills. Given the wide gulf between executive pay and pay in nearly any other profession
65 Murphy points out a problem with this methodology: firms with better managers would not be expected to offer
investors a permanently higher rate of return because of market forces. As soon as investors realized that a firm had
superior management, investors would bid up the current price of the firm’s stock. While the stock price was being bid
up, investors would experience a superior rate of return, but after this process was complete, future rates of return
would not be out of the ordinary because of the law of diminishing returns. An economist who compared these future
ordinary rates of return to the executives’ compensation might erroneously conclude that they were overpaid. Kevin
Murphy, “Executive Compensation,” in Orley Ashenfelter and David Card, eds., Handbook of Labor Economics, vol.
3B (Amsterdam: Elsevier Publishers, 1999), p. 2540.
66 BF define executive rent seeking as “additional value that managers obtain beyond what they would get in arm’s
length bargaining...” In Lucian Bebchuk and Jesse Fried, Pay Without Performance (Cambridge, MA: Harvard
University Press, 2004), p. 62.
(although some elite members of the financial, legal, and medical professions are comparably
paid), some might be skeptical that a pay cut that left a portion of the gulf intact would scare off
suitable candidates. For example, one economist recently pointed out that one CEO with a net
worth of $16 billion would need to spend $30 million per week just to keep his net worth from 67
rising. But most would agree that some reward is needed to compensate for job attributes such
as stress and long hours that are expected of executives.
Perhaps also implicit in efficiency concerns is the neo-classical assumption that higher pay leads
to higher levels of utility. In other words, individuals are motivated to work harder to earn more
because earning more makes them happier. Thus, policymakers looking to curb executive pay
may need to weigh the loss of welfare it would cause as one of the costs of such a proposal. But 68
recent empirical evidence on compensation and welfare suggests that this might not be the case.
This research suggests that after a certain point, people are not made happier by further increases
in absolute income. That is because when income increases, people quickly adapt to their new
circumstances, and any temporary rise in happiness caused by greater income dissipates. Rather,
the research suggests that people are made permanently happier by increases in relative income, a
sentiment captured in the saying “keeping up with the Joneses.” According to this theory, the
ratcheting up in executive pay in recent years has not made anyone better off since all executives’
pay is rising simultaneously. It also suggests that, hypothetically, if there were a costless way to
reduce executive pay across the board (admittedly, an unrealistic assumption), it might not make
anyone worse off (once individuals had adapted to their new circumstances) as long as enough of
the gulf between executive pay and other professions were maintained to leave a sufficient
superiority in relative income intact. Executives may also derive utility from non-material aspects
of their job, such as fame and power, that could be unaffected by an absolute decline in income.
Obviously, across-the-board changes would require some type of collective action; if any single
board decided to reduce executive pay, those executives would lose utility since their income
relative to their peers would fall.
A corporation’s bylaws lay out the general rules regarding a company’s corporate governance
protocol, including procedures for the determination of executive pay. In turn, the general
parameters of a firm’s corporate governance are principally dictated by the state corporate law
that prevails in the state in which a company is incorporated. By some estimates, about half of
publicly traded companies are incorporated in the state of Delaware, giving its corporate laws a
disproportionately large influence in this area. Delaware’s influence also extends to the Delaware
Chancery Court, which is widely viewed as the preeminent national legal forum for corporate
In spite of the primacy of state law in this area, the federal government and regulatory entities,
like the SEC and the nation’s securities exchanges, have also promulgated policies in this area.
The existing rules governing CEO pay have developed through a variety of legislation, executive
branch regulation, and regulation by independent, self-governing bodies. Most regulation
involves disclosure of pay, and does not set compensation levels. The following legislative and
67 Austin Goolsbee, “For the Super-Rich, Too Much Is Never Enough,” New York Times, March 1, 2007.
68 See, for example, Robert Frank, “The Frame of Reference as a Public Good,” Economic Journal, vol. 107, 1997, p.
1832; Richard Layard, Happiness (New York: Penguin Press, 2005).
regulatory developments have affected CEO pay, although that was not always their primary
focus. Starting with the earliest developments, they are listed in chronological order.
