CEO INCENTIVES - IT'S NOT HOW MUCH YOU PAY, BUT HOW

CEO INCENTIVES—IT’S NOT HOW MUCH YOU PAY, BUT HOW
Michael C. Jensen
Kevin J. Murphy
Harvard Business School
University of Rochester
mjensen@hbs.edu
Abstract
Paying top executives “better” would eventually mean paying them more. The
arrival of spring means yet another round in the national debate over executive
compensation. Soon the business press will trumpet answers to the questions it
asks every year: Who were the highest paid CEOs? How many executives made
more than a million dollars? Who received the biggest raises? Political figures,
union leaders, and consumer activists will issue now-familiar denunciations of
executive salaries and urge that directors curb top-level pay in the interests of
social equity and statesmanship.
The critics have it wrong. There are serious problems with CEO compensation,
but “excessive” pay is not the biggest issue. 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. On average, corporate America pays its
most important leaders like bureaucrats. Is it any wonder then that so many
CEOs act like bureaucrats rather than the value-maximizing entrepreneurs
companies need to enhance their standing in world markets?
We recently completed an in-depth statistical analysis of executive
compensation. Our study incorporates data on thousands of CEOs spanning five
decades. The base sample consists of information on salaries and bonuses for
2,505 CEOs in 1,400 publicly held companies from 1974 through 1988. We also
collected data on stock options and stock ownership for CEOs of the 430 largest
publicly held companies in 1988. In addition, we drew on compensation data for
executives at more than 700 public companies for the period 1934 through 1938.
© Jensen and Murphy 1990
Harvard Business Review, May-June 1990, No. 3, pp 138-153.
also published in
Foundations of Organizational Strategy, Michael C. Jensen, Harvard University Press, 1998.
This document is available on the
Social Science Research Network (SSRN) Electronic Library at:
http://papers.ssrn.com/sol3/paper.taf?ABSTRACT_ID=146148

---

CEO INCENTIVES—IT’S NOT HOW MUCH YOU PAY, BUT HOW
By Michael C. Jensen and Kevin J. Murphy*
Harvard Business Review, May-June 1990, No. 3, pp 138-153.
The arrival of spring means yet another round in the national debate over
executive compensation. Soon the business press will trumpet answers to the questions it
asks every year: Who were the highest paid CEOs? How many executives made more
than a million dollars? Who received the biggest raises? Political figures, union leaders,
and consumer activists will issue now-familiar denunciations of executive salaries and
urge that directors curb top-level pay in the interests of social equity and statesmanship.
The critics have it wrong. There are serious problems with CEO compensation,
but “excessive” pay is not the biggest issue. 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. On average, corporate America pays its most important leaders like
bureaucrats. Is it any wonder then that so many CEOs act like bureaucrats rather than the
value-maximizing entrepreneurs companies need to enhance their standing in world
markets?
We recently completed an in-depth statistical analysis of executive compensation.
Our study incorporates data on thousands of CEOs spanning five decades. The base
sample consists of information on salaries and bonuses for 2,505 CEOs in 1,400 publicly



* Michael C. Jensen is the Edsel Bryant Ford Professor of Business Administration at the Harvard Business
School. Kevin J. Murphy is an associate professor at the University of Rochester’s William E. Simon
Schol of Business Administration.

---

Jensen and Murphy
2
May 1990
held companies from 1974 through 1988. We also collected data on stock options and
stock ownership for CEOs of the 430 largest publicly held companies in 1988. In
addition, we drew on compensation data for executives at more than 700 public
companies for the period 1934 through 1938.
Our analysis leads us to conclusions that are at odds with the prevailing wisdom
on CEO compensation:
Despite the headlines, top executives are not receiving record salaries and
bonuses. Salaries and bonuses have increased over the last 15 years, but CEO pay levels
are just now catching up to where they were 50 years ago. During the period 1934
through 1938, for example, the average salary and bonus for CEOs of leading companies
on the New York Stock Exchange was $882,000 (in 1988 dollars). For the period 1982
through 1988, the average salary and bonus for CEOs of comparable companies was
$843,000.
Annual changes in executive compensation do not reflect changes in corporate
performance. Our statistical analysis posed a simple but important question: For every
$1,000 change in the market value of a company, how much does the wealth of that
company’s CEO change? The answer varied widely across our 1,400-company sample.
But for the median CEO in the 250 largest companies, a $1,000 change in corporate value
corresponds to a change of just 6.7 cents in salary and bonus over two years. Accounting
for all monetary sources of CEO incentives—salary and bonus, stock options, shares
owned, and the changing likelihood of dismissal—a $1,000 change in corporate value
corresponds to a change in CEO compensation of just $2.59.
Compensation for CEOs is no more variable than compensation for hourly and
salaried employees. On average, CEOs receive about 50% of their base pay in the form
of bonuses. Yet these “bonuses” don’t generate big fluctuations in CEO compensation.
A comparison of annual inflation-adjusted pay changes for CEOs from 1975 through
1988 and pay changes for 20,000 randomly selected hourly and salaried workers shows

