Performance Pay and Top-Management Incentives
Author(s): Michael C. Jensen and Kevin J. Murphy
Source:
The Journal of Political Economy,
Vol. 98, No. 2 (Apr., 1990), pp. 225-264
Published by: The University of Chicago Press
Stable URL: http://www.jstor.org/stable/2937665
Accessed: 15/10/2009 14:02
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=ucpress.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to The
Journal of Political Economy.
http://www.jstor.org
Performance
Pay
and Top-Management
Incentives
Michael
C. Jensen
Harvard University
Kevin
J. Murphy
University of
Rochester
Our estimates of
the
pay-performance
relation
(including
pay, op-
tions, stockholdings,
and dismissal)
for
chief
executive officers
indi-
cate that CEO
wealth
changes $3.25
for every $1,000
change
in
shareholder
wealth.
Although
the
incentives generated
by
stock
ownership
are
large relative
to
pay
and dismissal
incentives,
most
CEOs
hold
trivial fractions of
their firms'
stock,
and
ownership
levels
have
declined
over the
past 50 years.
We hypothesize
that public
and
private
political
forces impose constraints
that
reduce
the
pay-
performance
sensitivity.
Declines
in
both
the pay-performance
rela-
tion
and the level of CEO pay
since the
1930s
are consistent
with this
hypothesis.
The conflict of interest between shareholders of
a
publicly
owned
corporation
and
the
corporation's
chief executive officer
(CEO)
is a
We
have
benefited
from
the assistance
of
Stephanie
Jensen,
Natalie
Jensen,
Mary
Rojek,
and
Mike Stevenson
and from
the comments of Sherwin Rosen
(the
editor),
George
Baker,
Carliss Baldwin,
Ray
Ball, Gary Becker, Joseph Bower, James
Brickley,
Jeffrey
Coles,
Harry
DeAngelo,
Robert Gibbons,
Gailen
Hite,
Clifford Holderness,
Robert Kaplan,
Steven
Kaplan,
Edward
Lazear,
Richard Leftwich, John Long,
Jay
Lorsch, John
McArthur,
Paul
MacAvoy,
Kenneth
McLaughlin,
Kenneth Merchant,
Andrall Pearson,
Ronald
Schmidt,
G.
William
Schwert,
Robert
Simons,
Jerold
Warner,
Ross Watts,
and Jerold
Zimmerman.
This
research
is supported
by
the Division
of
Research,
Harvard
Business
School;
the Managerial
Economics
Research
Center,
Uni-
versity of Rochester;
and the
John
M. Olin
Foundation.
Journal of
Political
Economy,
1990,
vol.
98,
no. 2]
?
1990
by
The University
of
Chicago.
All rights
reserved. 0022-3808/90/9802-0006$01.50
225
226
JOURNAL
OF POLITICAL
ECONOMY
classic
example of a
principal-agent
problem. If
shareholders had
complete
information
regarding
the
CEO's activities and the
firm's
investment
opportunities,
they could
design
a
contract
specifying
and
enforcing the
managerial
action to
be taken
in
each
state of
the world.
Managerial actions and
investment
opportunities are
not,
however,
perfectly observable
by
shareholders;
indeed,
shareholders do not
often know
what actions
the
CEO
can
take or which of
these actions
will
increase shareholder
wealth. In
these
situations,
agency
theory
predicts that
compensation
policy
will be
designed
to give
the
man-
ager incentives to select
and
implement
actions that increase
share-
holder
wealth.
Shareholders want
CEOs to
take
particular
actions-for
example,
deciding
which
issue to work
on,
which
project to
pursue, and which
to
drop-whenever
the expected
return on
the
action exceeds
the
expected
costs. But
the
CEO
compares
only
his
private
gain
and
cost
from
pursuing a
particular
activity. If
one
abstracts
from
the effects
of
CEO risk
aversion,
compensation
policy
that
ties the
CEO's
welfare
to
shareholder
wealth
helps
align
the
private
and social costs
and
benefits of
alternative actions
and thus
provides incentives for
CEOs
to take
appropriate
actions.
Shareholder
wealth
is
affected
by
many
factors
in
addition
to the
CEO,
including
actions of other
executives
and
employees,
demand and
supply
conditions,
and
public
policy.
It
is
appropriate,
however, to
pay
CEOs on
the
basis of
shareholder
wealth since that
is the
objective of
shareholders.
There are
many
mechanisms
through
which
compensation
policy
can
provide
value-increasing
incentives,
including performance-
based
bonuses
and
salary
revisions,
stock
options,
and
performance-
based
dismissal
decisions. The
purpose of this
paper is to
estimate the
magnitude
of
the
incentives
provided
by each of
these mechanisms.
Our estimates
imply
that
each
$1,000
change
in
shareholder wealth
corresponds
to an
average
increase
in
this
year's and next
year's
salary
and
bonus of about
two cents.
We also
estimate the
CEO wealth conse-
quences associated with
salary
revisions,
outstanding
stock
options,
and
performance-related
dismissals;
our
upper-bound estimate of the
total
change
in
the
CEO's wealth
from these
sources that are
under
direct
control of the
board of
directors
is about
75?
per $1,000
change
in
shareholder wealth.
Stock
ownership
is
another
way
an executive's
wealth varies with
the value of the firm. In
our
sample
CEOs hold a
median
of
about
0.25 percent
of their firms'
common
stock,
including
exercisable stock
options
and
shares held
by
family
members or
connected trusts. Thus
the
value of the stock
owned
by
the median
CEO
changes
by $2.50
whenever the
value of the firm
changes by
$1,000.