Corporations can generally deduct employee pay, including executive pay, from their corporate
income subject to taxation. In 1993, in response to outrage at executive pay levels, OBRA (P.L.
103-66) added section 162 (m), titled “Certain Excessive Employee Remuneration,” to the
Internal Revenue Code. It imposes a $1 million cap on the deductibility of compensation that
applies to the CEO and the four next highest-paid officers. (Pay itself is not capped, only the
deduction of pay from corporate income.) No tax deduction for compensation above the $1
million limit is permitted except for “performance-based” pay, such as commissions or stock
options, where the ultimate compensation received by the executive depends on the stock price,
reported sales or profits, or some other financial indicator. To qualify for the exception, the goals
underlying the performance-based compensation must have been determined by a compensation
committee that is comprised solely of two or more outside directors. The terms under which the
performance-based compensation is to be paid, including the performance goals, must be
disclosed to shareholders and approved by a majority shareholder vote.
Executives’ cash compensation, the type of pay most directly affected by OBRA, increased
slowly after OBRA took effect. But this provision in OBRA is widely believed to have
contributed to the growing importance of stock options in CEO compensation in the mid and late 69
1990s. As a result of this trend, overall compensation grew even more quickly than before, so
OBRA may have had the unintended consequence of increasing CEO pay if stock options played 70
the enabling role in excessive pay that critics claim.
Enacted in the wake of accounting scandals at firms like Enron and WorldCom, the Sarbanes-
Oxley Act of 2002 (P.L. 107-204) contains a broad range of corporate governance and accounting
reforms, two of which are particularly relevant to executive pay:
• Prohibition on Personal Loans to Executives. Section 402 of the law makes it
unlawful for any public company, directly or indirectly, to extend credit, maintain
credit, or arrange for the extension of credit in the form of a personal loan to, or
for the benefit of, any director or executive officer.
• More Timely Reporting of Corporate Insider Stock-Based Transactions.
Section 403 of the law requires insiders (defined as officers, directors, and
shareholders owning at least 10% of outstanding stock) to file reports of their
trades of the issuer’s stock and stock options with the SEC before the end of the
69 Other factors, such as the wave of public offerings by cash-poor technology firms and the bull market itself, also
increased the popularity of options during the 1990s.
70 For example, see SEC Chairman Christopher Cox, “Testimony Concerning Options Backdating before the U.S.
Senate Committee on Banking, Housing and Urban Affairs,” September 6, 2006.
second business day on which the trade occurred. Previously, option grants did
not have to be disclosed until 45 days after the end of the fiscal year.
In 2003, the SEC approved changes to the listing standards for firms listed on the New York
Stock Exchange (NYSE) and the NASDAQ Stock Market that require shareholder approval of 71
almost all equity-based compensation plans. Firms must disclose the material terms of their
stock option plans prior to the shareholder vote. The required disclosures include the terms on
which stock options will be granted and whether the plan permits options to be granted with an
exercise price that is below the market value of the company’s stock on the date of the grant.
While the regulation requires shareholder approval of the overall compensation plan, it does not
require shareholder approval of the specific amount of compensation received by individual
In 2003, the SEC also approved other changes to the listing standards for firms listed on the 72
NYSE. The standards require companies listed on the NYSE to have a compensation committee 73
that is entirely composed of outside directors. In addition, if a compensation consultant is to be
used to assist in the evaluation of director, CEO, or senior executive officer compensation, the
compensation committee is required to have the sole authority to retain the consultant and
approve the consultant’s fees.
In 2004, the Financial Accounting Standards Board (FASB), a private sector entity that writes 74
accounting standards, released accounting directive FAS 123R, which requires companies to
“expense” (count as a cost) the value of employee stock option grants in their income statements
in their next fiscal year, beginning in June 2005 for large companies and December 2005 for
small companies. Recognition of the cost of options has the effect of reducing the corporation’s
71 Securities and Exchange Commission Release No. 34-48108, “Self-Regulatory Organizations; New York Stock
Exchange, Inc. and National Association of Securities Dealers, Inc.; Order Approving NYSE and Nasdaq Proposed
Rule Changes and Nasdaq Amendment No. 1 and Notice of Filing and Order Granting Accelerated Approval to NYSE
Amendments No. 1 and 2 and Nasdaq Amendments No. 2 and 3 Thereto Relating to Equity Compensation Plans,” June
72 Section 303A.05 of the New York Stock Exchange Listing Company Manual section, available at
73 Foreign private issuers, controlled companies (a company in which more than 50% of the voting power is held by an
individual, group or another company), limited partnerships, companies in bankruptcy proceedings, and NYSE
companies listing only preferred or debt securities are exempt from the requirement that they have a compensation
74 While federal securities law gives the SEC authority to set accounting standards for publicly traded companies, the
SEC generally defers to the standards established by FASB. For more information, see CRS Report RS21392, Stock
Options: The Accounting Issue and Its Consequences, by Bob Lyke and Gary Shorter.