---

Jensen and Murphy
3
May 1990
remarkably similar distributions. Moreover, a much lower percentage of CEOs took real
pay cuts over this period than did production workers.
With respect to pay for performance, CEO compensation is getting worse rather
than better. The most powerful link between shareholder wealth and executive wealth is
direct stock ownership by the CEO. Yet CEO stock ownership for large public
companies (measured as a percentage of total shares outstanding) was ten times greater in
the 1930s than in the 1980s. Even over the last 15 years, CEO holdings as a percentage
of corporate value have declined.
Compensation policy is one of the most important factors in an organization’s
success. Not only does it shape how top executives behave but it also helps determine
what kinds of executives an organization attracts. This is what makes the vocal protests
over CEO pay so damaging. By aiming their protests at compensation levels, uninvited
but influential guests at the managerial bargaining table (the business press, labor unions,
political figures) intimidate board members and constrain the types of contracts that are
written between managers and shareholders. As a result of public pressure, directors
become reluctant to reward CEOs with substantial (and therefore highly visible) financial
gains for superior performance. Naturally, they also become reluctant to impose
meaningful financial penalties for poor performance. The long-term effect of this risk-
averse orientation is to erode the relation between pay and performance and entrench
bureaucratic compensation systems.
Are we arguing that CEOs are underpaid? If by this we mean “Would average
levels of CEO pay be higher if the relation between pay and performance were stronger?”
the answer is yes. More aggressive pay-for-performance systems (and a higher
probability of dismissal for poor performance) would produce sharply lower
compensation for less talented managers. Over time, these managers would be replaced
by more able and more highly motivated executives who would, on average, perform

---

Jensen and Murphy
4
May 1990
better and earn higher levels of pay. Existing managers would have greater incentives to
find creative ways to enhance corporate performance, and their pay would rise as well.
These increases in compensation—driven by improved business performance—
would not represent a transfer of wealth from shareholders to executives. Rather, they
would reward managers for the increased success fostered by greater risk taking, effort,
and ability. Paying CEOs “better” would eventually mean paying the average CEO
more. Because the stakes are so high, the potential increase in corporate performance and
the potential gains to shareholders are great.
How Compensation Measures Up
Shareholders rely on CEOs to adopt policies that maximize the value of their
shares. Like other human beings, however, CEOs tend to engage in activities that
increase their own well-being. One of the most critical roles of the board of directors is
to create incentives that make it in the CEO’s best interest to do what’s in the
shareholders’ best interests. Conceptually this is not a difficult challenge. Some
combination of three basic policies will create the right monetary incentives for CEOs to
maximize the value of their companies:
• Boards can require that CEOs become substantial owners of company stock.
• Salaries, bonuses, and stock options can be structured so as to provide big
rewards for superior performance and big penalties for poor performance.
• The threat of dismissal for poor performance can be made real.
Unfortunately, as our study documents, the realities of executive compensation
are at odds with these principles. Out statistical analysis departs from most studies of
executive compensation. Unlike the annual surveys in the business press, for example,
we do not focus on this year’s levels of cash compensation or cash compensation plus
stock options exercised. Instead, we apply regression analysis to 15 years’ worth of data