Therefore,
our
final all-inclusive
estimate of
the
pay-performance
sensitivity-
TOP-MANAGEMENT
INCENTIVES
227
including compensation,
dismissal,
and
stockholdings-is
about
$3.25
per $1,000 change
in shareholder
wealth.
In large firms
CEOs
tend to own
less stock
and have
less
compensa-
tion-based
incentives
than
CEOs
in smaller
firms.
In particular,
our
all-inclusive
estimate
of
the pay-performance
sensitivity
for
CEOs
in
firms
in the top
half of
our sample
(ranked
by market
value)
is $1.85
per $1,000,
compared
to $8.05
per $1,000
for CEOs
in firms
in the
bottom
half of
our sample.
We
believe
that our results are inconsistent
with
the
implications
of
formal
agency
models
of
optimal
contracting.
The empirical
relation
between
the pay
of top-level
executives
and
firm
performance,
while
positive
and statistically
significant,
is small
for an
occupation
in
which
incentive pay
is expected
to play
an important
role.
In
addition,
our
estimates suggest
that dismissals
are
not
an
important
source of
mana-
gerial
incentives
since
the
increases
in
dismissal probability
due
to
poor
performance
and
the penalties
associated
with dismissal
are
both
small.
Executive
inside
stock
ownership
can provide
incentives,
but
these
holdings
are not
generally
controlled
by
the corporate
board,
and the
majority
of top
executives
have small
personal
stockholdings.
Our results
are consistent with several
alternative
hypotheses;
CEOs
may
be unimportant
inputs
in
the
production
process,
for ex-
ample,
or their actions
may
be easily
monitored
and evaluated by
corporate
boards.
We
offer an
additional
hypothesis
relating
to
the
role
of
political
forces
in the contracting
process that
implicitly
regu-
late executive
compensation
by
constraining
the type
of contracts
that
can
be written between
management
and
shareholders.
These
polit-
ical
forces, operating
both
in the
political
sector and
within
organiza-
tions,
appear
to be
important but
are difficult
to
document
because
they
operate
in informal
and indirect
ways.
Public
disapproval
of high
rewards seems
to have
truncated
the upper
tail of
the earnings
distri-
bution
of
corporate
executives.
Equilibrium
in the
managerial
labor
market
then
prohibits
large penalties
for
poor performance,
and as
a result the
dependence
of pay
on
performance
is
decreased.
Our
findings
that the pay-performance
relation,
the raw
variability
of pay
changes, and
inflation-adjusted
pay levels
have declined
substantially
since
the
1930s
are consistent
with
such implicit
regulation.
I.
Estimates
of the
Pay-Performance
Sensitivity
We
define
the
pay-performance
sensitivity, b, as
the dollar change
in
the CEO's wealth associated
with
a dollar
change
in
the
wealth
of
shareholders.
We interpret higher
b's as
indicating
a closer
alignment
of
interests
between
the CEO
and his
shareholders.
Suppose,
for
example,
that a CEO
is
considering a
nonproductive
but
costly "pet
228
JOURNAL OF POLITICAL
ECONOMY
project" that he values at $100,000 but that will
diminish the value of
his
firm's
equity by $10
million.
The CEO will avoid
this project
if
his
pay-performance sensitivity
exceeds b = .01
(through
some
combina-
tion of incentive compensation, options, stock
ownership, or probabil-
ity of being fired for poor stock price performance)
but will adopt the
project
if b
<
.01.
Incentives Generated by Cash Compensation
The pay-performance sensitivity is estimated by
following all 2,213
CEOs listed
in
the Executive
Compensation
Surveys published
in
Forbes from 1974 to 1986. These surveys include
executives serving
in
1,295 corporations, for a total of 10,400 CEO-years of
data.
We match
these compensation data to fiscal year corporate performance data
obtained from the data files of the Compustat and the Center for
Research
in
Security Prices (CRSP). After observations with
missing
data are
eliminated,
the final
sample
contains
7,750
yearly
"first
dif-
ferences"
in
compensation
and includes
1,688
executives from
1,049
corporations.
Fiscal
year
stock returns
are unavailable
for
219
of the
7,750 observations; calendar-year returns are used
in
these cases.
(Deleting
these
219
observations does not affect the
results.)
All
mon-
etary
variables are
adjusted
for inflation
(using
the consumer
price
index for the
closing
month of
the fiscal
year) and represent
thousands of 1986 constant
dollars.
Table
1
summarizes
estimates of the relation between
CEO
cash
compensation
and firm
performance
as measured
by
the
change
in
shareholder wealth. Column
1
of table
1
reports
estimated coeffi-
cients
from
the
following least-squares regression:
A(CEO
salary
+
bonus)t
=
a
+ bA(shareholder
wealth)t. (1)
The
change
in
shareholder
wealth variable is defined as
rt~tV
1,
where
rt
is the
inflation-adjusted
rate of
return on common stock realized
in
fiscal
year t,
and Vi
I
is the firm
value at the end of the
previous year.
Our
measure
of
firm
performance
is
subject
to two
qualifications.
First, performance should be evaluated before
compensation expense,
and
yet
rt~tV
1
is the
change
in firm
value after
compensation
expense;
the associated
bias
in
our estimates is small,
however, because CEO
pay changes
are
tiny
relative to
changes
in
firm
value.
Second,
our
measure of
performance ignores payments
to
capital.
When
capital
is
an
important input,
a better
performance
measure is
rt~tV
1
-
ftKt-
1,
where
ft
and
Kt
-1
are
the risk-free
interest rate for
period
t
and the
opportunity
cost
of
the
capital
stock at the
beginning
of
period
t. Since
f
and shareholder return r
tend to be
uncorrelated,
this
adjustment
will not
substantially
affect our
estimates. Fama and Schwert
(1977)