reported earnings, which may make granting options less desirable to companies. Previously,
most companies had simply noted the value of options grants in the footnotes to the financial
statements, which had no effect on earnings. Thus, FAS 123R may constrain executive pay even
though that was not its primary intent.
The requirement that publicly traded companies disclose how much they pay top executives dates
from the 1930s. The SEC has modified the disclosure format several times, as the forms of CEO
pay have become more varied and complex. In 1992, the SEC required that proxy statements
include tables setting out several categories of pay for the top five executives. These included
base salaries, bonuses, deferred, and incentive-based compensation, including stocks and stock
options. Corporations were required to place an estimated value on options granted to executives.
By 2006, the SEC had concluded the 1992 disclosure rules were, in the words of SEC Chairman
out of date.... [They] haven’t kept pace with changes in the marketplace, and in some cases
disclosure obfuscates rather than illuminates the true picture of compensation.... We want
investors to have better information, including one number—a single bottom line figure—for
total annual compensation.
In July 2006, for the first time since 1992, the SEC adopted major changes to executive disclosure 75
rules contained in public companies’ registration and proxy statements. The disclosure
requirements apply to the CEO, the chief financial officer (CFO), and the next three most highly
compensated executive officers. The rules require the disclosure of the executives’ total
compensation, the fair value of their stock option grants, estimates of potential post-employment
payments and benefits, and tabular disclosure of director pay. It requires that statements include a
Compensation Discussion and Analysis (CD&A), which is a narrative that must explain the
objectives and implementation of a company’s executive pay program. And in response to the
stock option grant backdating controversy, the rules require detailed information about a
company’s option grant practices in the both the CD&A and a supplemental table. The rules also
require the disclosure of directors’ compensation figures for the preceding fiscal year. The
disclosure requirements went into effect in 2007.
SEC officials have said that the central contribution of the disclosure reform is the provision of 76
enhanced transparency with respect to executive pay. They may also be intimating that the new
disclosure requirements could at least indirectly help improve shareholders’ ability to exert
pressure on management to temper executive pay:
By restraining executives from self-indulgent behavior—and using salary, bonuses, options,
long term benefits, and other financial incentives in very purposeful ways—compensation
committees acting on behalf of the shareholders can increase management’s incentives to
improve corporate performance. So our purpose in this very aggressive new executive
75 “Executive Compensation and Related Person Disclosure,” Federal Register, vol. 71, no. 174,
September 8, 2006, p. 53158. For an overview, see CRS Report RS22583, Executive Compensation: SEC Regulations
and Congressional Proposals, by Michael V. Seitzinger.
76 For example, see “Speech by SEC Chairman Christopher Cox: Proposed Revisions to the Executive Compensation
and Related Party Disclosure Rules,” January 17, 2006.
compensation rule is very straightforward: It is to protect and advance the interests of 77
But those who would hope that the reform will have a dampening effect on executive pay levels
may not be encouraged, at least so far: a survey of directors at 110 firms, conducted at the
conclusion of 2006 by Mercer Human Resource Consulting, found that 70% planned only
minimal changes to their executive compensation programs as a result of the new SEC rules.
Only 15% of the directors said that the reform would have a substantial impact on their approach 78
to executive compensation.
In the theoretical neo-classical world, executive pay would be determined by marginal product.
Since the outcome would already be economically efficient, any policy response that created a
wedge between executive pay and executives’ marginal product would reduce economic
efficiency. Those who believe that market forces are the best curb on excessive pay argue that
government should be doing less, not more, in this area because government cannot determine
appropriate levels or forms of executive pay more accurately than shareholders or boards.