---

Jensen and Murphy
5
May 1990
and estimate how changes in corporate performance affect CEO compensation and
wealth over all relevant dimensions.
We ask the following questions: How does a change in performance affect
current cash compensation, defined as changes in salary and bonus over two years? What
is the “wealth effect” (the present value) of those changes in salary and bonus? How
does a change in corporate performance affect the likelihood of the CEO being dismissed,
and what is the financial impact of this new dismissal probability? Finally, how does a
change in corporate performance affect the value of CEO stock options and shares,
whether or not the CEO exercised the options or sold the shares? (For a discussion of our
methodology, see the insert, “How We Estimate Pay for Performance.”)
The table “The Weak State of Pay for Performance” provides a detailed review of
our main findings for a subsample of CEOs in the 250 largest publicly held companies.
Together, these CEOs run enterprises that generate revenues in excess of $2.2 trillion and
employ more than 14 million people. The results are both striking and troubling. A
$1,000 change in corporate market value (defined as share price appreciation plus
dividends) corresponds to a two-year change in CEO salary and bonus of less than a
dime; the long-term effects of that change add less than 45 cents to the CEO’s wealth. A
$1,000 change in corporate value translates into an estimated median change of a nickel
in CEO wealth by affecting dismissal prospects. At the median, stock options add
another 58 cents worth of incentives. Finally, the value of shares owned by the median
CEO changes by 66 cents for every $1,000 increase in corporate value. All told, for the
median executive in this sub-sample, a $1,000 change in corporate performance translates
into a $2.59 change in CEO wealth. The table also reports estimates for CEOs at the
lower and upper bounds of the middle two quartiles of the sample. (For an extensive
review and comparison of the pay-for-performance relation for individual CEOs, see “A
New Survey of Executive Compensation” that follows this article.)

---

Jensen and Murphy
6
May 1990
THE WEAK STATE OF PAY FOR PERFORMANCE
ESTIMATES FOR CEOS IN THE
250 LARGEST COMPANIES
A $1,000 CHANGE IN SHAREHOLDER WEALTH CORRESPONDS TO
MEDIAN
MIDDLE 50%
Change in this year’s and next year’s salary and bonus
$0.067
$0.01 to $0.18
Present value of the two-year change in salary and bonus
0.44
0.05 to 1.19
Change in the value of stock options
0.58
0.16 to 1.19
Wealth effect for change in likelihood of dismissal
0.05
0.02 to 0.14
Total change in all pay-related wealth
$1.29
$0.43 to $2.66
Change in value of direct stockholdings
0.66

0.25 to 1.98

Total change in CEO wealth
$2.59
$0.99 to $5.87
Note: The median individual components do not add to the median total change in CEO wealth since sums of medians do
not in general equal the median of sums.
This degree of pay-for-performance sensitivity for cash compensation does not
create adequate incentives for executives to maximize corporate value. Consider a
corporate leader whose creative strategic plan increases a company’s market value by
$100 million. Based on our study, the median CEO can expect a two-year increase in
salary and bonus of $6,700—hardly a meaningful reward for such outstanding
performance. His lifetime wealth would increase by $260,000—less than 4% of the
present value of the median CEO’s shareholdings and remaining lifetime salary and
bonus payments.1
Or consider instead a CEO who makes a wasteful investment—new aircraft for
the executive fleet, say, or a spanking addition to the headquarters building—that benefits
him but diminishes the market value of the company by $10 million. The total wealth of
this CEO, if he is representative of our sample, will decline by only $25,900 as a result of
this misguided investment—not much of a disincentive for someone who earns on
average $20,000 a week.



1 The median CEO in our sample holds stock worth $2.4 million. The average 1988 salary and bonus for
the CEOs in our sample was roughly $1 million. At a real interest rate of 3%, the present value of the
salary and bonus for the next five years to retirement (the average for the sample) is $4.6 million. Thus
total lifetime wealth from the company is $7 million.

---

Jensen and Murphy
7
May 1990
One way to explore the realities of CEO compensation is to compare current
practices with the three principles that we outlined earlier. Let’s address them one at a
time.
CEOs should own substantial amounts of company stock. The most powerful link
between shareholder wealth and executive wealth is direct ownership of shares by the
CEO. Most commentators look at CEO stock ownership from one of two
perspectives—the dollar value of the CEO’s holdings or the value of his shares as a
percentage of his annual cash compensation. But when trying to understand the incentive
consequences of stock ownership, neither of these measures counts for much. What
really matters is the percentage of the company’s outstanding shares the CEO owns. By
controlling a meaningful percentage of total corporate equity, senior managers experience
a direct and powerful “feedback effect” from changes in market value.
Think again about the CEO adding jets to the corporate fleet. The stock-related
“feedback effect” of this value-destroying investment—about $6,600—is small because
this executive is typical of our sample, in which the median CEO controls only .066% of
the company’s outstanding shares. Moreover, this wealth loss (about two days’ pay for
the average CEO in a top-250 company) is the same whether the stockholdings represent
a big or small fraction of the CEO’s total wealth.
But what if this CEO held shares in the company comparable to say, Warren
Buffet’s stake in the Berkshire Hathaway conglomerate? Buffet controls, directly and
indirectly, about 45% of Berkshire Hathaway’s equity. Under these circumstances, the
stock-related feedback effect of a $10 million decline in market value is nearly $4.5
million—a much more powerful incentive to resist wasteful spending.
Moreover, these differences in CEO compensation are associated with substantial
differences in corporate performance. From 1970 through 1988, the average annual
compound stock return on the 25 companies with the best CEO incentives (out of the
largest 250 companies examined in our survey) was 14.5%, more than one-third higher