The managerial power critique makes the case that executive pay is not determined by arm’s
length contracting. It suggests a number of reforms to bring executive pay closer to an arm’s
length contracting ideal. These reforms could be promoted by shareholders, board members, or in
some cases, mandated by the government, through legislation or regulation. They fall under three
broad categories: improving the transparency of executive pay, strengthening board independence
to reduce the potential for board capture, and strengthening shareholder control over the board
and management. Proponents often argue that they are not trying to interfere with market forces,
but to level the playing field for shareholders in their interactions with management. Some of the
reforms they promote do not affect executive pay directly (e.g., not allowing the CEO to be
chairman of the board); rather, their stated intent is that reforms that strengthen shareholders’
rights or board independence should lead to lower pay (or pay that is more sensitive to
performance). But a few proposals do affect executive pay directly, and are analyzed below.
Some observers favor the status quo, arguing that Congress should continue to defer to more
specialized regulatory bodies, such as the SEC, with an expertise in corporate governance. The
regulatory bodies have focused on making occasional minor policy modifications to enhance
transparency and align boards’ incentives more closely with shareholders’ interests. In this view,
hasty policy changes in reaction to rising pay levels would risk undermining the current system
that, on the whole, is effective at rewarding good executive performance.
77 “Speech by SEC Chairman Christopher Cox: Remarks at Northwestern University School of Law, Chicago Ilinois,”
January 22, 2007.
78 “A Better Look At The Boss’s Pay: New SEC Rules Require Greater Disclosure, but Don’t Expect CEOs to Take a
Hit,” Businessweekonline.com, February 26, 2007.
In 1993, OBRA added Section 162(m) to the Internal Revenue Code, which limited a company’s
tax deduction for what it pays each of its top executives to $1 million, but exempted
“performance-based” pay like stock options. Supporters of the neo-classical model argue that the
million dollar deduction limit is a good example of how government intervention in markets can
have unintended consequences that cause a policy to backfire. Despite the deduction limit, overall
executive compensation continued to rise rapidly, and the limit may have contributed to the rapid 79
growth of executive stock options, which critics argue are at the heart of the corporate scandals.
Critics see this as an example of a policy with unintended consequences, in which government
regulation encourages behavior that circumvents the regulation’s original intent. Tax policy may
be too blunt a tool to effectively encourage these goals.
Since the deduction limit failed to curb the growth in executive pay, some would argue that it
should be eliminated or scaled back. Alternatively, if the purpose of the deduction limit is to curb
overall executive pay levels, others would argue the limit would be more effective if expanded to
cover all forms of pay (perhaps at a different deduction level). Some research found that the
deduction limit appears to have been a factor in the growth in executive stock options and overall 80
executive pay. But other research concluded that the statute had little to do with subsequent
increases in the sensitivity of overall executive pay to performance-based components like stock 81
options. Factors that may have played a larger role in the growth of stock options include the
bull market of the 1990s, not having to treat stock options as a corporate expense, and pressure on
firms to provide executive pay that better aligned their interests with those of shareholders.
Alternative research concluded that while the million dollar cap may have initially helped to
compress executive salaries around the $1 million level, it did not appear to have had a significant
impact on total compensation and other components of pay, such as bonuses and stock option 82
Some critics of executive pay have advocated changing Section 162(m) of the Internal Revenue
Code to prohibit businesses from taking tax deductions for compensation provided to executives
when the ratio of executive pay to that of its employees exceeds a certain level. The ratio could be
set relative to, say, the average employee’s pay or the lowest-paid employee’s pay. A very few
firms have voluntarily implemented such a policy. For example, the CEO of Whole Foods
Markets limits his pay to no more than 14 times the pay of the firm’s average employee.
79 If one argues that the purpose of the cap is solely to encourage performance-based pay, then the subsequent rise in
overall compensation levels may be irrelevant.
80 For example, see Brian Hall and Kevin Murphy, “The Trouble with Stock Options,” Journal of Economic
Perspectives, vol. 12, 2003, pp. 49-70, and Tod Perry and Marc Zenner, Pay for Performance Government Regulation
and the Structure of Executive Compensation, University of Michigan Working Paper, 2001.
81 Marilyn Johnson, Susan Porter, and Margaret Shackell, “Stakeholder Pressure and the Structure of Executive
Compensation,” May 1997, available at SSRN: http://ssrn.com/abstract=41780 or DOI: 10.2139/ssrn.41780.
82 Nancy L. Rose, and Catherine Wolfram, “Has the ‘Million-Dollar Cap’ Affected CEO Pay?,” American Economic
Review, vol. 90, no. 2, May, 2000, pp. 197-202.
Using the tax code to mandate such a policy might arguably help address some concerns with the
erosion of pay equity and growing income inequality. Supporters of this policy argue that
discouraging excessive pay through tax disincentives is preferable to—and less disruptive than—
prohibiting excessive pay directly. Critics could argue that the growing ratio between worker and
CEO pay owes itself in part to an unrelated development—the moribund growth in worker pay.
Critics could also argue that the reform could undermine the core investor concern of whether a
CEO receives compensation commensurate with his or her performance. Furthermore, wage
levels vary by company primarily because different companies hire different types of workers.
For example, the average pay at a software company is likely to exceed average pay at a chain of
fast food restaurants. Under this proposal, the fast food company would face higher taxes if it
wanted to pay its executives a comparable wage to the software company. As with complaints
about the OBRA cap, such a cap could result more in maneuvering by businesses to avoid the cap
than fulfillment of its intended goal—for example, any such cap would have to delve into the
complexities of what forms of employee and executive compensation should or should not fall
under the cap, creating incentives for stealth pay. Detractors could also cite research that
concluded that the tax deductibility of executive compensation tends to have a minimal impact on
firm’s ultimate profitability, raising additional questions about the ability of such a policy to help 83
constrain executive pay.
Outside compensation consultants are generally hired to help boards craft the firm’s executive
compensation packages. In a number of cases, the consultants are part of larger companies that
furnish additional consultation services to the firms. It could be argued that when firms provide
multiple consultancies to individual firms, there is a conflict of interest that makes it difficult for
their compensation consultant subsidiaries to resist pressure to recommend favorable executive
pay packages. Two possible options to address such concerns would be (1) an SEC requirement
that a firm’s proxy statements disclose all of the services it receives from companies that offer it
compensation consulting services; or (2) a law to ban outfits that furnish executive compensation
consulting services to a firm from offering other consultant services to them, similar to the
Sarbanes-Oxley Act of 2002’s proscription on auditors providing certain ancillary services to the
firms they audit. If the latter alternative were pursued, some companies might stop offering
compensation consulting services. These reforms might not lead to any change in behavior,
however, because the basic incentive to recommend high pay remains even if compensation 84
consulting is the firm’s sole business, for reasons discussed earlier.
83 Nancy L. Rose, and Catherine Wolfram, “Has the ‘Million-Dollar Cap’ Affected CEO Pay?,” American Economic
Review, vol. 90, no. 2, May, 2000, pp. 197-202.
84 On December 5, 2007, the House Committee on Oversight and Government Reform held a hearing to examine the
role played by compensation consultants in determining the pay packages of senior executives at the largest publicly
traded corporations. During the hearing, corporate governance experts, institutional investors, and compensation
consulting firms testified regarding the role of consultants in setting executive pay, efforts to prevent and manage
conflicts of interest, and the adequacy of the information available to shareholders and the public. Simultaneously, the
committee released a report, Executive Pay: Conflicts of Interest Among Consultants, which examined whether
compensation consultants who are hired by large publicly traded firms act independently. The report’s principal finding
was that “over 100 large publicly traded companies hired compensation consultants with substantial conflicts of interest
in 2006.... [And] that in many cases, the consultants who are advising on executive pay are simultaneously receiving
millions of dollars from the corporate executives whose compensation they are supposed to assess...” The study can be found
While directors must be approved by shareholders, the nominees are typically chosen by the
board or management. At present, shareholders can nominate directors, but the process is
arduous, expensive (estimates range up to $1 million), and thus rarely pursued. To ease the
process, the SEC proposed a rule in July 2003 that would have allowed shareholders with more
than 5% of a company’s voting securities to under certain conditions have their board nominees
included in a company’s proxy materials, which carry the management’s slate of board 85
nominees. A response to widespread concerns over the accountability of corporate directors
after a number of corporate scandals, the proposal received the support of various observers,
including some institutional investors. They argued that the integrity of corporate boards would
be enhanced because the reform would result in boards being populated with a greater number of
outside directors who are less beholden to management and better able to provide independent
oversight and scrutiny of executive compensation practices and excesses, such as backdating.