---

Jensen and Murphy
8
May 1990
than the average return on the 25 companies with the worst CEO incentives. A $100
investment in the top 25 companies in 1970 would have grown to $1,310 by 1988, as
compared with $702 for a similar investment in the bottom 25 companies.
As a percentage of total corporate value, CEO share ownership has never been
very high. The median CEO of one of the nation’s 250 largest public companies own
shares worth just over $2.4 million—again, less than 0.07% of the company’s market
value. Also, 9 out of 10 CEOs own less than 1% of their company’s stock, while fewer
than 1 in 20 owns more than 5% of the company’s outstanding shares.
It is unreasonable to expect all public-company CEOs to own as large a
percentage of their company’s equity as Warren Buffett’s share of Berkshire Hathaway.
Still, the basic lesson holds. The larger the share of company stock controlled by the
CEO and senior management, the more substantial the linkage between shareholder
wealth and executive wealth. A few companies have taken steps to increase the share of
corporate equity owned by senior management. Employees of Morgan Stanley now own
55% of the firm’s outstanding equity. Companies such as FMC and Holiday have used
leveraged recapitalizations to reduce the amount of outstanding equity by repurchasing
public shares, and thus allow their managers to control a bigger percentage of the
company. After FMC adopted its recapitalization plan, for example, employee ownership
increased from 12% to 40% of outstanding equity. These recapitalizations allow
managers to own a bigger share of their company’s equity without necessarily increasing
their dollar investment.
Truly giant companies like IBM, General Motors, or General Electric will never
be able to grant their senior executives a meaningful share of outstanding equity. These
and other giant companies should understand that this limitation on executive incentives
is a real cost associated with bigness.
Cash compensation should be structured to provide big rewards for outstanding
performance and meaningful penalties for poor performance. A two-year cash reward of

---

Jensen and Murphy
9
May 1990
less than 7 cents for each $1,000 increase in corporate value (or, conversely, a two-year
penalty of less than 7 cents for each $1,000 decline in corporate value) does not create
effective managerial incentives to maximize value. In most large companies, cash
compensation for CEOs is treated like an entitlement program.
There are some notable exceptions to this entitlement pattern. The cash
compensation of Walt Disney CEO Michael Eisner, whose pay has generated such
attention in recent years, is more than ten times more sensitive to corporate performance
than the median CEO in our sample. Yet the small number of CEOs for whom cash
compensation changes in any meaningful way in response to corporate performance
shows how far corporate America must travel if pay is to become an effective incentive.
Creating better incentives for CEOs almost necessarily means increasing the
financial risk CEOs face. In this respect, cash compensation has certain advantages over
stock and stock options. Stock-based incentives subject CEOs to vagaries of the stock
market that are clearly beyond their control. Compensation contracts based on company
performance relative to comparable companies could provide sound incentives while
insulating the CEO from factors such as the October 1987 crash. Although there is some
evidence that directors make implicit adjustments for market trends when they set CEO
pay, we are surprised that compensation plans based explicitly on relative performance
are so rare.2
The generally weak link between cash compensation and corporate performance
would be less troubling if CEOs owned a large percentage of corporate equity. In fact, it
would make sense for CEOs with big chunks of equity to have their cash compensation
less sensitive to performance than CEOs with small stockholdings. (For example,
Warren Buffet’s two-year cash compensation changes by only a penny for every $1,000
increase in market value.) In some cases, it might even make sense for pay to go up in



2 See Gibbons and Murphy [, 1990 #664, p. 30-S].

---