Although publicly listed firms are required to have outside directors, some critics have questioned
the independence of outside directors recommended by management. For example, one study
found that large numbers of outside directors inexplicably appear to have been the beneficiaries 86
of options manipulation. Opponents of shareholder access reform argue that it could potentially
result in antagonistic directors, and thus dysfunctional boards. Shareholders may also be less able
to identify the most qualified candidates for the position.
In the end, the SEC did not adopt the proposed 2003 shareholder proxy reform rule, a decision
that many ascribe to vehement business opposition. But the issue re-emerged in August 2006,
when the United States Court of Appeals for the Second Circuit reached a decision in American
Federation of State, County and Municipal Employees Pension Plan v. American International
Group, Inc. This ruling was the appeal’s courts response to an earlier petition by the American
Federation of State, County and Municipal Employees (AFSCME) to reverse the American
International Group’s (AIG) rejection of its effort to place a binding shareholder proposal in the
company’s proxy materials that would have changed its bylaws to facilitate shareholder
nomination of directors. Historically, the SEC has generally allowed firms to exclude shareholder
proposals relating to an election from their proxies, as it did in this case. However, the Second
Circuit found the SEC’s policy in this area to be historically inconsistent and asked the agency to 87
After the decision, the SEC basically chose not to rule on proxy access petitions, which Chairman
Cox said injected a note of uncertainty into the proxy process for 2007.
85 “Proposed Rule: Security Holder Director Nominations, Release nos. 34-48626; IC-26206; FILE NO. S7-19-03,
Securities and Exchange Commission,” October 17, 2003. At least one of two things would have had to have taken
place in the previous year’s board election to trigger the requirement: 1) 35% or more of shareholders voted to withhold
support for at least one director at the company’s annual meeting; or 2) a stockholder or a group of shareholders with at
least 1% of the company’s stock put a proposal on the proxy statement seeking the right to nominate a director, and the
proposal was approved by a majority vote.
86 Lucian Bebchuk et al., Lucky Directors, Harvard Law School working paper, December 2006.
87 “Remarks Before the Willamette Securities Regulation Institute by SEC Commissioner Roel Campos,” October 19,
In July 2007, the SEC proposed two quite divergent policy proposals. One proposal, the short
proposal, would essentially codify longstanding SEC practices of denying shareholder-proposed
candidates for board director positions to be included in company proxy statements (called proxy 88
access). The proposal, which the agency eventually adopted in late November of the year,
appears to have been a response to the uncertainty that prevailed after the AIG decision. The 89
second proposal, the so-called long proposal was not adopted. It would have allowed
shareholders or shareholder coalitions with greater than 5% of outstanding shares to propose
binding bylaw provisions that could permit specified shareholders to nominate directors and 90
require the company to include the nominees in the company’s proxy statement.
Generally, business interests applauded the agency’s decision to adopt the short proposal, while
shareholder interests derided it. Chairman Christopher Cox claimed that the agency’s vote would
create legal certainty for the upcoming annual proxy season in spring of 2008, but conceded that
investor advocacy groups and others would be disappointed. He has, however, indicated that the
agency will probably revisit the subject of shareholder proxy access in 2008 when the 91
Commission has its full complement of Democratic members.
After the SEC’s vote, Senate Banking Chairman Dodd said that he might try to offer legislation to
reverse the decision, saying that he did not think it was a fair decision. And Barney Frank,
Chairman of the House Financial Services Committee, expressed disappointment that the SEC
would deny shareholders the right to offer proxy-access proposals. Chairman Frank also stressed 92
that the SEC should have deferred action until it was at full strength.
While shareholders are required to vote on a company’s overall equity-based compensation
programs, they do not vote on pay packages for individual executives. The managerial power
critique has fueled growing interest in giving shareholders a non-binding vote on individual
88 A proxy statement has SEC-approved information that is to be voted on at an annual meeting.
89 The SEC did, however, adopt a subsidiary part of the long proposal. The proxy rules were amended to further the use
of electronic shareholder forums in which shareholders would have an additional means to communicate with each
other and the company. Historically, participation in such shareholder forums would potentially constitute a solicitation
subject to the current proxy rules. But as amended, such electronic communications would be exempt from most proxy
rules if the conditions to the exemption are satisfied. Forum participants would be able to rely on the exemption so long
as his or her communications occur more than 60 days prior to the date the company announces for its annual or special
meeting of shareholders, and the communicating party does not solicit authority while relying on the exemption. Also
under the change, management will be able to participate in such electronic shareholder forums without first providing
a proxy statement and a participant would not be liable for the statements or information provided by others
participating in the forum. A summary of the amendment is available at: http:[//www.sec.gov/news/press/2007/2007-
90 “Shareholder Proposals Related to the Election of Directors.” Securities and Exchange Commission Release No. 34-
56914, A summary is available at: http:[//www.sec.gov/news/press/2007/2007-247.htm].
91 Karey Wutkowski and Rachelle Younglai, “US SEC Restricts Shareholder Proxy Access,” Reuters, November 28,
2007. In early November 2007, Senator Christopher Dodd, Chairman of the Senate Committee on Banking, Housing,
and Urban Affairs and Senators Johnson, Reed, Schumer, Menendez, Akaka, Brown, Casey, and Tester sent a letter to
Chairman Cox urging him not to adopt either of two proposals. As an alternative, the letter recommended that the SEC
maintain current rules, which allow shareholders to propose procedures to elect directors. The letter went on to say that
“on this and other matters of great significance, we feel the Commission should proceed only after it has a full
complement of Commissioners...”
92 Wutkowski and Younglai, “US SEC Restricts Shareholder Proxy Access.”
executive pay packages. Along these lines, there is Representative Frank’s Shareholder Vote on
Executive Compensation Act (H.R. 1257), which was approved by the House on April 19, 2007,
and Obama’s companion and identically named bill (S. 1181).
Furthermore, in anticipation of the 2007 annual corporate meetings, activist investors have
submitted shareholder proposals at about 60 companies seeking the right to have a non-binding 93
vote on executive pay.
Proponents believe that votes against individual pay packages, or merely the threat, could raise
outrage costs, thus prompting directors to exercise greater restraint in pay setting and to be more
conscientious in linking pay to performance. They argue that the vote would not overly burden or
restrict the board since the vote would be non-binding. Several countries, including the United
Kingdom, Australia, and Sweden, have given their shareholders the right to such a non-binding
vote. Domestically, at least one firm, the insurer Aflac, has reportedly agreed to provide its 94
shareholders with such a vote. Many other firms, however, are publicly opposed to the idea.
Two main arguments are made in opposition to a mandatory non-binding shareholder vote on pay.
First, to the extent that current levels of executive pay are largely explained by legitimate market
forces, as some have argued, giving shareholders a non-binding vote on pay might inject
undesirable distortions into the pay setting process and the demand and supply of CEOs. Second,
the minutiae of CEO compensation packages can be difficult enough for corporate directors to
master. Thus, it has been argued that expecting shareholders with relatively limited resources
available for comprehending such things to be a knowledgeable presence in the pay setting
process would be unrealistic.
If the underlying concern with executive pay is equity, not efficiency, then the policy goal may be
to reduce inequality in the least economically costly way. A more progressive tax system is
widely considered to be the least costly way to redistribute income, in terms of lost economic
efficiency. Of course, the tax code cannot target executives specifically without also affecting
other high income individuals. Thus, while progressive taxation can be viewed as an effective
way to promote equity goals, it is not well targeted toward reducing potential efficiency losses
that result from the principal-agent problem that affects executive compensation.
Given the importance of stocks and stock options in executive pay, policymakers attempting to
increase the tax code’s progressivity could consider reducing the tax preference currently given to
capital income compared to labor income. For example, capital gains and dividends are taxed at a
lower marginal rate than labor income and a sizeable portion of capital income tax can be
deferred through tax-preferred savings vehicles. Stock options are taxed at regular income rates
but receive favorable treatment because tax liability is deferred until gains are realized.
Economists are divided over whether taxing labor and capital at the same rate would be
93 Erin White, and Aaron Patrick, “Shareholder Push for Vote on Executive Pay,” The Wall Street Journal, February
26, 2007, p. B-1.
94 PRNewswire, “Aflac Adopts Non-Binding ‘Say on Pay’ Shareholder Vote,” press release, February 14, 2007.
economically efficient,95 but given the unequal distribution of financial assets in the United
States, it would undoubtedly increase the tax code’s progressivity.
95 For an explanation, see CRS Report RL32603, The Flat Tax, Value-Added Tax, and National Retail Sales Tax:
Overview of the Issues, by Jane G. Gravelle.
A stock option allows the holder the right to buy a company’s stock at a predetermined fixed
price, called the strike price, regardless of the stock’s price at the date of purchase. The holder
exercises the option when he subsequently buys the stock. After the option is granted, the
company’s stock could either rise above or fall below the strike price. If the market price fell
below the strike price, the option would have no value and would not be exercised (because the
holder could buy the stock for less on the open market). An option with a strike price above the
market price is said to be under water. If the market price rose above the strike price, the value of
the option would be equal to the difference between the strike price and the actual price (because
the holder can buy the stock at the strike price and then sell it at the market price). For example, if
a person was granted an option to purchase company X’s stock in one year at $100, and the stock
turns out to be worth $125 in one year, then he would earn $25 by exercising the option in a year.
Alternatively, if the stock turns out to be worth $75 in one year, he would not exercise the option
and would neither gain nor lose any money. Thus, the option’s value can never fall below zero.
Because the future is uncertain, economists and accountants must use complex formulas to place
a value on an option when it is granted that takes into account how much value, if any, the option
is expected to have when it is ultimately exercised. There is a consensus among economists that
the expected value at issuance is the amount that should be included in measures of executive
pay. The value of the option when it is ultimately exercised is unlikely to be the same as its
expected value when it is issued, however.
Generally, stock options can be bought and sold on the open market by anyone. There are a subset
of stock options called employee stock options in which firms issue their own stock to their
workers and executives. Employee stock options cannot be sold to others, and often the holder
must wait until a vesting period is over before being able to exercise them in order to encourage
employees to stay with the firm. Usually, the strike price for employee stock options is set at the
firm’s current price. The firm does not have to outlay any cash when it grants employees stock
options. For this reason, employee stock options are particularly popular with start-up firms with
limited cash flow and high growth prospects. However, when the options are exercised, new stock
is created, which dilutes the ownership of the existing stockholders. The firm can offset this
dilution by buying back an equivalent amount of outstanding stock from the open market, which
would require a cash outlay at that point.
The prevalence of employee stock options is somewhat puzzling to economists because if
employees, firms, and shareholders acted rationally, they would each have reasons not to prefer
them. Employees should prefer cash wages to options since the options expose them to risk, and
people are generally risk averse. (Options do have tax advantages for the executives, however, 96
which are particularly valuable for executives facing high marginal tax rates). For firms, options
can be thought of as a loan from employees (in the form of forgone wages) that must be paid back
when the option is exercised. If capital markets are efficient, it should be cheaper for the firm to
borrow on the open market than through their employees. For shareholders, options are
undesirable since they dilute their ownership position. In light of these drawbacks, BF see the
popularity of options as evidence of their “managerial power” theory. Alternatively, some
economists have argued that the popularity of employee stock options is the result of the
96 Brian Hall and Jeffrey Liebman, The Taxation of Executive Compensation, National Bureau of Economic Research
working paper 7596, March 2000.
perceived cost of options being lower than their actual cost, because they require no cash outlay
and because until recently they did not have to be expensed (were not counted as a cost) on the 97
firm’s balance sheet.
Gary Shorter Marc Labonte
Specialist in Financial Economics Specialist in Macroeconomic Policy
firstname.lastname@example.org, 7-7772 email@example.com, 7-0640
97 Brian Hall and Kevin Murphy, “The Trouble with Stock Options,” Journal of Economic Perspectives, vol. 17, no. 3,